<%BANNER%>

UFIR



Template for Research Protocols in Advanced Epidemiology
https://sites.google.com/site/pharmayad ( Publisher's URL )
CITATION PDF VIEWER
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/IR00001472/00001
 Material Information
Title: Template for Research Protocols in Advanced Epidemiology
Series Title: Advanced Epidemiology, PBHL P601, Spring 2013, Richard M. Fairbanks School of Public Health, Indiana University
Physical Description: Course Material
Creator: Ayad K. Ali
Publisher: Ayad Ali (PharmAyad)
Place of Publication: USA
Publication Date: January 19, 2013
 Notes
Acquisition: Collected for University of Florida's Institutional Repository by the UFIR Self-Submittal tool. Submitted by Ayad K Ali.
Publication Status: Published
General Note: This template does not replace existing standardized templates, rather it serves as an introductory guide for students or junior faculty members who are involved in learning or teaching research methodologies in pharmacoepidemiology to facilitate the process of teaching and formulating research protocols and potentially graduate projects, such as theses and dissertations. The author recommends seeking further references for detailed information pertinent to relevant sections in this template. Additionally, the author welcomes any suggestions to improve the content of this document for future versions. The author encourages proper citation of this document when designing research protocols by including the following suggested wording in the text: “this research proposal was structured according to the template suggested by Dr. Ali, 2013” and including the suggested citation in the reference list. Suggested Citation: Ali AK. Template for Research Protocols in Advanced Epidemiology. January 19, 2013.
 Record Information
Source Institution: University of Florida Institutional Repository
Holding Location: University of Florida
Rights Management: Applicable rights reserved.
System ID: IR00001472:00001

Downloads

This item is only available as the following downloads:

Ayad's Template for Research Protocols_2013 Version ( PDF )


Full Text

PAGE 1

Richard M. Fairbanks School of P ublic Health, Indiana University Template for Research Protocols in Advanced Epidemiology, PBHL P601 Spring 2013 Ayad K. Ali, PhD ayadali@ufl.edu This template was abstracted from: Ali AK. Template for Pharmacoepidemiology Research Protocols (TPRP): A Suggested Guidance for Students and Academics. TPRP Version No. 2; 2013. Title Pa ge .................................................................................................................... .................................................. 2 Table of Co ntents ............................................................................................................. ............................................ 3 List of Tabl es ................................................................................................................ ................................................ 3 List of Figur es ............................................................................................................... ............................................... 3 Abstract ...................................................................................................................... .................................................. 4 1. Backg round ................................................................................................................. ............................................. 4 2. Study Objectives ........................................................................................................... ........................................... 4 2.1. Research Question(s) .............................................................................................................................. ...... 42.2. Specific Aim(s) .............................................................................................................................. .................. 52.3. Hypotheses ............................................................................................................................... ....................... 5 3. Methods .................................................................................................................... ................................................ 6 3.1. Research Ethics .............................................................................................................................. ................ 63.2. Study Type and Design .............................................................................................................................. ... 63.3. Data Source .............................................................................................................................. ...................... 63.4. Study Population and Setting ....................................................................................................................... 63.5. Variable Definition .............................................................................................................................. ............ 63.5.1. Exposures .............................................................................................................................. .................. 73.5.2. Outcomes .............................................................................................................................. .................. 73.5.3. Covariates .............................................................................................................................. ................. 73.6. Sample Size and Power Calculations ......................................................................................................... 73.7. Statistical Analysis Plan .............................................................................................................................. ... 83.7.1. Data Management .............................................................................................................................. .... 83.7.2. Descriptive Statistics .............................................................................................................................. 83.7.3. Inferential Statistics .............................................................................................................................. .. 83.7.4. Sensitivity Analyses .............................................................................................................................. 9 4. Study Li mitations .......................................................................................................... ........................................... 9 4.1. Study Design .............................................................................................................................. ..................... 94.2. Statistical Analyses .............................................................................................................................. .......... 94.3. Data Source .............................................................................................................................. ...................... 9 5. Discu ssion ................................................................................................................. ............................................... 9 Author Cont ributions .......................................................................................................... ......................................... 9 List of Re ferenc es ............................................................................................................ .............................................. 9

PAGE 2

Title Page The title page should include the research title and any subtitle. The title should be scientifically or clinically appea ling. In addition, this is fo llowed by the protocol version and date of last revision or update, names of all authors in the grou p that is assigned to design the project, affiliations and contact in formation. Since the a ssignment will be a team work, the affiliation for the whole group will be as it shown in the following is an example. The Correlation between Obesity and Asthma Is Newly Diagnosed Asthma among Obese Adults a Consequence of Obesity or a Manifestation of Respiratory Symptoms of Obesity? Version 1 January 18, 2013 Zac George georgez@iupui.edu James Michael michaelj@iupui.edu Jane Smith smithj@iupui.edu Daniel Smith smithd@iupui.edu Amy Thomas thomasa@iupui.edu Richard M. Fairbanks School of Public Health Indiana University Indianapolis, IN

PAGE 3

Table of Contents To enhance readability, a structured table of contents should be incl uded. The table of contents should be similar to the list on page 1 of this guidance document. This should be on a separate page and can extend to more than one page. List of Tables A list of tables that support the protocol shou ld be included in this section and should be on a separate page. Tables should be referenc ed in the order they appear in the text and should be numbered according to the se ction they appear in. This might include tables supporting the information in the background section or the methods. At a minimum, the team is expected to include the following tables: Table 3.1. Variable definitions. Table 3.2. Shell table fo r descriptive statistics e.g. characteristics of study sample. Table 3.3. Shell table fo r inferential statistics e.g. incidence rates and associations. List of Figures Listing figures should follow the same rules per tinent to tables. At a minimum, the team is expected to include the following figures: Figure 3.1. Study follow up. Figure 3.2. Study profile. Figure 3.3. Shell figure for sample disposition e.g. exclusion and final sample.

PAGE 4

Abstract A stand-alone one page summary of the protocol should be included as synopses for the rationale and significance for proposing this research study; research questions and objectives, including primary and secondar y objectives; study design, setting and populations, and data source and study variables. The author can adopt the structure of a journal article that is relevant to the fiel d of knowledge related to the research study. Since this is a proposal with no anticipated re sults, the abstract c an be structured into Background, Objectives, Met hods, and expected results. 1. Background The background section should include a descr iption of the proble m which stimulated the design of the proposed study. It should present a critical review of what is already known about the topic and what gaps in kno wledge the proposed study intends to fill. Also, the section should include the findi ngs from similar studies and the expected contribution to the fiel d by the proposed study. 2. Study Objectives Primary and secondary br oad objectives of the study can be included in this section, which are related to the research question a nd operationalized later in the specific aims and hypotheses sections. The following is an example of a study objective: “…to assess asthma-related morbidity and mort ality in patients exposed to inhaled longacting beta-agonist (LABA) bronchodilators as monotherapy and inhaled corticosteroids (ICS) combination therapy in the UK…” Ali AK dissertation; 2012. 2.1. Research Question(s) The question(s) the proposed st udy intends to answer should include the target population, exposures and primar y outcomes. Research questions can be of descriptive nature or analytical nature. It is important to give consi derable thoughts to construct sound research questions as the first step in research studies. This is because well structured questions determine st udy design, define study po pulation, guide sampling, determine data source, guide data collecti on and analysis plan, translate into measurable exposures and ou tcomes, and guide interpreta tion and extrapolation of findings. Below is an exampl e of a research question: “Research Question No. 1: is there a difference in the incidence of asthma-related morbidity in terms of asthma-related hos pitalizations between asthmatic patients exposed to inhaled LABA monot herapy and ICS/LABA combination therapy in the UK?” Ali AK dissertation; 2012.

PAGE 5

2.2. Specific Aim(s) Study objectives are achieved by accomplishi ng specific aims which are translated from the proposed research questions. Each res earch question should be followed with a specific aim and corresponding hypotheses. Specific aims usually start with infinitive phrases that begin with “to” When the study intends to answer more than one question with more than one aim, I recommended specif ying each aim separa tely. The following is an example: “Specific Aim No. 1: to examine the association between LABA monotherapy and asthma morbidity rates compared with ICS/ LABA combination therapy in asthmatic patients in the UK” Ali AK dissertation; 2012. 2.3. Hypotheses Hypotheses are usually the theor etical and statistical constr ucts for the specific aims. They usually follow the aims, and should be distinguished for each specific aim. This statement should define what your study will target. T he following are examples of a hypothesis for the previous specific aim: “Hypothesis No. 1: there is no difference in asthma morbidity rates between inhaled LABA monotherapy and ICS/LABA combination t herapy in asthmatic patients in the UK “Hypothesis No. 2: at there is a differenc e in asthma morbidit y rates between inhaled LABA monotherapy and ICS/LABA combination t herapy in asthmatic patients in the UK

PAGE 6

3. Methods This section is probably the mo st important part of a study pr oposal. The choice of study designs, analytical strategies and other met hodological issues are critical parts in epidemiological studies. This section can be divided into the following subsections: 3.1. Research Ethics A statement about study review and approval by independent ethics and review board is recommended to be included. At the earlier stages of s ubmitting the proposal, the authors can state that the res earch proposal will follow due process of review by the Institutional Review Board (IRB) at the instit ution where the study w ill be conducted, e.g. The IRB at Indiana University Research Pr otection Program (HRPP). Note that some studies may not require IRB approval and some require varying levels of review. This level of detail should be captured here. 3.2. Study Type and Design State the type of study and the proposed epidemiological a pproach to address study objectives. Highlight the strengths and weaknesses of the adopted design with justifications for selecting the proposed appr oach given the research question(s). 3.3. Data Source From every research study, data will be generated and subsequently translated to conclusions that might have public heal th impacts. The authors should discuss the source of the data, the st rengths and limitations, how t he data was collected/generated, and any procedure for data linkage to other s ources or data validat ion. Reasons why this particular data source is deemed appropr iate to answer the research questions should be stated. 3.4. Study Population and Setting In terms of person, place and time, this se ction should elaborate on what the units of analyses are, sampling method, and should incl ude inclusion and excl usion criteria with rationale for every criterion. T he impact of inclusion/exclusion criteria on the sample size for analysis. Internal and exte rnal validity should be discu ssed in this section. Study duration, selection of comparison groups, and types of comparison groups should be stated as well. In case of utilizing databas es, e.g. electronic medical records or healthcare insurance claims, how the pati ents are identified in these databases and whether there is a potential for misclassi fication and selection bias should be considered. Study profile graph can be incl uded in this section (e.g. Figure 3.2.) 3.5. Variable Definition This section is important to guide the implementation of statistical analysis plan. Identified independent (expos ure), dependent (outcome), and covariates (confounders or effect modifiers) should be stated withi n the corresponding section, and the process

PAGE 7

of identifying each variable should be di scussed, especially ways to prevent misclassification bias, measurement bias, and confounding. The choice of confounders and effect modifiers should be ju stified. I recommend includin g a variable definition table (e.g. Table 3.1.) which facilitates the pr ocesses of reviewing, development and completion of the protocol An example is below: Domain Variable Values Exposure Bronchodilator 0: ICS/LABA 1: LABA Outcome Asthma-hospitalization 0: No 1: Yes Comorbidity Weight status, BMI (kg/m 2 ) 0: <18.5 (normal) 1: 18.5-24.9 (underweight) 2: 25.0-29.9 (overweight) 3: 30 + (obese) 9: Unknown Comedication Antibiotics 0: No 1: Yes Demographics Age (years) 13-65 Sex 0: Female 1: Male Asthma severity Oral cortic osteroids 0: Not prescribed 1: Prescribed Number of asthma drug classes prescribed >0 Inhaled short-acting beta-agonists 0: Not prescribed 1: Prescribed Ali AK dissertation; 2012. 3.5.1. Exposures How the exposures are identified, wh at are the comparison groups, and how misclassification bias is minimiz ed or prevented should be discussed. 3.5.2. Outcomes What are the outcomes of interest and how they are measured and whether they are clinical end points or surrogat e outcomes should be discussed. 3.5.3. Covariates Why these factors are deemed important covariates and which are confounders and which are effect modifiers should be discussed. In addition, the possibility of residual confounding should be stated and whether thes e covariates minimize such confounding or not. 3.6. Sample Size and Power Calculations Whenever possible a power calc ulation and sample size determination is encouraged to be included.

PAGE 8

3.7. Statistical Analysis Plan Formulating a statistical analys is plan is a crucial part of the protocol and should mirror the epidemiological appr oach. I recommend dividing the sect ion into the following parts: 3.7.1. Data Management Include how the data will be pr ocured, where it will be st ored, and what statistical analysis software will be employed to conduc t data cleaning and analysis. I also recommend mentioning the type of statistical testing and the le vel of type-1 error that is specified. For example: “…two-sided tests with alpha=0.05 a priori level of statistical significance are used throughout the analysis procedures…”Ali AK dissertation; 2012. 3.7.2. Descriptive Statistics This part includes types of statistical te sts that will be applied to compare baseline characteristics of the comparison groups by describing the frequenc ies and patterns of patients, exposures and outcomes. Appropria te tests should be used for each data type, e.g. Chi-square test for categor ical data, and t-test for co ntinuous data. In addition, Use descriptive measures that are consistent with data ty pe, e.g. proportions and 95% confidence intervals to describe categorical data, and means and standard deviations to describe continuous data. Type s of graphs used for the description should also be mentioned, e.g. bar graphs and pie charts for categorical data, and histograms and scatter plots for continuous data. The interpreta tion of the statistical significance of the differences between exposure groups in te rms of baseline variables should be discussed here, e.g. what woul d a p-value of 0.05, <0.05 and >0.05 means? In addition, I recommend including shell tables for the expected results corresponding to characteristics of units of analyses, e.g. patients (e.g. Table 3.2.). 3.7.3. Inferential Statistics This part includes the procedures that in tend to answer the research questions by testing the hypotheses for corresponding specific aims. Usually, this part aims to test associations between exposures and outcomes a fter controlling for potential and actual confounders and other covariates. Authors should discuss the strengths and limitations of every statistical test. I recommend pr oviding an operationalized annotations and mathematical equations to support this part. In addition, identified measures of associations should be stated and an overvi ew of the interpretation of the estimated results should be discussed, e.g. what woul d a relative risk of 1.30 means, what would an odds ratio of 0.40 means, and how a 95% confidence interval be interpreted? Furthermore, Shell tables for expected re sults from relevant analyses should be included here (e.g. Table 3.3.).

PAGE 9

3.7.4. Sensitivity Analyses Sensitivity analyses can be applied to evaluat e the uncertainty of study results to changes in study parameters, e.g. c hanging inclusion/excl usion criteria. 4. Study Limitations The weaknesses and limitations of the proposed study should be discussed in this section and I recommend dividing it into t he following parts. Although the limitations might be stated in the corresponding parts in the methods section, the authors can restate them here to facilit ate readability and grading when this document is used as part of a teaching course. 4.1. Study Design 4.2. Statistical Analyses 4.3. Data Source 5. Discussion This section should include a description of the potential impact of the proposed study and any necessary follow-up in order to addr ess the research pr oblem identified. Author Contributions Individual contribution and res ponsibilities for every secti on or part of the protocol should be listed in table format, e.g. A uthor Contribution Remarks Abstract Zac George, James Michael Background All Consulted with Dr. Ali Study Objectives James Michael Study Type and Design James Michael Data Source Zac George, James Michael Study Population and Setting Jane Smith Variable Definition Amy Thomas Statistical Analysis Plan James Michael, Jane Smith Study Limitations Daniel Smith, Jane Smith Expected Findings All Tables James Michael, Jane Smith Figures Amy Thomas List of References Any source of information referred to in t he protocol should be listed in numerical or alphabetical order according to bibliography citation conventions to allow easy retrieval.

PAGE 10

Additional Resources International Society for Pharmacoepidemiol ogy. Guidelines for Good Pharmacoepidemiology Practices. Pharmacoepidemiology and Drug Safety 2008;17(2):200-208. Berger ML et al. Good Research Practices for Comparative Effectiveness Research: Defining, Reporting and Interpreting Nonrandomized Studie s of Treatment Effects Using Secondary Data Sources: The ISPOR Good Research Practices fo r Retrospective Database Analysis Task Force Report—Part I. Value in Health. 2009;12(8):1044-1052. Cox E et al. Good Research Practices for Comparative Effectiveness Research: Approaches to Mitigate Bias and Confounding in the Design of Nonr andomized Studies of Treatment Effects Using Secondary Data Sources: The ISPOR Good Research Practices for Retrospective Database Analysis Task Force Report—Part II. Value in Health .2009;12(8):1053-1061. Johnson ML et al. Good Research Practices fo r Comparative Effectiveness Research: Analytic Methods to Improve Causal Inference from Nonr andomized Studied of Treatment Effects Using Secondary Data Sources: The ISPOR Good Research Practices for Retrospective Database Analysis Task Force Report—Part III. Value in Health 2009;12(8):1062-1073. von Elm E et al. The Strengthening the Report ing of Observational Studies in Epidemiology (STROBE) Statement: Guidelines fo r Reporting Observational Studies. Journal of Clinical Epidemiology. 2008;61:344-349. Checklist: www.strobe-statement.org Moher D et al. Preferred Reporting Items for Sy stematic Reviews and Meta-Analyses: The PRISMA Statement. British Journal of Medicine. 2009;339:332-336. Checklist: www.prisma-statement.org Stoup DF et al. Meta-Analysis of Observational Studies in Epidemiology: A Proposal for Reporting. Journal of American Medical Association. 2000;283:2008-2012. Chan AW et al. SPIRIT 2013 Statement: Defining Standard Protocol Items for Clinical Trials. Annals of Internal Medicine. 2013 [Epub Ahead of Print]. Checklist: www.spirit-statement.org Enhancing the Quality and Transparency of Health Research (EQUATOR) Network Initiative. List of Reporting Guidelines by Study Type and Guideline Source: http://www.equator-network.org/resourcecentre/library-of-heal th-research-reporting The European Network of Centres for Pharmacoepidemiology and Pharmacovigilance. Guide on Methodological Standards in Pharmacoepidemiology. Revision 1. EMA/95098/2010. London, UK: European Medicines Agency. July 2012. Gliklich RE, Dreyer NA, eds. Registries for Evaluating Patient Outcomes: A User’s Guide. AHRQ Pub. No. 07-EHC001-1. Rockville, MD: Agency for H ealthcare Research and Quality. April 2007.