<%BANNER%>

Pressure Ulcer Prevention Research

Permanent Link: http://ufdc.ufl.edu/UFE0024391/00001

Material Information

Title: Pressure Ulcer Prevention Research
Physical Description: 1 online resource (133 p.)
Language: english
Creator: Peterson, Matthew
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: bed, bioimpedance, buttock, care, clinical, decubitus, edema, error, feedback, icu, impedance, interface, mapping, medical, medicare, monitoring, patient, pillow, pressure, prevention, risk, sacral, sacrum, skin, sore, turning, ulcer, vap, ventillator, wedge, xsensor
Biomedical Engineering -- Dissertations, Academic -- UF
Genre: Biomedical Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Pressure ulcers are painful, debilitating, and severely compromise an individual's health by increasing morbidity and mortality in terms of increased length of stay, risk of infection, and the need for additional surgical procedures. They are also considered to be preventable and incidence is used to indicate quality of care. Numerous devices have been developed to prevent their occurrence and protocols that involve regular turning of patients is a standard of care. Despite the many interventions used to prevent pressure ulcers, there is still a high prevalence. Research comprising this dissertation involved: identifying clinically relevant biomechanical factors likely to increase the risk of pressure ulcer formation in the sacral area, investigating bioimpedance measurements as a feasible method to detect or identify the onset of pressure ulcer formation or skin damage, monitoring the interface pressures of patients at risk for pressure ulcer formation to determine what pressure thresholds are experienced and how they are affected by body position, and to develop a protocol that is safest for patients by helping prevent pressure ulcers formation. Results from healthy subject studies demonstrated that increasing the head of bed increased the interface pressures and the areas subjected to interface pressures of 32 mm Hg or greater. Additionally, lateral turning was not adequate in unloading areas at risk for pressure ulcer formation and these areas always remained at risk despite repositioning. The bioimpedance methods did not succeed as a reliable method to detect pressure ulcer formation. Patient studies confirmed what was observed in healthy subjects, and patients were shown to be at even greater risk by experiencing higher interface pressures and larger areas at risk for pressure ulcer formation. Tissue areas were also shown to be always at risk for several (5+) consecutive hours despite 2-hour turning protocols. Solutions to reduce pressure ulcer incidences involve reducing or eliminating high interface pressures, improving patient positioning, and reducing or eliminating areas that are always at risk for pressure ulcer formation. These goals can be achieved through a number of methods: improved turning protocols, better support devices for lateral turning, providing feedback to caregivers, and continuous interface pressure monitoring.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Matthew Peterson.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Van Oostrom, Johannes H.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024391:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024391/00001

Material Information

Title: Pressure Ulcer Prevention Research
Physical Description: 1 online resource (133 p.)
Language: english
Creator: Peterson, Matthew
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: bed, bioimpedance, buttock, care, clinical, decubitus, edema, error, feedback, icu, impedance, interface, mapping, medical, medicare, monitoring, patient, pillow, pressure, prevention, risk, sacral, sacrum, skin, sore, turning, ulcer, vap, ventillator, wedge, xsensor
Biomedical Engineering -- Dissertations, Academic -- UF
Genre: Biomedical Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Pressure ulcers are painful, debilitating, and severely compromise an individual's health by increasing morbidity and mortality in terms of increased length of stay, risk of infection, and the need for additional surgical procedures. They are also considered to be preventable and incidence is used to indicate quality of care. Numerous devices have been developed to prevent their occurrence and protocols that involve regular turning of patients is a standard of care. Despite the many interventions used to prevent pressure ulcers, there is still a high prevalence. Research comprising this dissertation involved: identifying clinically relevant biomechanical factors likely to increase the risk of pressure ulcer formation in the sacral area, investigating bioimpedance measurements as a feasible method to detect or identify the onset of pressure ulcer formation or skin damage, monitoring the interface pressures of patients at risk for pressure ulcer formation to determine what pressure thresholds are experienced and how they are affected by body position, and to develop a protocol that is safest for patients by helping prevent pressure ulcers formation. Results from healthy subject studies demonstrated that increasing the head of bed increased the interface pressures and the areas subjected to interface pressures of 32 mm Hg or greater. Additionally, lateral turning was not adequate in unloading areas at risk for pressure ulcer formation and these areas always remained at risk despite repositioning. The bioimpedance methods did not succeed as a reliable method to detect pressure ulcer formation. Patient studies confirmed what was observed in healthy subjects, and patients were shown to be at even greater risk by experiencing higher interface pressures and larger areas at risk for pressure ulcer formation. Tissue areas were also shown to be always at risk for several (5+) consecutive hours despite 2-hour turning protocols. Solutions to reduce pressure ulcer incidences involve reducing or eliminating high interface pressures, improving patient positioning, and reducing or eliminating areas that are always at risk for pressure ulcer formation. These goals can be achieved through a number of methods: improved turning protocols, better support devices for lateral turning, providing feedback to caregivers, and continuous interface pressure monitoring.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Matthew Peterson.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Van Oostrom, Johannes H.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024391:00001


This item has the following downloads:


Full Text

PAGE 1

1 PRESSURE ULCER PREVENTION RESEARCH By MATTHEW JAMES PETERSON A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Matthew J. Peterson

PAGE 3

3 To my parents

PAGE 4

4 ACKNOWLEDGMENTS I would first like to thank my parents for all of their support and for believing in me in all of my endeavors. I als o want to thank my family and friends for their support and helping me get through graduate school. In particular, I thank my fianc, Upohar, for without her help I could not have achieved everything that I have accomplished. I thank my advisor, Johannes H van Oostrom, for guiding my research and encouraging me throughout every step towards the completion of my dissertation. I thank Richard Melker, David Paulus, and Brian Sorg for serving on my graduate committee and providing valuable feedback. I thank Wi lhelm Schwab for all his help with the data collection and for always trying to keep my mathematical skills sharp. I thank Nikolaus Gravenstein for his encouragement and guidance of my clinical education and research. I thank Lawrence Caruso for his assist ance and guidance of the clinical studies. I thank the J. Crayton Pruitt Family Department of Biomedical Engineering, the Anesthesiology Department, and the Self Insurance Program for funding my research and education, as well as their respective staffs, i n particular, Kelly Spaulding for her willingness to help with anything and Anita Yeager and Hope Olivo for their help with editing and submitting publications. I thank Jack DiGiovanna for always being able to discuss our research and for his assistance wi th Matlab. I would also like to thank Sungho Oh for his help and assistance with Comsol and Rosalind Sadleir for her expertise with bioimpedance. I also thank everyone else who has helped me along the way to achieve my degree.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................................... 4 LIST OF TABLES ................................................................................................................................ 9 LIST OF FIGURES ............................................................................................................................ 10 ABSTRACT ........................................................................................................................................ 13 CHAPTER 1 SPECIFIC AIMS AND BACKGROUND ................................................................................ 15 Specific Aims .............................................................................................................................. 15 Pressure Ulcer Background ........................................................................................................ 16 Pressure Ulcers ..................................................................................................................... 16 Pressure Ulcer Prevalence and Costs ................................................................................. 17 Etiology and Risk Factors of Pressure Ulcers .................................................................... 18 Pressure Ulcer Locations of Occurrence and Stages of Development ............................. 20 Pressure Ulcer Prevention Methods ................................................................................... 21 Organization of Chapters ............................................................................................................ 23 2 TECHNIQUES USED TO INVESTIGATE THE PRESSURE ULCER PROBLEM: INTERFACE PRESSURE MEASUREMENTS AND BIOIMPEDANCE ............................ 25 Interface Pressure Measurements ............................................................................................... 25 Origins and Appl ications ..................................................................................................... 25 Measurement Devices/Systems .......................................................................................... 27 Review of early methods ............................................................................................. 27 Comparison and analysis of different systems ........................................................... 28 Methods ................................................................................................................................ 29 Proper use ..................................................................................................................... 29 Variability ..................................................................................................................... 30 Limitations .................................................................................................................... 31 Device Evaluation and Setup .............................................................................................. 31 Bioimpedance .............................................................................................................................. 34 Origins and Applications ..................................................................................................... 34 Measurements ...................................................................................................................... 35 Methods ................................................................................................................................ 37

PAGE 6

6 3 IDENTIFY THE CLINICALLY RELEVANT BIOMECHANICAL FACTORS THAT ARE LIKELY TO INCREASE THE RISK OF PRESSURE ULCER FORMATION ......... 39 Clinical Applications ................................................................................................................... 39 Ventilator -Associated Pneumonia ...................................................................................... 39 Turning Effectiveness .......................................................................................................... 40 Interface Pressure Measurement and Other Equipment ........................................................... 41 Subjects ........................................................................................................................................ 42 HOB Study ................................................................................................................................... 43 HOB Protocol and Data Collection .................................................................................... 43 HOB Data Analysis ............................................................................................................. 43 HOB Results ........................................................................................................................ 44 Interface pressures ........................................................................................................ 44 At -risk areas .................................................................................................................. 44 Effect of body habitus .................................................................................................. 45 HOB Discussion .................................................................................................................. 46 HOB Conclusion .................................................................................................................. 49 Lateral Turning Study ................................................................................................................. 50 Lateral Turning Protocol and Data Collection ................................................................... 50 Lateral Turning Data Analysis ............................................................................................ 51 Lateral Turning Results ....................................................................................................... 51 Interface pressures and at risk areas ........................................................................... 51 Triple jeopardy at risk areas ........................................................................................ 52 Effect of body habitus .................................................................................................. 55 Pillows vs. wedges ....................................................................................................... 55 Lateral Turning Discussion ................................................................................................. 56 Lateral Turning Conclusion ................................................................................................ 59 HOB and Lateral Turning Limitations and Future Research ................................................... 59 Limitations ........................................................................................................................... 59 Future Research ................................................................................................................... 60 4 INVESTIGATE THE USE OF BIOIMPEDANCE MEASUREMENTS AS A METHOD TO ASSESS OR DETECT SKIN DAMAGE ........................................................ 62 Research Applications ................................................................................................................ 62 Localized Bioimpedance Analysis ..................................................................................... 62 Detecting Edema with High Resolution Ultrasound ......................................................... 65 Application and Goals ......................................................................................................... 67 COMSOL Modeling ................................................................................................................... 68 Skin ....................................................................................................................................... 68 Skin Model ........................................................................................................................... 69 Results .................................................................................................................................. 73 With and without edema .............................................................................................. 73 Varying edema size ...................................................................................................... 73 Edema displacement .................................................................................................... 76 Discussion ............................................................................................................................ 76

PAGE 7

7 Limitations ........................................................................................................................... 77 Conclusion............................................................................................................................ 77 Dermal Phase Meters .................................................................................................................. 80 Equipment ............................................................................................................................ 81 Product Evaluation ............................................................................................................... 81 F orearm Measurements Sites .............................................................................................. 81 Measurement Protocols ....................................................................................................... 82 Results .................................................................................................................................. 83 DPM 9003 standard vs. articulated probe ................................................................... 83 DPM 9003 vs. Petite .................................................................................................... 84 Moisturizing lotion ....................................................................................................... 84 Blanching ...................................................................................................................... 85 Indenting ....................................................................................................................... 86 Discussion ............................................................................................................................ 86 Forearm measurements ................................................................................................ 86 Moisturizing lotion ....................................................................................................... 87 Blanching and indenting .............................................................................................. 87 Conclusions .......................................................................................................................... 88 Bioimpedance Experiments with Basic Electrode Setups ........................................................ 89 Equipment ............................................................................................................................ 90 Measurement of Electrode Impedance ............................................................................... 90 Protocol ......................................................................................................................... 90 Results and Discussion ................................................................................................ 91 Interface Pressure Detection ............................................................................................... 92 Protocol ......................................................................................................................... 92 Results ........................................................................................................................... 92 Initial electrode setup ................................................................................................... 92 Tetrapolar electrode setup ............................................................................................ 93 Discussion ..................................................................................................................... 93 5 MONITORING THE INTERFACE PRESSURE PROFILES OF PATIENTS AT -RISK FOR PRESSURE ULCER FORMATION ................................................................................ 94 Interface Pressure Measurement and Other Equipment ........................................................... 94 Power Analysis ............................................................................................................................ 95 Protocol ........................................................................................................................................ 95 Subjects ........................................................................................................................................ 96 Data Analysis ............................................................................................................................... 97 Results .......................................................................................................................................... 97 Observations ......................................................................................................................... 97 Data collection .............................................................................................................. 97 HOB elevation .............................................................................................................. 98 Interface Pressures ............................................................................................................... 98 At -Risk Areas ..................................................................................................................... 100 Always At -Risk Areas ....................................................................................................... 102 Triple Jeopardy Areas ........................................................................................................ 104 Discussion .................................................................................................................................. 104

PAGE 8

8 Limitations and Future Research .............................................................................................. 108 Conclusion ................................................................................................................................. 109 6 FINAL DISCUSSION AND CONCLUDING REMARKS .................................................. 110 Recap and Discussion ............................................................................................................... 110 Solutions .................................................................................................................................... 113 Improved Turning Protocols ............................................................................................. 113 Improved Support Devices ................................................................................................ 114 Providing Feedback to Caregivers .................................................................................... 115 Interface Pressure Monitoring ........................................................................................... 116 Pressure magnitudes ................................................................................................... 116 Areas over time .......................................................................................................... 118 System adjustments .................................................................................................... 120 Concluding Remarks ................................................................................................................. 124 LIST OF REFERENCES ................................................................................................................. 126 BIOGRAPHICAL SKETCH ........................................................................................................... 133

PAGE 9

9 LIST OF TABLES Table page 2 1 Comparison of commercially available interface pressure measuring systems. ................ 29 3 1 Comparison of interface pressure measurements between studies. .................................... 50 3 2 Triple jeopardy areas (cm2). .................................................................................................. 55 3 3 Mean turn angles and standard deviations comparing the pillow and wedge techniques. .............................................................................................................................. 56 3 4 Comparison of maximum interface pressure measurements (mm Hg) from 3 studies. ..... 58 4 1 Electrical conductivity values and thicknesses for skin model components. ..................... 70 4 2 Electrode impedance results. ................................................................................................. 91 5 1 Triple jeopardy and always at risk areas. ........................................................................... 103 5 2 Comparison of interface pressures (mm Hg) between healthy subjects and at risk patients. ................................................................................................................................. 107 5 3 Comparison of at -risk areas (cm2) between healthy subjects and at risk patients. .......... 107 5 4 Comparison of triple jeopardy and always at -risk areas between healthy subjects and at risk patients. ..................................................................................................................... 108 6 1 Triple jeopardy and always at risk area results for various resolutions. .......................... 123

PAGE 10

10 LIST OF FIGURES Figure page 1 1 Pressure ulcer locations of occ urrence. ................................................................................. 20 2 1 Allowable pressure vs. time of application for tissue under bony prominences. ............... 26 2 2 Xsensor pad on bed. ............................................................................................................... 33 2 3 Xsensor pad connected to computer setup in clinical setting. ............................................. 33 2 4 Xsensor pressure mapping system in use with display. ....................................................... 34 2 5 Equivalent circuit for biological tissue. ................................................................................ 35 2 6 Homogeneous cylinder. ......................................................................................................... 36 2 7 Tetrapolar technique. ............................................................................................................. 37 2 8 Diagram of an impedance locus depicting the phase angle and its relationship with resistance, reactance, impedance, and frequency. ................................................................ 38 3 1 Peak sacral area interface pressures at selected HOB angles of elevation. ........................ 44 3 2 Interface pressure profiles at various head of b ed (HOB) positions for one subject. ........ 45 3 3 At -risk sacral areas mm Hg at selected HOB angles of elevation. ............................. 46 3 4 Peak peri -sacral area interface pressures. ............................................................................. 52 3 5 At -risk peri -sacral areas. ........................................................................................................ 53 3 6 Peri -sacral area interface pressure profiles .......................................................................... 53 3 7 Triple jeopardy at risk areas using pillows. .......................................................................... 54 3 8 Triple jeopardy at risk areas using wedges. ......................................................................... 54 4 1 Body resistances. .................................................................................................................... 63 4 2 Localized abdominal BIA electrode location and spacing. ................................................. 64 4 3 Dependence of impedence spectra based on subcutaneous fat layer thickness. ................ 64 4 4 SFL impedance correlation from total -population data (model t) at 50 kHz. ................. 64 4 5 High resolution ultrasound images of deep and superficial edema. ................................... 66

PAGE 11

11 4 6 High resolution ultrasound images demonstrating the 3 phases of pressure ulcer development. ........................................................................................................................... 66 4 7 Skin anatomy. ......................................................................................................................... 69 4 8 Skin layers from the COMSOL skin model. ........................................................................ 70 4 9 Skin model shown with electrodes on the skin surface and edema located within the deep ti ssue layer. .................................................................................................................... 71 4 10 Skin model with current density arrows. .............................................................................. 74 4 11 Skin model with current density lines. .................................................................................. 74 4 12 Voltage difference measurements with varying edema size in thin deep tissue layer. ...... 75 4 13 Voltage difference measurements with varying e dema size in thick deep tissue layer. .... 75 4 14 Current density lines with varying edema depth. ................................................................. 78 4 15 Voltage difference measu rements with varying edema depth and lateral placement in thin deep tissue layer. ............................................................................................................. 79 4 16 Voltage difference measurements with varying edema depth and lateral placement in thick deep tissue l ayer. ........................................................................................................... 79 4 17 DPM 9003 standard vs. articulated probe. ............................................................................ 83 4 18 DPM 9003 vs. Petite. ............................................................................................................. 84 4 19 DPM measurements before and after moisturizing lotion. .................................................. 85 4 20 DPM measurements before and after blanching and indenting. ......................................... 86 4 21 Electrode impedance measurement circuit. .......................................................................... 91 5 1 Interface pressure results. ...................................................................................................... 99 5 2 Peak in terface pressures over time for one patient. .............................................................. 99 5 3 Peak interface pressures over time for one patient. ............................................................ 100 5 4 At -risk area results. .............................................................................................................. 101 5 5 At -risk areas over time for one patient. .............................................................................. 101 5 6 At -risk areas over time for one patient. .............................................................................. 102 5 7 Stacked image demonstrating triple jeopardy areas. .......................................................... 105

PAGE 12

12 5 8 Stacked image demonstrating triple jeopardy areas. .......................................................... 106 6 1 Peak interface pressures over time for one patient. ............................................................ 119 6 2 Stacked triple jeopardy images demonstrating changes in resolution for one patient. .... 122

PAGE 13

13 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy PRESSURE ULCER PREVENTION RESEAR CH By Matthew James Peterson May 2009 Chair: Johannes H. van Oostrom Major: Biomedical Engineering Pressure ulcers are painful, debilitating, and severely compromise an individuals health by increasing morbidity and mortality in terms of increased leng th of stay, risk of infection, and the need for additional surgical procedures. They are also considered to be preventable and incidence is used to indicate quality of care. Numerous devices have been developed to prevent their occurrence and protocols tha t involve regular turning of patients is a standard of care. Despite the many interventions used to prevent pressure ulcers, there is still a high prevalence. Research comprising this dissertation involved: identifying clinically relevant biomechanical fac tors likely to increase the risk of pressure ulcer formation in the sacral area, investigating bioimpedance measurements as a feasible method to detect or identify the onset of pressure ulcer formation or skin damage, monitoring the interface pressures of patients at risk for pressure ulcer formation to determine what pressure thresholds are experienced and how they are affected by body position, and to develop a protocol that is safest for patients by helping prevent pressure ulcers formation. Results from healthy subject studies demonstrated that increasing the head of bed increased the interface pressures and the areas subjected to interface pressures of 32 mm Hg or greater. Additionally, lateral turning was not adequate in unloading areas at risk for pre ssure

PAGE 14

14 ulcer formation and these areas always remained at risk despite repositioning. The bioimpedance methods did not succeed as a reliable method to detect pressure ulcer formation. Patient studies confirmed what was observed in healthy subjects, and pati ents were shown to be at even greater risk by experiencing higher interface pressures and larger areas at risk for pressure ulcer formation. Tissue areas were also shown to be always at risk for several (5+) consecutive hours despite 2 -hour turning protoco ls. Solutions to reduce pressure ulcer incidences involve reducing or eliminating high interface pressures, improving patient positioning, and reducing or eliminating areas that are always at risk for pressure ulcer formation. These goals can be achieved t hrough a number of methods: improved turning protocols, better support devices for lateral turning, providing feedback to caregivers, and continuous interface pressure monitoring.

PAGE 15

15 CHAPTER 1 SPECIFIC AIMS AND BA CKGROUND Specific Aims Preventing pressure u lcers and reducing their incidence is an ongoing challenge. A pressure ulcer is costly in regards to the patients health as well as monetarily for treatment, so prevention is the goal. Various factors play a role in pressure ulcer formation, but high pres sure applied over a length of time will inevitably cause tissue damage ( Reswick and Rog ers, 1976; Stekelengburg Oomens, and Bader 2005; Swain, 2005). The research in this dissertation seeks to identify clinically relevant biomechanical factors that are likely to lead to an increased risk of pressure ulcer formation near the sacrum and surrounding tissues (buttocks and trochanters) and to identify or develop a method(s) to mitigate these factors. The focus is on the sacrum and surrounding tissues because t hese are the sites of most frequent pressure ulcer occurrence (Amlung, Miller, and Bosley, 2001; Dealey, 1991). Techniques, consisting of interface pressure mapping and bioimpedance measurements were used and investigated, respectively, in attempting to d etermine the clinically relevant biomechanical factors. Clinically relevant studies implementing interface pressure mapping were conducted with healthy subjects. Bioimpedance measurements were investigated as to whether or not they could be a reliable method for measuring damage, or changes in or beneath the skin surface, to help detect the onset of pressure ulcer formation. In addition to the healthy subject studies, patients at risk for pressure ulcers were monitored to determine what types of interface p ressures actual patients are subjected to during their stay in the hospital. Ultimately, findings from these investigations were then applied to develop the best method(s) for preventing pressure ulcers with the strict aim of focusing on patient safety.

PAGE 16

16 T he specific aims of for this dissertation are defined below: Specific Aim 1: Identify clinically relevant biomechanical factors that are likely to increase the risk of pressure ulcer formation around the sacrum and surrounding tissues. Specific Aim 2: Inve stigate the use of bioimpedance measurements as a method to assess or detect skin damage, or changes in or beneath the skin surface, to help detect the onset of pressure ulcer formation. Specific Aim 3: Obtain interface pressure data from patients at risk for pressure ulcer formation to determine what interface pressures are actually experienced and how they are affected by body position. Specific Aim 4: Develop a device, system, or other protocol that is safest for at risk patients in aims to help prevent the onset of pressure ulcer formation. Pressure Ulcer Background Pressure Ulcers Pressure ulcers are painful, debilitating, and severely compromise an individuals health by increasing morbidity and mortality in terms of increased length of stay, risk of i nfection, and the need for additional surgical procedures. Also known as pressure sores, decubitus ulcers, and bedsores, pressure ulcers are a common problem in hospitals and nursing homes. They increase the nursing workload by 50% (Barratt, 1987), and, in patients with clinically relevant pressure ulcers, the length of hospital stay increases by an average of 11 days (Lapsley and Vogels 1996). Generally, pressure ulcers are thought to be preventable, making this an important patient -safety and risk manage ment issue. It is a concern for patients, but also for hospitals and caregivers alike, as the incidence of pressure ulcers is used as an indicator of quality of care (Hobbs, 2004). Unfortunately, there are few objective data regarding methods of prevention, and current recommendations are based largely on expert opinion ( American Thoracic Society 1992; Wound Ostomy and Continence Nurses Society 2003).

PAGE 17

17 Pressure Ulcer Prevalence and Costs The prevalence of pressure ulcers can range anywhere from 0% to over 38% depending on the sector, from general or university hospitals to home care to nursing homes (Agostini, Baker, and Bogardus, 2000; Bennett et al., 2000; Haalboom, 2005; Hobbs, 2004; Hofman et al., 1994; Lyder, 2003; Milne and Pagnamenta, 2004; Quintaval le et al., 2006; Reddy, Gill, and Rochon, 2006; Stimler and Pryor, 1999; Weststrate and Bruining, 1996). Among intensive care unit (ICU) patients, prevalence has been reported to be roughly 14% in short -stay units (average length of stay 4.5 days) and 42% in long-stay units (average length of stay 12.8 days) (Weststrate and Bruining, 1996). At Shands Hospital UF, of the hospital acquired pressure ulcers documented from January 2005 through September 2006, the majority occurred in ICU patients (Shands, 2005, 2006). Not only are pressure ulcers a serious health threat to the patient, but they are extremely costly to the patient and the hospital or caregiver as well. It has been estimated that $11 billion is spent on pressure ulcer treatment each year in the U. S. (Reddy et al., 2006). Similar high treatment costs are seen around the globe (Clark and Price, 2004; Gebhardt, 2004; Haalboom, 2005; Hofman et al., 1994; Milne and Pagnamenta, 2004). The cost to manage one full -thickness ulcer can be as much as $70,000 (Reddy et al., 2006). In addition to the costs, pressure ulcers are considered preventable and have become a liability for hospitals and caregivers (Agostini et al., 2000; Clark and Price, 2004). Incidence of pressure ulcers is an indicator of quality of c are as outlined by programs conforming to Medicare and Medicaid requirements (Agostini et al., 2000; Hobbs, 2004). Several lawsuits have been filed against hospitals and caregivers by patients that have developed pressure ulcers. Many of these lawsuits hav e resulted in the favor of the afflicted plaintiffs under the claim of malpractice or negligence (Bennett et al., 2000; Milne and Pagnamenta, 2004). In the fifteen year span from 1983 to 1997, settlements in the U.S. ranged in awards of $2,200 to $65 milli on, with an average

PAGE 18

18 award of $250,000 (Agostini et al., 2000; Bennett et al., 2000). Furthermore, the Centers for Medicare and Medicaid Services (CMS) have recently made a significant change in policy and have identified eight hospital acquired conditions in their final rule, FY2008, in which they will no longer pay, one of which is pressure ulcers (Medicare Program, 2007). CMS notes that pressure ulcers meet the burden criteria, as they are a high cost, high volume problem. The effects of this change are y et to be realized. Etiology and Risk Factors of Pressure Ulcers By definition from the National Pressure Ulcer Advisory Panel (NPUAP), a pressure ulcer is localized injury to the skin and/or underlying tissue usually over a bony prominence, as a result of pressure, or pressure in combination with shear and/or friction. A number of contributing or confounding factors are also associated with pressure ulcers; the significance of these f actors is yet to be elucidated (NPUAP, 2007) High pressures can lead to i schemia of the affected tissue (Colin et al. 1996) resulting in a lack of oxygen transport and nutrients to the tissue. Oxygen delivery is compromised when the pressure applied over an area is greater than the capillary -closing pressure. Normal capillary pressures in the body range from about 10 to 30 mm Hg (Guyton, 2000). Lower capillary-closing pressure values have been observed in unhealthy patients (Dealey, 1995). Thirtytwo mm Hg is the capillary closing pressure initially measured by Landis (1930) an d has become the commonly quoted value (Bouten et al. 2003; Swain, 2005). When interface pressures, or the pressure between the patients body and his/her supporting surface, exceed these limits, reduced blood flow, accumulation of metabolites via lymphat ic occlusion, and impairment of tissue reperfusion may occur, all of which can damage the tissue (Gebhardt, 2005; Reddy, 1990; Rithalia and Gonsalkorale 1998).

PAGE 19

19 Not only is the magnitude of the pressure a factor, but its duration is as well (Colin et al. 1996; Rithalia and Gonsalkorale 2000). The longer pressure is applied to one area, the more likely pressure ulcers will develop. One of the first studies demonstrating that pressure and time affected pressure ulcer formation was by Kosiak (1961) This phe nomenon was further studied by Reswick and Rogers (1976) where they measured skin -cushion interface pressures on human volunteers and patients that demonstrated that there is an inverse relationship between pressure and time (Stekelengburg et al. 2005; Sw ain, 2005); the greater the pressure, the less time needed for pressure ulcer formation. Additional contributing factors for pressure ulcer formation are shear and frictional forces. Rubbing, which occurs between a patient and his/her bed, is a frictional force that can lead to additional forces on underlying tissues. Shearing forces can occur when the patients skin sticks to his/her support surface and underlying tissue movement results. This shearing causes additional stresses on the underlying tissue ma king it more susceptible to damage. S tudies on pigs by Dinsdale (1974) showed that the presence of friction significantly increased the formation of pressure ulcers, and studies by Goldstein and Sanders (1998) showed that skin breakdown occurred more rapidly as shear was increased. Consequently, individuals who are at risk of developing pressure ulcers are those that are in an altered state of consciousness and those that are physically unable to reposition themselves. Groups of patients that immediately fa ll into this risk are those with spinal cord injuries, paralysis, burn patients (Still et al. 2003), mechanically ventilated patients (Pender and Frazier 2005), and in some instances the weak or elderly. Patients that have high pain thresholds or those o n pain medication can also be at risk, as they may not feel the need to periodically reposition themselves.

PAGE 20

20 Pressure Ulcer Locations of Occurrence and Stages of Development Pressure sores, as mentioned above, are caused by unrelieved pressure. Pressure ulc ers are most commonly found over and around bony prominences, locations where interface pressures are the greatest (Colin et al. 1996; Rithalia and Gonsalkorale 2000). The majority of all pressure sores are found in the gluteal and sacral regions (Amlung et al., 2001; Dealey, 1991), notably around the sacrum, coccyx, and ischial tuberosities (Figure 1 1) The next most -prevalent locations of pressure ulcers are found in the lower extremity, primarily the back of the heel, with the remaining found elsewher e on the body such as on the back, upper extremities, or the head (Amlung et al. 2001; Dealey, 1991). Figure 1 1. Pressure ulcer locations of occurrence. Figures from surveys by Dealey. Source: Dealey, 1991. T here are six stages of pressure ulcer c lassification (traditionally four), ranging from suspected injury and initial redness to full -thickness tissue loss and necrosis. Despite the numbered stages of severity, the development and the healing of pressure ulcers do not

PAGE 21

21 necessarily pass consecutiv ely from one stage to the next. All six stages are defined a s follows (Source: NPUAP, 2007): Suspected Deep Tissue Injury: Purple or maroon localized area of discolored intact skin or blood -filled blister due to damage of underlying soft tissue from pressure and/or shear. The area may be preceded by tissue that is painful, firm, mushy, boggy, warmer or cooler as compared to adjacent tissue. Stage I: Intact skin with non-blanchable redness of a localized area usually over a bony prominence. Darkly pigmented skin may not have visible blanching; its color may di ffer from the surrounding area. Stage II: Partial thickness loss of dermis presenting as a shallow open ulcer with a red pink wound bed, without slough. May also present as an intact or open /ruptured ser um filled blister. Stage III: Full thickness tissue loss. Subcutaneous fat may be visible but bone, tendon or muscle are not exposed. Slough may be present but does not obscure the depth of tissue loss. May include undermining and tunneli ng. Stage IV: Full thickness tissue loss with exposed bone, tendon or muscle. Slough or eschar may be present on some parts of the wound bed. Often inc lude undermining and tunneling. Unstageable: Full thickness tissue loss in which the base of the ulcer is covered by slough (yellow, tan, gray, green or brown) and/or eschar (tan, br own or black) in the wound bed. Pressure Ulcer Prevention Methods There are many contributing factors that tend to make a patient more susceptible to pressure ulcers. By noting these factors, sever al assessment tools, besides age, have been developed to try to prioritize those patients who are at higher risk. The Braden scale and the Norton scale are widely used for pressure ulcer prevention (Agostini et al. 2000; Bergstrom, 2005). The Braden scale assesses six parameters: activity, dietary intake, mobility, sensory perception, friction, and skin moisture; the Norton scale assesses five parameters: activity, physical condition, mobility, mental status, and incontinence (Agostini et al. 2000). These tools

PAGE 22

22 can be good predictors of at risk patients but have not been prove n to be any more effective than clinical judgment (Gebhardt, 2004) and are by no means fail -safe. Once the at risk patients have been identified, there are several options to help pro vide the best possible care for these patients. There are a wide variety of devices and methods that are used to prevent pressure ulcers, from pressure reducing and alternating air pressure mattresses to turning practices The following device guide, from Dealey (1995), identifies and describes various types of pressure reducing and relieving devices. Starting with the air support systems, various devices include mattresses, beds, and overlays. Alternating pressure air systems are the most common and consis t of a dynamic pressure relieving system that alternates inflation among a variable number of cells of a mattress or overlay. Some systems are also sensitive to the patients weight and adjust accordingly. Other pressure reducing or support systems include fiber, foam, gel, and water. The fiber support systems are overlays made from silicon hollow core fiber that look like a duvet and are divided into diagonal segments. The foam support systems are either overlays or mattresses that are made of multiple lay ers of foam with various densities that allow the patient to sink into the foam to distribute his/her weight. The gel systems are overlays that contain thick polymer gels or mattresses that consist of a gel that flows from cell to cell. Lastly, the water s upport systems consist of beds and overlays that attempt to use the buoyant effec t of water to equalize interface pressures. An alternative to pressure relieving or reducing devices, is manual turning or repositioning of patients. Turning patients regularl y every two hours, q2 hour turning (q2h), to reduce and redistribute interface pressures is one of the most common pressure ulcer prevention methods and is considered a standard of care (Bergstrom, 2005; Colin et al. 1996; Defloor De Bacquer, and Grypdo nck 2005; Dini, Bertone, and Romanelli 2006; Lyder, 2003; Ri thalia and

PAGE 23

23 Gonsalkorale, 1998; Vanderwee et al., 2007). However, the evidence -base to support this practice of turning frequency is unclear as the origin of this basis is considered by some to o riginate from the length of time it took nurses to work their way around the hospital ward during the Crimean War (18531856) ( Hagisawa and Ferguson -Pell 2008)1. Organization of Chapters The research presented in this dissertation is organized into six chapters. Chapter 1 describes the specific research aims and background information about the pressure ulcer problem as well as providing an overview of the document. Chapter 2 provides an introduction to the research methods that were conducted, including i nterface pressure measurements and bioimpedance techniques. Chapter 3 includes research that was conducted on healthy subjects to identify the clinically relevant biomechanical factors that are likely to increase the risk of pressure ulcer formation. Chapt er 4 describes how bioimpedance techniques were investigated as a potential method to assess or detect skin damage. Chapter 5 entails the research involving patients at risk for pressure ulcer formation. Chapter 6 consists of the final discussion and concl uding remarks which summarize the findings of this dissertation and provide potential solutions and recommendations to help prevent the onset of pressure ulcer formation. Various parts of this dissertation have been published or are in the process of publi cation. The background information about the pressure ulcer problem provided in Chapter 1, along with the preliminary results from the research detailed in Chapter 3, has been published as a book chapter. The research involving healthy subjects as describe d in Chapter 3 has been published in 1 In a review of early literature (prior to 1978), Hagisa wa and FergusonPell investigated the origin of the two hour turning regimen. They conclude that no evidence supporting the twohour turning frequency could be clearly identified, but suggest that it could have come from Guttmans articles about rehabilita ting the injured spinal cord or from animal experiments by Husain and Kosiak. No compelling scientific evidence could be found in early literature as to why two hour turning was optimal for preventing pressure ulcer formation in humans.

PAGE 24

24 one journal and submitted to another. For the work regarding ICU patients as described in Chapter 5 a publication is in preparation. Peterson M J Fitzgerald K E ., Caruso L J ., 2008. Prevention of p ressure i nj uries in the intensive care unit. In: Gabrielli A Layon A J Yu M ( E ds) Civetta, T aylor, and Kirbys Critical Care, 4th ed. Lippincott Williams & Wilkins, Philadelphia, pp 24952502. Peterson M Schwab W McCutcheon K van Oostrom J H Gravenst ein N Caruso L ., 2008. Effects of elevating the head of bed on interface pressure in volunteers. Crit ical Care Med icine 36, 303842. Peterson M Schwab W McCutcheon K van Oostrom J H Gravenstein N Caruso L ., 2009. Effects of turning on s kin-bed interface pressures in healthy adults Journal of Advanced Nursing, submitted.

PAGE 25

25 CHAPTER 2 TECHNIQUES USED TO INVESTIGATE THE PRESSURE ULCER PROBLEM: INTERFACE PRESSURE MEASUREMENTS AND BIOIMPEDANCE To identify clinically relevant biomechanical fact ors that are likely to increase the risk of pressure ulcer formation, interface pressure mapping and bioimpedance measurement techniques were implemented. Pressure is defined as force per unit area and i nterface pressure is the pressure that exists between two surfaces. Essentially, interface pressure measurements measure the contact forces from two objects that act on one another. Interface pressure mapping, thus, takes individual measurements over a specific area and specifies the pressures over its entir e surface. Bioimpedance is the application of electrical impedance to biological systems. Electrical impedance is a measure of the overall opposition to electric current. In direct current (DC) circuits, the opposition to current flow is electrical resista nce. Electrical impedance is the resistance equivalent in alternating current (AC) circuits, i.e. applying Ohms Law to AC circuit analysis. Ohms law states that current is directly proportional to voltage and inversely proportional to resistance. However impedance, unlike resis tance, can be a complex number. This chapter will introduce the methods used to address the first two specific aims of this dissertation by providing an overview of their history, applications, and the devices used for each of the techniques. Interface Pressure Measurements Origins and Applications Even though there is no number that can define the absolute point at which pressure ulcers will or will not occur, most researchers agree that high interface pressures are a primary facto r of pressure ulcer formation. Kosiak (1961) was one of the first to attribute interface pressures to pressure ulcer formation. His study subjected rat muscle to several loading experiments, which consisted of either constant pressure or alternating pressure (5 minute intervals), over several

PAGE 26

26 time ranges to monitor the rate of pressure ulcer formation. Results showed that both continuous and alternating pressure damaged the muscle; the severity of the damage increased with increased pressure and with increa sed duration. Tissue under alternating pressure was more tolerant than under continuous pressure to the extent that a longer period of time was necessary to cause damage. His previous research with dogs also demonstrated this inverse relationship between t ime and the amount of pressure that can be tolerated before tissue damage is observed. Reswick and Rogers (1976) expanded on Kosiaks inverse relationship which resulted in their well known pressure vs. time curve (Figure 2 1). This chart is based off of a combination of 980 observations on patients and volunteers. However, the authors note that this figure should only be used as a guideline, based on much experience as the conditions over which the observations were obtained were all somewhat different, a s well as the patients themselves. Very few of the observations were obtained from controlled measurements. Figure 2 1. Allowable pressure vs. time of application for tissue under bony prominences. The curve gives general guidelines and should not be t aken as absolute. Source: Reswick and Rogers, 1976.

PAGE 27

27 The primary applications for interface pressure measurements are: clinical education for proper seating and cushion selection for wheelchair patients, and in research for product development and evaluatio n of pressure reducing systems. Patients confined to wheelchairs are at high risk for pressure ulcer formation. Fortunately, seating clinics have been shown to be effective in helping prevent pressure ulcers (Swain, 2005). These clinics show patients how t heir body positions and postures affect their interface pressures and also help them choose proper cushions. Even though a particular cushion may seem more comfortable to a patient, it does not necessarily mean that it is providing the best pressure reduct ion (Stinson and Porter -Armstrong, 2008). Measurement Devices/Systems Review of early methods A review by Swain (2005) explains how interface pressure measurement devices were evaluated initially by FergusonPell, Bell, and Evans (1976). At this time, devices primarily fit into 3 general groups: thin sheets of various materials treated with inks or chemicals, devices composed of air cells with electrical contacts inside them, and strain gauge diaphragm transducers. The chemical/ink pressure-measuring device s were able to provide a map of the pressure distribution but were hard to quantify. They were also subject to in -plane forces and were sensitive to loading rates and temperatures. For the air devices, the inflation pressure was measured and increased unti l the cell surfaces separate d, typically identified by an indicator bulb. The resulting inflation pressure was taken as the interface pressure. The last group was strain gauge diaphragm transducers, some of which could be used for fluid measurements. Durin g this period, research was being devoted to create capacitive, resistive, and inductive sensors. Even though Swains review was for the early methods of interface pressure measurement, the conclusions from Perguson-Pell et al. (1976) still hold true today Notably, the device must

PAGE 28

28 provide a reliable output from 10 250 mm Hg; the device should not be time dependent, but at the very least the time dependence should be minimal, well -defined, and repeatable; the sensor must be smaller than the radius of curvat ure of the object being measured; and, when calibrating, the properties of the transducer must be taken into consideration as well as not to subject it to forces that are not in the normal direction. Comparison and analysis of different systems Only a few studies have compared different interface pressure measurement systems, and while various researchers may recommend some systems over others, there is no gold standard. Systems of note reported about from Swains review (2005) include : the Talley SA500 Pressure Evaluator and the Tally Pressure Monitor (Talley Medical, Romsey, Hants, UK), the DIPE (Next Generation, CA, USA), the Force Sensing Array (Vista Medical, Winnipeg, Canada), and the Tekscan system (Boston, MA, USA). Diesing et al. made the only re cent account of a system comparison at a European Pressure Ulcer Advisory Panal meeting in 2002 (Swain, 2005). There they compared the Force Sensing Array, Novel (Munich, Germany) and Xsensor (Calgary, Canada) systems, and concluded that all systems under estimated the applied force when in was confined to a small contact area, about 5 20 cm2. The researchers thought all 3 systems were appropriate for clinical use but felt measurements between systems could not be compared and that the systems would tend to underestimate interface pressures under smaller bony prominences. A comparison of interface pressure measurement devices manufactured by Xsensor Technology Corp., Vista Medical Ltd., Tekscan Inc., Talley Medical, Cleveland Medical Devices Inc., and Novel can be seen in Table 2 1. These devices were readily available in the United Kingdom as of 2005 (Swain, 2005). As for what is currently available in the US, this list remains accurate, minus Talley Medical, with all companies still manufacturing devices th at are

PAGE 29

29 commercially available. In addition to these companies, Pressure Profile Systems Inc. and Sensor Products Inc. are companies that can provide customized interface pressure mapping equipment. Table 2 1. Comparison of commercially available interface pressure measuring systems. Source: Swain, 2005. Methods Proper use Swain (2005) offers several practical suggestions for the proper use of interface pressure measurements. While most suggestions are obvious, many variables can affect the output of the s ensor. Clearly, correct positioning of the sensor is necessary to obtain the proper results. Results can also be affected by any objects located between the patients skin and the support surface, particularly if they are inflexible. Large pressure arrays allow for finding the point of maximal pressure in a given area, but if they are inflexible, they will adversely affect the surface being measured. If smaller, individual sensors are used, the potential error affecting the surface can be reduced if they ar e smaller than the area of interest; however, they must be accurately positioned

PAGE 30

30 to record the pressure exerted on the particular boney prominence or area of interest. If the individual sensor is too large, the sensor could act as an additional support, es pecially if it is an inflatable sensor. It should also be recognized that these interface pressure measurements are only from a particular moment in time at a particular position. Variability Upon implementing the use of interface pressure measurements, du e to the nature of the experiments themselves, several sources of variability present themselves. Subjects will differ not only by height, weight, and body mass index (BMI), but also by their underlying skeletal structure, musculature, and subcutaneous fat These factors result in differences in an individuals bony prominences and ultimately in their interface pressure measurements. Patients at risk for pressure ulcer formation tend to be elderly, lack good muscle tone, and/or are subject to an underlying pathology, such as a spinal cord injury, all of which result in different tissue (muscle and skin) properties than healthy volunteers and can lead to higher interface pressures. Another major source of variability is subject positioning. It is apparent tha t the positioning of the subject will have a major impact on the interface pressures. When lying supine (flat on ones back), ones body weight is supported by the entire body surface. However, upon sitting up, the weight of the upper body is supported by less surface area, primarily the buttocks, thus increasing its interface pressure. Another example of how patient positioning affects the interface pressure is when one slouches in a chair or sits in a semi -recumbent position (head of bed elevated 30 to 45 degrees). In these situations, the subjects body weight gets shifted primarily to the sacrum and increases the interface pressure. Shifting ones weight on or off a bony prominence will have a major effect on the observed interface pressures. Another fac tor that needs to be considered is clothing and/or any other material that comes in contact with the subject or the sensor. Subjects wearing rigid clothing, or jeans with rivets for

PAGE 31

31 example, can result in different pressures than if they were wearing a hos pital gown or scrubs. Additionally, the cushion or other support surface that the subject rests upon will factor in as well. This should be evident as interface pressure measurements are typically used to evaluate mattresses, cushions, and surfaces. Limita tions When measuring tissue interface pressures, it should also be acknowledged that this is not a direct measurement of the internal pressures experienced by the various internal tissues and vessels of the body (Bouten et al. 2003; Swain, 2005). Interface pressure mapping is, however, the best available method to noninvasively measure pressures exerted on the skin. Though not internal, these interface pressure measurements provide a good representation of the pressures exerted on the tissue just below the surface of the skin. Device Evaluation and S etup Evaluating the differences between the various commercially available interface pressure measuring and mapping devices led to three qualities that were considered most important. First, pressure mapping was preferred over individual sensors for its larger number of sensors, larger measurement surface area, and the inherent improved resolution. In comparison to individual sensors, while they may be appropriate for measuring pressures at discrete locations suc h as the sacrum, buttock, or trochanter, interface pressure changes in surrounding areas go undetected. Interface pressure mapping can very precisely measure pressures that occur anywhere over an entire weight supporting tissue with the same amount of effo rt as measuring a discrete location. The second desired quality was a flexible, non rigid, sensor so that it could conform between a human body and a hospital bed. Using a flexible device prevents the device itself from affecting the interface pressure mea surement and avoids the concern that the large sensor array could act as an additional support. Finally, real -time data acquisition and visualization was preferred so

PAGE 32

32 that real -time feedback was possible. To satisfy these research quality demands was just as important as the fact that such a device was available for use from the Department of Anesthesiology, University of Florida. Interface pressure measurements and profiles were obtained using an Xsensor pressure mapping system. The pressure -sensing pad is very thin and flexible and consists of 48 x 48 half inch sensors comprising a 24inch x 24inch square array. This array of 2,304 independent sensors uses proprietary capacitive technology to discretely measure the pressures applied to the pad. Two rigid cables, originating from the pads top -left corner, connect the pad to an interface box. The interface box relays the individual pressure information from each sensor and sends the data to a personal computer for real time visualization and recording. Cali bration is performed by placing the sensor array between two air bladders, held together in a metal frame, and inflated to specific pressures. Accuracy is 10%. The device was calibrated to measure pressures from 0 200 mm Hg. The Xsensor interface box was mounted to a mobile cart to allow for transport to and from the hospital for clinical use Along with the interface box, a computer with a keyboard and mouse, a monitor, and a battery backup were mounted to the cart, and the entire setup was s afety checke d by clinical engineering prior to use in the hospital. The device can been seen in use in a clinical setting in Figures 2 2, 2 3, and 2 4. The first generation Xsensor system that was used for this research was corrected for creep compensation with help f rom the manufacturer. Creep can be a concern for some interface pressure measurement systems where the interface pressure measured for an object or subject increases over time despite no actual change in the object being measured. Additionally, the

PAGE 33

33 Xsensor equipment was interfaced through customized software that allowed for data acquisition and export to be tailored to the specific needs of the research protocols. Figure 2 2. Xsensor pad on bed. Figure 2 3. Xsensor pad connected to computer setup in cl inical setting.

PAGE 34

34 Figure 2 4. Xsensor pressure mapping system in use with display. Bioimpedance Origins and Applications Bioimpedance measurements originated in 1940 where clinical changes in hydration were correlated with total body resistance and capacit ive changes. It was at this time that Nyboer (1959) developed research, referred to as impedance plethysmography, which measured electrical impedance changes over various body compartments to detect dynamic blood volume changes. Two decades later, Thomaset t reported a relationship between whole -body electrical impedance measurements and total body water; these results were further expanded upon by Hoffer, Meador, and Simpson (1969). Since then, bioimpedance measurements have been validated and are regularly used for non invasive measurements of body composition. They are safe, repeatable, and results can be obtained quickly. Bioimpedance measurements can determine many whole -body parameters,

PAGE 35

35 such as total body water (TBW), extracellular water (ECW), intracel lular water (ICW), fat mass (FM), fat free mass (FFM), body cell mass (BCM), total body potassium (TBK), and total body muscle mass (TBMM) (Kaysen et al., 2005; Kotler et al., 1996; Kyle et al., 2004; Lukaski et al., 1985). To demonstrate their correlations, these parameters were verified with the appropriate reference methods: deuterium dilution (TBW), bromide dilution (ECW), total body potassium (ICW), densitometry (FM), dual energy X ray absorptiometry (FFM), total body potassium (BCM), total body counti ng (TBK), and magnetic resonance imaging (TBMM). Measurements Bioimpedance measurements are impedance measurements taken on biological systems. Biological systems differ from purely electrical systems in that the resistance to current flow by their conduct ors (fluids) is not the lone contributor to impedance, but capacitive impedance exists as well. The inherent nature of the biological environment is the reason for the capacitive component of impedance measurements. Each cell in the body is enclosed in a p oor -conducting lipid bilayer membrane that separates itself from its environment. Th e lipid membranes, along with tissue interfaces result in the capacitive portion of the bodys impedance. A typical equivalent circuit for biological tissues (Figure 2 5) consists of a resistor representing the extracellular fluid in parallel with a resistor and a capacitor in series representing the intracellular fluid (Cornish, Thomas, and Ward, 1993; Grimnes and Martinsen, 2006; Kyle et al., 2004). Figure 2 5. Equivale nt circuit for biological tissue. Source: Kyle et al., 2004.

PAGE 36

36 In regard to AC circuits, the impedance measurements themselves can be defined by a resistance and a reactance. The impedance (Z) of the system is a two -dimensional vector quantity, expressed as both a resistance (R) and a reactance (X) or as a magnitude (Ohms) and a phase angle (degrees). Impedance can be represented by a complex number with the resistance corresponding to the real part and the reactance corresponding to the imaginary part. React ance is caused by the presence of inductors or capacitors in the circuit and also results in a phase shift between voltage and current. The principle of bioimpedance stems from the resistance of a homogeneous cylinder Figure 2 6 Resistance is defined by R = L / A (2 1) where is the resistivity of the conducting material, L is the length, and A is the cross sectional area. The volume (V) can be calculated similarly R = L / A = L / A (L / L) = L2 / V, (2 2) thus, V = L2 / R. (2 3) Figure 2 6. Homogeneous cylinder. Source: Kyle et al., 2004. R = L / A

PAGE 37

37 Although the human body is not a perfect cylinder or homogeneous, Equations 2 1 and 2 3 can be used as an empirical model for this equivalent cylinder (Cornish et al. 1993 ; Kyle et al. 2004; Lukaski et al. 1985). These types of calculations can be made to help determine the various whole -body variables such as TBW and FFM. Methods There are a few different methods to obtain bioimpedance measurements: single frequency bi oelectrical impedance analysis (SF BIA), multi -frequency bioelectrical impedance analysis (MF -BIA), and bioelectrical impedance spectroscopy (BIS) (Kyle et al. 2004). All three techniques involve passing an alternating current through the body. Injected c urrent s can range from 200800 A (Cornish, 2006). For whole -body measurements, current is typically passed from hand to foot via skin surface ( electrocardiogram ECG ) electrodes. The typical tetrapolar arrangement (Figure 2 7) places electrodes on the han d, wrist, ankle, and foot, with the distal electrodes carrying the current and the proximal electrodes detecting the voltage drop (Cornish, 2006; Grimnes and Martinsen, 2006; Kyle et al. 2004; Lukaski et al. 1985). Figure 2 7. Tetrapolar technique. I1 and I2 are the current carrying electrodes, and E1 and E2 are the voltage measuring electrodes. Source: Lukaski et al. 1985.

PAGE 38

38 Current flow in biological tissue is frequency dependent (Cornish, 2006). At zero or low frequency (R0), the curren t only passes t hrough the extra cellular fluid due to the capacitance (cell membranes) of the intracellular fluid, while at infinite or very high frequency (R), the reactance decreases and the curren t passes through both the extra cellular fluid and the intracellular fluid (Cornish et al., 1993; Cornish, 2006). SF BIA passes an alternating current at a fixed frequency, usually 50 kHz, through the body. Results from this method are based on a combination of mixture theories and empirical equations and can be used to determine TBW and FFM (Kyle et al. 2004). MF BIA also uses empirical models but differs from SF -BIA in that it uses impedance measurements from mult iple frequencies. MF -BIA frequencies can range from 0 to over 1 MHz and can be used to determine TBW, FFM, ICW, and ECW (Kyle et al. 2004). BIS differs from MF -BIA in that it predicts R0 and Rand then develops empirically derived prediction equations rather than using mixture modeling. R0 and R are predicted by measuring impedance at logarithmically spaced frequencies in a range from, for example, 5 kHz 1 MHz (Cornish, 2006; Kyle et al. 2 004). The resistance and reactance of these measurements are plotted to form a semicircular locus (Cornish, 2006; Kyle et al., 2004), as shown in Figure 2 8 The R0 and Rvalues can then be extrapolated from the data to approximate the ideal frequencies ( zero and infinity ). Figure 2 8. Diagram of an impedance locus depicting the phase angle and its relationship with resistance, reactance, impedance, and frequency. Source: Kyle et al., 2004.

PAGE 39

39 CHAPTER 3 IDENTIFY THE CLINICALLY RELEVANT BIOMECHANICAL FACTORS THAT ARE LIKELY TO INCR EASE THE RISK OF PRESSURE ULCER FORMATION Biomechanical factors, such as pressure and duration of loading, are known to have a role in pressure ulcer development. Research addressing Aim 1 involved measuring interface pressures of healthy subjects Pressures and areas are currently thought to be the most important factors that need to be monitored or controlled in regard to adjusting the patients head of bed (HOB ). The refore, interface pressure s between subject s and an ICU bed were studied as the HOB was incrementally elevated from 0 to 75 degrees. By setting a pressure threshold, the affected area s can be observed as well Redistributing the pressures and affected areas are currently thought to be the most important factors that ne ed to be monitored or controlled in regards to turning. So a dditionally, interface pressure s were studied as subjects were laterally turned as in q2h turning, the standard of care for pressure ulcer prevention. ICU patients are at particular risk for press ure ulcers and ventilator associated pneumonia (VAP) due to multiple predisposing factors. The research is this chapter was tailored to identify the biomechanical factors as they apply to the clinically relevant situations of VAP and q2h turning. Clinical Applications Ventilator -Associated Pneumonia VAP is a common complication in mechanically ventilated patients. Approximately 9% to 28% of all mechanically ventilated patients will develop VAP (Chastre and Fagon, 2002; Rello et al., 2002), with the risk inc reasing with duration of ventilation (Cook et al., 1998). Attributable mortality is difficult to pinpoint but may be as high as 50% in some patient populations (Bercault and Boulain, 2001; Heyland et al., 1999; Papazian et al., 1996; Rello et al., 1993). Current guidelines for the prevention of VAP suggest several methods to decrease the incidence, including rapid ventilator weaning, continuous aspiration of subglottic secretions, and

PAGE 40

40 placing the patient in a semi recumbent position (American Thoracic Socie ty, 2005). Elevation of the HOB is designed to decrease regurgitation and aspiration of gastric contents. A reduction in aspiration and VAP when the HOB is elevated has been demonstrated in a number of studies (Davis et al., 2001; Drakulovic et al., 1999; Orozco Levi et al., 1995; Torres et al., 1992). Prone positioning has been demonstrated to reduce the incidence of lung worsening (Beuret et al., 2002) and may decrease VAP incidence (Dodek et al., 2004; Guerin et al., 2004) however, there is a lack of fe asibility for universal application (Dodek et al., 2004), early enteral nutrition is poorly tolerated (Reignier et al., 2004), and it increases intracranial pressure (Beuret et al., 2002) In the study by Guerin et al. (2004), despite an improvement in oxy genation and a decrease in VAP incidence, the prone position did not improve mortality, and was associated with significantly greater occurrences of selective intubation, endotracheal tube obstruction, and pressure ulcers incidences compared to the supine position. Although there is good evidence that the semi recumbent position has pulmonary benefits, the effect on pressure ulcer risk has not been defined. Given that ventilated patients are often at risk of pressure ulcers due to immobility and altered con sciousness, it was hypothesized that elevating the HOB to decrease the risk of VAP would, as an unintended consequence, increase the interface pressures in the sacral area, and potentially increase the risk of developing pressure ulcers. Therefore, a study was designed to examine the effects of elevating the HOB on skin bed interface pressures in the sacral and buttock regions. Turning Effectiveness Recent studies have attempted to determine the effect of turning frequency on the incidence of pressure ulcer s. Defloor et al. (2005) compared frequent turning (2 -hours or 3 hours) on a standard mattress with less frequent turning (4 hours or 6 -hours) on a pressure reducing mattress. Of these 4 groups, there was no statistical difference in incidence of grade I

PAGE 41

41 p ressure ulcers. However, their results demonstrated a significantly lower incidence of grade II IV pressure ulcers when a patient lying on a pressure reducing mattress was turned every 4 hours compared to turning every 2 or 3 hours on a standard mattress. Vanderwee et al. (2007) studied an experimental turning routine of alternating 2-hour lateral positions with a 4 hour supine position in comparison to a routine of repositioning every 4 hours. The experimental group demonstrated a lower, but not statistica lly significant, incidence of grade II IV pressure ulcers, suggesting that more frequent turning does not necessarily lower pressure ulcer incidence. Hobbs (2004) also demonstrated this finding upon implementing a turn team. Results showed no reduction i n pressure ulcer incidence despite a decrease in the average length of stay. Therefore, despite the research efforts, it remains unclear which turning protocols are best, or even for that matter if the act of turning is, in fact, substantially protective a gainst pressure ulcer development. Given that several studies using turning as a chief intervention strategy have failed to reduce the incidence of pressure ulcers, the effectiveness of lateral turning in unloading at risk tissue was questioned. Although m aintaining the skin -support surface interface pressure below a capillary closing pressure of 32 mm Hg would be expected to reduce pressure ulcer risk, the effectiveness of turning patients to accomplish this remains to be established. Therefore, a study wa s designed to examine the effects of lateral turning on skin-bed interface pressures in the sacral, trochanteric, and buttock regions, and its effectiveness in unloading at risk tissue. Interface Pressure Measurement and Other Equipment Interface press ure measurement profiles were obtained using an Xsensor pressu re -mapping system. The pressure sens or pad is very thin and flexible and consists of 48 x 48 half -inch independent sensors comprising a 24 inch x 24 inch square array. Additional, details about thi s device can be found in Chapter 2.

PAGE 42

42 A modern ICU bed was used f or all measurements (Total Care Hill -Rom, Batesville, IN). The HOB was adjusted with the side rail controls and the angle of the HOB elevation was measured with the beds built -in ball bearing measurement located in the side rail of the bed. Specifically marked HOB angles included 0, 10, 20, 30, 45, 60, and 75 degrees. A CXTA01 tilt sensor (Crossbow Technology Inc., San Jose, CA) was used to confirm the HOB angles by also taking measurements at the subjects sternum s The sensor outputs a voltage response, 0 5 volts, which is proportional to the sine of the tilt angle and has a full angular range of 75 and 75-degree HOB measurements, the tilt sensor was offset 90 degrees prior to measurement to keep the instrument well within i ts angular range. The tilt sensor was also used to measure the angle of the lateral turn. The lateral turn angle was measured in the pelvic region by placing the sensor on an adjustable, rigid, U shaped device that rested on the subjects anterior superior iliac spines. Standard foa m pillows or wedges (triangle -shaped) were used for turning support as would be done for a patient.1 Subjects Fifteen healthy adults, 14 men and 1 woman, provided informed consent and participated in the studies. The same subjects participated in both the HOB elevation and lateral turning studies and were aged from 23 to 54 years (36.7 7.3). The subjects heights ranged from 1.70 m to 1.85 m (1.80 m 0.04 m), and they weighed between 65.8 kg and 122.5 kg (87.2 kg 17.7 kg). The resulting BMIs ranged from 20.3 to 38.7 (26.8 5.5). For the lateral turning study, e ight subjects underwent the experimental turning protocol with both pillows and wedges to determine any differences between the two supporting devices 1 W hen nurses turn their patients, they must support them to keep them correctly positioned. The equipment used at Shands Hospital at the University of Florida is either foam pillows or wedges. These devices are used d epending on what is available or the nurses preference.

PAGE 43

43 with regard to the patients safety. This gr oup was all male, aged from 23 to 43 years (33.3 6.9). The subjects heights ranged from 1.70 m to 1.85 m (1.79 m 0.05 m), and they weighed between 65.8 kg and 114.7 kg (85.6 kg 15.9 kg). The resulting BMIs for this group ranged from 22.7 to 35.3 (26.6 4.7). HOB Study HOB Protocol and Data Collection After IRB approval and informed consent, subjects dressed in hospital scrubs were positioned supine with their sacrum centered on the pressure sensor pad on the bed. A calibrated, interface pressure pro file was acquired for each subject in the supine position (0 -degree HOB) and a HOB measurement was made with the tilt sensor. The HOB was then adjusted to the next position and measured. Once the bed was adjusted and the subject was settled, another interf ace pressure profile was obtained. This procedure was repeated until all HOB positions were recorded: supine, 10, 20, 30, 45, 60, and 75 degrees. The entire protocol took no more than 10 minutes to complete for each subject. At every HOB position, an inter face pressure profile was captured. Each profile consisted of 10 measurement readings that were recorded sequentially. The average of these profile measurements was used in the data analysis for the corresponding HOB position. The reliability of the averag ed profiles is evident in that 99.2% of the individual sensors had a standard deviation of less than 2 mm Hg across all volunteers and positions. HOB Data Analysis Matlab (Mathworks, Natick, MA) and Excel (Microsoft, Redmond, WA) were used to plot, image, analyze, and compare the interface pressure profile data. Each pressure profile provided the interface pressure (mm Hg) at each of the 2,304 discrete sensors. The maximal pressures were determined, and the at risk area that was subjected to various pressur e thresholds was calculated. SPSS 15.0 (SPSS Inc., Chicago, IL) was used for all statistical analyses. A repeated

PAGE 44

44 measures analysis of variance was used to determine statistical differences in the peak interface pressures and at risk areas at the various H OB positions. A p < 0.05 was significant. HOB Results Interface p ressures Peak sacral interface pressures increased with HOB elevation (Figure 3 1). A repeated measures analysis of variance demonstrated that the HOB positions of 30, 45, 60, and 75 degrees all had statistically significant peak interface pressures that were greater than the supine measurement (p < 0.02, Bonferroni adjustment). Additionally, the HOB positions of 30, 45, 60, and 75 degrees were all statistically different from all other HOB po sitions. A visual example of how the interface pressure profile change s due to HOB elevation can be seen in Figure 3 2. Figure 3 1. Peak sacral area interface pressures at selected HOB angles of elevation. The data points represent the mean SD of the peak interface pressures measured for the subjects at each HOB position. *p < 0.02 vs. 0 degree HOB elevation. At -risk areas At risk areas, defined as skin surfaces over which an interface pressure greater than 32 mm Hg was observed, increased upon elevat ing the HOB (Figure 3 3). A repeated measures

PAGE 45

45 analysis of variance demonstrated that the at risk areas in the HOB positions of 45, 60, and 75 degrees were all significantly larger than the supine measurement (p < 0.0001, Bonferroni adjustment). This phenom enon can also be seen in Figure 32. With the same statistical test, the at risk areas of the 45 60 and 75-degree HOB positions were all statistically different from all other HOB positions. Additionally, the 30-degree HOB elevation position was differ ent from the 10and 20 degree HOB elevation positions (p < 0.02, Bonferroni adjustment). Figure 3 2. Interface pressure profiles at various head of bed (HOB) positions for one subject The color intensifies with an increase in pressure and demonstrates how the interface pressure increases over the sacral r egion when the HOB is elevated. The color bars units are in mm Hg with color denoting the areas where the skin -bed interface pressure exceeds 32 mm Hg Effect of body habitus Since the subjects represe nted various body types, the data were analyzed to see if height, weight, or BMI affected the results. Due to the small sample size, only qualitative remarks could be made. No trend emerged upon analyzing peak interface pressures as height, weight, or BMI increased. However, a trend of increasing at risk area appeared to be emerging as weight and

PAGE 46

46 BMI increased, but not for height. When viewing the interface pressure profiles, there did not appear to be any noticeable differences between varying heights, wei ghts, or BMIs apart from the relative size of the subject. F i gure 3 3. Atrisk sacral areas mm Hg at selected HOB angles of elevation. The data points represent the mean SD of the at risk area s measured for the subjects at each HOB position. p < 0.0001 vs. 0 -degree HOB elevation. HOB Discussion The results clearly demonstrate that raising the HOB to 30 degrees or higher significantly increases the skin -support surface interface pressure, and a 45 degree HOB angle or greater significantly increase s the at risk area. These findings demonstrate the law of unintended consequences, in which an attempt to decrease the risk of VAP may increase the risk of pressure ulcer formation. Raising the HOB places the subject in a more seated position (compared to lying supine) and places more of the patients upper body weight above the sacrum and buttocks resulting in higher interface pressures. While an increase in sacral area interface pressure with increased HOB elevation is somewhat intuitive, it does not nece ssarily follow that the at risk

PAGE 47

47 skin surface area also increases. Because this study shows that elevating the HOB increases the maximum pressures experienced on the sacral and buttock skin regions and also increases the total area at -risk for developing a pressure ulcer, the standard practice of elevating the HOB may increase the risk of skin breakdown. The literature involving the effects of elevating the head of bed on skin -support surface interface pressures is limited and somewhat mixed. Sideranko et al (1992) measured interface pressures over the sacrum on three mattress overlays. Sacral interface pressures were significantly higher in the semiFowler (45-degree HOB elevation) compared to the supine position, a factor of about 1.5, with the alternating air mattress overlay. This factor agrees with the current study, as interface pressures increased by 1.52 from the supine position (58.1 mm Hg) to the 45 degree HOB position (88.4 mm Hg). In two studies by Allen, Ryan, and Murray (1993, 1994), localized i nterface pressures were measured on a variety of mattresses and overlays at six anatomical locations, including the sacrum and buttocks. Their results demonstrated a slight decrease in interface pressure at the sacrum but an increase at the buttocks upon r aising the HOB to 60 degrees. The pressures at the buttocks increased by a factor of 1.7 on three different mattresses (Allen et al., 1993) and 1.6 on a continuous airflow mattress overlay (Allen et al., 1994). These factors of increased pressure correlate well with the current study as the peak interface pressure increased by a factor of 1.75 from the supine position (58.1 mm Hg) to the 60degree HOB position (101.7 mm Hg). In contrast to the current study, Whittemore et al. (1993) did not demonstrate a si gnificant increase in sacral interface pressures at the 45 -degree HOB position compared to supine. Their results likely differ from the current study because they measured sacral interface pressures with a 2 inch diameter plastic inflatable bladder whereas the current study measured pressures

PAGE 48

48 beneath the subjects entire seating surface. It may be that the interface pressures strictly confined to the sacrum do not significantly change from supine to 45-degree HOB elevation, but other weight bearing tissues do experience a significant increase in interface pressure with this change of bed position. With regard to the magnitudes of the interface pressure measurements themselves, the results are very widely. Table 3 1 lists the approximate interface pressure va lues measured at various HOB positions in 8 studies, including this one. These studies all measured tissue interface pressures at the sacrum and/or buttock. Every other study reported lower interface pressures, with the majority reporting results that were significantly lower, than the peak pressures that this study recorded. These differences in interface pressure magnitude may be due to many variations among the studies: different beds/mattresses, different subjects, and different pressure measurement ins truments. Those experiments used discrete sensors on the sacrum, buttocks, or trochanter ( Allen et al., 1993, 1994 28 mm sensor pad; Harada, Shigematsu, and Hagisawa, 2002 three 38 mm x 35 mm sensors; Maklebust, Mondoux, and Sieggreen, 1986 discrete electropneumatic sensor; Sideranko et al., 1992 4.5 x 3.5 bag; Whittemore et al., 1993 2 inch diameter bladder). While these techniques may be appropriate for measuring pressures at these discrete locations, high interface pressures in surrounding ar eas will go undetected. Defloor (2000) used pressure mapping to measure interface pressures in various supine positions on two mattresses. Nonetheless, the current study demonstrated higher peak interface pressures in similar HOB positions. However, more i mportantly, this studys results demonstrate a statistical difference between the 0 and 30-degree HOB positions, whereas the results of Defloors supine measurements only showed statistical significance between the 0 and 90-

PAGE 49

49 degree HOB positions. This dif ference is likely attributed to this studys instrumentation having a much greater resolution in the sacral area. Defloor used a measurement device that covered the entire mattress with 684 sensors, whereas this study used a device with 2,304 sensors around only the sacral region. With lower resolutions, high localized interface pressures will be spread over a larger sensor area resulting in a lower observed pressure. By using higher resolution instrumentation, this study was able to demonstrate that the in terface pressures do increase significantly over the sacral area upon elevating the HOB. Changing the HOB elevation, from a supine to a seated position, results in a shift of pressure points. This phenomenon has been discussed by Braden and Bryant (1990), who describe that raising the HOB above 30 degrees increases pressure and shear forces. This studys results demonstrate an increase in interface pressures when increasing the HOB above 30 degrees. Additionally, this studys results suggest that interface pressures are higher than once believed, as previous discrete and low resolution pressure measurements were not able to simultaneously measure interface pressures over the entire weight -bearing tissue. Attempts were made to measure various body types (weights and BMIs) to determine if interface pressures and at risk areas differ due to these factors. The results show that regardless of body type, interface pressures and at risk areas increase with HOB elevation. The emerging trend can only suggest, due to t he small sample size, that larger subjects (weight and BMI) will have greater at risk areas upon HOB elevation. HOB Conclusion Raising the HOB to 30 degrees or higher on a modern ICU bed increases the peak interface pressure between the skin and support surface in healthy subjects. At 45 degrees or higher, the at risk area of skin with a skinICU bed interface pressure greater than 32 mm Hg also increases.

PAGE 50

50 Both observations strongly suggest that increasing the HOB beyond 30 degrees increases the risk for pr essure ulcer formation. Table 3 1. Comparison of interface pressure measurements between studies Pressures measured at the sacrum unless marked otherwise. HOB Position Supine 30 degrees 45 degrees 60 degrees 90 degrees Interface Pressure Measurements ( mm Hg) Current Study 58.1 a 73.0 a 88.4 a 101.7 a Sideranko et al., 1992 42.6 64.3 Allen et al., 1993 19.0, 22.0 b 12.0, 39.1 b Allen et al., 1994 14.6, 20.9 b 13.4, 32.6 b Whittemore et al., 1993 24 44 29 45 Maklebust et al., 1986 12 19 Ha rada et al., 2002 3 15 Defloor, 2000 39.5 a,c 27.7 a,d 38.4 a,c 26.9 a,d 37.4 a,c 28.7 a,d 48.4 a,c 39.0 a,d a peak pressure measurement over entire sacral area b measurement taken at the buttocks c standard mattress d pressure -reducing mattress Lateral Turning Study Lateral Turning Protocol and Data Collection With IRB approval and informed consent, subjects dressed in hospital scrubs were positioned supine with their sacrum centered on the pressure sensor pad on the bed. A calibrated interface pressure profile was acquired for each subject in the supine position (0 -degree HOB), and the HOB and turn angle measurements were made with the tilt sensor. An experienced ICU nurse turned each subject (approximately 30 degrees) to a lateral left or la teral right position using supporting pillows or foam wedges beneath the subject and the pressure sensor to permit interface pressure measurement between the subject and the pillow/wedge. Once the subject was settled into position, the computer recorded t he interface pressure profile along with the

PAGE 51

51 corresponding turn angle. The HOB was then elevated to 30 degrees, to mimic what is commonly used for ICU patients to reduce aspiration risk, and another interface pressure profile and tilt sensor measurement we re recorded in the elevated left or elevated right turned position. This procedure was repeated for both the left and right side. The entire protocol was completed in less than 10 minutes for each subject. The interface pressure profiles from every pos ition consisted of 10 measurement readings recorded successively. The average of these profile measurements was used for data analysis of the corresponding supine or turned position. The reliability of the averaged profiles is evident in that 99.1% of the individual sensors had a standard deviation of less than 2 mm Hg across all volunteers and positions. Lateral Turning Data Analysis Matlab and Excel were used to plot, image, align, analyze, and compare the interface pressure profile data. Each profile provided the interface pressure (mm Hg) at each of the 2,304 discrete sensors. The maximal pressures were determined, and the at -risk area that was subjected to various pressure thresholds was calculated. The supine images obtained before lateral turning were aligned using 2D cross -correlation to determine how the turning altered the interface pressure profiles. SPSS 15.0 (SPSS Inc., Chicago, IL) was used for all statistical analyses. A repeated measures analysis of variance was used to determine any statistic al differences in the peak interface pressures and at risk areas between the successive turned positions for each side. Paired t tests were used to compare pillows to wedges at the various positions. A p < 0.05 was significant. Lateral Turning Results Inte rface pressures and at -risk areas Peak peri -sacral area interface pressures were not significantly affected by lateral turning, but did increase upon elevating the HOB to 30 degrees (Figure 3 4). At risk areas were again

PAGE 52

52 defined as skin surfaces over which an interface pressure greater than 32 mm Hg was observed (Figure 3 5). At risk areas between the supine and laterally turned (left or right) positions did not differ. The at risk areas in the elevated turned positions were, however, significantly greater than their corresponding supine and laterally turned positions. It was readily visible that, despite turning, areas intended to be unloaded from pressure still experienced significant levels of interface pressure and remained at risk (Figure 3 -6). Figure 34. Peak peri -sacral area interface pressures. The data points represent the mean SD of the peak interface pressures measured for the 6 selected positions. *p < 0.05 compared to previous corresponding positions. Triple jeopardy at -risk areas Every posi tion exhibited specific at risk areas of skin that overlapped with all other positions in all but one subject, i.e. 14/15. The tissue area at risk in all positions was termed the triple jeopardy area because the skin is at risk in all 3 positions: supine left, and right. In this case, the pressure is never relieved via turning as intended; see Figures 3 7 and 38 for examples. The mean triple jeopardy area was 60 54 cm2 with a range from 0 198 cm2 (Table 3 -2).

PAGE 53

53 Figure 3 5. At risk peri -sacral areas. T he data points represent the mean SD of the a trisk peri sacral areas measured for the 6 selected positions. *p < 0.05 compared to previous corresponding positions. Figure 3 6. Peri -sacral area interface pressure profiles Typical interface pressure profiles f rom a single subject. The color intensifies with an increase in pressure and demonstrates how the interface pressures change upon lateral turning. The color bar units are in mm Hg with color denoting interface pressures exceeding 32 mm Hg.

PAGE 54

54 Figure 3 7. Tr iple jeopardy at risk areas using pillows. Interface pressure profiles of one subject in all positions turned with pillows. Color denotes interface pressures exceeding 32 mm Hg; maroon, in particular, denotes the triple jeopardy area where the skin bed int erface pressure exceeded 32 mm Hg throughout all positions. Figure 3 8. Triple jeopardy at risk areas using wedges. Interface pressure profiles of one subject in all positions turned with wedges. Color denotes interface pressures exceeding 32 mm Hg; maro on, in particular, denotes the triple jeopardy area where the skin bed interface pressure exceeded 32 mm Hg throughout all positions.

PAGE 55

55 Table 3 2. Triple jeopardy areas (cm2). The second pillows column is repeated to show the mean SD for comparison to wedg es. Subject Pillows Pillows Wedges 1 69 69 223 2 52 52 69 3 37 37 202 4 35 35 137 5 35 35 29 6 2 2 76 7 150 150 335 8 0 0 155 9 10 10 45 11 37 12 68 13 103 14 198 15 63 Mean SD 60 54 48 47 153 99 Effect of body habi tus Since the subjects represented various body types, the data were analyzed to see if height, weight, or BMI affected the results. No trend emerged upon analyzing peak interface pressures as height, weight, or BMI increased. However, for at -risk areas, a n increasing trend began to emerge as weight and BMI increased, but not height. As for the triple jeopardy areas, again an increasing trend began to emerge as weight and BMI increased, but not height. When viewing the interface pressure profiles, t here did not appear to be any noticeable differences between varying heights, weights, or BMIs apart from the relative size of the subject. Pillows vs. wedges Eight subjects were studied with sequential turning support by foam pillows or wedges. In this group, the re was no significant difference in peak interface pressures at any of the positions

PAGE 56

56 between the two techniques. However, the at risk areas were significantly larger with the wedges at the lateral left and elevated left positions. The triple jeopardy area increased dramatically using wedges (153 cm2 99 cm2, range: 29335 cm2) when compared to pillows (48 cm2 47 cm2, range: 0 150 cm2) (p < 0.05) (Table 3 2). Seven of eight subjects demonstrated an increase in triple jeopardy area using wedges compared to their triple jeopardy area measurement with pillows. The turn angle measurements for the pillow and wedge techniques can be seen in Table 3 3. The turn angles using either technique (turned left and right) were: reference supine angles ranged from 0 2 degrees, lateral turn angles ranged from 31 40 degrees, and head elevated turn angles ranged from 28 33 degrees. Significant differences in turn angles were demonstrated between the pillows and wedges only in the lateral right and elevated right positions. Ta ble 3 3. Mean turn angles and standard deviations comparing the pillow and wedge techniques. *p < 0.05 for turn angle differences between pillows and wedges. Position Pillows Wedges Turn Angle SD Turn Angle SD Supine 1.1 1.8 0.7 2.0 Lateral Left 3 4.3 8.4 35.0 5.4 Elevated Left 30.7 7.6 30.7 5.8 Supine 0.8 2.7 1.9 1.2 Lateral Right 33.3 6.2 40.2 7.8 Elevated Right 28.8 7.9 33.3 7.2 Lateral Turning Discussion Regular turning of patients is routinely used to ostensibly de crease the risk of pressure ulcers and is considered a standard of care. As shown by these results, standard turning by an experienced ICU nurse does not reliably relieve the elevated skin-bed interface pressures as intended. In this sample, 14 of 15 subje cts demonstrated a significant triple jeopardy area suggesting that the current turning process is not reliably effective at unloading skin subjected to

PAGE 57

57 pressures believed to put it at risk for pressure ulcer formation. Essentially, there were areas of ski n, predominantly near or around the sacrum, that were always at risk and were not relieved by turning in all but one of the subjects. Lateral turning with pillows did not increase the skin -support surface interface pressure or the at risk area unless the s ubjects HOB was elevated to 30 degrees. Elevating the HOB is an important component of modern best practice care in mechanically ventilated patients, and has been demonstrated to reduce the risk of ventilator associated pneumonia (Drakulovic et al., 1999). However, as a consequence of increased HOB elevation, increased skin -bed interface pressures result. Although intuitively HOB elevation should increase peri -sacral area interface pressures, it does not necessarily follow that the at risk skin surface a rea also increases. Because this study demonstrates that elevating the HOB increases the maximum pressures experienced on the peri -sacral and trochanteric skin regions and also increases the total area at risk for developing a pressure ulcer, the standard practice of elevating the HOB while turning patients may increase the risk of skin breakdown. Larger at risk areas were also measured when turning and using wedges for support compared to pillows, dependent on the magnitude of the turn. The significantly l arger turn angles observed for the lateral right and elevated right positions (Table 3 3) resulted in relatively more relieved tissue and less at risk area for right lateral turns than for the left. The difference in turn angles results from variation in nursing technique. For either support device, the turn angles achieved were somewhat less in the elevated (left and right) positions than the lateral (left and right) positions because of the subjects natural tendency to roll more supine with the HOB eleva ted. In addition to the increased at risk areas using wedges, the triple jeopardy at risk area increased more than three -fold.

PAGE 58

58 The limited studies regarding the effects of lateral turning on skin-support surface interface pressures do not describe monitori ng the at risk areas subjected to various pressure thresholds. Lateral turning studies by Maklebust et al. (1986) and Defloor (2000) both report interface pressure results (see Table 3 4). Maklebust et al. measured interface tissue pressures with discrete electropneumatic sensors over the sacrum of subjects on a hospital mattress and three different coverings in the supine position. Defloor used pressure mapping to measure interface pressures in various positions on two mattresses. Four of these positions i ncluded a supine position, a 30 degree lateral position, a 90 degree lateral lying position lying on the shoulder, and a 90 degree lateral -lying position not lying on the shoulder. Defloors results showed that on a standard hospital mattress a 30 degree l ateral position had significantly lower peak pressures than the 90 degree lateral lying positions. On a pressure reducing mattress, the 30 degree lateral position was only significantly lower than the 90 degree lateral position lying on the shoulder. There was no statistical difference between the two side lying positions on either mattress. Table 3 4. Comparison of maximum interface pressure measurements (mm Hg) from 3 studies. Patient Position Study Supine Lateral ly Turned Turned with HOB Elevation Cur rent Study 68.6 19.5 (L), 65.8 11.7 (R) 69.2 12.8 (L), 64.8 9.1 (R) 84.5 1 7.5 (L), 8 0.4 1 1.4 (R) Maklebust et al ., 1986 12 19 (1) Defloor 2000 39.5 7.0 ( 2 ), 27.7 4.1 ( 3 ) 51.4 15.4 ( 2 ), 38.6 8.5 ( 3 ) L prior to, or turned to lef t side R prior to, or turned to right side (both directions on a modern ICU bed) 1 standard mattress or covering 2 standard mattress 3 polyethylene urethane mattress A comparison of the current study to these other studies can be s een in Table 3 4. Maklebust et al. reported lower interface pressures than the current study or those measured by Defloor. The magnitudes of interface pressure from the current study were also considerably

PAGE 59

59 higher than Defloors. The differences in interfac e pressure magnitude between studies could partially be attributed to different beds/mattresses, subjects, and pressure measurement instruments Differences between this studys results and Defloors for corresponding positions can potentially be attribute d to this studys instrumentation having a much greater resolution in the peri -sacral area which allows for more precise measurement of interface pressures and changes that occur during lateral turning. A comparison of resolutions between the devices used for the current study and by Defloor was previously discussed in the HOB Discussion section. Attempts were made to measure various body types (weights and BMIs) to determine if tissue gets unloaded differently due to these factors. The results showed that regardless of body type, triple jeopardy areas were still observed. The emerging trends can only suggest, due to the small sample size, that larger subjects (weight and BMI) will have greater at risk areas and triple jeopardy areas, which would be expected because of their larger skin surface areas. Lateral Turning Conclusion Standard turning by experienced ICU nurses does not reliably unload all areas of high skin bed interface pressures. These areas remain at risk for skin breakdown and help explain why p ressure ulcers occur despite implementation of standard preventive measures such as scheduled turning of patients. Additionally, support materials for maintaining lateral turned positions can influence tissue unloading and triple jeopardy areas and need to be further evaluated to improve care. HOB and Lateral Turning Limitations and Future Research Limitations Healthy subjects, as used in these studies, likely have better gluteal muscle tone and bulk than typical ICU or other patients at risk for pressure u lcer formation. Better muscle tone and bulk can elevate the sacrum above the support surface and aid in pressure relief (Maklebust et al.,

PAGE 60

60 1986), and greater bulk can redistribute pressure within tissues from surface pressure points. Consequently, even hig her peak interface pressures could be expected in at -risk patients (Sideranko et al., 1992). Additionally, patients are more likely to have bare skin against the bed as opposed to the hospital scrubs worn by the healthy subjects. This difference is not exp ected to significantly affect interface pressures, although shearing forces may be affected. These studies only measured subjects on a single brand of a modern ICU bed, but the results are expected to be qualitatively similar regardless of mattress. The in terface pressures at the various positions are expected to differ somewhat on different beds/mattresses, but the relative trends of increased peak pressures, at risk areas, and triple jeopardy areas should hold true. Tissue interface pressures do not direc tly assess internal tissue and capillary pressures (Bouten et al. 2003). Interface pressure mapping presently provides the best available method to noninvasively measure pressures exerted on the skin. Though not internal, these interface pressure measurem ents should provide a good representation of the pressures exerted on the tissue just below the surface of the skin. Qualitatively, it should be recognized that increasing or decreasing interface pressures will consequently increase or decrease the resulti ng internal pressures experienced by the underlying tissues. For this and other reasons, the traditional threshold of 32 mm Hg for tissue damage has been disputed (Bouten et al., 2003; Swain, 2005). Interface pressures greater than the capillary closing pr essure can be tolerated for some time before ischemia results (Bader, 1990; Bouten et al. 2003). It is expected, however, that avoiding pressures exceeding capillary closing pressure should substantially reduce the risk for tissue ischemia and resultant p ressure ulcers. Future Research Further study is needed to determine whether or not the increased peak interface pressures and at risk areas due to raising the HOB actually increase the incidence of pressure ulcer

PAGE 61

61 formation. Further study is also needed to establish how to achieve optimum positioning, along with appropriate supporting materials, to favorably impact and even eliminate the triple jeopardy areas, the peak interface pressures, and the overall at -risk areas. In addition, clinical studies are nee ded to determine whether patients with large areas of triple jeopardy are, in fact, more likely to develop a pressure ulcer or if more serious pressure ulcers develop.

PAGE 62

62 CHAPTER 4 INVESTIGATE THE USE OF BIOIMPEDANCE MEASUREMENTS AS A METHOD TO ASSESS OR D ETECT SKIN DAMAGE The history, background, and current uses for bioimpedance analyses were discussed in Chapter 2 Due to the success and reliability of these previous methods, it seem ed promising that applying bioimpedance techniques to measure changes in the skin and its underlying tissues may help detect the onset of pressure ulcer formation. Previously mentioned applications used whole body bioimpedance measurements to determine certain physiological parameters, such as total body water and fat free mas s However, whole -body measurements may not be appropriate or sufficient for certain applications. T he torso of the human body, f or example only represents only about 5 10% of the total body impedance when measured from either handto -foot or even foot to -foot (Scharfetter et al., 2001, 2005). The majority of the measured impedance is due to the narrower -diameter body segments such as the arms and legs and even more so towards their extremities (Figure 4 1) Therefore, whole body impedance methods are not necessarily appropriate for extracting information about the torso or other individual body parts. However, the following research by Scharfetter et al. (2001) demonstrate d how bioimpedance measurements can be used in a localized area. Additiona lly, resear ch by Quintavalle et al. (2006) provide d some evidence for how pressure ulcers begin to develop which can hopefully be exploited by this localized bioimpedance technique. Research Applications Localized Bioimpedance Analysis Research by Scharfetter et al. (2001) w as able to correlate abdominal fatness with local bioimpedance analysis ( BIA ). They developed a variation of the tetrapolar setup, which placed 4 surface ECG electrodes in a symmetrical line on either side of the umbilicus a t the height of the iliac crest ( Figure 4 2 ). A Xitron 4000B (Xitron Technologies Inc., San Diego, CA) was used for

PAGE 63

63 impedance measurements. BIS analysis was used with a constant current of 250 A and samples were taken at 50 logarithmically spaced intervals between 5 500 kHz. 12 males and 12 females underwent the BIS procedure and abdominal magnetic resonance imaging (MRI) for reference. As mentioned in other studies (Kyle et al. 2004), measurements below 5 kHz and above 200 kHz are not as reproducible and therefore not as relia ble. For this reason and to reduce statistical evaluation, only four frequencies were analyzed: 5 kHz, 20 kHz, 50 kHz, and 204 kHz These frequencies represented the highest and lowest frequencies of the reliable range, while also using the most com monly used frequency of 50 kHz. Figure 4 1. Body resistances. Source: Grimnes and Martinsen, 2006. The impedance spectra dependence on the subcutaneous fat layer (SFL) thickness can be seen in Figure 4 3. Figure 44 shows this relationship at 50 kHz for model t (total population). Male and female populations were statistically different except at the 50 kHz frequency. A highly linear correlation (r2 = 0.98 or better ) resulted between the SFL and the bioimpedance measurements made across the waist.

PAGE 64

64 Figure 4 2. Localized abdominal BIA e lectrode location and spacing. Source: Scharfetter et al. 2001. Figure 4 3. Dependence of impedence spectra based on subcutaneous fat layer thickness. Data from 5 select male subjects. Source: Scharfetter et al. 2001. Figure 4 4. SFL impedance correlation from total population data (model t) at 50 kHz. The figure depicts the regression line and the 95% prediction intervals. One outlier is labeled with its subject number (12). Source: Scharfetter et al. 2001.

PAGE 65

65 Detecting Edem a with High -Resolution Ultrasound Research by Quintavalle et al. (2006) continued the investigation of pressure ulcer etiology They used high -resolution ultrasound (H R US) to image the skin of at -risk patients (Braden score of 18 or less) in comparison to those of healthy volunteers. The Longport Digital Scanner (EPISCAN I 200, Glen Mills, PA) was used in this study for its 65-micron resolution and depth of penetration (2 cm). Weak reflective patterns in HRUS images indicate increased fluid content or edema in the tissue. Images were obtained from 119 long-term residents The results s howed that over half of the 1139 readable images (630, 55.3% ) were abnormal with many images showing evidence of deep subdermal edema (541, 47.5% ), and some images demonstratin g superficial edema just b elow the epidermis (89, 7.8%). Data from their study suggested different etiologies for deep ver sus superficial pressure ulcers. Therefore, a healthy volunteer was subjected to two forms of pre ulceration and was imaged prior to and at various intervals during the intervention. One area was rubbed with a gauze pad for 7 minutes and another area was subjected to lying on a hard object for 1 hour. The friction (gauze) demonstrated edema directly below the epidermis with none in the deep tissue (Pattern 2 in Figure 4 5) The pressure from lying on the hard object resulted in pockets of deep edema with no superficial edema (Pattern 1 in Figure 4 5) These pre ulcerative changes were consistent with the images obtained from the study group. T he images demonstrating deeper areas of weak reflection were divided into 3 subgroups based on the extent and location of the weak reflective pattern (Figure 4 6). The first subgroup (91 images, 16.8%) showed pockets of weak reflections between the bone and the dermal layer but was normal in the dermal layer, the subepidermal layer, and the intact epidermis. The second subgroup (177 images, 32.7%) had more weak reflections in the subdermal tissue while extending into the dermal layer. The third subgr oup (273 images, 50.5%) had a layer of weak

PAGE 66

66 reflections just beneath the epidermis and distinctly less reflections in the subdermal layer and the dermal layer. Figure 4 5. High resolution ultrasound images of deep and superficial edema Source: Quintaval le et al. 2006. Figure 4 6. High resolution ultrasound images demonstrating the 3 phase s of pressure ulcer development. Source: Quintavalle et al. 2006. The HRUS findings in this study indicated that pressure ulcers are forming before the clinical sign s of erythema are observed the standard of care for skin assessment A significant number of images, 79.3%, had tissue changes with no presence of erythema; therefore, HRUS can be used to detect skin damage sooner than would be evidenced by a visual assess ment. The results support the notion that ulcers due to pressure originate in deeper tissue whereas ulcers due

PAGE 67

67 to friction occur near the epidermis. Additionally, for ulcers developing in deeper tissue, the results suggest 3 phases of pressure ulcer develo pment that progress consecutively from subgroup 1 to subgroup 3. Application and Goals The combination of methods from Scharfetter et al. (2001) and results and observations from Quintavalle et al. (2006) inspired the investigation of trying to use localiz ed BIA measurements to detect fluid changes (edema) in local tissue, specifically in the sacral area. It is known that tissue bioimpedance is affected by its intracellular and extracellular components; changes in fluid levels should alter the electrical pr operties of the tissue, which would thus change the impedance. Therefore, by using impedance measurements, edema could be detected, and ideally, at risk tissue would be able to be identified before it develops into a pressure ulcer, as was done with HRUS, but without the need for an experienced technician or elaborate procedure. As mentioned by Quintavalle et al., edema was present in many individuals before clinical signs were apparent (erythema). If this method proves successful, this measurement could be used in addition to, or in replacement of, a skin assessment, such as the Braden scale, so that rather than just identifying potential at risk patients, it could be determined which patients are actually already developing a pressure ulcer. Additionally, a nonvisual pressure ulcer detection method would be very helpful to assess those with darker pigmented skin, as erythema is harder to detect in darker skin tones. Alternatively, bioimpedance measurements could be very useful as well if they could be used in place of an interface pressure mapping system. Three preliminary studies were performed to determine the viability of bioimpedance measurements. The rest of this chapter will describe in detail the three preliminary steps that were undertaken: computer modeling, experimental measurements and evaluation of a brand of skin impedance instruments, and bioimpedance experiments with basic electrode arrangements.

PAGE 68

68 COMSOL Modeling As demonstrated by Quintavalle et al. (2006), edema appears to be a precursor to p ressure ulcer formation. If edema could be detected beneath the skin surface with a bioimpedance measurement, additional prevention methods could be implemented to stop the onset of pressure ulcer formation. The first study implemented computer modeling to model the skin and run various simulations. COMSOL Multiphysics 3.4 (COMSOL, Inc., Los Angeles, CA) is commercially available software based on the finite element method. This method solves partial differential equations by dividing the model into small e lements and approximating the solution with a function or assuming a constant value for the element (Pavselj, Pr e at, and Miklavcic 2007). A graduated triangular mesh was used for the model which resulted in a denser mesh near the electrodes and thinner sk in layers. The electromagnetics module was used for all simulations. Skin The skin is the largest organ of the human body and serves many functions, from acting as an external barrier, to temperature regulation, to transmission of sensory information. The skin is composed of many layers, but there are two principal layers, the epidermis and the dermis (Figure 4 7). The epidermis, the outermost layer of skin, is avascular and plays the most important role in the electrode -skin interface (Wang and Sanders, 2005; Webster 1998). The stratum corneum forms the top layer of the epidermis and is comprised of dead keratinized cells. The epidermis thickness varies between 60 m to 100 m over most areas of the body but can reach 600 m in the plantar and palmar regions (Wang and Sanders, 2005). The dermis accounts for the majority of the skins thickness, ranging between 1 mm and 4 mm (Wang and Sanders, 2005). It is vascularized and contains nervous components, lymph vessels, sweat glands, hair follicles, sebaceous gl ands, and smooth muscle (Wang and Sanders, 2005; Webster 1998). Apart

PAGE 69

69 from the sweat glands, the deeper layers of skin are similar to other body tissues and do not provide any distinctive electrical characteristics to the skin (Webster 1998). Figure 4 7. Skin anatomy. Source: Wang and Sanders, 2005. Skin Model A two -dimensional (2D) model was created to represent the body or the skin. If the 2D model proved successful, further steps would be taken to move to a three dimensional (3D) model. A circular or elliptical shape was considered as it could represent a cross -section of the body, but a rectangular shape was chosen to represent the various skin layers and tissue depth. The overall size of the skin model and electrode setup was ad hoc, but the propert ies and dimensions of the skin layers themselves were chosen to be as accurate as possible. To create the skin model, a geometrically correct cross -section of skin was needed. The model also required the electrical properties of the various skin layers as well as their relative dimensions. The skin layers included in the model consisted of a stratum corneum layer, an

PAGE 70

70 epidermal layer, a dermal layer, a subcutaneous (fat, connective) tissue layer, and a deep tissue layer (Figure 4 8). A recent study by Pavsel j et al. (2007), which utilized COMSOL to model the skin, was used to guide the models tissue properties and dimensions (Table 41), primarily the electrical conductivity values (S/m) of the various skin layers. In their study, a thicker stratum corneum w as used to avoid numerical problems, but was compensated for by making it more conductive (6 times thicker, but 6 times more conductive). Figure 4 8. Skin layers from the COMSOL skin model. Deep tissue layer not visible. Table 4 1. Electrical conductivit y values and thicknesses for skin model components. Skin Model Component Electrical Conductivity Value (S/m) Thickness Stratum Corneum 0.0005 100 m Epidermis 0.2 250 m Dermis 0.2 1 mm Subcutaneous Tissue 0.05 650 m Deep Tissue 0.05 2.5 cm or 5 cm Electrode 5.99e7 1 mm Edema 1.5 variable Since the COMSOL software is unit less in regard to dimensions, for this model, 1 unit was chosen to equal 1 mm. In regard to the coordinate system, the top of the deep tissue layer began at y = 0, and its depth proceeded in the negative y-direction. All subsequent skin layers were placed on top of the layer below it, with thicknesses proceeding in the positive y direction, beginning with the subcutaneous tissue layer. The overall width of the model was 150 units (15

PAGE 71

71 cm). The thickness of all skin layers combined, not including the deep tissue layer, was 2 mm thick. The deep tissue layer was either 2.5 cm or 5 cm thick and will hereafter be referred to as the thin or thick deep tissue layers, respectively. A te trapolar electrode setup, similar to that used by Scharfetter et al. (2001), was modeled. Four 1 cm wide electrodes were placed on top of the skin. The left -hand x -coordinate for each electrode was: 30, 50, 90, and 110. This spacing placed the outermost el ectrodes 3 cm from the edges of the model, provided 1 cm between the pairs of electrodes, and allowed for 3 cm between the innermost electrodes. The outermost electrodes were the current carrying electrodes, and the innermost, or pickup, electrodes measure d the voltage difference. The amount of current modeled was 250 A as was used in the study by Scharfetter et al. (2001). The electrical conductivity of copper was used for the electrodes (Table 41). In hindsight, silver may have been a better choice sinc e silver/silver chloride electrodes are common but the two metals have very similar conductivity values. A picture of the entire model can be seen if Figure 4 9. Figure 4 9. Skin model shown with electrodes on the skin surface and edema located within t he deep tissue layer. Boundary conditions were needed to define the model in order to run the simulations. All internal boundary conditions (between skin layers, etc.) were set to continuity. The bottom edge boundary of the deep tissue layer was defined as electric insulation. This is appropriate assuming that this is a tissue -bone boundary, as current will not as readily pass through bone and pressure ulcers form over bony prominences. The top skin boundary and the side, edge boundaries of the

PAGE 72

72 model were a lso set to electric insulation. The top skin boundary condition is appropriate as the stratum corneum has a large electrical impedance Despite the fact that some of the applied current could be expected to flow towards the edges of the skin model, this as sumption provides a simplification that would provide a best case scenario, with all current flow passing through the edematous tissue. For the current carrying electrodes, the boundary conditions were set to current flow, one positive and one negative. Ed ema was added to the model to determine whether the electrode setup could detect its presence. In order to model an appropriate amount, the size of edema that was introduced to the model needed to be determined. An ellipse was selected as the 2D object to represent a pocket of edema beneath the skin. Despite using a 2D model, the volume that the ellipse would represent as a 3D ellipsoid was calculated to determine the amount of edema. The volume of an ellipsoid (V) is determined by (4 1) where a, b, and c represent the radii of the ellipsoid in all three dimensions. With the xand y axes representing a 2D Cartesian coordinate system, assume the third dimension, z, will add depth. By setting the y and z axes to be equal (B = C) the ellipsoid volume equation becomes V = (4/3) 2. (4 2) For example, with A = 10 and B = 5, the ellipsoid volume equals 1047.20 mm3. Converting cubic millimeters to milliliters, 1000 mm3 = 1 cm3 = 1 mL, (4 3) the volume for the example represents just over 1 mL of edema. The value used for the electrical conductivity of edema, 1.5 S/m, was chosen to be similar to that of the electrical conductivity of blood plasma (Mohapatra and Hill, 1975).

PAGE 73

73 Simulations were run with and without edema present. Simulations were also run var ying the edema size and depth, with the horizontal ellipse radius, A, ranging from 2 to 18 mm by 4 mm intervals (2:4:18), and the vertical ellipse radius, B, ranging from 2 to 8 mm by 2 mm intervals (2:2:8). The edema was centrally located at x = 75 and y = 15. The equivalent edema volumes ranged from just greater than zero to over 4.8 mL. The edema was then moved horizontally and vertically in the tissue, but remained below or between the pickup electrodes. With a fixed -size ellipse of A = 10 and B = 5, t he simulations were run with the edema position ranging from 60 to 90 by 5 mm intervals for the x-coordinate (60:5:90), and from 7 to 19 by 4 mm intervals for the y -coordinate ( 7: 4: 19). Both sets of simulations were run with both the thin and thick d eep tissue layers to see how the edge effects at the bottom of the tissue layer were affected. Results With and without edema The current flow through the model with and without the presence of edema is shown as current density arrows (Figure 4 10) and cur rent density lines (Figure 4 11). The edema was removed by giving it the same electrical properties as the deep tissue layer. It can be seen in these figures how the current is drawn towards the more conductive medium, the edema, when present. The voltag e difference measurement between electrodes with no edema present dropped from a magnitude of about 0.07 V for the thin deep tissue layer to about 0.05 V for the thick deep tissue layer. Varying edema size The voltage difference measurements can be seen as the size of the edema varies in both the x and y -directions. Simulation results are shown for the thin (Figure 412) and thick (Figure 4 13) deep tissue layers. The magnitude of the voltage difference measurement decreased with

PAGE 74

74 an increase in edema size the greater the edema, the smaller the measured signal. However, the greater the edema size, the greater the signal difference compared to having no edema present. Figure 4 10. Skin model with current density arrows. Model shown with (upper) and without (lower) edema. Figure 4 11. Skin model with current density lines. Model shown with (upper) and without (lower) edema.

PAGE 75

75 Figure 4 12. Voltage difference measurements with varying edema size in thin deep tissue layer. Figure 4 13. Voltage difference measurements with varying edema size in thick deep tissue layer.

PAGE 76

76 Edema displacement The current flow through the model while varying the depth of edema is shown as current density arrows in Figure 4 14. The voltage difference measurements can be seen as t he placement of the edema varies in depth and lateral placement; simulation results are shown for thin (Figure 4 15) and thick (Figure 4 16) deep tissue layers. The magnitude of the voltage difference measurements decreased slightly when centrally located between both pickup electrodes resulting in a larger signal difference when compared to having no edema present. Shallower edema placement also resulted in smaller signal magnitudes but greater signal differences compared to having no edema present. Discussion The skin -edema model appeared to behave consistently and predictably throughout the simulations. However, the magnitude of the voltage differences between the pickup electrodes was relatively small, on the order of tens of millivolts, and was even sma ller when comparing edema to no edema. Also, as the deep tissue layer increased in thickness, the magnitudes of all signals, with and without edema, diminished. The thick deep tissue layer was only used to see how the bottom boundary condition and potentia l edge effects of the model might affect the results. The model provided consistent and predictable results for both deep tissue layer values. Voltage differences for increasing edema size (0 to 4.8 mL) resulted in a net change ranging from about zero for no edema to about 35 mV in the thin deep tissue layer. When modeling the thick deep tissue layer, the net change ranged from about zero to 20 mV. For the lateral and vertical edema displacements, the further away the edema was from the electrodes (deeper o r horizontally), the voltage difference between edema and without diminished. The lateral placement of the edema did not affect the results as much as the depth. In the thin deep tissue layer, for a given depth, there was no more than a 3 5 mV difference u pon varying the

PAGE 77

77 lateral placement of edema; varying the vertical placement from the shallowest to deepest position resulted in no more than a 8 mV difference. In the thick deep tissue layer, there was only a 1 2 mV difference upon varying the lateral place ment of edema for a given depth; varying the vertical placement from the shallowest to deepest position resulted in no more than a 5 mV difference. Lateral placement of edema, as long as it was beneath or between the pickup electrodes, had variations in me asurements that were within 5 10% of the measured signal for either thickness of deep tissue. For various edema depths, the variation between measurements ranged up to 1316% of the measured signal. Limitations The computer model was limited to two dimensi ons, but this is an appropriate approach to begin simulations. Skin properties and skin layer thicknesses can vary based on location of the body and from person to person, but starting values need to be chosen. The frequency of the injected current was a n ot parameter that could be defined whereas it can be controlled when using BIA or BIS. No simulations were run with the edema placed outside the bounds of the pickup electrodes which could occur in practice as the exact location of edema would not likely b e known. Conclusion After running several simulations varying the presence, size, and location of edema in a skin model, the resulting voltage difference measurements were minimal. The largest measurement difference had the magnitude of tens of millivolts, which would make obtaining accurate and reliable measurements very difficult in practice. Additionally, the results of these simulations were discussed with Dr. Rosalind Sadleir, an expert in the field of electrical impedance tomography (EIT) from the Dep artment of Biomedical Engineering, University of

PAGE 78

78 Florida.1 In her opinion, she believed that the signal levels were too low to obtain the desired results. For these reasons, this approach was no longer pursued. Figure 4 14. Current density lines wi th varying edema depth. Model shown with shallow (first), central (second), and deep (third) edema placement in thin deep tissue layer. Model shown with central edema placement in thick deep tissue layer (fourth). 1 EIT is an imaging technique that produces an image of conductivity or permittivity based on a combination of surface electrode measurements.

PAGE 79

79 Figure 4 15. Voltage difference measurem ents with varying edema depth and lateral placement in thin deep tissue layer. Figure 4 16. Voltage difference measurements with varying edema depth and lateral placement in thick deep tissue layer.

PAGE 80

80 Dermal Phase Meters Recent research by Bates Jensen et al. (2007) claimed to have some success in predicting stage I pressure ulcers by using a dermal phase meter (NOVA Technology Corp., Manchester, MA) that measured subepidermal moisture in dermal phase meter units (DPM).2 Their results demonstrated that, on average, DPM values increased from normal skin, to erythema/stage I pressure ulcer, to stage 2+ pressure ulcer. The odds ratio for pressure ulcer prediction was 1.26 for every 100 DPM increase, or 26% of erythema/stage I pressure ulcers that developed were visible by a skin assessment the following week. NOVA Technology Corp. was contacted to learn more about their equipment and their unique DPM unit measurements. The company explains on their webpage3 that the DPM unit is a measure of impedance. They not e that impedance is a complex quantity consisting of a real and an imaginary part that comprise the resistive and capacitive components of the impedance measurement. However, their devices output a relative value or reading, typically in the range of 90999 DPM units. It can only be speculated that this is some magnitude value combining the real and imaginary parts of impedance measurement. Speaking directly with a company representative did not clarify the representation of the DPM unit, but only confirmed that this information is proprietary. After several conversations and research discussions, a device evaluation was arranged for two of their devices, the DPM 9003 and the Petite (DPM 9020). The goal was to use their equipment as an impedance device to he lp detect the onset of pressure ulcer formation by potentially detecting edema beneath the skin surface. 2 In the study by Bates Jensen et al., they use the term DPU for dermal phase units. Since NOVA Technology Corp. uses the term DPM, that is the term used in this chapter. 3 http://www.novatechcorp.com/dpm.html

PAGE 81

81 Equipment The DPM 9003 device has been commercially available since 1990 and consists of an instrument box with a removable sensor probe. The sensor on the tip of the probe consists of two concentric brass ring electrodes with an isolater between them. Probes are available with various sensor diameters, primarily ranging from about 4 mm to 9 mm (outer diameter). Articulated probes/tips are also available that place the sensor at a particular angle to the probe handle. The Petite device, which was used in the Bates Jensen et al. study, was commercialized in 2001. This device is more of an all in -one device with no instrument box; essentially, it is a large r probe with built in components. The probe interfaces with a personal digital assistant (PDA) to record and store the DPM data. The device not only outputs its impedance value but also measures skin temperature and has force compensation. However, the Pet ite is no longer commercially available from NOVA because the company was unable to support the device with continuous PDA manufacturer updates. Product Evaluation Various tests were conducted to compare the two devices on two healthy adult males, aged 26 and 45 years, to determine whether or not these impedance devices would be appropriate for trying to detect pressure ulcer formation. They were evaluated based on experimental results, reliability, and feasibility to use on patients in an ICU environment. Experimental protocols consisted of taking several measurements on the forearm, as well as before and after the use of moisturizing lotion, blanching the skin, and using an indenter. Forearm Measurements Sites To compare the two NOVA devices, DPM measureme nts were taken at several sites on both forearms. Forearm sites included the volar and dorsal sides, as well as along the ulna. For the dorsal forearm, four measurement sites were marked with a pen at 2 intervals (2, 4, 6,

PAGE 82

82 and 8) from wrist to elbow. For the volar and ulnar forearm, four measurement sites were marked with a pen at 2 intervals from elbow to wrist. For all forearm measurements, the subject sat in an armchair with his lower arm at rest on the armrest with the elbow bent approximately 90 degrees. The DPM measurements were not made on the pen marks but next to them and care was taken not to take measurements on any well -defined superficial veins. Measurement Protocols Forearm measurements were taken with both the DPM 9003 (standard and arti culated probes) and the Petite at the four measurements sites on the volar, dorsal, and ulnar locations of each forearm. The measurements were taken again at the dorsal and volar forearm sites after using moisturizing lotion (Quench Body Lotion, Olay Body, Cincinnati, OH). DPM measurements were also taken before and after blanching the skin. The subject firmly pressed on the skin with one finger or thumb at the measurement site for 2 minutes. The measurement sites were the 2 and 4 dorsal, volar, and ulnar forearm sites for each arm. The subject positioning did not change for the blanching measurements. For the indenter measurements, the measurement site was marked with a pen on the left upper arm, 2 from the point of the elbow. The subject sat (perhaps sl ouched) in an armchair with his upper arm at rest on the armrest, with the lower arm held vertical. For the before measurement, the subject sat in the prescribed position for 2 minutes. To access to the measurement site, the upper arm was kept on the armre st but allowed to rotate as the lower arm was rotated inwards (about 90 degrees). The subject then rested between measurements with his lower arm on the armrest for at least 2 minutes before performing the indentation. For the indention measurement, the su bject sat in the prescribed position for 2 minutes with a marble

PAGE 83

83 placed under the arm at the measurement site. The arm was to rest, but not press down, on the marble.4 After 2 minutes, the arm was rotated to take the DPM impedance measurement. Results DPM 9003 standard vs. articulated probe The forearm measurement protocol was completed with the DPM 9003 with both the standard probe and the articulated probe. For all 24 measurements, 12 per arm, the articulated probe output a DPM value that was significantl y lower than the standard probe. The comparison was made by taking the value of the articulated probe as a percentage of the standard probe. Over the entire dataset, the mean articulated probe DPM measurement was 69.0% of the standard probe measurement (me an 0.690, SD 0.094) while ranging from 52.7% to 85.5% of the standard probe value (Figure 4 17). Figure 4 17. DPM 9003 standard vs. articulated probe. 4 The indenting procedure (use of marble, location, and duration) came from a study by Rajendran et al. (2006) in which they were trying to improve the detecti on of Stage I pressure ulcers by enhancing digital color images.

PAGE 84

84 DPM 9003 vs. Petite The forearm measurement protocol was completed twice with both the DPM 9003 and the Petite device. For all 24 measurements, 12 per arm, the Petite probe output a DPM value that was significantly lower than the DPM 9003. Again, the comparison was made by taking the value of the Petite measurement as a percentage of the DPM 9003. Over the entire dataset, the mean Petite DPM measurement was 34.1% (day 1) and 39.3% (day 2) of the DPM 9003 measurement (day 1: mean 0.341, SD 0.135; day 2: mean 0.393, SD 0.153) while ranging from 14.1% to 70.8% (day 1) and 11.9% to 72.9% (day 2) of the standard probe value (Figure 4 18). Figure 4 18. DPM 9003 vs. Petite. Moisturizing lotion Forearm measurements were taken at the dorsal and volar sites with both the DPM 9003 and Petite devices before and after using moisturizing lotion. For all 16 measurements, 8 per arm, the DPM values after using the lotion were tremendously larger than beforehand. The

PAGE 85

85 comparison was made by taking the value of the after lotion measurement as a percentage of the before lotion measurement. For the DPM 9003, over the entire datas et (1 before -lotion data point did not record), the mean after lotion DPM measurements was 300% larger then the before -lotion measurement (mean 3.00, SD 1.09) while ranging from 195% to 623% of the before lotion measurement (Figure 4 19). For the Petite, over the entire dataset, the mean after lotion DPM measurement was 810% larger then the before -lotion measurement (mean 8.10, SD 6.84) while ranging from 209% to 2,600% of the before lotion measurement (Figure 4 19). Figure 4 19. DPM measurements before a nd after moisturizing lotion. Blanching Forearm measurements were taken on both subjects at the 2 and 4 dorsal, volar, and ulnar sites with the DPM 9003 before and after blanching the skin. Of the 12 before and after measurements, 6 per arm, 10 of the DP M measurements increased in value after blanching for subject one, and 11 of 12 measurements increased for subject two (Figure 4 -20).

PAGE 86

86 The Petite device was used to measure before and after blanching on subject two at the 2 and 4 dorsal and volar sites. O f the 8 before and after measurements, 4 per arm, 7 of 8 DPM measurements increased after blanching (Figure 420). Indenting Measurements were taken on two subjects with the DPM 9003 before and after using an indenter 2 above the elbow on the left arm. Af ter using the indenter, the DPM values increased by a factor of 1.38 (142 to 196) for subject one and by 1.32 (386 to 509) for subject two (Figure 4 20). Figure 4 20. DPM measurements before and after blanching and indenting. Discussion Forearm measureme nts The results demonstrated that different skin types (forearm locations) have different DPM measurements. The DPM values generally increased from the dorsal, to the ulnar, to the volar sites. Of the 4 measurements on a particular skin type, the particula r location did not matter, as

PAGE 87

87 there was no correlation between DPM value and the distance from the wrist or elbow. The forearm DPM measurements varied day to day, but the trend of increasing DPM value, from dorsal, to ulnar, to volar, held. The results also demonstrated that the DPM 9003 output values that were much larger than the Petite for the very same locations on the skin. Also, the standard probe output DPM values that were much higher than the articulated probe. The magnitudes of the DPM value seeme d to correlate somewhat with the size of the sensor itself since the measurements were made on the same subjects, on the same days, under the same conditions, and the only difference between measurements was the probe and/or the device used to measure them The probes sensors increased in outer diameter from the Petite, to the articulated probe of the DPM 9003, to the standard probe of the DPM 9003. Moisturizing lotion The NOVA representatives explained that the testing environment needed to be controlled (temperature, humidity) when taking DPM measurements. In particular, they mentioned that the measurement was sensitive to sweat, as it can be used to measure skin moisture. To investigate how the skin would be affected by moisture, moisturizing lotion was applied to the skin after taking the initial forearm measurements. The results demonstrated a tremendous change after lotioning the skin. The DPM measurements increased by an average of 300% for the DPM 9003 and 810% for the Petite. Blanching and indenting Blanching the skin was investigated because nonblanchable redness is an indicator of stage I pressure ulcers. During pressure ulcer formation, the skin can become blanched due to external pressures, forcing blood flow away from the tissue, which can even tually lead to tissue breakdown. It seemed reasonable that blanching would move fluid away from the skin and thus

PAGE 88

88 increase its impedance value. However, not knowing precisely what the DPM unit represents, the only certainty was that the DPM values increase with moisture as evidenced by the moisturizing lotion trials. Recalling the study by Bates Jensen et al. (2007), their results demonstrated that pressure ulcers, on average, had higher DPM values than normal skin, even when controlled for incontinence. It is curious then, if just referring to DPM values, one would not be able to distinguish healthy, moisturized skin from skin with a stage 2+ pressure ulcer. However, it may be possible, as they tried to demonstrate, that the measurements could be useful whe n monitored over time. The majority of DPM measurements increased after blanching. However, since the protocol instructed the subjects to use a finger or thumb to press on their arm, it seemed possible that sweat or moisture from one hand was transferring to the other arm and resulting in a larger DPM reading. This cannot be definitively answered from the blanching measurements, but the indenter measurements provide some insight since the indenter trial did not involve any exertion by the subject or touching with their other hand. Results showed that for both subjects the DPM value increased after using the indenter, suggesting that moisture transfer from handto arm was not the cause of the increased DPM values during the blanching trials. Conclusions The p revious experiments were useful in evaluating two of NOVA Technology Corp.s DPM impedance devices, the DPM 9003 and the Petite. They both seemed to reliably provide consistent DPM output values, however, it was never clear as to what these values actually represented or how the impedance is represented as a DPM value. The devices were very sensitive to moisture, which at first seemed to be an attribute, but then again, the skin surface is not the area of immediate interest, but the tissue below the surface It was not clear whether or not the devices were providing information based solely on the surface of the skin or actually

PAGE 89

89 identifying what was happening deeper, beneath the surface. Due to the large measurement fluctuations depending on skin moisture, t he devices did not seem like they would be reliable, or at least feasible to use, in the ICU environment with incontinence concerns and other external moisture factors. The size of the sensors was also a concern. If DPM measurements were to be taken on the sacrum of a patient, for example, with such a small sensor (outer diameter less than 1 cm), repeating the measurement on a patient multiple times, or let alone on different patients, it did not seem realistic that the same precise location would be measur ed. Even if the skin were marked so that the same location could be measured, to say that one particular mark on the sacral skin is a better choice to measure than another would be a matter of guesswork. It is understood how a good protocol is vital to obt aining appropriate and consistent results. However, working directly with a target population of at risk patients, performing the necessary protocols would have its challenges. For these reasons, and for all of the concerns about the devices and their meas urements, the use of these devices was no longer persued. Bioimpedance Experiments with Basic Electrode Setups Since neither the computer modeling nor the skin impedance devices showed promise for detecting the onset of pressure ulcer formation, it was exp lored as to whether or not a basic or tetrapolar electrode setup could detect changes in interface pressures applied to the skin. To investigate this idea, a bioimpedance laboratory exercise was performed, followed by an investigation into how interface pr essures affect a skin impedance measurement. The purpose of the laboratory exercise was to measure the effective impedance of two different types of electrodes and to become re -familiarized with some of the electrical equipment.

PAGE 90

90 Equipment The electrical eq uipment used for the following experiments consisted of a 4 MHz Function Generator (182A, Wavetek, San Diego, CA ), a 100 MHz oscilloscope (2235, Tektronix, Beaverton, OR), a 444 resistor (used in place of a decade box set to 500 ), various leads, ECG el ectrodes (Meditrace 535, Kendall, Chicopee, MA ), and patient return electrodes (Infant REM PolyHesive II, Valleylab, Boulder, CO ). The ECG electrodes had 1 diameters with additional foam backing that made the overall diameter 1.75. The patient return ele ctrode is built as a 2 -in 1 electrode comprised of 2 rectangular (1.25 x 4) electrodes divided and surrounded by foam padding (overall size is 3.75 x 4.75) The patient return electrode was cut in half to separate the two electrodes in order to use th em individually. A n electrical safety check for leakage current and ground integrity was performed on the oscilloscope and signal generator before use. The leakage current was confirmed to be less 50 A and the ground integrity was less than 0.5 Measurement of Electrode Impedance Protocol The amplitude of the function generator was set to 4 volts peak-to -peak (Vpp) at 50 Hz using the oscilloscope. The frequency of the generator was adjusted as needed to make one cycle of the sine wave exactly 10 horizont al divisions on the oscilloscope display. The circuit shown in Figure 4 21 was then constructed. After the circuit was assembled, two ECG electrodes were placed approximately 6 inches apart on a clean forearm. Alligator clips connected the electrodes to th e signal generator. The oscilloscope was set to dual trace mode and the amplitude of channel 2 was measured. The phase difference between channels 1 and 2 was measured as accurately as possible. The voltage across

PAGE 91

91 the resistor was a complex number and was converted to root mean squared voltage (Vrms). By subtraction, the voltage drop across the subject (and the electrodes) was determined. The current in the circuit was then calculated using the voltage drop measured across the resistor. The current value wa s used to calculate the impedance of the subject and the electrodes. Then, the impedance of the patient (assumed to be a resistive load of 500 ) was subtracted to leave the impedance of only the two electrodes. Finally, half of this value equaled the impe dance of just one electrode. The protocol was then repeated for the patient return electrodes. Figure 4 21. Electrode impedance measurement circuit. A resistor was used in place of the decade box. Results and Discussion The step -by -step results for calculating the electrode impedance can be found in Table 4 2. The measurement of electrode impedance was determined for two types of electrode pairs, ECG and patient return electrodes. The impedance of one patient return electrode was about 4 0% of one ECG electrode. Table 4 2. Electrode impedance results. ECG Electrodes Patient Return Electrodes Amplitude 20 mV pp 100 mV pp Phase difference between channels 2.5 ms 4 ms V rms across resistor 0.14142 V 0.316228 V Voltage drop across subject (a nd electrodes) 3.85858 V 3.68377 V Current in circuit 318.52 A 712.22 A Impedance of subject and electrodes 12,114.2 12,114 30.285i 5,172.2 5,172.2 20.689i Impedance of electrodes (2) 11,614 30.285i 4,672.2 20.689i Impedance of electrode (1) 5,807.1 15.143i 2,336.1 10.344i

PAGE 92

92 Interface Pressure Detection Protocol The same electrical circuit as in the previous experiment was setup (Figure 4 21). The patient return electrodes were again placed approximately 6 inches apart on a clean forearm. The amplitude of the function generator was set to 4 volts peak-to -peak (Vpp) at 50 Hz using the oscilloscope. Pressure was then applied to the skin between and around the electrodes to see if any change in the signal could be detected. T he frequency of the function generator was then increased and pressure was again applied to determine any signal changes. Observations were made for signal frequencies of 50 Hz, 500 Hz, 5 kHz, 50 kHz, and 500 kHz. The protocol was then repeated with the el ectrode setup arranged on the upper leg (quadriceps). The above protocol was performed again with a tetrapolar electrode setup. Two patient return electrodes were placed outside the first two. The signal generator was connected to the two outermost electro des, and channels 1 and 2 of the oscilloscope were connected to the innermost electrodes. Pressure was again applied to the skin between and around the electrodes to see if any change in the signal could be detected. The procedure was also repeated with th e tetrapolar setup arranged on the upper leg. Results Initial electrode setup A s the frequency of the signal generator increased, the peak to peak voltage of channel 2 increased. At 5 kHz and higher, the peak-to -peak voltage of channel 1 decreased by about 10%. The phase difference between channels decreased with an increase in frequency. However, at all frequencies, applying pressure to the arm or leg did not affect the magnitude or phase difference of the signal.

PAGE 93

93 Tetrapolar electrode setup The peak to -pea k voltage of channel 2 was just slightly less than channel 1 for all frequencies on the arm and leg. The was no measurable phase difference between the 2 signals at any of the frequencies. Additionally, no signal changes were observed while applying signif icant pressure to the skin and underlying tissue when the electrode setups were arranged on the arm or the leg. Discussion Application of large interface pressures did not change the measured signal on either the forearm or the upper leg when using the two-electrode setup from the electrode impedance experiments or when using a basic tetrapolar setup. Therefore, it seems unlikely that a basic electrode arrangement and impedance measurement can be used to determine interface pressures applied to the skin.

PAGE 94

94 CHAPTER 5 MONITORING THE INTERFACE PRESSURE PROFILES OF PATIENTS AT -RISK FOR PRESSURE ULCER FORMATION Healthy subject studies, as described in C hapter 3, demonstrated that raising the HOB to 30 degrees or higher on an ICU bed increases the peak interface pressure between the skin and support surface. A t 45 degrees or higher, the area of skin with a skinICU bed interface pressure greater than 32 mm Hg also increases. These studies also showed that standard turning by experienced ICU nurses does not reliabl y unload all areas of high skin bed interface pressur es resulting in areas always at risk regardless of position, a triple jeopardy area. These results were novel and clinically relevant. However, the limitation of both studies was that the subjects were h ealthy adults and not patients at risk for pressure ulcer formation. Therefore, the study outlined in this chapter addressing Aim 3 studied patients at risk for pressure ulcer formation. Not only were interface pressure measurements taken on at risk pati ents, but the patients were monitored over time to determine what sort of pressures they experienced during their care which involved q2h turning. Currently, t here is no data in the literature that has monitor ed the interface pressures of at risk patients Interface Pressure Measurement and Other Equipment Interface pressure measurement profiles were obtained using an Xsensor pressure -mapping system. The pressure sensor pad is very thin and flexible and consists of 48 x 48 half -inch, independent sensors co mprising a 24 inch x 24 inch square array. Additional, details about this pressure mapping device can be found in Chapter 2. Before using the equipment in the ICU environment, the Xsensor pad was wrapped in thin (1 mil) plastic sheeting to protect it from any bodily fluids or other contaminations. The sensor pad was also placed beneath the patients underpads/chux when in use. The pad was disinfected after each use and re -wrapped prior to the next patient. A modern ICU bed was used for all measurements (Tot al Care or Total Care

PAGE 95

95 SpO2RT, Hill Rom, Batesville, IN). The HOB elevation was observed from the beds built in ball bearing measurement located in the side rail of the bed. Specifically marked HOB angles included 0, 10, 20, 30, 45, 60, and 75 degrees. Pow er Analysis As in the lateral turning study (Chapter 3), the tissue area subjected to prolonged interface pressure greater than 32 mm Hg was considered at risk tissue. When the same tissue was at risk in all three turned positions, supine, left, and right, it was termed a triple jeopardy area. G*Power 3.01 was used for the power analysis. The a priori null hypothesis was that at -risk patients undergoing q2h turning would not demonstrate a triple jeopardy area. To achieve an appropriate power of 80%, with an effect size of 0.8 on a one -tailed test with an error probability of 5%, a sample size of 12 was needed. Protocol After IRB approval and informed consent, the Xsensor pad was placed beneath the patient. The process usually required the nurses to roll the patient to one side and then the other, similar to the lateral turning procedure, so that the pad could be appropriately positioned beneath the sacrum. Apart from positioning the patient on the pad, no other patient interventions were made on account of this research as this was purely an observational study. Interface pressure measurements were continuously recorded from the patients as they lay in their beds and received their routine care. When the patient was turned, the nurses were asked to place any p illows or wedges, which were used to maintain the turned position, beneath the pad in order to measure the interface pressures between the patient and the supporting device. For every turn, it was confirmed that the patient was centered on the pad. It was necessary at times to have the 1 http://www.psycho.uni duesseldorf.de/abteilungen/aap/gpower3/

PAGE 96

96 nurses straighten out the pad if it got bunched up or to reposition the patient if he or she had moved or slid off the pad. Raising the patients HOB occasionally resulted in the patient sliding down the pad. Repositioning was only conducted when the patient was already in the process of being turned to limit intervention. Interface pressure profiles were recorded typically for a 4 6 hour period. This time frame was chosen to theoretically allow observation of all 3 of the t urned positions: supine, left, and right. An interface pressure profile was recorded by the computer every 30 seconds, twice a minute. The researcher did not sit at the patients bedside for the entire study, but regularly took notes about the general posi tioning of the patient (direction of turn and HOB elevation). The data was also intermittently saved over the course of the study. Subjects Twenty three patients participated in the interface pressure monitoring study. Patients were enrolled in the study u ntil at least 12 patients were monitored in all three positions: supine, left, and right. The inclusion criterion was for ICU or intermediate care (IMC) patients that would undergo lateral turning as part of their routine care. Physicians identified potent ial candidates for the study during regular care of their patients. The total population consisted of 14 men and 9 women aged from 32 to 84 years (63.3 12.7). The subjects heights ranged from 1.40 m to 1.85 m (1.70 m 0.11 m), and they weighed between 55.0 kg and 124.7 kg (85.9 kg 21.9 kg). The resulting BMIs ranged from 20.1 to 40.1 (29.3 5.6). The triple jeopardy population (n = 13) that was observed in all 3 positions consisted of 8 men and 5 women aged from 32 to 84 years (60.1 15.2). The subj ects heights ranged from 1.57 m to 1.85 m (1.71 m 0.09 m), and they weighed between 62.0 kg and 124.7 kg (89.6 kg 23.3 kg). The resulting BMIs ranged from 22.7 to 38.0 (30.3 5.5).

PAGE 97

97 Data Analysis Matlab and Excel were used to plot, image, align, analy ze, compile, and compare the interface pressure profile data. Each profile provided the interface pressure (mm Hg) at each of the 2,304 discrete sensors. The maximal pressures were determined, and the at risk area that was subjected to various pressure thr esholds was calculated. Patient movies were generated, which consisted of viewing the interface pressure profiles one after the next, to allow visualization of the entire monitoring period. Over the course of the monitoring period, some interface pressure profiles were unusable and were not included in the analysis. These removed profiles consisted of interface pressure profiles that were obtained during patient repositioning or when the patients sacral area had moved off the pad. The positions obtained for each patient were aligned anatomically to determine how turning altered the interface pressures. 2D cross correlation was used to help align the observed positions for a given patient but were manually adjusted when necessary. Wilcoxon rank -sum tests and Wilcoxon signed rank test s, when appropriate, were used to compare interface pressures and at -risk areas between populations, different positions, and to compare ICU patients to healthy subjects. A p < 0.05 was significant. Results Observations Data colle ction Over the entire study, 15,784 interface pressure profiles were recorded over a period of more than 131.5 hours. After the removal of unusable profiles, the total number of profiles analyzed was 1 4 5271 21 hours of monitoring time. For the triple jeopardy population, 8 028 profiles w ere analyzed 66.9 hours of monitoring time.

PAGE 98

98 HOB elevation Patient positioning observations were made throughout the study, which included HOB elevations for each of the positions. For the supine position, the average HOB el evation was 30 degrees. All supine HOB angles ranged from 15 45 degrees with the exception of one patient who was positioned in a sitting position once and had a HOB elevation of 65 degrees. For the lateral left position, the average HOB elevation was 26 degrees, and the HOB angles ranged from about 18 40 degrees. For the lateral right position, the average HOB elevation was also 26 degrees, and the HOB angles ranged from 15 45 degrees. Interface P ressures The peak interface pressures (Figure 5 1 ) did not d iffer much between differ ent positions or between the total population and the triple jeopardy population. Since data w ere recorded over a period of time, the peak interface pressures for a patient could be monitored over time as well. It can be seen in Fi gures 5 2 and 5 3 how the peak interface pressures are susceptible to change upon changes in position, but just because a patient is repositioned does not mean that interface pressures will change significantly (Figure 5 2). Conversely, in Figure 5 3, upon turning to the left, the peak interface pressures increased nearly three -fold, likely due to positioning that shifted that patients body weight directly over a bony prominence, in this case the trochanter. The mean of the peak interface pressures observ ed for the different positions can also be seen in Figure 5 1 The means of the peak interface pressures did not differ much by position or between the total population and the triple jeopardy population. The mean of peak interface pressures were consisten tly about 7 3 8 0 % of the peak interface pressure value s for all positions and the difference was significant (p < 0.0001)

PAGE 99

99 Figure 5 1. Interface pressure results. The data points represent the mean SD for the peak interface pressures and the mean of pea k interface pressures for the total population in the supine, left, and right positions. For the supine, left, and right positions, n = 24 2 3 and 2 3 respectively. *The mean of peak interface pressures are statistically different from the peak interface p ressures at each of the positions ( p < 0.0001) Figure 5 2. Peak interface pressures over time for one patient. This patient was initially supine, turned right at frame 13, turned supine at frame 423, and turned left at frame 652.

PAGE 100

100 Figure 5 3. Peak inte rface pressures over time for one patient. This patient was initially turned to the right, turned supine at frame 294, and turned left at frame 535. At -Risk Areas The at risk areas were monitored over time for the three positions for both the total and triple jeopardy populations. The average amount of at -risk area over time for a given position was determined; the amount of specific tissue area that was at risk throughout the entire duration of each position was determined; and, the amount of specific tiss ue area that was at risk during at least 95% of the entire duration of each position was also determined. The results of these 3 situations were all statistically significant from one another (p < 0.0001) and can be seen in Figure 5 4. There was little dif ference and no statistical significance between populations. T he at risk area s could be monitored across all positions experienced by each individual patient as well. It can be seen in Figures 5 5 and 5 6 how the amount of at -risk area changed over the cou rse of the patients monitoring period, but just because a patient is repositioned does not

PAGE 101

101 necessarily result in a significant change of at risk area (Figure 5 5). However, in Figure 5 6, it can be seen again how positioning plays a significant role in how much tissue area is at risk. Figure 5 4. At risk area results. The data points represent the mean SD for the average amount of at risk area, the same area always at -risk, and the same area always at risk for at least 95% of the duration for the total population in the supine, left, and right positions. For the supine, left, and right positions, n = 2 4 2 3 and 2 3 respectively. *The average amount of area at risk, the same area always at -risk, and the same area always at risk for at least 95% of the duration are all statistically different from one another at each position ( p < 0.0001) Figure 5 5. At risk areas over time for one patient. This patient was initially supine and was turned left at frame 168.

PAGE 102

102 Figure 5 6. At risk areas over time for one patient. This patient was initially turned to the right, turned left at frame 145, turned right at frame 403, and turned left at frame 664. Always At -Risk Areas Always at risk area is a term that describes specific areas of skin that were at risk throughout the entire observation period, regardless of the positions experienced. This term includes those patients who were not necessarily observed in all three positions but nonetheless had specific areas of skin always at risk. For the total population, 19 of 23 patients demonstrated always at risk areas across every interface pressure profile. The mean always at -risk area for all patients was 135 cm2 with a range from 0 561 cm2 (Table 5 1) Considering areas that were always at risk for at least 95% of the t otal observation period, all patients, 23 of 23, demonstrated always at risk areas. The mean always at risk areas for at least 95% of the patients monitoring period was 206 cm2 with a range from 2 613 cm2 (Table 5 1) The always at risk area for at least 95% of the total observation period increased for all subjects compared to the absolute always at risk area and the difference was significant (p < 0.0001)

PAGE 103

103 Table 5 1. Triple jeopardy and always at -risk areas. The positions are: supine (S), left (L), righ t (R), and sitting (Sit). *The always at risk and triple jeopardy areas > 95% of the entire duration are both significantly larger than their absolute measure ( p < 0.002) Number of Always At Risk Area Triple Jeopardy Area Always At Risk Area Triple Jeop ardy Area Patient Positions Positions absolute absolute > 95% of monitoring time > 95% of monitoring time [cm2] [cm2] [cm2] [cm2] 1 R, L, R, L 4 240 297 2 R, L 2 0 5 3 S, L 2 350 469 4 S, L, R, L 4 0 0 8 8 5 S, L, S 3 8 73 6 S, R, S, L 4 16 16 108 108 7 S, L, R 3 68 68 119 119 8 S, L 2 131 169 9 S 1 2 261 10 R 1 221 247 11 S, R, L, R 4 5 5 110 110 12 R, S, L 3 166 166 456 456 13 S, R, L 3 561 561 613 613 14 S, Sit, S, R 4 418 516 15 S, L 2 361 427 16 R, L, R, L 4 68 105 17 S, L, S, R 4 40 40 76 76 18 R, S, L 3 3 3 15 15 19 R, L, R, S 4 135 135 195 195 20 S, L, R 3 200 200 248 248 21 S, R, L 3 0 0 2 2 22 S, L, R 3 0 0 15 15 23 S, L, R 3 118 118 194 194 avg 3.0 135 101 206 166

PAGE 104

104 Triple Jeopardy Area s The only difference between a triple jeopardy area and an always at risk area is whether or not the patient was observed in all 3 positions. The triple jeopardy areas are the areas that are at risk throughout the entire observation period To view the sp ecific areas of skin that were always at risk for a patient, a stacked image was created. This image stacks up all of the at risk areas from each interface pressure profile from the patients entire monitoring period. Two examples of this can be seen in Figure s 5 7 and 5 8 that demonstrate how the at -risk skin area s were affected over time Of the 13 at -risk patients in the triple jeopardy population 10 exhibited areas of triple jeopardy across all interface pressure profiles The mean triple jeopardy ar ea for the group was 101 cm2 with a range from 0 561 cm2 (Table 5 1). Considering areas that were always at risk for at least 95% of the total observation period, all patients, 13 of 13, demonstrated triple jeopardy areas. The mean triple jeopardy area at risk during at least 95% of the patients monitoring period was 166 cm2 with a range from 2 613 cm2 (T able 5 1) The triple jeopardy area for at least 95% of the total observation period increased for all subjects compared to the absolute triple jeopardy a rea and the difference was significant (p < 0.002) Discussion Regular turning of patients every two hours is the standard of care used to reduce the risk of pressure ulcer formation by unloading and redistributing interface pressures. However, despite thi s practice, the results clearly demonstrate that at risk patients experience significant areas of skin that do not get unloaded and remain at risk. This observation was not an isolated example, but was observed for all 23 patients as they all demonstrated always at -risk areas throughout at least 95% of their entire monitoring period. Even for patients observed in all 3 positions, the entire triple jeopardy population also demonstrated triple jeopardy areas throughout at least 95% of their entire monitoring period. The greater than 95% term was implemented to give a more

PAGE 105

105 realistic representation of what the patients actually experienced. For example, if a patient momentarily rolled to one side and then back, the pressure sensor would have recorded an interf ace pressure profile that suggested particular tissue areas were unloaded, though just briefly. Most would agree that unloading an area for only a few minutes every two hours in not adequate It was confirmed that only one patient (no. 4) from the triple j eopardy population could have had a triple jeopardy area observed due to just two of the three positions because the 5% was longer than one of the position durations Figure 5 7. Stacked image demonstrating triple jeopardy areas. The image demonstrates a reas of skin that were at risk over the course of all positions experienced by this one patient. The color bar indicates the amount of time (in hours) specific skin areas were at risk. Areas at risk for the maximal amount of time means that area was never unloaded. The absolute triple jeopardy area for this patient was 68 cm2. Patients were monitored, on average, over 5 consecutive hours and specific tissue areas were at risk for the entire duration. Based on these results, one can sp eculate that these area s are at risk for the majority of the patients stay in the hospital if not the entire time The results mirrored those observed in the healthy subject study that discovered the triple jeopardy area

PAGE 106

106 phenomenon (Chapter 3), and they confirm that patients e xperience always at risk areas of skin for several continuous hours despite turning. Figure 5 8. Stacked image demonstrating triple jeopardy areas. The image demonstrates areas of skin that were at risk over the course of all positions experienced by thi s one patient. The color bar indicates the amount of time (in hours) specific skin areas were at risk. Areas at risk for the maximal amount of time means that area was never unloaded. The absolute triple jeopardy area for this patient was 561 cm2. Comparin g at -risk patients to healthy subjects, peak interface pressures, at -risk areas, and triple jeopardy areas were all higher for the patients. The peak interface pressures were all considerably higher in the patients regardless of the position (p < 0. 00002) (Table 5 2). Even the average peak pressure over the entire duration of each position was larger than the healthy subjects data despite not being statistically significant. The average across the duration of the position was used to give an idea of the hi ghest continual interface pressure experienced by patients rather than just a one time measurement. The at risk areas for patients were also significantly higher than for healthy subjects (Table 5 3). Patient values were significantly larger than the supin e and laterally turned positions of the healthy subjects (p < 0.002), and were larger than the laterally turned positions with HOB elevation but not significantly. The triple jeopardy

PAGE 107

107 areas and always at risk areas, were also much larger in patients (Table 5 4). Patient values were significantly larger for the always atrisk areas > 95% duration than the healthy subject triple jeopardy areas (p < 0. 006). Perhaps due to age, frailty, medical condition, nursing reservation not to move the patient too much due to pain or other medical reason, at risk patients do experience higher interface pressures and are more susceptible to larger at risk areas and always at risk areas than healthy subjects. Table 5 2. Comparison of interface pressures (mm Hg) between health y subjects and at risk patients. *Patient values significantly larger than healthy subjects at corresponding supine and laterally turned positions (p < 0. 000 002). ^Patient values significantly larger than healthy subjects at corresponding turned with HOB e levation position (p < 0. 0000 2). Subject Position Supine Left Right Healthy subject s : 68.6 19.5 (L) 65.8 11.7 (R) Laterally turned 69.2 12.8 64.8 9.1 Turned with 30 HOB elevation 84.5 1 7.5 8 0.4 1 1.4 At risk patients with HOB elevati on: Peak pressures *^ 126.3 45 8 136.0 42 9 122.2 32 6 Mean of peak pressures 101.1 34.6 99.7 29 9 89.5 16 8 L prior to turn ing to left side R prior to turn ing to right side All measurements taken on the same brand of modern ICU bed Table 5 3. Comparison of at risk areas (cm2) between healthy subjects and at risk patients. *Patient values significantly larger than healthy subjects at corresponding supine and laterally positions (p < 0.002). Subject Position Supine Left Right H ealthy subject s : 470 170 (L) 480 1 70 (R) Laterally turned 468 151 43 4 147 Turned with 30 HOB elevation 569 1 92 55 8 1 59 At risk patients with HOB elevation 716 290 742 304 744 287 L prior to turn ing to left side R prior to t urn ing to right side All measurements taken on the same brand of modern ICU bed

PAGE 108

108 Table 5 4. Comparison of triple jeopardy and always at risk areas between healthy subjects and at risk patients. *Patient value significantly larger than healthy subjects (p < 0.006). Triple Jeopardy or Always At Risk Area (cm 2 ) Healthy subject s : 60 At risk patients: Entire monitoring time > 95% of monitoring time Triple jeopardy area 101 166 Always at risk area 135 206 All measurements taken on the same brand of modern ICU bed Despite q2h turning being a standard of care, the turning practice itself did not appear to be strictly followed and was not the same for each patient. M any different turning protocols were observed and two subjects were not turned at all throughout the ir entire monitoring period (Table 5 1) It was unclear as to why the turning procedure s w ere so different However, in regard to patient turning, if the concern is really about unloading tissue to prevent pressure ulcers, it does not matter how many positions the patient experiences, but whether or not the tissue gets regularly relieved from pressure that puts it at risk. Regardless, of the turning practices observed, none were effective in unloading all at risk tissue. To put things in pers pective, the area of a standard 8.5 x 11 sheet of paper is just over 600 cm2. With an average of over 200 cm2 always at risk for each patient, roughly an area a third of the size of a sheet of paper is not getting unloaded and is at risk for several conse cutive hours Limitations and Future Research Active patients move around in bed, and at times slide down the pad with changes in HOB position, resulting in unusable interface pressure profiles. The active movements are likely to benefit the patient by hel ping the patient redistribute their weight as healthy people do on a regular basis, i.e. when sitting in a chair. However, extra movements could lead to additional shearing and frictional forces that could be damaging in regard to pressure ulcer formation. Significant changes in position on the sensor pad require more effort on the data analysis end to

PAGE 109

109 determine which areas get unloaded and which do not. Additionally, the Xsensor pad needed to be placed beneath the patients chux to protect it from the pati ent and additional contaminants, but also so it would not be used in place of the chux to help turn the patient. The chux are thin and towel -like but could aid in slight pressure relief. Further study is needed to establish how to favorably impact and even eliminate peak interface pressures, at risk areas, and always at risk (including triple jeopardy) areas, in patients at risk for pressure ulcer formation. Additionally, it needs to be determined whether or not patients with always at risk areas are, in fa ct, more likely to develop a pressure ulcer, or if more serious pressure ulcers develop, and whether or not any developing pressure ulcers form at the suspected tissue area locations. Conclusion ICU and IMC patients at risk for pressure ulcer formation exh ibit areas of skin that are always at risk throughout the duration of their hospital stay despite the preventative measure of turning. Turning practices appear to vary, yet none were successful in unloading all tissue areas at risk for skin breakdown Addi tionally, at risk patients are susceptible to higher interface pressures, larger at risk areas, and larger always at risk and triple jeopardy areas in comparison to healthy subjects.

PAGE 110

110 CHAPTER 6 FINAL DISCUSSION AND CONCLUDING REMARKS Recap and Discussion The cost of pressure ulcer formation has been reiterated numerous times. The consequences are severe in regard to the patients health as well as the cost of medical care. Not only have pressure ulcer incidences been used as a meter for quality of care, b ut they can be considered a medical error. Additionally, Medicare no longer reimburses the cost of hospital acquired pressure ulcers. There are many conditions under which pressure ulcers can develop, but it is known through years of clinical experience and observation, and confirmed in animal studies, that pressure over time will cause tissue damage. However, due to the complex nature of the tissue injury, there is no absolute threshold that will or will not cause pressure ulcer formation in patients. Many protocols and devices have been developed to help identify those patients who are at greatest risk and to alleviate high pressures over time, but despite their practice, pressure ulcers still develop at a high rate. Since there is still a high prevalence of pressure ulcers despite adherence to protocols that are implemented to help prevent them, this standard of care was investigated. The investigation focused on the sacrum and surrounding areas because this is the site of most frequent occurrence. The fou r specific aims of this research, as described in this dissertation, consisted of identifying clinically relevant biomechanical factors that are likely to lead to pressure ulcer formation, investigating the use of bioimpedence measurements as a method to i dentify pressure ulcer formation, and studying patients at risk for pressure ulcer formation. The final aim was to determine what would be best for patients, in terms of preventing pressure ulcers, from the findings of the previous aims.

PAGE 111

111 The first healthy subject study addressing Aim 1 investigated the effects of raising the HOB on interface pressures in the sacral area. Results clearly demonstrated that elevating the HOB significantly increased interface pressures near the sacrum and buttock areas. Additio nally, the amount of skin area subjected to interface pressures greater than 32 mm Hg increased with HOB elevation. These areas were termed at risk areas because they were subjected to pressures that are believed to potentially lead to pressure ulcer for mation. The practical importance of this finding is in how it relates to the clinical care of the patient. For mechanically ventilated patients, for example, protocols recommend elevating the HOB to decrease the risk of VAP and aspiration. However, this po ses a clinical dilemma as raising the HOB will decrease the risk of one condition (VAP) but will increase the risk of another (pressure ulcer formation). For the second healthy subject study for Aim 1, the lateral turning protocol performed in q2h turning was investigated as to how the interface pressures and resulting at risk areas would be affected. The results showed, again, that increasing the HOB increased the interface pressures and at risk areas, this time for subjects turned to either their left or right. Considerably more profound than those results, it was observed that some of the same skin areas at risk in one position were actually at risk in all three positions: supine, left, and right. This tissue area was termed a triple jeopardy area becau se of it being at risk in the three aforementioned positions. Observing triple jeopardy areas means that skin areas are always at risk regardless of position and demonstrates that the tissue does not get unloaded during the turning process. Several bioimpe dance methods were explored for Aim 2 as to whether or not they could be exploited to detect the formation of pressure ulcers. Previous research had shown that a localized tetrapolar impedance measurement correlated well with the subcutaneous fat layer thi ckness in the abdomen (Scharfetter et al., 2001). Other research suggested that there are different etiologies

PAGE 112

112 for pressure ulcer formation. Prolonged pressure lead to edema development, which was believed to be a precursor to pressure ulcer formation, and could be seen with HRUS before other clinical signs were apparent (Quintavalle et al., 2006). A combination of these studies was envisioned to help detect pressure ulcer formation by using a tetrapolar impedance measurement to detect edema beneath the ski n. However, basic computer modeling results, along with counsel from an expert in the field of EIT, suggested that the signals would be too minute when attempting to detect edema and that the desired results would not be possible. Further study continued w ith two commercially available impedance devices, one of which was claimed to have some success in predicting the onset of stage I pressure ulcers (Bates Jensen et al., 2007). However, after evaluating the devices and running some skin measurement experi ments, it was not clear what the measurements actually represented and neither device seemed appropriate for use in the ICU environment. Finally, a basic electrode experiment was conducted and it was determined that a neither a basic or a tetrapolar electr ode arrangement could be used to determine interface pressures on the skin with a bioimpedance measurement. Unfortunately, none of the attempted bioimpedance methods were successful in detecting changes or damage in the skin and could not be used as a meth od for pressure ulcer detection. Aim 3 involved studying actual patients at risk for the formation of pressure ulcers. The study was strictly observational but continuously measured the interface pressure profiles of patients over the course of about 6 hours to monitor how interface pressures changed throughout their hospital stay. Specific areas of skin that were at risk for the entire monitoring period were called always at risk areas. The results demonstrated that patients were subjected to higher inte rface pressures, larger at risk areas, and larger triple jeopardy and always at -risk areas than were observed for healthy subjects. This confirms the notion that at risk patients are at even

PAGE 113

113 greater risk for pressure ulcer formation than healthy subjects. Since triple jeopardy and always at risk areas were seen in actual patients, this crucial observation could provide a key insight to the understanding of hospital acquired pressure ulcers. Despite following standard turning protocols, it is evident that tu rning, in the manner that it is currently performed, is not adequate at unloading and redistributing interface pressures of at -risk tissue. Solutions What needs to be done to ensure patients are protected from the very beds in which they lie? The concise answer is that triple jeopardy areas and always at risk areas need to be reduced as much as possible, or preferably, eliminated altogether. Furthermore, high interface pressures need to be eliminated. When it comes to turning, primarily, it is when the pati ent rolls back onto the supporting device (pillow or wedge) that areas from the previous position continue to be loaded, or the turn itself was not sufficient in unloading said areas. When it comes to interface pressures, tissue can withstand smaller magni tudes for longer periods of time before tissue damage occurs than for larger magnitudes. Significantly reducing always at -risk areas and large interface pressures would undoubtedly reduce pressure ulcer incidence. The following methods should be implemente d, based on the data collected in this dissertation, to improve patient safety and hopefully lower the incidence of pressure ulcers. The following methods will aim to achieve one or both of the needs for patient safety: reducing or eliminating triple jeopa rdy and always at risk areas and/or reducing or eliminating high interface pressures. While potential solutions may not be appropriate for all at -risk patients, these methods aim to improve on the current standard of care. Improved T urning P rotocols The re sults from healthy subjects in Chapter 3, and from at risk patients in Chapter 5, clearly illustrate that triple jeopardy areas exist, and naturally, these areas need to be eliminated

PAGE 114

114 to achieve proper tissue unloading. For those patients demonstrating tri ple jeopardy areas, the specific areas were found primarily on the buttocks, in particular, over the bony prominences of the sacrum and the ischial tuberosities, and to a lesser extent over the trochanters. One way to reduce or eliminate triple jeopardy ar eas is to improve upon the current turning protocol. The practice of q2h turning is not without benefit, but its origin, as mentioned in Chapter 1, is somewhat arbitrary. Considering the prevalence of pressure ulcers despite its implementation, it is not u nreasonable to try to improve upon this practice. One potential solution is to completely unload the sacrum and opposite buttock. For this to occur, supporting devices likely need to be placed at the lower back and at the upper thigh to provide enough support to maintain the turn. It should be verified during the process that the sacral area is clear of any supporting device. For this to be achieved, the patient will likely have to be turned at a more -extreme lateral angle than the current practice, which may not be appropriate for some patients, and may introduce even higher interface pressures near the trochanteric region. Essentially a new turning protocol could be investigated in which instructions for placement of the support device are explicitly descr ibed so that the sacrum remains unloaded throughout the lateral positions. Improved Support Devices Another way to reduce triple jeopardy areas is to develop a better pillow, wedge, or other supporting device that will do a better job of pressure relief while turning. The midline and opposite buttock (ischial tuberosity) areas should not become re -loaded when turning from one side to another, but this is exactly what happens when patients roll back onto the support devices. Basically, the concept would be t o create a support device for the specific application of patient turning, rather than to have caregivers use whatever materials are available. It was shown in Chapter 3 that even though the use of foam wedges did not result in larger interface pressures i n

PAGE 115

115 comparison to pillows, the at risk areas increased significantly when using wedges. More importantly, the triple jeopardy areas for the same population of subjects increased more than three -fold, 153 99 cm2 compared to 48 47 cm2, using wedges rather than pillows. These results clearly demonstrate that the support device can either play a major role in prevention or cause of pressure ulcer formation. To improve upon the current method, a support device should be created that has a strategically located cut -out or indentation targeted at reducing the triple jeopardy area near the sacrum, while still providing enough support to maintain a lateral position. Additionally, a more ergonomically shaped device has the potential to aid in pressure relief. A ne w turning support would provide a novel device to the current turning practice that arbitrarily stuffs pillows under a patient to maintain a turned position. Not only would the cut -out aid in pressure relief of the target at risk area, but it would provide a landmark on the support device itself, which would result in more -consistent placement of support devices and implementation of turning protocols by caregivers. In the patient study, there appeared to be a general protocol for turning, but variations we re observed as evidenced by the turning sequences (positions) shown in Table 5 1. Additionally, if a separate device was needed for turning to the left than to the right (turn specific), the potential error of turning a patient to the same side they were m ost recently could be avoided. Providing Feedback to Caregivers Providing feedback to caregivers about how their turning affects their patients can only be of benefit. One method to provide this would be to use an indicator on the at risk skin areas. Wheth er or not a thin film can or should be applied to the skin of an at -risk patient is up for

PAGE 116

116 debate. However, if a low magnitude pressure indicating film1 could be placed on a patients at risk area, i.e. centered across the sacrum between turns, the nursing staff would know whether or not their turn was successful in unloading the side as intended. Although the results would be obtained after the most recent turn, the information could be used to improve care and the ongoing turning practice for the remainde r of the patients stay. This film could be used on its own as a feedback mechanism about the current turning procedure as well as a pressure indicator. It could also be used to supplement either of the previous suggestions, improved turning protocols and/ or improved support devices, to help evaluate their success. Interface Pressure Monitoring Interface pressure monitoring with an interface pressure mapping system is a feedback mechanism in its own right, but multiple aspects of its implementation will be discussed. An interface pressure mapping system can be used track pressure magnitudes as well as areas subject to various pressure thresholds. Variations and adjustments to the system used for this research will be proposed. Pressure magnitudes Despite the fact that there is no absolute pressure threshold that leads to pressure ulcer formation, it is widely accepted that minimizing high interface pressures is of utmost importance in regard to at risk patients. Currently, patients interface pressures are no t regularly monitored. Providing constant monitoring would allow caregivers the opportunity to avoid circumstances when interface pressures are higher than they should be, which can vary from patient to patient. A great example of this was observed during the ICU patient study. When the pressure sensor was positioned beneath a patient prior to beginning data collection for the monitoring study, the 1 Similar products are commercially available from Tekscan and Sensor Products Inc. that can detect interface pressures from 10 0 to 2.25e6 mm Hg.

PAGE 117

117 protocol was to ensure correct patient positioning on the sensor and to be certain the pad was not improperly creased or folded. For one patient, it was obvious from looking at the display that there was a hot spot (a location of high interface pressure) that did not appear to fit the normal pressure profile for the patient. Therefore, the cause of the hot spot was investigated to see what was causing it or to see if something was wrong with the sensor. The cause of the hot spot was a syringe cap that had been used during the patients care and the patient was actually lying on it. Finding this cap likely prevent ed tissue damage and would not have been found without the interface pressure monitoring system. This may be an extreme case, but without even trying to, a pressure ulcer was likely avoided. Another example of how unnecessarily high interface pressures can be avoided with monitoring comes from a more general observation from the ICU patient study. Every time a patient gets repositioned (turned), their interface pressure profile changes as a whole. Ideally, areas that were loaded get unloaded and the patient s weight gets shifted to areas that have not recently been loaded. However, even if a patient was lying completely on their left side and then turned to the right, caregivers can only know that they unloaded a specific tissue area but not the magnitude of interface pressure that is experienced. It was shown in Figure 5 1 that there was no significant difference between the maximal interface pressures of at -risk patients between the supine, left, and right positions. There was also no significant difference between positions when the mean of the peak interface pressures was taken over each positions specific monitoring time. Since the HOB is known to influence interface pressures, it was also investigated whether or not any differences could have affected t he results. For the total population, the average HOB position and range of elevations were similar for all three positions. These results verify that, as a whole, just because a patient gets

PAGE 118

118 turned from a supine position to a lateral position (or vice ver sa), or from one side to the other, the interface pressures do not need to increase. Both sides of this explanation can be seen for the patient observed in Figure 6 1. This patient initially experienced mean peak interface pressures of about 80 mm Hg while turned to the right. Upon getting turned to the left, the peak interface pressures immediately shot up and the mean peak pressure remained at about 150 mm Hg for the remainder of this position. Subsequent turns to the right and left, however, had mean pea k pressures that were closer to 85 mm Hg and 100 mm Hg respectively. This demonstrates, for the same patient, that turning to either side could easily have had lower mean peak interface pressures than the 150 mm Hg that was experienced during the first obs erved turn to the left. If interface pressure monitoring were utilized, then this patient could have been repositioned on the left side in a way that the high interface pressures (150 mm Hg compared to 100 mm Hg) would not have been experienced. For the total population, there were 10 patients that had dramatic, sustained changes in peak interface pressures (higher or lower) due solely to repositioning. The magnitude of these pressure changes were typically around 4050 mm Hg, but ranged from about 20120 m m Hg. This is a tremendous amount when considering that the mean of peak interface pressures is around 100 mm Hg for any of the positions, and a 50 mm Hg increase is a 50% increase in peak interface pressure over the duration of the turned position. Areas over time Discovering the syringe cap and preventing unnecessarily high interface pressures are just two examples that demonstrate how monitoring interface pressure magnitudes can improve patient safety. However, interface pressure monitoring has the potential to provide additional information regarding skin areas under pressure. Utilizing pressure mapping over time allows for specific areas to be monitored. It is known that as the applied pressure to tissue increases, the amount of time until tissue damage occurs decreases. Monitoring can track the amount of

PAGE 119

119 pressure specific locations of skin experience throughout a patients stay and can alert caregivers when tissue needs to be unloaded. Figure 6 1. Peak interface pressures over time for one patient. Th is patient was initially turned to the right, turned left at frame 145, turned right again at frame 403, and turned left again at frame 664. As previously mentioned, there is no absolute value that is known to prevent or cause tissue damage over time as th ere are too many confounding factors to consider. However, to err on the side of patients safety, an appropriate starting point would be to use the 32 mm Hg interface pressure threshold and the q2h turning time duration. This starting point would essentia lly allow tissue to experience 32 mm Hg over a two hour time period. Once a skin area exceeded this pressuretime threshold, caregivers could be alerted to the area at risk. The pressure time relationship may not be linear, but it would be an appropriate s tarting point. By tracking tissue over time, it would be known which tissue areas are likely at a greater risk for pressure ulcer formation by knowing which areas that have experienced the most pressure. It may be, depending on the size and weight of the p atient, that keeping specific tissue areas at or

PAGE 120

120 below the 32 mm Hg threshold for 2 hours is not possible, but if for nothing else, the at risk areas will be identified and care can be taken to monitor them. Since all patients demonstrated always at risk a reas > 95% of their monitoring duration, the easy answer is to say that the patients should have been turned more frequently, especially for those patients who were not turned at all or not frequently enough. However, if the at risk area is sufficiently sm all, slight positional adjustments may be all that is necessary to keep at risk tissue below the pressure time threshold. Additionally, since triple jeopardy patients experienced always at risk areas, more frequent turning (in the same capacity that was pe rformed) would not have solved the problem since the same area was at risk regardless of position. Therefore, even though maintaining 32 mm Hg or less over time would be desired, maybe a more realistic starting point would be to use the low end of the ranges of the mean of peak interface pressure data, which was just below 60 mm Hg for any of the positions. System adjustments For interface pressure monitoring to become a realistic application, a few alterations would have to be made to the pressure mapping system used for the research described in this dissertation. To provide feedback to caregivers, alarms or other alerts could be customized to indicate when a patient is subjected to high interface pressures or when abnormalities in the interface pressure p rofiles are experienced. A potential improvement to the device itself would be to use a larger pressure mapping sensor array so that the patients entire weight bearing tissue can be monitored as well as providing some room for error with patient movement. Mattress sized sensor arrays are commercially available. However, practical challenges would present themselves, if for example the sensor was built into the bed or as a bed sheet, because then, upon patient turning or repositioning, the sensor would not be between the patient and the support device.

PAGE 121

121 A lthough a larger sensor array, or using smaller -sized sensors over the same area, would improve the resolution of the instrumentation, for economical and computational reasons, the triple jeopardy and always at risk area results were re analyzed to determine whether or not similar results could be achieved after decreasing the resolution of the array or the size of the array since many sensors were not loaded. The resolution was reduced by averaging adjacent s ensors using block processing procedures in Matlab. The interface pressure profiles were divided into m -by-n sections, or blocks, and the average interface pressure for each section was determined. The original 48x48 resolution was reduced to 48x24, 24x48, 24x24, 16x16, and 12x12. Examples of how the stacked triple jeopardy images changed with a decrease in resolution can be seen in Figure 6 2. The actual triple jeopardy and always at risk area results can be seen in Table 6 1. Though not quite as precise, it appears feasible that increasing the sensor size from half -inch square sensors (48x48) to one -inch square sensors (24x24) would still allow for reasonable triple jeopardy area detection while reducing the numbers of sensors by a factor of 4. As for the dimensions of the sensor array itself, based on observations of the relative size of the sacral area for the larger patients, the height of the sensor array could be reduced by 1/4, or 6 inches, and still achieve the same results as long as the patients we re positioned appropriately. Unless the sensor was affixed to the patient, the width of the array cannot readily be reduced due to the nature of the lateral turning procedure. By implementing these modifications, the 24-inch x 24 inch ( 48 x 48 half -inch se nsors ) array could be reduced to an 18 inch x 24inch (1 8 x 24 one -inch sensors ) array and still achieve relatively similar results for the sacral area with a reduction in sensors from 2,304 to 432.

PAGE 122

122 Figure 6 2. Stacked triple jeopardy images demonstrat ing changes in resolution for one patient. The color bar indicates the amount of time (in hours) specific skin areas were at risk. The resolutions are 48x48 (top left), 48x24 (bottom left), 24x48 (top center), 24x24 (bottom center), 16x16 (top right), and 12x12 (bottom right).

PAGE 123

123 Table 6 1. Triple jeopardy and always at -risk area results for various resolutions. The triple jeopardy population is in blue. Always At Risk Always At Risk Area > 95% of monitoring time [cm 2 ] Patient Area [cm 2 ] Number of Sensors in Sensor Array (row x column) absolute 48x48 24x48 48x24 24x24 16x16 12x12 1 240.32 296.77 229.03 254.84 212.9 174.19 206.45 2 0.00 4.84 0 0 0 0 0 3 350.00 469.35 438.71 425.81 400 337.42 335.48 4 0.00 8.06 6.45 0 0 0 0 5 8.06 72.58 67.74 58.06 58 .06 29.03 25.81 6 16.13 108.06 54.84 58.06 32.26 14.52 0 7 67.74 119.35 90.32 90.32 70.97 43.55 25.81 8 130.64 169.35 138.71 138.71 109.68 116.13 103.23 9 1.61 261.29 229.03 248.39 225.81 174.19 154.84 10 220.97 246.77 209.68 180.64 161.29 87.1 77.42 11 4.84 109.68 70.97 74.19 51.61 29.03 25.81 12 166.13 456.45 380.64 370.97 335.48 333.87 258.06 13 561.29 612.90 558.06 545.16 509.68 450 490.32 14 417.74 516.13 474.19 483.87 458.06 450 335.48 15 361.29 427.42 367.74 335.48 296.77 203.23 180.64 16 67.74 104.84 70.97 80.65 51.61 43.55 25.8 17 40.32 75.81 51.61 41.94 38.71 29.03 51.61 18 3.23 14.52 3.23 6.45 0 0 0 19 135.48 195.16 151.61 138.71 103.23 29.03 51.61 20 200.00 248.39 170.97 193.55 148.39 145.16 103.23 21 0.00 1.61 0 0 0 0 0 22 0.00 14.52 9.68 6.45 0 0 0 23 117.74 193.55 125.81 119.35 90.32 43.55 0 avg 135.27 205.54 169.56 167.46 145.86 118.81 106.59 Another adjustment to the system, and perhaps the primary challenge, would be accounting for patient movement while monitoring to al low for pressure time feedback for specific tissue areas. While many at -risk patients are relatively immobile, any monitoring software would need to account for patient movement (perhaps some form of image recognition software) so that the cumulative press ures and durations would be assigned to the appropriate skin locations. A solution to both challenges could be achieved by placing or attaching a sensor directly to the skin, but then that could introduce added skin care challenges and the possible introduction of, or additional, frictional or shear forces that could be detrimental to the pressure ulcer prevention process.

PAGE 124

124 Concluding Remarks Patient safety can be improved by lowering the incidence of pressure ulcers, which can be achieved by reducing or eli minate always at risk and triple jeopardy areas, and by reducing or eliminating high interface pressures. Improved turning protocols could aid in reducing or eliminating triple jeopardy areas by not allowing patients to roll back onto the side from which t hey were just turned. Developing a support device specifically for turning could aid in reducing triple jeopardy areas and interface pressures and would likely result in a more -consistent implementation of turning protocols. Providing feedback to caregiver s could also help by informing them about the quality of their turning practice and would allow them to adjust their practice from patient to patient when needed. Interface pressure monitoring could alert caregivers when interface pressures get too high, w hen specific skin areas need to be unloaded, or just to provide feedback about how repositioning affects the interface pressures a patient experiences. Interface pressure monitoring is clearly the most powerful of the proposed solutions due to all of the i ssues it can address. However, taking simple steps to improve turning protocols or support devices may prove to be just as effective as they could be readily implemented, and they would require only a slight modification to the current q2h turning practice and would not require additional positional interventions or adjustments to be made by caregivers. All of these solutions have the potential to improve patient safety and all of them have the potential to help lower the incidence of pressure ulcer format ion. Future work should be performed to validate their utility, but all proposed solutions were based on addressing actual clinical observations. Investigating their use can only lead to a better understanding of why and how pressure ulcers form in clinica l and extended -care settings and

PAGE 125

125 addressing them will augment the current state -of the art for the standard care of preventing pressure ulcers.

PAGE 126

126 LIST OF REFERENCES Agostini, J.V., Baker, D.I., Bogardus, S.T., 2000. Prevention of pressure ulcers in older patients. In: Shojania, K.G., Duncan, B.W., McDonald, K.M., Wachter, R.M. (Eds), Making Health Care Safer: A Critical Analysis of Patient Safety Practices. Rockville, Agency for Healthcare Research and Quality, pp. 301 306. Allen, V., Ryan, D.W., Murray, A ., 1993. Potential for bed sores due to high pressures: influence of body sites, body position, and mattress design. British Journal of Clinical Practice 47, 195197. Allen, V., Ryan, D.W., Murray, A., 1994. Measurements of interface pressure between body sites and the surfaces of four specialised air mattresses. British Journal of Clinical Practice 48, 125129. American Thoracic Society, Infectious Diseases Society of America, 2005. Guidelines for the management of adults with hospital acquired, ventilat or associated, and healthcare associated pneumonia. American Journal of Respiratory and Critical Care Medicine 171, 388416. Amlung, S.R., Miller, W.L., Bosley, L.M., 2001. The 1999 National Pressure Ulcer Prevalence Survey: a benchmarking approach. Advanc es in Skin & Wound Care 14, 297301. Bader, D.L., 1990. The recovery characteristics of soft tissues following repeated loading. Journal of Rehabilitation Research and Development 27, 141 150. Barratt, E., 1987. Pressure sores. Putting risk calculators i n their place. Nursing Times 83, 6570. Bates Jensen, B.M., McCreath H E ., Kono, A., Apeles N C., Alessi, C., 2007. Subepidermal moisture predicts erythema and stage 1 pressure ulcers in nursing home residents : a pilot study Journal of the American Ger iatrics Society 55, 11991205. Bennett, R.G., O'Sullivan, J., DeVito, E.M., Remsburg, R., 2000. The increasing medical malpractice risk related to pressure ulcers in the United States. Journal of the American Geriatrics Society 48, 7381. Bergstrom, N., 2 005. Patients at risk for pressure ulcers and evidence based care for pressure ulcer prevention. In: Bader, D., Bouten, C., Colin, D., Oomens, C. (Eds), Pressure Ulcer Research: Current and Future Perspectives. Springer, Berlin, pp. 35 50. Beuret P Cart on M J Nourdine K Kaaki M Tramoni G Ducreux J C. 2002. Prone position as prevention of lung injury in comatose patients: a prospective, randomized, controlled study. Intensive Care Medicine 28, 564569. Bercault, N., Boulain, T ., 2001. Mortal ity rate attributable to ventilator associated nosocomial pneumonia in an adult intensive care unit: a prospective case -control study. Critical Care Medicine 29, 23032309.

PAGE 127

127 Bouten, C.V., Oomens, C.W., Baaijens, F.P., Bader, D.L., 2003. The etiology of pres sure ulcers: skin deep or muscle bound? Archives of Physical Medicine and Rehabilitation 84, 616 619. Braden, B.J., Bryant, R., 1990. Innovations to prevent and treat pressure ulcers. Geriatric Nursing 11, 182186. Chastre, J., Fagon, J.Y., 2002, Ventilato r associated pneumonia. American Journal of Respiratory Critical Care Medicine 165, 867903. Clark, M., Price, P.E., 2004. Is wound healing a true science or a clinical art? Lancet 364, 13881389. Colin, D., Abraham, P., Preault, L., Bregeon, C., Saumet, J.L., 1996. Comparison of 90 degrees and 30 degrees laterally inclined positions in the prevention of pressure ulcers using transcutaneous oxygen and carbon dioxide pressures. Advances in Wound Care 9, 35 38. Cook, D.J., Walter, S.D., Cook, R.J., Griffith L.E., Guyatt, G.H., Leasa, D., Jaeschke, R.Z., Brun Buisson, C., 1998. Incidence of and risk factors for ventilator associated pneumonia in critically ill patients. Annals of Internal Medicine 129, 433 440. Cornish, B., 2006. Bioimpedance analysis: scientific background. Lymphatic Research and Biology 4, 4750. Cornish, B.H., Thomas, B.J., Ward, L.C., 1993. Improved prediction of extracellular and total body water using impedance loci generated by multiple frequency bioelectrical impedance analysis. Phys ics in Medicine and Biology 38, 337346. Davis, K.J., Johannigman, J.A., Campbell, R.S., Marraccini, A., Luchette, F.A., Frame, S.B., Branson, R.D., 2001. The acute effects of body position strategies and respiratory therapy in paralyzed patients with acu te lung injury. Critical Care 5, 81 87. Dealey, C., 1991. The size of the pressure -sore problem in a teaching hospital. Journal of Advanced Nursing 16, 663670. Dealey, C., 1995. Mattresses and beds. A guide to systems available for relieving and reducin g pressure. Journal of Wound Care 4, 409412. Defloor, T., 2000. The effect of position and mattress on interface pressure. Applied Nursing Research 13, 2 11. Defloor, T., De Bacquer, D., Grypdonck, M.H., 2005. The effect of various combinations of turni ng and pressure reducing devices on the incidence of pressure ulcers. International Journal of Nursing Studies 42, 3746. Dini, V., Bertone, M,, Romanelli, M., 2006. Prevention and management of pressure ulcers. Dermatologic Therapy 19, 356 364.

PAGE 128

128 Dinsdale, S.M., 1974. Decubitus ulcers: role of pressure and friction in causation. Archives of P hysical M edicine and R ehabilitation 55, 147152. Dodek P Keenan S Cook D Heyland D Jacka M Hand L Muscedere J Foster D Mehta N Hall R. Brun Buisson C ., 2004. Evidence based clinical practice guideline for the prevention of ventilator associated pneumonia Annals of Internal Medicine 141, 305313. Drakulovic, M.B., Torres, A., Bauer, T.T., Nicolas, J.M., Nogue, S., Ferrer, M., 1999. Supine bo dy position as a risk factor for nosocomial pneumonia in mechanically ventilated patients: a randomised trial. Lancet 354, 18511858. Ferguson -Pell M.W. Bell, F., Evans, J.H., 1976. Interface pressure sensors: existing devices, their suitability and limitations. In: Kenedi, R.M., Cowden, J.M., Scales, J.T. (Eds), Bedsore Biomechanics. Macmillan, London pp. 189197. Gebhardt, K.S., 2004. Pressure ulcer research: where do we go from here? British Journal of Nursing 13, S14 S18. Gebhardt, K.S., 2005. Resea rch in biomedical engineering: an overview of recent literature. Journal of Tissue Viability 15, 17 18. Goldstein, B., Sanders, J., 1998. Skin response to repetitive mechanical stress: a new experimental model in pig. Archives of Physical Medicine and Reha bilitation 79, 265272. Grimnes, S., Martinsen, O.G., 2006. Bioimpedance. In: Akay, M. (Ed), Encyclopedia of Biomedical Engineering. John Wiley, Hoboken, pp. 438447. Guerin C Gaillard S Lemasson S Ayzac L Girard R. Beuret P Palmier B. L e Q V Sirodot M Rosselli S Cadiergue V Sainty J M Barbe P Combourieu E Debatty D Rouffineau J Ezingeard E Millet O Guelon D Rodriguez L Martin O Renault A Sibille J P Kaidomar M ., 2004. Effects of systematic prone positioning in hypoxemic acute respiratory failure: a randomized controlled trial. Journal of the American Medical Association 292, 23792387. Guyton, A.C., Hall, J.E., 2000. Textbook of Medical Physiology, 10th ed. Philadelphia, WB Saunders Company pp. 163174. Haalboom, J., 2005. Medical perspectives in the 21st century. In: Bader, D., Bouten, C., Colin, D., Oomens, C. (Eds), Pressure Ulcer Research: Current and Future Perspectives. Springer, Berlin, pp. 1121. Harada, C., Shigematsu, T., Hagisawa, S., 2002. The effect of 10-degree leg elevation and 30 degree head elevation on body displacement and sacral interface pressures over a 2 hour period. Journal of Wound, Ostomy, and Continence and Nursing 29, 143148. Hagisawa, S., Ferguson Pell, M., 200 8. Evidence supporting the use of two-hourly turning for pressure ulcer prevention. Journal of Tissue Viability 17, 76 81.

PAGE 129

129 Heyland, D.K., Cook, D.J., Griffith, L., Keenan, S.P., Brun Buisson, C., 1999. The attributable morbidity and mortality of ventilator associated pneumonia in the critically ill patient. the Canadian critical trials group. American Journal of Respiratory and Critical Care 159, 12491256. Hobbs, B.K., 2004. Reducing the incidence of pressure ulcers: implementation of a turnteam nursing program. Journal of Gerontological Nursing 30, 46 51. Hoffer, E.C., Meador, C.K., Simpson, D.C., 1969. Correlation of whole -body impedance with total body water volume. Journal of Applied Physiology 27, 531534. Hofman, A., Geelkerken, R.H., Wille, J., H amming, J.J., Hermans, J., Breslau, P.J., 1994. Pressure sores and pressure decreasing mattresses: controlled clinical trial. Lancet 343, 568571. Kaysen, G.A., Zhu, F., Sarkar, S., Heymsfield, S.B., Wong, J., Kaitwatcharachai, C., Kuhlmann MK, Levin NW, 2005. Estimation of total -body and limb muscle mass in hemodialysis patients by using multifrequency bioimpedance spectroscopy. American Journal of Clinical Nutrition 82, 988 995. Kosiak, M., 1961. Etiology of decubitus ulcers. Archives of Physical Medici ne and Rehabilitation 42, 19 29. Kotler, D.P., Burastero, S., Wang, J., Pierson, R.N. J., 1996. Prediction of body cell mass, fat free mass, and total body water with bioelectrical impedance analysis: effects of race, sex, and disease. American Journal of Clinical Nutrition 64 (3 Suppl), 489S 497S. Kyle, U.G., Bosaeus, I., De Lorenzo, A.D., Deurenberg, P., Elia, M., Gomez, J.M., Heitmann, B.L., Kent Smith, L., Melchior, J.C., Pirlich, M., Scharfetter, H., Schols, A.M., Pichard, C., 2004. Bioelectrical impe dance analysis --part I: review of principles and methods. Clinical Nutrition 23, 12261243. Landis, E., 1930. Micro-injection studies of capillary blood pressure in human skin. Heart 15, 209228. Lapsley, H.M., Vogels, R., 1996. Cost and prevention of pre ssure ulcers in an acute teaching hospital. International Journal of Quality in Health Care 8, 61 66. Lukaski, H.C., Johnson, P.E., Bolonchuk, W.W., Lykken, G.I., 1985. Assessment of fat -free mass using bioelectrical impedance measurements of the human bo dy. American Journal of Clinical Nursing 41, 810817. Lyder, C.H., 2003. Pressure ulcer prevention and management. Journal of the American Medical Association 289, 223226. Lyder, C., 2005. Medicolegal implications. In: Bader, D., Bouten, C., Colin, D., Oomens, C. (Eds), Pressure Ulcer Research: Current and Future Perspectives. Springer, Berlin, pp. 2334.

PAGE 130

130 Maklebust, J.A., Mondoux, L., Sieggreen, M., 1986. Pressure relief characteristics of various support surfaces used in prevention and treatment of pres sure ulcers. Journal of Enterostomal Therapy 13, 85 89. Medicare Program; Changes to the Hospital Inpatient Prospective Payment Systems and Fiscal Year 2008 Rates; Final Rule: 72 Fed Reg 473794742872 (22 August 2007) (to be codified at 42 CFR Parts 411, 412, 413, and 489). Milne, J., Pagnamenta, F., 2004. An online ordering system for therapy beds and mattresses. British Journal of Nursing 13, S38-S42. Mohapatra, S.N., Hill, D.W., 1975. The changes in blood resistivity with haematocrit and temperature. E uropean Journal of Intensive Care Medicine 1, 153162. National Pressure Ulcer Advisory Panel, 2007. Updated pressure ulcer staging. Available at: http://www.npuap.org/pr2.htm. Accessibility verified May 7, 2007. Nyboer, J., 1959. Electrical Impedance Plet hysmography. Thomas, Springfield, IL. Orozco -Levi, M., Torres, A., Ferrer, M., Piera, C., el -Ebiary, M., de la Bellacasa, J.P., Rodriguez Roisin, R., 1995. Semirecumbent position protects from pulmonary aspiration but not completely from gastroesophageal r eflux in mechanically ventilated patients. American Journal of Respiratory and Critical Care 152, 13871390. Panel for the Prediction and Prevention of Pressure Ulcers in Adults, 1992. Pressure ulcers in adults: prediction and prevention. Clinical practic e guideline number 3. Agency for Health Care Policy and Research, Public Health Service, US Department of Health and Human Services. AHCPR Publication No. 92 0047, Rockville. Papazian, L., Bregeon, F., Thirion, X., Gregoire, R., Saux, P., Denis, J.P., Peri n, G., Charrel, J., Dumon, J.F., Affray, J.P., Gouin, F., 1996. Effect of ventilator associated pneumonia on mortality and morbidity. American Journal of Respiratory and Critical Care 154, 9197. Pavselj N Pr e at V Miklavcic D. 2007. A numerical mod el of skin electropermeabilization based on in vivo experiments. Annals of Biomed ical Eng ineering 35, 2138-2144. Pender, L.R., Frazier, S.K., 2005. The relationship between dermal pressure ulcers, oxygenation and perfusion in mechanically ventilated patients. Intensive and Critical Care Nursing 21, 2938. Quintavalle, P.R., Lyder, C.H., Mertz, P.J., Phillips Jones, C., Dyson, M., 2006. Use of highresolution, high -frequency diagnostic ultrasound to investigate the pathogenesis of pressure ulcer development Advances in Skin & Wound Care 19, 498505. Rajendran P.J., Leachtenauer, J., Kell, S., Turner, B., Newcomer, C., Lyder, C., Alwan, M., 2006. Improving the detection of stage I pressure ulcers by enhancing digital color images. Engineering in Medicine and Biology Society, EMBS '06. 28th Annual International Conference of the IEEE, 52065209.

PAGE 131

131 Reddy, M., Gill, S.S., Rochon, P.A., 2006. Preventing pressure ulcers: a systematic review. Journal of the American Medical Association 296, 974984. Reddy, N.P., 1990. Effects of mechanical stresses on lymph and interstitial fluid flows. In: Bader, D.L. (Ed.), Pressure Sores: Clinical Practice and Scientific Approach. Macmillan, London, pp. 203220. Reignier J. Thenoz Jost, N., Fiancette, M., Legendre, E., Lebert, C., Bontemps, F., Clementi, E., Martin Lefevre, L., 2004. Early enteral nutrition in mechanically ventilated patients in the prone position. Critical Care Medicine 32, 9499. Rello, J., Ausina, V., Ricart, M., Castella, J., Prats, G., 1993. Impact of previous antimicrobial therapy on the etiology and outcome of ventilator associated pneumonia. Chest 104, 12301235. Rello, J., Ollendorf, D.A., Oster, G., Vera -Llonch, M., Bellm, L., Redman, R., Kollef M H 2002. Epidemiology and outcomes of ventilator associat ed pneumonia in a large US database. Chest 122, 21152121. Reswick, J.B., Rogers, J.E., 1976. Experience at Rancho Los Amigos Hospital with devices and techniques to prevent pressure sores. In: Kenedi, R.M., Cowden, M., Scales, J.T. (Eds), Bedsore Biomech anics. Macmillan Press, London, p 301310. Rithalia, S.V., Gonsalkorale, M., 1998. Assessment of alternating air mattresses using a time based interface pressure threshold technique. Journal of Rehabilitation Research and Development 35, 225230. Rithalia S.V., Gonsalkorale, M., 2000. Quantification of pressure relief using interface pressure and tissue perfusion in alternating pressure air mattresses. Archives of Physical Medicine and Rehabilitation 81, 13641369. Scharfetter, H., Brunner, P., Mayer, M., Brandstatter, B., Hinghofer -Szalkay, H., 2005. Fat and hydration monitoring by abdominal bioimpedance analysis: data interpretation by hierarchical electrical modeling. IEEE Transactions on Biomedical Engineering 52, 975982. Scharfetter, H., Schlager, T., Stollberger, R., Felsberger, R., Hutten, H., Hinghofer -Szalkay, H., 2001. Assessing abdominal fatness with local bioimpedance analysis: basics and experimental findings. International Journal of Obesity and Related Metabolic Disorders 25, 502511. Sha nds Hospital, University of Florida. Decubitus ulcer events (All Shands Healthcare Facilities). 2005. Shands Hospital, University of Florida. Decubitus ulcer events (All Shands Healthcare Facilities). 2006.

PAGE 132

132 Sideranko, S., Quinn, A., Burns, K., Froman, R.D., 1992. Effects of position and mattress overlay on sacral and heel pressures in a clinical population. Research in Nursing and Health 15, 245251. Stekelenburg, A., Oomens, C., Bader, D., 2005. Compressioninduced tissue damage: animal models. In: Bader, D., Bouten, C., Colin, D., Oomens, C. (Eds), Pressure Ulcer Research: Current and Future Perspectives. Springer, Berlin, pp. 187204. Still, J.M., Wilson, J., Rinker, C., Law, E., Craft -Coffman, B., 2003. A retrospective study to determine the incidence o f pressure ulcers in burn patients using an alternating pressure mattress. Burns 29, 505507. Stimler, C., Pryor, V., 1999. Impact Results of a Statewide Effort to Improve Pressure Ulcer Prevention in New York State. IPRO HCQIP Publication No. 99 05. Stin son M D Porter -Armstrong A P. 2008. Seating and pressure support needs of people with cancer in the cervix or rectum: a case series on the clinical usefulness of pressure mapping assessment. European J ournal of Cancer Care 17, 298305. Swain, I., 2005. The measurement of interface pressure. In: Bader, D., Bouten, C., Colin, D., Oomens, C. (Eds), Pressure Ulcer Research: Current and Future Perspectives. Springer, Berlin, pp. 5171. Torres, A., Serra Batlles, J., Ros, E., Piera, C., Puig de la Bellacasa J., Cobos, A., Lomena, F., Rodriguez Roisin, R., 1992. Pulmonary aspiration of gastric contents in patients receiving mechanical ventilation: the effect of body position. Annals of Internal Medicine 116, 540543. Vanderwee, K., Grypdonck, M.H., De Bacqu er, D., Defloor, T., 2007. Effectiveness of turning with unequal time intervals on the incidence of pressure ulcer lesions. Journal of Advanced Nursing 57, 5968. Wang, Y.N., Sanders, J., 2005. Skin model studies. In: Bader, D., Bouten, C., Colin, D., Oom ens, C. (Eds), Pressure Ulcer Research: Current and Future Perspectives. Springer, Berlin, pp. 263285. Webster, J.G., 1998. Medical Instrumentation: Application and Design, 3rd ed. New York, John Wiley and Sons, Inc., pp. 196200. Weststrate, J.T.M., Brui ning, H.A., 1996. Pressure sores in an intensive care unit and related variables: a descriptive study. Intensive Critical Care Nursing 12, 280284. Whittemore, R., Bautista, C., Smith, C., Bruttomesso, K., 1993. Interface pressure measurements of support surfaces with subjects in the supine and 45 degree fowler positions. Journal of ET Nursing 20, 111115. Wound Ostomy and Continence Nurses Society, 2003. Guideline for Prevention and Management of Pressure Ulcers. WOCN Clinical Practice Guideline Series.

PAGE 133

133 BIOGRAPHICAL SKETCH Matthew James Peterson was born in 1981 and was raised in Decatur, IL. His father is a mechanical engineer and his mother is a registered nurse. Matthew graduated magna cum laude from Saint Louis University in 2004 with a Bachelor of S cience in Biomedical Engineering. Later that year, he enrolled in the Ph.D. program of the J. Crayton Pruitt Family Department of Biomedical Engineering at the University of Florida and was named a University of Florida Alumni Fellow. In December 2007, Mat thew earned his Master of Engineering degree from the J. Crayton Pruitt Family Department of Biomedical Engineering at the University of Florida. Under the guidance of Johannes H. van Oostrom, Matthew graduated with his Doctorate of Philosophy in Biomedica l Engineering in 2009. He has produced one book chapter, submitted three journal publications, participated in an invention disclosure, and helped develop 2 IRB medical research protocols of which he was a sub investigator. Matthews research interests aim to bridge the gap between physicians and engineers to solve clinically relevant problems. Matthew is engaged to Upohar Haroon, a Ph.D. candidate in Political Science at the University of Florida.