Surgically correcting velopharyngeal insufficiency


Material Information

Surgically correcting velopharyngeal insufficiency a retrospective analysis comparing the combined palatal pushback with pharyngeal flap and the teflon augmentation pharyngoplasty
Physical Description:
ix, 151 leaves : ill. ; 28 cm.
Eisenbach, Charles R ( Charles Robert ), 1944-
Publication Date:


Subjects / Keywords:
Velopharyngeal insufficiency -- Surgery   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1983.
Includes bibliographical references (leaves 140-150).
Statement of Responsibility:
by Charles R. Eisenbach, II.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001038718
notis - AFC1235
oclc - 10116773
System ID:

Full Text







Copyright 1983

Charles R. Eisenbach, II

To all those individuals, past and present,

who have influenced me in their role as teacher.


Appreciation and gratitude are extended to the chair-

man of my committee, Dr. William N. Williams, who not only

provided direction and support throughout this undertaking,

but also has provided the opportunity for me to become

skilled in the techniques and well-versed in the body of

knowledge prerequisite to attempting to use cinefluorography

as a research tool.

I would like to thank the other members of my committee

for giving generously of their time and assistance. Dr. Thomas

Abbott has also been an advisor and friend since I began my

graduate study. Dr. Kenneth Bzoch,in 1973, first planted

in me the seed of interest in the area of cleft palate.

Dr. Leonard LaPointe has provided me with invaluable clini-

cal insights and the opportunity to improve my clinical

skills during my graduate study. Dr. Wallace Mealiea, Jr.,

graciously joined the committee in the eleventh hour and

has been a source of inspiration to those who know and work

with him. Thank you, gentlemen.

Special thanks are in order for Dr. Leonard Furlow,

who served as an ad hoc committee member; for Dr. Wiley

Rasbury, who coordinated my minor in psychology and was an

original member of my committee prior to his leaving the

University of Florida; and for Dr. G. Paul Moore, also an

original committee member before his retirement.

Thanks also to Dr. Randy Carter for help with the

statistical analyses and to Mrs. Diane Fischler for typing

the manuscript.

Thanks Hutch, Leslie, and you too, Sherry, for always

having the time and just the right word of encouragement.










S. . . .viii

Introduction 1
Statement of Purpose 12

Velopharyngeal Kinesiology .. .14
Surgical Correction of Velopharyngeal
Insufficiency. . .. .25
Research Questions and Hypotheses. ... 57

Selection of Subjects. .... .. 66
Procedure for Obtaining Medical Record
Data . . 69
Procedure for Obtaining Data from the
Cinefluorographic Films. ... 74
Method of Analysis . ... 82

RESULTS . ... 85
Preoperative Group Homogeneity .. .85
Morphological Changes Following Surgery. 88
Effectiveness in Eliminating Hyper-
nasality and Nasal Emission. .. .90
Persisting Postoperatively Induced
Hyponasality . 91
Operative Time, Length of Hospital Stay,
and Immediate Postoperative Morbidity. 93
Differences in Preoperative and Post-
operative Measures of Velar Kinesiology
Between the Groups . ... .98
Differences in Preoperative and Post-
operative Measures of Velar Kinesiology
Within the Groups. . .. .102
Summary .. 102











Discussion . .
Conclusions . .











. 106
* 106
* 117

* 120

. 123

. 127

. 132

. 135

. 136

. 138

. 140

. 151

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



Charles R. Eisenbach, II

April, 1983

Chairman: William N. Williams
Major Department: Speech

In the area of surgically treating velopharyngeal insuf-

ficiency (VPI) there are few comparative studies. The in-

tent of this study was to retrospectively compare the speech

outcome, hospital utilization data, and effects on velar

function during closure for speech of the Teflon augmenta-

tion pharyngoplasty and the combined pushback with supe-

riorly based pharyngeal flap.

The Teflon group consisted of 20 patients: 11 males and

9 females. The pushback group consisted of 25 patients: 17

males and 8 females. Nine patients in the pushback group

and seven in the Teflon group had both preoperative and

postoperative cinefluorographic records available for analy-

sis of their velopharyngeal function.

The groups were statistically compared on age, sex dis-

tribution, reported etiology of their VPI, and on the pre-

operative consistency of their symptoms of hypernasality and

nasal emission. No statistically significant differences


were found at the 1% level of confidence. Preoperative meas-

ures of velar stretch, velar height (elevation), and velar

rise rate were not found to be significantly different be-

tween the groups. It was concluded that each group repre-

sented an independent sample drawn from the same clinical


In eliminating the primary speech symptoms of VPI the

Teflon procedure was effective in 80%, the combined push-

back in 92% of the patients. The difference was not sig-

nificant. The Teflon procedure was found to take signifi-

cantly less time surgically (x = 28 mins. vs. 119 mins.)

and resulted in a significantly shorter hospital stay

(x = 2 days vs. 4 days) than did the combined pushback

with pharyngeal flap procedure. The frequency and severity

of postoperative complications were not found to differ sig-

nificantly between the groups. Both procedures produced

similar postoperative decreases in velar stretch and velar

height. Velar rise rate appeared to be unaffected by either

procedure. The decrease in velar height postoperatively was

significant for the Teflon group. It was concluded that

velar stretch and velar rise rate to not have sufficient

clinical relevance to warrant continued study in the clini-

cal population.

These findings suggest that both procedures accomplish

treatment objectives without major consequential differences.

However, the Teflon procedure was shown to be the potentially

more economical and surgically conservative.



The subject of the study reported here is an analysis

of the speech outcome, morbidity, and effect on velar (soft

palate) function between two surgical procedures used to

correct velopharyngeal insufficiency.

Velopharyngeal insufficiency (VPI) is a condition,

detrimental to normal speech sound production and voice

quality, that results when the velopharyngeal mechanism

fails to adequately separate the nasopharynx from the oro-

pharynx during speech. The primary components or symptoms

of VPI are an excessively nasal voice quality (hypernasality)

and the inappropriate loss of air through the nose (nasal

emission) during speech.

When VPI exists because of some morphological abnor-

mality, the individual is often referred for surgical re-

construction or modification of his or her velopharyngeal

mechanism. Of the many surgical procedures that have been

devised to correct VPI this thesis is concerned with only

the combined palatal pushback with a high-based pharyngeal

flap (Dibbell, Laub, Jobe, & Chase, 1965) and the Teflon

augmentation pharyngoplasty (Lewy, Cole, & Wepman, 1965).


The development of these two surgical procedures is described

in detail in the following chapter.

As this study is concerned with comparing two surgical

procedures, a brief overview of how surgical techniques have

been and are currently being evaluated will be presented.

Evaluating New Surgical Techniques

Following the introduction of ether anesthesia in 1846

and the development of antiseptic methods in the 1860's,

surgeons were able to report their results and innovations

in the form of the series (a succession of similar opera-

tions following one another in time, generally without any

attempt at comparison). Bunker, Barnes, and Mosteller (1977)

observed that from the "rough and ready" (p. 107) beginnings

of surgery there arose a tradition that placed a "premium

on effective, simple measures directly tested by the sur-

geon's immediate observations" (p. 107).

Writing about surgical innovation by trial and error,

Barnes (1977) reviewed articles advocating new or improved

methods of surgical therapy that appeared in the Trans-

actions of the American Surgical Association from 1880

through 1942 (in that time period probably the most re-

spected and prestigious publication of its kind). While

the primary objective of his paper was to obtain examples

of surgical innovations and therapies which were subse-

quently judged to be worthless, the author concluded that,

in general, the concept of controls (different groups used

for comparison in order to systematically rule out other

possible causes of the effect being studied) appeared to

be "totally unknown" (p. 119) to surgeons of that period.

He lists five major reasons for the prolonged acceptance of

some ineffectual operations of the past:

(1) uncritical acceptance of established dogma
or conventional wisdom in surgery as dic-
tated by established leaders in the field

(2) primitive comprehension of standards for ade-
quate follow-up including such details as
sample bias, duration of follow-up and ob-
jectivity of observations

(3) casual acceptance by the surgeon that a dis-
crete disease requiring treatment was present
when, in fact, there was none

(4) unjustified confidence of some influen-
tial surgeons in their understanding of human

(5) virtual absence of ethical constraints
demanding the greatest economy in human mor-
bidity and mortality in the development and
application of new treatments. (pp. 120-121)

Barnes (1977) tempers his criticism of the surgeons of that

era with the recognition that they faced their challenging

tasks with a strong, emotional orientation and a fervent,

personal involvement as a probable consequence of their

more solitary position in the medical community. These

surgeons also worked in a time when many ancillary medical

sciences were either in their infancy or as yet unknown and

when many diagnostic aids and procedures, which are rou-

tinely used today prior to surgery, had yet to be invented.

The call for improved methods for evaluating new surgi-

cal procedures and techniques appears to have begun in the

early 1960's (Beecher, 1961; Moore, 1963) in response to

the short life span of the internal mammary artery ligation

procedure for the treatment of angina pectoris which was,

initially, enthusiastically received by the medical com-

munity but was completely discredited within two years

(Beecher, 1961; Cobb, Thomas, & Dillard, 1959). Also, the

applicability of drug studies to the technique of the ran-

domized clinical trial has led at least one author to

strongly advocate the use of randomized clinical trials

in evaluating other therapies, particularly surgery

(Spodick, 1973; Spodick, 1975). In response to Spodick's

(1975) criticism of surgeons' unobjective approach to sci-

ence, Love (1975) points out that surgical operations un-

like drugs are conceptual plans which require execution

rather than prepackaged preparations. Further, the techni-

cal aspects of a given operation generally change over time

unlike the uniform chemical compounds that make up a given

drug. In addition, there is a certain skill factor in-

volved in surgical procedures, unlike the use of drugs, all

of which suggest that using the same techniques to evaluate

both is inappropriate.

The randomized clinical trial for surgical procedures,

as described by Chalmers (1976), begins with accepting the

proposition that the relative merits of the treatment, or

treatments, are currently unknown. Patients who meet pre-

stated eligibility requirements for inclusion in the trial

are subsequently randomly assigned to one of the treatment/

control groups. Strict ethical standards regarding in-

formed consent of the patient must be followed, and patients

must be free to remove themselves from the trial at any

time. Treatment/control groups must then be followed for

an equal period of time following surgery, preferably by

someone who does not know which treatment the patient re-

ceived. The major advantages of randomization are sum-

marized by Byar, Simon, Friedwald, Schlesselman, DeMets,

Ellenberg, Gail, and Ware (1976). Chiefly, bias is elimi-

nated from the assignment of treatments, groups are bal-

anced in covariates, statistical tests are valid, and the

findings are generalizable.

The need for these types of data on many surgical pro-

cedures currently in wide use has been underscored by nu-

merous writers (Beecher, 1961; Chalmers, 1976; Crile, 1978;

Doran, 1964; Moore, 1963). Medical journals have also been

criticized for publishing uncontrolled surgical studies

(Spodick, 1979) and for not applying the same standards

for published trials of medical and surgical therapies

(Spodick, Aronow, Barber, Blakburn, Boyd, Conti, LeGerfo,

Lown, Mathur, McIntosh, Preston, Selzer, & Takaro, 1978).

Like Love (1975), not all authors advocate the strict

use of randomized clinical trials in the evaluation of new

therapies, surgical or other. Gehan and Freireich (1974),

writing about clinical trials in cancer research, state

that, "quantitative, comparative clinical trials can some-

times be better accomplished with techniques other than

randomization" (p. 198). Two specific situations in which

randomization is not necessary, in their opinion, are when

the primary purpose of the study is to estimate the effec-

tiveness of a treatment when large differences are expected

and when a proposed therapy can be compared to a standard

from a recent previous trial in a sequence of studies.

Another situation in which they do not recommend randomiza-

tion is when only a small number of patients are available

for study.

In the actual practice of surgery randomized clinical

trials are not as easily carried out as it might be in-

ferred. New operations are, generally, introduced tenta-

tively and often change dramatically over time making re-

sults invalid because of changes in technique (Bonchek,

1979a). Bonchek (1979b) also points out that randomiza-

tion does not change the operative risk to the patient but

rather dictates only that chance and not the patient will

determine exposure to the risk. Byar et al. (1976) acknow-

ledge that a given physician may not be able to continue

with patient randomization if evidence accumulates during

the course of the trial that one treatment is superior.

Gehan and Freireich (1974) also question the ethics of con-

tinuing any randomized clinical trial past the suspicion

of such doubts. Randomized clinical trials may be diffi-

cult to carry out on a given procedure because the rarity

of the condition that necessitates the surgery may make it

impractical to randomize. The skill, or lack thereof, of the

surgeons) with one or another of the treatment techniques

may also bias any results in favor of the more established

technique or the one that is technically more simple (van

der Linden, 1980).

How the surgical community, in general, has responded

to the call for utilizing randomized clinical trials to

evaluate surgical techniques is revealed by Gilbert, McPeek,

and Mosteller (1977) who reviewed 107 published papers from

the medical literature in surgery and anesthesia which ap-

peared between 1964 and 1972. All the papers were concerned

with appraising surgical and anesthetic treatments as used

on human subjects. They found three basic types of studies

reported: randomized controlled trials, nonrandomized con-

trolled trials, and series. The authors use the term "ran-

domized controlled trial" when the investigator compared

two or more treatment groups, and patients were assigned to

the groups by a formal randomization process. The nonran-

domized controlled trials did not have such a formal randomi-

zation process and varied from comparing groups treated

concurrently in the same institution to comparing patients

treated previously by one method with patients treated

currently with another. The papers reporting on series

described a set of patients treated in some specified manner

but with no specific referent for comparison except pos-

sibly other reports in the literature dealing with similar

patients. The 107 studies reviewed included 36 randomized

controlled trials, 15 nonrandomized controlled trials, and

56 papers reporting findings on a series of patients. Thus,

just a decade ago only one-third of new or modified surgical

and anesthetic treatments used on humans were being evaluated

using randomized controlled trials. More current informa-

tion on this controversy was not found in the literature.

Evaluation of Surgical Techniques in the Area of Cleft
Palate and Velopharyngeal Insufficiency

Practitioners of cleft palate surgery do not appear to

have enthusiastically embraced the idea of randomized clini-

cal trials in actual practice. In a 1972 editorial, Grabb

urged greater use of the "experimental method" (p. 563) in

clinical studies done by plastic surgeons. Further, he

suggested the creation of full-time research positions in

plastic surgery at major academic institutions. In response,

Johnson (1973) seemed to belittle Grabb's (1972) proposals

when he suggested that future great discoveries in the field

of plastic surgery will not come from research plastic sur-

geons and that good clinical research can be done without

large research facilities and without substantial support.

A MEDLARS II computer search of the English language

literature from January, 1966, through December, 1981, using

the key words "cleft lip," "cleft palate," 'velopharyngeal

insufficiency," "cleft palate prosthesis," "surgery," "ran-

dom," and "prospective" revealed only three articles per-

taining to cleft palate surgery that report randomization

of subjects (Marsh & Wray, 1980; Whitaker, Randall, Graham,

Hamilton, & Winchester, 1972; Wray, Dann, & Holtmann, 1979).

Further, a 22-year review of the major English language

literature pertinent to cleft palate and VPI including

Cleft Palate Bulletin/Journal, Plastic and Reconstructive

Surgery, and the British Journal of Plastic Surgery failed

to reveal any other articles on randomized clinical trials

not already uncovered by the MEDLARS II search. With re-

spect to surgery pertaining to cleft palate and VPI, the

evaluation of the outcome of new techniques would appear

to continue to consist almost exclusively of reporting re-

sults by means of describing a series.

Why more controlled, unbiased studies have not been

undertaken in this field is speculative. Certainly with

the relatively high incidence of VPI and with the emergence

of numerous cleft palate teams and craniofacial centers to

centralize treatment in the United States there should not

be a lack of clinical material upon which to undertake ran-

domized clinical trials. Perhaps, as Crile (1978) suggests,

many surgeons continue to cling to the established methods

of correcting certain conditions. Possibly the consumers

of cleft lip and/or palate management (the patient and his

or her family) have not been educated to the point where

they are aware that there are often several procedures

available to treat a particular problem.

The reasons that more definitive answers to questions

about why we do not know more about which operations opti-

mally will benefit a given patient probably encompass, in

part, all of the previously stated possibilities. However,

in the literature pertaining to the surgical correction of

VPI there exists a single, glaring deficit that perhaps

overshadows all other problems in this area. That is, to

date, there have not emerged any universally established

or accepted rationales for selecting a particular opera-

tive procedure over others available to treat VPI. Health

care professionals who work in the area of cleft lip and/or

palate were alerted to this problem almost 20 years ago

when, in 1963, at a University of Iowa conference on the

physical management of cleft lip and palate, Dr. H. L.

Morris, program chairman, noted that many of the confer-

ence participants had difficulty identifying the diagnos-

tic criteria they used in making specific management deci-

sions. Even when diagnostic criteria were identified,

the participants had difficulty in describ-
ing, in operational terms, the nature of observa-
tions which they make or the relationship of the
observations to the criteria and to the conse-
quent management decisions. (p. 78)

Yules (1970) observed that if investigators were to

gather more objective data preoperatively and postopera-

tively on patients undergoing secondary procedures for the

correction of VPI, that these data might provide a basis

for preoperative procedure selection.

In 1971, Yules, Chase, Blocksma, and Lang stated un-

equivocally that "there is no established method for select-

ing an individual patient for a given operation" (p. 457)

in cases of secondary procedures for correcting VPI. They

further noted that many medical centers use a single opera-

tion for all patients or subjectively select a given proce-

dure over others. These authors attribute this imprecision

to a lack of objective, quantitative preoperative and post-

operative evaluation data.

In reporting on 98 patients with VPI without overt

cleft palate, 48 of whom had undergone surgery, Minami,

Kaplan, Wu, and Jobe (1975) observed that one surgical pro-

cedure was more frequently employed to treat the patients

regardless of the underlying reasons for the VPI. They

stated that this practice "has often been dictated by local

tradition" (p. 585). They did not recommend such an ap-

proach and suggested that surgical treatment be tailored to

the individual needs of a given patient.

Hogan and Schwartz (1977) insist that surgical proce-

dures for the treatment of VPI should only be judged by

their speech results (elimination of hypernasality and

nasal escape), and that such results must be measured ob-

jectively, be reproducible and translatable between differ-

ent cleft palate centers.

Krischer (1980), in presenting a quantitative method-

ology (decision analysis) for evaluating treatment alterna-

tives in cases of VPI, states that more research is needed,

including more definitive descriptions of outcomes and their

quantification, before decision analysis can be used to

model team decision-making dynamics.

Crile (1978) maintains that when a condition is not

totally understandable scientifically, treatment will be

based empirically on observed results rather than on ac-

cepted theories. Crile (1978) further suggests that the

surgeon, given several equally effective choices to treat a

given condition, generally will make a decision based on

the side effects, complications, and discomforts of the

different procedures. Thus, if several surgical techniques

used to treat VPI remain under debate, it would be reason-

able to conclude that a surgeon would choose the procedure

that he or she is most comfortable with in terms of opera-

tive technique and possible postoperative complications.

Despite the assertion of Converse, Hogan, and McCarthy

(1977) that surgery for cleft lip and palate began to em-

brace the "scientific approach" (p. 1932) in the 1930's

with the work of Victor Veau, it would appear that the field

continues to rely, almost exclusively, on the reporting of

a series for dissemination of new ideas and techniques. The

numerous authors cited earlier who have called for more ob-

jective data gathering when reporting outcomes appear to

have gone unheard.

Statement of Purpose

If surgeons, in particular, and cleft palate teams, in

general, are to make more fully informed management deci-

sions that should more optimally benefit a given patient,

then they must be provided with more than empirical

observations of surgical outcomes. The broad purpose of

this study is to retrospectively compare, on a variety of

variables, two very different surgical procedures used to

correct VPI. The two procedures, combined palatal pushback

with superior high-based pharyngeal flap (Dibbell et al.,

1965) and the augmentation pharyngoplasty using injectable

Teflon (Lewy et al., 1965), both appear to have been used

on patients from the same clinical population, thus making

them excellent candidates for a comparative analysis. Spe-

cific research questions and hypotheses are offered in the

following chapter.


Velopharyngeal Kinesiology

A normally functioning velopharyngeal mechanism is es-

sential for acceptable speech sound production and for

normal vocal resonance. In the patient with a cleft palate

it is the formidable task of the surgeon to carefully re-

construct the mechanism so that proper function is achieved.

In the approximately 33% of cases in which normal velopha-

ryngeal function is not achieved following surgical recon-

struction and in the much smaller (1 in 120 with a sub-

mucous cleft plate and 1 in 1,459 following removal of

tonsils and adenoids) population which manifests the symp-

toms of VPI without an overt cleft of the palate (Weatherley-

White, Sakura, Brenner, Stewart, & Ott, 1972; Gibb, 1958),

the professionals involved in the habilitation process must

first discern the underlying reason or reasons for the VPI

and then decide which course of management will most benefit

the patient.

Normal Velopharyngeal Function

For an excellent overview of the anatomy of the velo-

pharyngeal sphincter the reader might consider either


Bateman (1977), Millard (1980) or the older but still useful

Zemlin (1968). The consensus of the literature suggests

that the elevator palatine muscle is the primary muscle re-

sponsible for velopharyngeal closure. Upon contraction

this muscle elevates and pulls the velum posteriorly. Ele-

vation of the middle third of the velum is also thought to

be substantially assisted by the uvular muscles (Dickson,

1972). The middle third of the velum consistently has been

shown to be that portion of the velum acting as the func-

tional valve during velopharyngeal closure (Calnan, 1953,

1956, 1961a; Aram & Subtelny, 1959). The elevator palatine

also is thought to be primarily responsible for the lateral

pharyngeal wall motion which contributes to velopharyngeal

closure (Dickson, 1972). This has been disputed (Shprintzen,

McCall, Skolnick, & Lencione, 1975; Zagzebski, 1975), and

the issue remains unsolved. It is possible that the superior

constrictor does play a major role in velopharyngeal closure

for nonspeech activities (Minifie, Abbs, Tarlow, & Kwaterski,

1974). The superior fibers of the superior constrictor do

appear to be responsible for the formation of Passavant's

ridge; however, the functional significance of this ridge,

when present, is not clear. Normal velopharyngeal closure

is also known to vary with age, sex, and task.

Variation in normal closure by age. Aram and Subtelny

(1959) in a lateral cephalometric study of normal velopha-

ryngeal function found that the degree of movement from

rest to closure increased with increasing age until the age

of 17 years. In the adult the degree of movement decreased

and stablilized to approximate the degree of movement seen

in the 12 to 14 year age group studied. These investigators

also found changes in the site of closure with respect to

age. In the younger age groups the velum most often ap-

proximated the superior aspect of the nasopharynx. In the

older subjects the velum contacted the posterior pharyngeal

wall at a lower level. The amount of velar tissue effect-

ing closure also was found to change with age, decreasing

by about 50% from the youngest age group (4-5 years) to the

oldest (18-20 years). Calnan (1961a) observed that normal

children effected velopharyngeal closure at a lower level

than normal adults. He attributed this to the frequently

present adenoids which would limit the "full elevation of

the soft palate" (p. 35). Calnan (1961a) studied only 46

normal children and did not group them according to increas-

ing age as did Aram and Subtelny (1959) who studied twice

as many children. Bjork (1961) studied 30 adults and 21

children (5-8 year olds) using cineradiography and sound

spectography and concluded that the children, on the average,

had faster velar movement than did the adults, but "the dif-

ference was not significant" (p. 60). He found no signifi-

cant differences in the amount of closure but did observe

that the upper point of closure in children was lower than

that seen in the adults. This finding is in agreement with

Calnan (1961a) but was just the opposite of the findings of

Aram and Subtelny (1959). Yules and Chase (1968), using

cineradiography to study velar function in both normal and

VPI children, found that the ascent rate of the velum was

faster in normals than in the experimental groups. Further,

the velar ascent rate increased with age in normals, and

the "knee" (p. 124) elevatorr eminence) ascended faster than

the tip (uvula) of the velum. Velopharyngeal closure was

always observed to be effected by the knee of the velum in

the controls, and compensatory movements of the tongue or

posterior pharyngeal wall were never observed in the normal


Variation in normal closure by sex. In 1969, McKerns,

in a cineradiographic study of 40 young adults (20 males

and 20 females) with normal speech, hearing, and occulusion,

identified that velar configuration during closure for speech

is different between the sexes in adults. In the male pat-

tern the amount of velar contact with the posterior pharyn-

geal wall is less and the height of elevation is greater

than in the female. In the male the inferior point of velar

contact is typically above the palatal plane, and the velum

forms an acute angle during closure. In the female the

amount of velar contact is greater and the height of velar

elevation is less. Velar contact typically takes place at

or below the palatal plane, and the velum tends to form a

right angle during closure (see Figure 1). Sex differences

in velar configuration during closure for speech have not

been demonstrated in children. However, Owsley, Chierici,

Miller, Lawson, and Blackfield (1967) did show that common




Figure 1. The configuration of the velum during closure
for speech in young adults.

measurements of velopharyngeal parameters (hard palate

length, soft palate length, and depth of the nasopharynx)

do not differ by sex in children from 6 to 12 years of age.

Variation in normal closure by task. Warren and Hof-

mann (1961) studied 37 normal speaking adults using cineradi-

ography and found no relationship between soft palate length

and pharyngeal depth; however, they did find significantly

different levels of velar elevation during speech and non-

speech acts (higher for blowing than for phonation of /a/).

The findings for Warren and Hofmann (1961) were confirmed

later by Moll (1965) who analyzed speech and nonspeech ac-

tivities using cinefluoroscopy on ten normal adults. Moll

(1965) further observed that puffing the cheeks and sucking

can be performed without velopharyngeal closure by utiliza-

tion of lingual-palatal valving.

The incomplete descriptions of velar function in chil-

dren found in the literature (Aram & Subtelny, 1959; Bjork,

1961; Calnan, 1961a; Yules & Chase, 1968) suggest that velo-

pharyngeal closure in children closely resembles the pattern

described by McKerns (1969) for the young adult female.

That is, the velum typically makes contact with the posterior

pharyngeal wall at or below the level of the palatal plane

during normal velopharyngeal closure in children.

More recent studies of velopharyngeal function have

tended to exclude children because of increasing concern

over needless irradiation. They have focused on improving

and expanding existing techniques such as multiview

videofluoroscopy (Skolnick, 1970), the submentrovertical

view in radiographic analysis of the velopharyngeal mecha-

nism (Zwitman, Gyepes, & Sample, 1973), and raising issues

to be considered when using cine or videofluoroscopy

(Shelton & Trier, 1976). More recent publications have also

introduced new instrumentation for assessing individuals

with VPI such as ultrasound (Minifie, Hixon, Kelsey, &

Woodhouse, 1970), nasendoscopy (Pigott, Bensen, & White,

1969), airflow measurement devices (Warren, 1967), and

the oral telescopy (Zwitman, Gyepes, & Ward, 1976). Other

investigators have defined patterns of velopharyngeal clo-

sure in normals and in subjects with VPI (Shprintzen,

Lencione, McCall, & Skolnick, 1974; Shprintzen et al.,

1975; Skolnick, McCall, & Barnes, 1973; Zwitman, Sonder-

man, & Ward, 1974). And several writers have pointed out

the potential danger in making inferences from normative

data to the clinical population (Peterson, 1973; Williams &

Eisenbach, 1981).

Abnormal Velar Function

Hypernasality and nasal emission are the primary speech

components of VPI. Other problems resulting from VPI such

as the use of glottal stops and/or pharyngeal fricatives,

nasal grimacing, and breathiness are compensatory habit

patterns, usually habituated unconsciously over time and,

as such, must be considered as secondary components of VPI.

The distinction becomes important in assessing the outcome

of a surgical procedure done to correct VPI.

The causes of abnormal velopharyngeal function result-

ing in VPI are many. To attempt to individually list every

known or potential cause of VPI would be unproductive and

probably impossible. Calnan (1959, 1961b)presented and re-

vised a classification scheme based on what he described as

defects in mechanics (anatomy), dynamics (motor power), and

kinetics (velopharyngeal disproportion). Randall, Bakes,

and Kennedy (1960) elaborated on the neurologic deficiencies

which may produce VPI in reporting a small series of nine

patients who had been surgically treated for VPI. Minami

et al. (1975) modified the classifications of Calnan (1959)

and Randall et al. (1960) to emphasize what they described

as "the big four" (p. 574) causes of VPI. The four included

palatopharyngeal disproportion, abnormal anatomy of the

elevator palatine muscles, palatal paresis, and functional

abnormalities. Most investigators are careful to note that

the causes of VPI may be present in combination in a given

patient. While the diagnosis is relatively easy (by ear),

the underlying causes must be carefully and systematically

investigated if the patient is to receive the optimal treat-


Table 1 outlines potential causes of VPI and is a syn-

thesis of the three previously cited classification schemes.

It is intended to be comprehensive yet not all inclusive.

In summary, VPI is well-recognized to occur in approxi-

mately one-third of the cases of cleft palate following the

primary palatal repair (Converse, 1965; Yules et al., 1971)

>1 4

I 0400
I C 0 0
I ri r H ,CrA
S-H04-J 4JO
0>0 0
I z 0 a 0o
a -)-H ) 4-4 -
QI C 0 z 0H
rd1)0 r-H CI --4
H- r 1 4 C
0 0I >W 0)C 03
0 0 I0 O3 0 c 03

0. Ir .
0) i0 .0 4- 4 04.

-4 I
4-4 I H4c4 r N m

0 I

rd U) I1 ((dH r g
C I0 C 4-i) ) d 4d
* 7II Hcd> Qa Cd>
SdH 0
t7)1 1 0 1 0

C I d0 X
10 ) 44 <-U (U u O -

0 I ()0 r -H 0 4
UI Hr- 40 >r4 H
> 44 41 4-0 0 : () 0 44 Q
0) m .r r- O U 0) 0 -H
Cdl C0d -H o -0c)0-
4-4 u-I 0 0 HU 4-i C n (00
O -HI -0H C0 C d >34M0d r c
(0 U00U00 0 Z, Zr
U) W0 I QJsir-H-H 0- ;: 0 -H
S0 I 0) ( ) r -44 : rd u E 0.4,0
r. I Q m 5 ra. O q) 0
O C 7 rCQ L 0 DC--- 0O
C UI1 -4 O3 O i U I 4C ( 0 u

0 I
-o u o uuI
4-1 I *

0 I
*H I

-H I

U) I H

I 0) 0

I (-0 C)
Cd I t ,C
0u o0

C I )

0 I Q
C) I
I cJ
S W < C



m0 r



0 0)

,-4 rd
(d a)
0 0
>I 0

X: p
M -H

04 0

*H H
4-I 4-I

W4 Cd 0C
t4- K


mO I
0 -1

0r 0
04& 0 r

E a )

Cd co -
Q4 U) '

rI CN m n r-


0 -4

E m
O 0
r P4

S 0-
4 -rW


4-I) 02

0 4-
-4 a)

-P M
Ed W

0< a) 04
oZ Cd U a)
Z-lr ( ( n
0 Q S:l- &
H **





*> -r



a) -4

) U
aU f

r4 NCM



r-l 0

A 0


(1 0


W (zd 04W) 4J
-4 -( O rn -1 -
cd Cd k -d -- 0
-I r-4 i iC 1 H
.-4 M 2 00--I 02U
I 02 -H-n m(m 4r 4-
U +J U 4- H >.0 rq *0
m a i "A 4o4
l 4 m 104 -'- -1 r. O
1 0) 4 -Hl *H- 4 4 Z c a ) n (a)
CI >i>10 M a 02 a) o 4- 0 >i r
S1 r-I 4: Ei O U r(I Znj >1 p p O
I r 0 4-) p 4 C r i E4 a) 0 4O
r.3 o-ri a aO) > u i- >4 E 4J0-H U
U01 -P O E U -40 O 0 0 )-H U
r4 c4 r4 r4 c S4u c4 &< a4 i

I rim HcN r-4INrm H N

l I
0 I

Cd I0 U -
4) U) .d -, (.
- H o Cd0
0) I 0i H-H

S I > 0 0) E0

- I >1 -i -> 0 1 fo0 p
IJ 0*H H0 0 p

0 1 >1 >1 4 O Q)
I rCd > 10 C4

C r4 M Cd -H 4
-H~I O 0a)

3 u *a
r)1 Cd >1
01 C >i1, -H-HOW

( 4 1 >1 1-1r E-4i

aI H H Nti) H NC
I 0

I s r 0)
I Cd H -H

I rn C) a- -

0 I


and, as Table 1 implies, may occur in the noncleft popula-

tion for a variety of reasons. Some writers have initialed

the presence of VPI in the absence of a cleft palate as CPI

(congenital velopharyngeal incompetence) (Pruzansky, Peterson-

Falzone, Laffer, & Parris, 1977). For the purposes of this

study these cases will be called congenital VPI to distin-

guish them from individuals with VPI following a palatal


Surgical Correction of Velopharyngeal Insufficiency


Since normal speech sound production and voice quality

require a functioning velum and mobile pharyngeal walls,

the individual with VPI secondary to an organic factor or

factors will require some form of direct management to

achieve a functioning velopharyngeal mechanism. In these

cases speech therapy cannot correct VPI although it can re-

duce or eliminate the secondary components of VPI and is

frequently necessary following the physical management of

VPI. Currently, to correct the primary components of VPI,

there are essentially only two options available: surgery

and dental prosthetics. Dental prosthetics will not be con-

sidered in this study.

The techniques of surgically correcting VPI are desig-

nated as either secondary surgical procedures in cases of

previously repaired clefts of the palate or primary surgical

procedures in cases of congenital or acquired VPI. In actual

practice a given surgical procedure often can be used as a

primary procedure or as a secondary procedure or in combina-

tion with another operation in either a primary or secondary

surgical correction. Yules (1970) classified the surgical

procedures for the correction of VPI into four groups:

palatal pushbacks, pharyngeal flaps, pharyngoplasties, and

muscle transfers. He did not consider that pharyngeal wall

implants were true surgical procedures as they were "simpler

to perform" (p. 234). However, both Converse (1965) and

Hogan and Schwartz (1977) consider pharyngeal wall implants

as surgical procedures. Hogan and Schwartz (1977) simply

call implants augmentation pharyngoplasties. A pharyngo-

plasty is any surgical procedure designed to reduce the size

of the pharynx, and Converse (1965) and Hogan and Schwartz

(1977) consider muscle transfer procedures to be pharyngo-

plasties. The pharyngeal wall implant, as other surgical

procedures, involves admission of the patient to a hospital

and is typically done under general endotracheal anesthesia.

Therefore, for the purposes of this study, pharyngeal wall

implants will be considered as a surgical procedure.

In the review of the surgical and speech literature

that follows, the palatal pushback with pharyngeal flap pro-

cedure will be more extensively reviewed than either of the

two component operations, and the Teflon augmentation pharyngo-

plasty will be more extensively reviewed than will the earlier

types of pharyngoplasties which predate it.

The Pharyngeal Flap Procedure

The pharyngeal flap operation consists of elevating a

flap of muscle and overlying mucosa from the posterior

pharyngeal wall and inserting it into the soft palate.

Gustav Passavant is universally recognized as the first

surgeon to attempt to decrease hypernasality by surgically

joining the velum to the posterior pharyngeal wall (Millard,

1980). In 1876, Karl Schoenborn of Germany was the first

to describe the elevation of a caudally inferiorlyy) based

flap from the posterior pharyngeal wall to be sutured into

the velum. The idea apparently did not take hold until 1924,

when Wolfgang Rosenthal revived and "popularized" (Converse,

1965, p. 72) the inferiorly based pharyngeal flap procedure.

In 1930, Padgett introduced both the superiorly and

inferiorly based pharyngeal flap procedures in the United

States. He seemed to prefer the inferiorly based procedure

because it was technically easier to perform. Millard

(1980) states that both Schoenborn and Padgett, in their

time, came to prefer the superiorly based pharyngeal flap.

However, it is Sanvenero-Rosselli of Milan, Italy, who is

generally credited with popularizing the superiorly based

pharyngeal flap procedure (Hogan & Schwartz, 1977).

In 1966, Owsley, Lawson, Miller, and Blackfield re-

ported a modification of the superiorly based pharyngeal

flap which included elevating the flap as high toward the

cranial base as possible, splitting the velum down the mid-

line and elevating mucous membrane flaps from the nasal side,

suturing the pharyngeal flap into the midanterior aspect of

the velum and lining the raw surface of the pharyngeal flap

with the mucous membrane flaps from the velum to prevent

the flap from contracting and narrowing. This procedure

created lateral velopharyngeal apertures (lateral ports) at

a level in the nasopharynx more consistent with the level

of normal velopharyngeal closure and was thought by the

authors to "put to effective use the existing palatine elevator

function and allowed the palate to close at the point of con-

tact with the pharyngeal wall comparable to that seen in the

normal person" (p. 241).

In 1973, Hogan introduced the concept of lateral port

control, based, in part, on the work of Warren and Devereux

(1966). This technique takes care to surgically create

lateral ports of approximately 10 square millimeters each,

and this approach appears to be the current thinking when

using a pharyngeal flap alone as a secondary procedure to

treat VPI.

Palatal Lengthening Procedures

Surgical procedures designed to lengthen the velum re-

sulted from an awareness of the general anatomical shortness

of the velum following surgical repair. Millard (1980) notes

that early lengthening procedures (1800's) involved only the

velum leaving the defect in the hard palate either to be

closed in a subsequent operation or to be obturated with a

prosthesis. The concept of treating the velum and the hard

palate as essentially separate entities reached its peak

in the 1920's with the work of the English team of plastic

surgeon, Sir Harold Gillies and dentist, Sir Kelsey Fry.

The approach of Gillies and Fry (1921) aimed for "perfect

speech, perfect mastication; normal nasal respiratory func-

tion; and, in so far as a displaced maxilla causes deformity,

a normal bony contour" (pp. 335-336). Basically, their treat-

ment technique involved surgical reconstruction of the velum

after separating it from the hard palate, insertion of a

temporary dental device to prevent contraction of the velum

during healing, and finally, obturation, with a dental pros-

thesis, of the defect in the hard palate and the gap between

the hard palate and the velum. Gillies and Fry (1921) cri-

ticized the then current techniques of total palatal closure

as destructive of toothbuds and alveolar arches (Brophy tech-

nique), producing too much scarring and tight tissue (Lane

technique), and resulting in a short immobile velum (von

Langenbeck technique). Of these three procedures for re-

constructing both the hard and soft palates only the von

Lengenbeck, in modified form, survives. The Gillies-Fry

procedure was, itself, eventually discarded as more effec-

tive palatal pushback procedures were developed.

The von Langenbeck procedure is the oldest cleft palate

operation widely in use today (Lindsay, 1971), but it is not,

and never was a true lengthening procedure. Von Langenbeck's

major contribution was in demonstrating that mucuperio-

steal flaps could be elevated and moved on the oral side

of the hard palate without immediate detriment to the tissue

or to the patient.

Hogan and Schwartz (1977) state that lengthening of the

total palate began in the 1920's with Veau's modification

of the von Langenbeck procedure. Where von Langenbeck had

elevated and moved medially bipedical oral mucoperiosteal

flaps, Veau elevated and retrodisplaced oral and nasal mucu-

periosteum, emphasizing the need to always cover raw sur-

faces whenever possible, for closure of the hard palate.

In 1925, Dorrance described a palatal lengthening proce-

dure in which he elevated large U-shaped mucoperiosteal flaps,

divided the palatine aponeurosis at the posterior border of

the heard palate, fractured the hammular process to free the

tensor palatine, divided the posterior neurovascular bundle,

and retropositioned the entire soft tissue mass. In a com-

prehensive treatise on congenital velopharyngeal insuffi-

ciency (VPI in the absence of an overt cleft ot the palate),

Dorrance (1930) refined his technique introducing a horse-

shoe-shaped circumalveolar incision (see Figure 2) and pro-

nounced that his research had "solved the problem of correct-

ing congenital insufficiency of the palate" (p. 248). In

1932, Dorrance coined the term "push-back" (p. 1228) to

describe his operation and advocated its use in all cases

of short palate either congenital or postoperative. He re-

treated from his 1930 statement acknowledging that the push-

back was not the "last word in cleft palate surgery" (p. 1228)

and noting that it would undoubtedly be improved upon.

shaped incision

Postoperative view

Figure 2. The Dorrance palatal pushback.


In 1937, two English surgeons, William Wardill and Pom-

fret Kilner, working independently, revised the palatal

lengthening procedure described by Veau. The Wardill-Kilner

operation, known as a V-Y advancement or retropositioning,

like the Dorrance technique, remains in use today. Figure 3

shows the basic outline of this surgical procedure. The two

surgeons differed in that Wardill (1937) cut the neurovascular

bundle at the greater palatine foramen and also performed a

pharyngoplasty by making a transverse incision in the pos-

terior pharyngeal wall and suturing it closed vertically.

Kilner did not cut the neurovascular bundle and did not con-

sider that a pharyngoplasty was necessary in all cases, and

Kilner's technique is more popular today than that of Wardill

(Millard, 1980). Peet (1961) further refined the Wardill-

Kilner procedure, and it has subsequently been called the

Oxford technique.

A concern expressed about the methods of Dorrance and

Wardill revolved around their cutting of the neurovascular

bundles which required that the operation be performed in

at least two stages. After the neurovascular bundles were

cut, several weeks were allowed to pass before the rest of

the operation was carried out in order to allow time for

collateral circulation to develop in the mucoperiosteal

flaps. It was well known that the fixed neurovascular

bundles coming from their bony foramina had a restraining

effect against successful pushback procedures.

Millard (1980) reveals that in 1927, Alexander Limberg

of Leningrad decided that bone was more expendable than blood


Postoperative view

Figure 3. The V-Y palatal pushback.


supply and advocated freeing the neurovascular bundle from

its bony canal using an osteotomy and moving the bundle to-

gether with the mucoperiosteum posteriorly and medially.

Brown (1936) suggested that the Dorrance pushback could be

done in a single stage by preserving the major palatine ar-

teries. He attempted to elongate them by "carefully stretch-

ing them from their foramina and slightly separating them

from the raw surface of the palate" (p. 769). Conway (1947)

felt that, in many cases, retropositioning of the velum could

not be obtained by the method described by Brown (1936). He

advocated a technique similar to Limberg's: removing a small

portion of the posterior wall of the pterygopalatine canal,

thus freeing the neurovascular bundle so that it could be

moved posteriorly with its mucoperiosteal flap.

Broadbent and Hochstrasser (1959) noted that palatal

paresis had never been observed following the cutting of

the greater palatine neurovascular bundle. However, they

observed that for two 20-month old patients who had had

the neurovascular bundle sectioned on the left side only,

the velum was longer on the unsectioned (right) side when

measured 112 days postoperatively. Further, these investi-

gators found decreased secretary activity on the sectioned

side of the velum. From these observations they concluded

that the neurovascular bundle "should be preserved to avoid

atrophy of mucous glands, shortening of the palate and de-

creased speech efficiency" (p. 614).

Summarizing the current thinking on the question of

whether or not to cut the neurovascular bundles, Millard

(1980) wrote:

Although the neurovascular bundles do tend to
tether the soft palate to the hard palate, there
seems to be no real justification for dividing
them. It is not likely that their division will
be catastrophic for either blood supply or growth.
Yet if they can be freed from the foramen or the
flap, and if necessary the restricting posterior
wall of the foramen can be removed, there is no
need to divide them. It takes a little more time
and skill, but these vessels can be of great value
under certain circumstances. If they have been
destroyed at leisure, they are alas no longer
available for important duty in an emergency!
(p. 472)

The presumed reason why the early palatal lengthening

procedures often failed to result in velopharyngeal closure

adequate for normal speech and voice production is that the

raw area left on the nasal surface by the pushback contracts

and pulls the velum forward to its preoperative position,

thus negating the intent of the surgery. To overcome this

problem Padgett (1930) advocated lining the raw upper sur-

face of the palate with flaps of cheek tissue. Dorrance

and Bransfield (1943) reported using a split-skin graft to

line the palatal mucoperiosteum flap. In 1957, Cronin used

flaps of mucosa taken from the floor of the nasal cavity to

cover the raw nasal surface of the mucoperiosteal flaps

during pushback procedures. Millard (1962) took portions

of one of the mucoperiosteal flaps containing the neuro-

vascular bundle and flipped it over to cover the raw surface

of the retropositioned velum. He called the technique an

"island flap for nasal lining" (p. 49). None of these ap-

proaches appears to have gained any outstanding popularity

outside of their originators.

Combined Pushback with Pharyngeal Flap

In 1950, Marino and Segre described combining an in-

feriorly based pharyngeal flap with a Dorrance type pushback.

Their objective was to anchor the pushback during healing

to prevent contracture, and they suggested that the pharyn-

geal flap be divided after healing was complete.

Conway (1951) elevated either a superiorly or inferiorly

based pharyngeal flap and attached it to either the nasal or

oral side of the velum respectively, following a V-Y type

pushback. He reasoned that the healing of the raw area at

the base of the pharyngeal flap would counteract the healing

of the raw area on the nasal aspect of the retropositioned

velum and thus there would not be contracture.

In 1955, Conway and Stark reported on 81 patients, 30

of whom had been operated on using a pushback with an in-

feriorly based pharyngeal flap. The remaining 51 were

treated with a pharyngeal flap alone. The authors reported

that of the total of 81 patients 79% exhibited "normally ac-

ceptable speech" (p. 670).

Conway and Goulian (1960) reported that 92 patients

achieved "acceptable speech" (p. 594) out of 98 treated

with a pushback and a pharyngeal flap either inferiorly (97

cases) or superiorly (1 case) based. They also reported no

postoperative complications in any of the 98 patients.

Millard (1962) criticized Conway's (1951) procedure

stating that "two raw wrongs do not make a right" (p. 49).

His major criticism was that the pharyngeal flap anchored

the free mobility of the velum.

In 1965, Dibbell et al. described a modification of the

combined pushback with pharyngeal flap operation. They

elevated a long superiorly based pharyngeal flap for use as

a lining and sutured it to the nasal side of the velum in

such a manner as to cover the raw surface of the retroposi-

tioned velum (see Figure 4). They used a Dorrance pushback

in cases of velar involvement only and recommended a modi-

fied V-Y pushback in cases of combined hard and soft palate

clefts. They described the advantages of their technique

as follows:

(1) The raw nasal defect is resurfaced, minimiz-
ing postoperative scarring with shortening
and narrowing.

(2) There is no tethering of the palate because
the flap does not attach at the uvular margin
and therefore cannot impede free motion of
that part.

(3) The flap acts as a horizontal buttress against
which the soft palate can brace and unroll,
thereby gaining stability and some additional
posterior length.

(4) Anterior placement of the flap on the palate
may aid the action of the elevator palatii
muscles. (p. 169)

Having performed the procedure on 13 patients, the authors

reported 9 were "considerably improved" (p. 170) and 4 were

"only somewhat improved" (p. 170). They reported no major

complications from the procedure but noted that most pa-

tients have hyponasality for some time after surgery.

Figure 4. The combined palatal pushback with pharyngeal
flap (after Dibbell et al., 1965).

Using the operative method described by Dibbell et al.

(1965), Buchholz, Chase, Jobe, and Smith (1967) reported on

a series of 47 patients. In this series 26 patients had

little or no detectable nasal escape postoperatively and 20

had improved speech but were too early postoperatively for

final assessment. One patient was unimproved. No major

complications were reported. Some of the patients developed

serious otitis and in others rhinitis and excessive naso-

pharyngeal mucous were observed in the first several weeks

postoperatively. Many patients experienced pain and stiff-

ness in the neck which, generally, lasted less than a week.

The patients in this series had been followed over a range

from six months to four years and the authors reported that

speech improvement had been noted in several patients be-

tween the first and second postoperative years.

Kaplan (1973) reported that since the article by Buch-

holz et al. (1967), an additional 125 procedures had been

performed. He stated that the patients' speech was rated

on a 5-point scale and that most patients had improved about

2 points on the scale, postoperatively. Further, he wrote

that "we have no evidence that the pharyngeal flap actively

pulls the palate toward the pharynx, necrotizes the palate

muscles, or is an effective obturator" (p. 207). He thus

retracted several of the original suggested advantages of

this procedure.

Dixon, Bzoch, and Habal (1979) compared preoperative

and postoperative speech evaluations and several other

variables on 15 subjects who underwent a combined pushback

with pharyngeal flap procedure in the manner described by

Dibbell et al. (1965). They reported no difference in out-

come related to the subjects' age, sex, etiology of the VPI

(diagnosis), or whether the operation was performed as a

primary or secondary procedure. Elimination of nasal emis-

sion was achieved in 13 (87%) of the subjects. Ten of the

subjects (67%) had hypernasality completely eliminated and

three of the remaining five had only very minimal hyper-

nasality which was "not considered significant" (p. 80).

Twelve of the 15 (80%) showed no evidence of hyponasality

at the time of their postoperative evaluation, which averaged

10.7 months (range = half month to 24 months). The authors

concluded that the amount of time from the surgery to the

postoperative speech evaluation is not an important factor

once the postoperative reaction has subsided.

In summary, the combined pushback with pharyngeal flap

procedure has undergone several major modifications in the

past three decades. The method described by Dibbell et al.

(1967) appears to have gained wide acceptance and has proved

to be a safe procedure with a speech success rate of ap-

proximately 81% (see Table 2).

Augmentation Pharyngoplasties

The surgeon Rutenberg is credited with being the first

(1876) to attempt to narrow the nasopharynx to aid in velo-

pharyngeal closure (Converse, 1965; Millard, 1980). He

simply made a transverse incision in the posterior pharyngeal

wall and closed the wound longitudinally. In 1878, Passavant

elevated a flap from the posterior pharyngeal wall and folded

it on itself to create a shelf-like projection. However,

the shelf flattened out over time (Millard, 1980). Wardill

(1928) advocated the use of a pharyngoplasty, similar to

that described by Rutenberg, as the first stage in any palate


In 1950, Hynes described raising two vertical mucomus-

cular flaps, each containing salpingopharyngeus muscle, and

rotating them 90 degrees to be sutured into a horizontal de-

fect created across the posterior pharyngeal wall of the

pharynx. Hynes clarified his technique in 1967, emphasizing

that the lateral pharyngeal flaps were to be elevated from

the entire width and thickness of the lateral pharyngeal

walls, and not from the lateral margins of the posterior

pharyngeal wall as, apparently, some of his imitators had


The relative simplicity of narrowing the nasopharyn-

geal isthmus, in contrast to lengthening the palate, has

led many surgeons to investigate the feasibility of implant-

ing materials, other than adjacent muscle, into the poste-

rior pharyngeal wall to correct VPI. Gersuny in 1900,

using soft paraffin and Eckstein in 1904, using hard paraf-

fin, are credited with the first implants in the posterior

pharynx (Converse, 1977; Millard, 1980; Yules et al., 1971).

Cartilage augmentation pharyngoplasty. Hagerty, a sur-

geon, pioneered the augmentation pharyngoplasty concept in

the United States (Millard, 1980); however, Hagerty and Hill

(1961) credited a Russian, Lando, with the first homogenous

graft using the transoral route in 1950. Hagerty and Hill

(1961) used a transoral approach to implant both homogenous

and autogenous cartilage grafts in 20 patients with VPI.

They reported improved speech and retention of three-fourths

of the grafts at one year postoperatively. In 1969, Hagerty,

Mylin, and Hess reported on a series of 64 cases using viable

homogenous cartilage to correct VPI in postoperative cleft

palate patients. Complete speech records were only avail-

able on 31 of the patients,and the authors concluded, with-

out providing data, that the procedure had a beneficial ef-

fect on speech. No cartilage graft was extruded; however,

in three cases there was complete absorption of the cartilage

within the first year. Millard (1980) writes that difficulty

in maintaining a cartilage bank caused Hagerty to change to

silicone in 1971 for his more recent augmentation pharyngo-


Silicone augmentation pharyngoplasty. In 1960, Brown,

Fryer, and Ohlwiler reported on the use of a variety of

synthetic materials that demonstrated promise for use as

subcutaneous prostheses. They concluded that the silicones

and Teflon were superior to other materials in that they

were easier to procure, more easily stored, could be shaped,

caused minimal tissue reaction, did not tend to erode,

calcify or bend, were less expensive, and could be re-

peatedly sterilized. However, they cautioned that opera-

tions using these materials "are new and difficult; they

possibly may not stand the test of time" (p. 227).

The use of silicone has, apparently, not stood the

test of time. Blocksma (1963) described a series of 23 pa-

tients with VPI who were treated with various forms of sili-

cone implanted or injected into the posterior nasopharynx

above the level of the tubercle of the atlas. One patient

was treated with a sponge implant, four with solid implants,

nine with liquid injections, and nine with shredded implants.

The author reported an "excellent" (p. 272) result in four

patients and a "good" (p. 272) result in 13 others. Some

shredded Silastic was extruded in two patients and the

single sponge implant became infected and had to be removed.

Blocksma discontinued the use of shredded and sponge Silastic

as implant materials as a result of this experience.

In 1969, Blocksma and Braley reported on the current

status of implant pharyngoplasty based on information ob-

tained from a sample of mailed questionnaires to plastic

surgeons in the United States. They found that few sur-

geons had developed a "sound method" (p. 243) for select-

ing patients for pharyngeal wall implants and that the

preoperative and postoperative speech data often reported

did not allow for the comparison of the effectiveness of

pharyngeal wall implants with other surgical procedures be-

cause of a lack of "commonly accepted" (p. 246) evaluation

methods. In a summary of the long term retention rates of

pharyngeal wall implants, these authors found that 50% of

the solid silicone implants were lost either by erosion or

absorption, 20% of the injected silicone implants were

lost, and 0% of the injected Teflon implants were lost.

Teflon augmentation pharyngoplasty. Teflon (tetra-

fluorocarbon) is an inert, long chain polymer (plastic) de-

veloped by the Du Pont Company. Arnold (1962) reported that

an early use of Teflon powder was as a lubricant in machin-

ery. Teflon is probably universally recognized as a non-

stick coating on cookware.

Arnold (1962) and Lewy (1963) both reported the use of

a Teflon-glycerine mixture injected into vocal cords to im-

prove phonation. Both authors had also used a tantalum

powder-glycerine mixture for the same purpose (Arnold, 1961;

Lewy, 1963). However, the Teflon-glycerine mixture was re-

ported as easier to use and capable of being prepared in ad-

vance and, thus thought to be preferable to tantalum (Lewy,


Ward and Wepman (1964) studied the effects of injecting

Teflon-glycerine and tantalum-glycerine into the posterior

pharyngeal walls of cats. They found that both were equally

effective in creating and maintaining an artificial Passa-

vant's ridge. They also found that there was greater tissue

reaction to larger Teflon particles (50-100 microns) than to

smaller particles, and they concluded that this tissue reac-

tion might have the beneficial effect of helping to stabilize

the Teflon implant through greater stromal deposition (the

natural process of creating a framework of connective tis-

sue). They stated that the simplicity and safety of the

procedure "commends its cautious trial in human patients

with velopharyngeal insufficiency" (p. 454).

In 1965, Lewy, Cole, and Wepman reported the first use

of injected Teflon in a human. The velopharyngeal space of

a 15-year-old female was reduced using the technique (see

Figure 5) and the treatment "effectively eliminated nasal

emission of air and hypernasality" (pp. 878-879).

Ward, Goldman, and Stoudt (1966) reported a pilot study

in which Teflon injection was carried out on five patients

with severe VPI. They reported "normal speech" (p. 270) in

two patients who had had two injections and "good improve-

ment" (p. 270) in two other patients who had had a single

injection. The other patient in the study was not de-

scribed. These authors reported no significant complica-

tions with the procedure, and they concluded that the tech-

nique "appears to offer real promise" (p. 272) in the manage-

ment of VPI.

In 1967, Ward, Stoudt, and Goldman reported on the re-

sults of a series of 18 patients who had received Teflon

injections for VPI. They reported no major complications

from the procedure. Several of the patients had mild swell-

ing and pain in the neck; others had an elevated tempera-

ture (102 degrees Fahrenheit) postoperatively. These com-

plaints subsided quickly without sequelae. Patients had

Injected Teflon

Figure 5. The Teflon augmentation pharyngoplasty (after
Lewy et al., 1965).

been followed for two years, and the authors reported an

"excellent" or "good" (p. 929) result in 14 of the 18 (78%)

cases. "Excellent" was defined as "normal speech" (p. 929),

and "good" was defined as "demonstrated 100% closure by

manometric measurements with marked improvement in speech

but with persistence of severe articulatory abnormalities"

(p. 929). They concluded that the most suitable cases for

Teflon injection were individuals with VPI secondary to

adenoidectomy or to the presence of a submucous cleft palate.

They suggested that the least favorable results would be

obtained in patients with severe muscular palsy or paresis

of the palatal musculature. They recommended against the

procedure in mentally retarded individuals and in cases with

severe congenital defects whose VPI was accompanied by a

severe articulation disorder.

In 1968, Bluestone, Musgrave, and McWilliams reported

on a four-year series of 27 patients who had had a Teflon

injection procedure. The patients had been followed for

two to 35 months postoperatively with 20 of the patients

having been followed over 12 months. The authors reported

elimination of hypernasality in 21 (78%) of the cases, five

of which required reinjection. They also reported no

serious complications from the procedure, but many of their

patients had postoperative complaints similar to those re-

ported by Ward et al. (1967). The authors concluded that

the procedure was indicated in patients "with good elevator

activity and a small velopharyngeal gap" (p. 563).

Sturim and Jacob (1972) described a series of 23 pa-

tients treated by Teflon injection and followed for one

year postoperatively. Three of the patients were treated

following unsatisfactory pharyngeal flap procedures. The

remaining 20 had repaired clefts of the palate or congenital

VPI of varying etiologies. In two cases there were "signifi-

cant" (p. 182) postoperative complications reported. One

case had a severe stiff neck persisting three weeks and re-

quiring the wearing of a supportive collar. Another case

had a severe stiff neck present for two and one-half months,

requiring additional hospitalization for orthopedic traction.

Both cases had complete remission of their postoperative

symptoms. Of the 20 patients without a previous pharyngeal

flap procedure, the authors reported a "good result (verbal

communication which did not call attention to itself)"

(p. 182) in 12 (60%) cases, and seven cases (35%) were de-

scribed as "improved (hypernasality decreased but

some residual degree of distortion was still apparent)"

(p. 182). One patient was unchanged following the proce-

dure. Several advantages of the Teflon injection procedure

were cited including limited morbidity, immediate changes in

verbal communication (in many cases), ability to easily re-

peat the procedure (if necessary), less surgical trauma,

short hospital stay, less cost to the patient, and no re-

ported deaths. The authors felt that the thickness of the

Teflon pad did not appear to change over the course of one

year, and they emphasized that careful patient selection

was necessary to insure a high rate of success. They treated

patients with velopharyngeal gaps no greater than 1 cm.

In 1977, Smith and McCabe reported on a series of 80

patients treated by Teflon injection. Subjects were selected

who had good palatal movement and a velopharyngeal gap of

5 mm or less. An additional nine subjects with failed

pharyngeal flaps were described. The authors reported no

complications other than a mild-to-moderate stiff neck for

one week postoperatively. Of the 80 patients with VPI, 60%

were described as having "good voice quality" (p. 562) and 18%

were "improved" (p. 562).

Kuehn and Van Demark (1978) analyzed the preoperative

and postoperative speech and x-ray data on 69 patients who

had undergone a Teflon pharyngoplasty. They concluded that

speech and velopharyngeal function improved over time and

that final evaluation of the procedure should not be under-

taken until at least three months following the Teflon in-

jection. Further, they stated that patients with either

very mild or severe VPI were poor candidates for the Teflon

procedure. They summarized their findings by stating that

patients with mild-to-moderate VPI, a gap of 2 mm or less

and good palate mobility are "excellent candidates for

treatment with the Teflon injection procedure" (pp. 148-


Furlow, Williams, Eisenbach, and Bzoch (1982) reported

on a series of 35 patients (8 with failed pharyngeal flaps)

who had been followed postoperatively for an average of

three years. These authors reported that all of their pa-

tients reported a sore throat and stiff neck postoperatively

and several ran a low grade fever. One patient extruded

the injection Teflon two weeks after the procedure. One pa-

tient demonstrated some of the symptoms of malignant hyper-

thermia but recovered without sequelae. One patient de-

veloped objectionable snoring and obstructive sleep apnea

postoperatively. Later resection of the Teflon pad in-

ferior to the atlas relieved the patient's symptoms without

altering the velopharyngeal competence in this patient.

These investigators judged the injection procedure as a

success only if there was "total elimination of hypernasal

voice quality and total elimination of inappropriate nasal

air emission during speech" (p. 52). In this series 21 of

the 27 patients (78%) with VPI were successfully treated, and

five of the eight (62%) cases of failed pharyngeal flap were

successfully treated. These authors also reported a higher

success rate overall in children under 11 years (88%) as

compared to those patients over 12 years (60% success rate).

Using serial postoperative cinefluoroscopic analysis, they

also demonstrated that the Teflon implant thickness remains

stable over time. They did not treat any patient with a

velopharyngeal gap greater than 1 cm and felt that the size

of the velopharyngeal gap (1 cm or less) did not appear to

be critical to a successful outcome.

In summary, the Teflon augmentation pharyngoplasty ap-

pears to be a safe procedure with a speech success rate of

approximately 67% (see Table 2). However, despite the wide

reporting of this procedure, the underlying concept of the

augmentation pharyngoplasty does not appear to have gained

wide acceptance in the surgical community.


Table 2 summarizes the reported rates of operative suc-

cess in terms of speech outcome of the major series reviewed

under the combined pushback with pharyngeal flap and the

Teflon augmentation pharyngoplasty. Only the various authors'

strictest definition of success (when reported) was used to

compile Table 2. It is difficult, if not impossible, to

translate the various criteria used to define a successful

outcome across investigators. Rarely is there across study

similarity in the definition of the terms "excellent, good,

or acceptable" speech. Judgements of improvement are often

made on rating scales which would be impossible to duplicate

from study to study. Despite these inadequacies, the over-

all rates of success for these two procedures appear to be

close, with the combined pushback with pharyngeal flap being

somewhat more successful. In viewing Table 2 one must keep

in mind that the early work of Conway and his associates was

a significantly different operation from that used by Dibbell

et al. (1965) and later writers. This might explain the

greater variability in the reported rates of success for

the combined pushback with pharyngeal flap as compared to

the Teflon procedure, which has remained fairly constant, in

terms of operative technique, since it was introduced.

Table 2
Speech Success after Combined Pushback with Pharyngeal Flap
and after Teflon Augmentation Pharyngoplasty

Combined Pushback with Pharyngeal Flap

Year Author No. Cases Total % Success

1955 Conway & Stark 30 24/30* 79%

1960 Conway & Goulian 98 92/98 94%

1965 Dibbell et al. 13 9/13 69%

1967 Buchholz et al. 47 26/47 55%

1979 Dixon et al. 15 13/15 87%

N = 203 164/203 = 81%

Teflon Augmentation Pharyngoplasty

Year Author No. Cases Total % Success

1967 Ward et al. 18 14/18 78%

1968 Bluestone et al. 27 21/27 78%

1972 Sturim & Jacob 20 12/20 60%

1977 Smith & McCabe 80 48/80* 60%

1982 Furlow et al. 27 21/27 78%

N = 172 116/172 = 67%

= Number of successes inferred from the original
article. Actual number was not stated.

In selecting suitable candidates for either procedure

there is general agreement that velar mobility is an im-

portant variable and that younger patients appear to be

treated more successfully than older patients. When an

opinion was stated, it was more often suggested that the

evaluation of an outcome of a combined pushback with pharyn-

geal flap should not be judged for at least a year postopera-

tively. Dixon et al. (1979) disagreed and stated that so

long as the postoperative reaction had subsided, it did not

appear to matter when the speech outcome was evaluated post-

operatively. In the Teflon series, when an opinion was

stated, it was suggested that evaluation of the speech out-

come could be safely made approximately three months post-

operatively. When discussed, sex, etiology of the VPI, or

whether the procedure was a primary or secondary operation

did not appear to influence the speech outcome of either


It seems reasonable to conclude from this review of

the literature that when similar criteria are used for pa-

tient selection, the combined pushback with pharyngeal flap

and the Teflon injection augmentation pharyngoplasty are

both effective and similar, in terms of success rate, with

respect to speech outcome. They also appear to be safe

procedures, in terms of morbidity, to correct VPI either

secondary to a repaired cleft palate or existing as a

primary complaint.

Comparison Studies of Secondary Surgical Procedures

A search of the literature reveals that comparative

studies in the entire field of cleft palate are rare. Com-

paring secondary procedures is rarer still, and the intro-

duction of statistical control in comparative studies has

been found in only three studies.

In 1962, Williams and Woolhouse descriptively compared

the pharyngeal flap (both inferiorly and superiorly based)

and the Hynes (1950) pharyngoplasty (one of the earliest

secondary pharyngoplasty procedures) on 24 consecutive pa-

tients with VPI following repair of clefts of the palate.

These two procedures were divided evenly between the patients.

The authors' reported "consistently better results" (p. 41)

were obtained with the pharyngeal flap (83% excellent or

good speech) than were obtained with Hynes pharyngoplasty

(67% excellent or good speech).

Bzoch (1964) described several groups of patients with

VPI treated either with a speech appliance or by a pharyn-

geal flap procedure. He reported correction of VPI in

92.6% of the patients treated with a speech appliance and

in 95.5% of the patients treated with a pharyngeal flap

procedure. He noted that, regardless of the approach

chosen to treat VPI, speech therapy is often necessary be-

fore the potential benefits of the treatment can be ob-


Whitaker et al. (1972) reported a prospective and ran-

domized series which compared the superiorly and inferiorly

based pharyngeal flap. They found no significant difference

in speech outcome, hearing acuity, short and long term com-

plications, or length of hospital stay between the two pro-

cedures. Interestingly, this study covered a five year

period during which only 35 out of 142 patients who under-

went the pharyngeal flap procedure met the criteria for in-

clusion in the study which included "the anatomic possibility

of equal choice between a superiorly and inferiorly based

flap" (p. 305). To this writer, it would seem more reason-

'able to conclude that if your selection criteria are strin-

gent enough, then you will not observe differences between

the inferiorly and superiorly based pharyngeal flap proce-

dures, rather than to have concluded that there were no sig-

nificant differences in the outcomes between the two proce-


Although not a comparative study, Shprintzen, Lewin,

Croft, Daniller, Argamaso, Ship, and Strauch (1979) demon-

strated that the outcome of secondary procedures incorporat-

ing a pharyngeal flap could be improved by assigning pa-

tients with VPI to procedures according to the degree of

lateral pharyngeal wall motion seen preoperatively.

Schprintzen et al. (1979) emphasized the previously stated

concerns (Chapter I) for more objective and quantifiable

data regarding preoperative evaluation and development of

objective criteria for selecting a patient for a given pro-

cedure over other available procedures. This is a notable

study as it attempted to define criteria for choosing the

most appropriate surgical procedure on a case by case basis.

In 1980, Hathaway retrospectively compared two groups

of subjects postoperatively, one with vertical pharyngeal

flaps and the other with transverse pharyngeal flaps, on per-

ceived nasality and EMG (electromyographic) activity in the

pharyngeal flaps. No significant differences on either vari-

able were identified relative to either of these two tech-


Marsh and Wray (1980) randomly assigned 39 patients

with moderate-to-severe VPI to groups treated by either a

pharyngeal flap procedure or a prosthetic speech appliance.

The final speech evaluation ranged from six months to three

years after entry into the study. They found that prosthetic

management and pharyngeal flap surgery were equally effective

in treating VPI. Further, they noted that the cost of pros-

thetic management was only about 40% that of surgery in 1977

(when the study began); however, there was a 29% noncompli-

ance rate in those patients selected for prosthetic manage-

ment. No patient selected for surgical treatment failed to


The other randomized study in the field of cleft palate

or VPI referred to earlier (Wray et al., 1979) compared the

in-hospital morbidity of three different techniques of pri-

mary palate reconstruction and is not germain to this thesis.

In this review of the small number of comparative studies

on secondary procedures, one could still agree with Yules

(1970) who a decade ago, after reviewing the literature on

secondary procedures for correcting VPI, concluded that

"which technique will prove most effective in each of these

patients is a point which remains to be determined" (p. 242).

It is also evident that there still have not emerged any uni-

versally accepted rationales for selecting a particular treat-

ment technique over others, nor is there any consistency in

defining or judging an acceptable treatment outcome. It

would appear that these issues will remain unresolved unless

professional associations, supported by their members, offer

the appropriate guidelines or until the randomized clinical

trial becomes the universally accepted method of assessing

treatment methods and procedures.

Research Questions and Hypotheses

The review of the literature has revealed that, al-

though the patient selection criteria have been somewhat

similar in the reported series, the pushback with pharyngeal

flap and the Teflon injection augmentation procedures have

never been directly compared as to speech outcome or on any

other quantifiable variables. The need for more objective

and quantifiable preoperative and postoperative evaluation

protocols has also been stressed. If, in fact, the speech

outcomes are similar regardless of the technique employed

then other variables of importance to the patients, their

families, and individuals making management decisions should

be investigated.

At the University of Florida's J. Hillis Miller Health

Center there exist records on a number of patients who have

had VPI treated by either the pushback with a high based

pharyngeal flap or by the injectable Teflon augmentation

procedure. The essential factor that makes these two groups

candidates for analysis is the fact that the same criteria

were used to select either of the surgical procedures.

These criteria are the following:

(1) the existence of VPI during connected speech
as evidenced by the presence of a hypernasal
voice quality and by inappropriate nasal emis-
sion during the production of pressure conso-

(2) a mobile soft palate during connected speech
as demonstrated by cinefluoroscopy

(3) movement pattern of the soft palate appropri-
ate to the speech samples employed, as demon-
strated by cinefluorographic analysis, and

(4) a residual velopharyngeal gap of one centi-
meter or less at the individual's optimal at-
tempt at closure during speech, as measured
from cinefluoroscopy.

Generally, patients selected for either of the surgical pro-

cedures did not demonstrate the habituation of gross articu-

lation errors glottall stops or pharyngeal friccatives) as

these habit patterns could potentially affect both the clini-

cal testing for hypernasality and nasal emission and the

cinefluoroscopic analysis. Patients who met the above cri-

teria and were recommended for surgical management were done

so without regard to age, sex, or etiology of the VPI.

In the late 1960's the J. Hillis Miller Health Center

was selected as a test site for the use of injectable Teflon

to treat VPI. Patients who met the above criteria were

selected for treatment using the injectable Teflon

augmentation pharyngoplasty. When injectable Teflon did not

gain final Federal Drug Administration (FDA) approval for

use as an implant in the retropharyngeal space, similar pa-

tients were treated using a combined pushback (either a Dor-

rance or a V-Y retropositioning procedure) with a high-based

pharyngeal flap. Thus, chronology was the only factor which

distinguished the inclusion of a given patient to either of

the two groups.

A pilot study in which five individuals were randomly

selected from the pool of patients who had undergone either

of the two surgical procedures revealed no significant dif-

ferences between the two groups in the preoperative con-

sistency of their hypernasality or in the preoperative con-

sistency of their inappropriate nasal emission. This lack

of significant difference in speech and voice symptoms to-

gether with the fact that each patient met the other criteria

for selection for surgical treatment suggested that the pa-

tients in both groups came from the same clinical population.

If there is a clinical population that can be effec-

tively treated by either a combined pushback with a high-

based pharyngeal flap or by a Teflon augmentation pharyngo-

plasty, then there is justification for comparing the two

surgical procedures on a variety of parameters in order to

determine if either procedure merits consideration over the

other on factors other than speech outcome.

The strategy to be employed in carrying out this study

is as follows. First, demonstration of the preoperative

equality of the two groups of patients in terms of age, sex,

etiology of their VPI, and consistency of their speech and

voice symptoms will be undertaken. If areas of inequality

are uncovered, any influence these may exert on the vari-

ables directly related to the surgical procedures will be

discussed in the appropriate section of this thesis. Second,

data will be gathered and subjected to statistical analysis

in order to validate the two surgical procedures using pre-

operative and postoperative cinefluoroscopic measurements,

that is, to objectively demonstrate that each of the two

surgical procedures does what it is intended to do. Only

one investigation (Cronin, Brauer, Alexander, & Taylor, 1964)

has objectively attempted to demonstrate that their surgical

technique does retroposition the velum. No investigator has

objectively demonstrated that the Teflon augmentation pharyngo-

plasty does, in fact, decrease the depth of the nasopharynx

(the presumed intent of the procedure). Furlow et al. (1982)

did demonstrate that the thickness of injected Teflon remains

stable over time. However, Da Pra (1981), in an unpublished

Master of Arts thesis, using data gathered from the same pa-

tients studied by Furlow et al. (1982), concluded that pre-

operative and postoperative measurements of the depth of

the nasopharynx were unchanged in the nine patients she

studied. Apparently, most investigators are content to let

the postoperative speech results infer that the physiologi-

cal objective of the operation has been achieved when it

could be easily argued that, in the absence of demonstrating

that the surgical procedure's physiological objective has

been achieved, any number of other variables might account

for changes in a patient's speech and voice symptoms. Fi-

nally, the two surgical procedures will be directly compared

on the following variables: speech outcome, duration of

the operation in minutes, length of hospital stay in days,

morbidity, and effects on velar rise rate, velar stretch,

and velar height above or below the palatal plane during

closure or attempted closure.

To carry out this strategy the following research ques-

tions have been generated together with appropriate hypotheses

for answering the questions.

(1) Do the two groups of patients to be studied come
from the same clinical population?

Null hypothesis (Ho): The average age of the
patients does not differ between the two groups.

Alternate hypothesis (Ha): The average age of
the patients does differ between the two groups.

Ho: The ratio of males to females does not dif-
fer between the two groups.

Ha: The ratio of males to females does differ
between the two groups.

Ho: The etiologies (diagnosed causes) of the VPI
do not differ between the two groups.

Ha: Theetiologies of the VPI do differ between
the two groups.

Ho: The average preoperative consistency of the
patients' hypernasal voice quality does not dif-
fer between the two groups.

Ha: The average preoperative consistency of the
patients' hypernasal voice quality does differ
between the two groups.

Ho: The average preoperative consistency of the
patients' nasal emission does not differ between
the two groups.

Ha: The average preoperative consistency of the
patients' nasal emission does differ between the
two groups.

Ho: The average preoperative frequency of gross
articulation errors does not differ between the
two groups.

Ha: The average preoperative frequency of gross
articulation errors does differ between the two

(2) Do the surgical procedures accomplish what they
are designed to accomplish?

Ho: Within the group treated by a combined push-
back with a pharyngeal flap preoperative and
postoperative measurements of velar length do
not differ.

Ha: Within the group treated by a combined push-
back with a pharyngeal flap preoperative and
postoperative measurements of velar length do
differ with the postoperative measure being

Ho: Within the group treated by injectable Teflon
augmentation pharyngoplasty preoperative and post-
operative measurements of the depth of the naso-
pharynx do not differ.

Ha: Within the group treated by injectable Teflon
augmentation pharyngoplasty preoperative and post-
operative measurements of the depth of the naso-
pharynx do differ with the postoperative measure
being smaller.

(3) Are the two surgical procedures equally effective
in eliminating hypernasality and nasal emission?

Ho: The frequency of persisting postoperative
hypernasal voice quality does not differ between
the two groups.

Ha: The frequency of persisting postoperative
hypernasal voice quality does differ between the
two groups.

Ho: The frequency of persisting postoperative
nasal emission does not differ between the two

Ha: The frequency of persisting postoperative
nasal emission does differ between the two groups.

(4) Does either surgical procedure induce persisting
hyponasality postoperatively?

Ho: The frequency of postoperatively induced
hyponasality does not differ between the two groups.

Ha: The frequency of postoperatively induced
hyponasality does differ between the two groups.

(5) Are the two surgical procedures similar with re-
spect to operation time, length of hospital stay,
and morbidity?

Ho: The average time, in minutes, required to
perform each of the two operations does not differ.

Ha: The average time, in minutes, required to
perform each of the two operations does differ.

Ho: The average patient hospital stay, in days,
does not differ between the two groups.

Ha: The average patient hospital stay, in days,
does differ between the two groups.

Ho: The frequency of acute postoperative compli-
cations does not differ between the two groups.

Ha: The frequency of acute postoperative compli-
cations does differ between the two groups.

Ho: The severity of acute postoperative compli-
cations does not differ between the two groups.

Ha: The severity of acute postoperative compli-
cations does differ between the two groups.

(6) Are there differences in preoperative and post-
operative measures of velar kinesiology between
the two groups?

Ho: The average preoperative velar rise rate
does not differ between the two groups.

Ha: The average preoperative velar rise rate
does differ between the two groups.

Ho: The average postoperative velar rise rate
does not differ between the two groups.

Ha: The average postoperative velar rise rate
does differ between the two groups.

Ho: The average amount of preoperative velar
stretch does not differ between the two groups.

Ha: The average amount of preoperative velar
stretch does differ between the two groups.

Ho: The average amount of postoperative velar
stretch does not differ between the two groups.

Ha: The average amount of postoperative velar
stretch does differ between the two groups.

Ho: The average amount of preoperative velar
height above/below the palatal plane does not
differ between the two groups.

Ha: The average amount of preoperative velar
height above/below the palatal plane does differ
between the two groups.

Ho: The average amount of postoperative velar
height above/below the palatal plane does not
differ between the two groups.

Ha: The average amount of postoperative velar
height above/below the palatal plane does differ
between the two groups.

(7) Are there differences in preoperative and post-
operative measures of velar kinesiology within
the two groups?

Ho: Within the group treated by a combined push-
back with pharyngeal flap preoperative and post-
operative measurements of velar rise rate do not

Ha: Within the group treated by a combined push-
back with pharyngeal flap preoperative and post-
operative measurements of velar rise rate do

Ho: Within the group treated by injectable Teflon
augmentation pharyngoplasty preoperative and post-
operative measurements of velar rise rate do not

Ha: Within the group treated by injectable Teflon
augmentation pharyngoplasty preoperative and post-
operative measurements of velar rise rate do differ.

Ho: Within the group treated by a combined push-
back with pharyngeal flap preoperative and post-
operative measurements of velar stretch do not

Ha: Within the group treated by a combined push-
back with pharyngeal flap preoperative and post-
operative measurements of velar stretch do differ.

Ho: Within the group treated by injectable Teflon
augmentation pharyngoplasty preoperative and post-
operative measurements of velar stretch do not

Ha: Within the group treated by injectable Teflon
augmentation pharyngoplasty preoperative and post-
operative measurements of velar stretch do differ.

Ho: Within the group treated by a combined push-
back with pharyngeal flap preoperative and post-
operative measurements of velar height above/below
the palatal plane do not differ.

Ha: Within the group treated by a combined push-
back with pharyngeal flap preoperative and post-
operative measurements of velar height above/below
the palatal plane do differ.

Ho: Within the group treated by injectable Teflon
augmentation pharyngoplasty preoperative and post-
operative measurements of velar height above/below
the palatal plane do not differ.

Ha: Within the group treated by injectable Teflon
augmentation pharyngoplasty preoperative and post-
operative measurements of velar height above/below
the palatal plane do differ.

The rationales for establishing, and the operational def-

nitions of, the measurements to be made are described in the

following chapter.


Selection of Subjects

Potential subjects for inclusion in this study were

initially identified using the records of the Oral-Facial

Clinic (now the Craniofacial Center) at the J. Hillis

Miller Health Center, University of Florida, Gainesville,

Florida. Following the identification of a subject pool

who had undergone either of two operations under investiga-

tion in this thesis, individual medical records were re-

trieved and examined in order to ascertain that a given pa-

tient had, as a minimum, preoperative speech data, opera-

tive and hospitalization data, and at least one postopera-

tive speech evaluation. No restrictions were placed on pa-

tient age, sex, or etiology of the VPI. In fact, it was

the intent of this investigator to identify and include as

many patients as possible into each of the surgical groups.

The only restriction placed on potential subjects, other

than the availability of the minimum criteria mentioned

above, was that they be Caucasian. This restriction resulted

in the loss of only one potential subject. As dental and

facial characteristics show racial influence (Graber, 1966),

the effects (if any) of which might confound the variables

under investigation, it was felt to be prudent to study pa-

tients of only one race.

Forty-five subjects were identified in this manner in-

cluding 20 who had undergone a Teflon augmentation pharyngo-

plasty and 25 who had undergone a palatal pushback with

pharyngeal flap procedure.

Teflon Augmentation Pharyngoplasty Group

The Teflon group was made up of 11 males and nine fe-

males (1.2:1 ratio) ranging in age from 53 months (4 years,

5 months) to 254 months (21 years, 2 months) at the time of

their first operation. The average patient age of this group

at the time of first operation was 133 months (11 years, 1

month). Sixteen of the patients underwent one Teflon injec-

tion procedure; four patients underwent two procedures; and

one patient had undergone three procedures. The distribu-

tion of probable causes of patients' VPI is shown in Table 3.

Teflon pre/post cinefluorographic group. Nine of the

patients who had undergone a Teflon augmentation pharyngo-

plasty also had a preoperative cinefluorographic analysis

as well as one or more postoperative cinefluorographic films.

This subgroup was similar to the larger Teflon group in that

the mean age of these patients at the time of first opera-

tion was 137 months (11 years, 5 months); however, the male

to female ratio was 2:1 (6 males, 3 females).


Table 3
Possible Causes of the VPI
in the Two Groups under Study

Diagnosis Group

Teflon Pushback

Cleft Palate (CP) 7 5

Unilateral CL/P 1 3

Bilateral CL/P 2 2

Submucous CP 3 2

Velar Cleft Only 1 0

Congenital VPI 6 13

20 25


Pushback with Pharyngeal Flap Group

The group was made up of 17 males and eight females

(2.1:1 ratio) ranging in age from 34 months (2 years, 10

months) to 31b months (26 years, 4 months) at the time of

their operation. The average patient age at the time of

operation was 120 months (10.0 years). No patient in this

group had more than one pushback with pharyngeal flap opera-

tion. The distribution of the probable causes of the pa-

tients' VPI is shown in Table 3.

Pushback pre/post cinefluorographic group. Seven of

the patients who had undergone the pushback with pharyngeal

flap operation also had a preoperative cinefluorographic

analysis as well as one or more postoperative cinefluoro-

graphic films. This subgroup was similar to the larger

pushback group in that the mean age of these patients at

the time of their operation was 126 months (10 years, 6

months) and the male to female ratio was 2.5:1 (5 males, 2


Procedure for Obtaining Medical Record Data

Utilizing both individual medical records and Cranio-

facial Center files, data were obtained on each patient's

preoperative and postoperative speech evaluations, number

of days hospitalized for the surgical procedure, length of

the particular surgical procedure in minutes, and frequency

and severity of medical complications which occurred in the

acute postoperative period. Appendix A lists the minimum

amount of information necessary to retrieve or verify any

data of interest to the reader.

Speech Data

Patients being followed by the Craniofacial Center have

their speech and voice quality screened at every visit. Ty-

pically, the screening procedure encompasses five broad

areas. First, the presence of audible or visible abnormal

nasal air flow during the articulation of pressure consonant

sounds is determined by having the patient repeat 10 bi-

syllabic words containing either the /p/ or /b/ consonant.

The presence or absence of inappropriate nasal air emission

is observed using some convenient and easily reproducible

device such as the oral-nasal air paddle (Bzoch, 1979) and

is recorded as some ratio of ten. Second, hypernasal reso-

nance distortion is determined by alternately pinching the

nares closed and then leaving them open while the patient

produces consonant-vowel-consonant (CVC) words in which

only nonnasal oral consonants are used. Any shift in reso-

nance on a given word is evaluated as hypernasal resonance

distortion. Carrying out this simple procedure on 10 dif-

ferent words allows the examiner to record the consistency

of the patient's hypernasality as a ratio of ten. Third,

hyponasal resonance distortion is determined in the manner

described for determining hypernasal resonance distortion

with the patient producing CVC words which contain at least

one nasal consonant. In this case if a shift in resonance

is not detected on a given production, it is counted as hypo-

nasal resonance distortion. Using ten different words a ratio

of ten is recorded. Fourth, dysphonia characterized by the

use of an aspirate voice or by hoarseness is screened for.

The patient is requested to sustain the phonation of several

vowel sounds while the examiner measures the phonation time

and carefully judges the vocal quality. Finally, articulatory

proficiency is screened by having the patient repeat 31 words

taken from the Bzoch Error Pattern Diagnostic Articulation

Test (Bzoch, 1971). This procedure examines 23 individual

consonant sounds occurring in the medial position of single

words. Additionally, eight consonant blends are tested in

the initial position of single words. Interjudge reliabili-

ties on these five measures have been reported by Dixon et al.

(1979) and are sufficiently high for research purposes. The

screening form that is typically used at the University of

Florida's Craniofacial Center is reproduced in Appendix B.

Of primary interest in this study were the preoperative

and postoperative nasal emission and hypernasality ratios as

these were used both to help determine the equivalence of pa-

tients' symptoms preoperatively and to assess the success or

failure of the surgical procedure postoperatively. Of secon-

dary interest were the articulation screening, used to pre-

operatively compare the number of gross articulation errors

glottall stops and pharyngeal or velar fricatives) and the

hyponasality ratio which was used to determine the presence

of surgically induced hyponasality. The preoperative and

postoperative data for the patients in both surgical groups

are presented in Appendix C.

Operation and Hospitalization Data

From the medical records of an individual patient the

length of each operation in minutes was recorded by subtract-

ing the starting or "opening" time from the finishing or

"closing" time. In the Teflon group one patient had three

separate procedures and four patients had two separate pro-

cedures. Thus, there were potentially 26 operation time data

points available for this group. However, one data point on

a patient who had undergone two Teflon procedures was miss-

ing from the patient's medical record leaving a total of 25

data points for analysis. The 25 patients in the pushback

group had had only a single such operation and all had opera-

tive times recorded in their respective medical records.

The length of hospital stay in days was also taken di-

rectly from information available in individual medical rec-

ords. As is the common practice in hospital billing, the

day the patient was admitted to the hospital was counted,

and the day he or she was discharged was not counted in de-

termining the length of stay. As was the case with opera-

tive time, there were 26 potential hospital admissions to

be counted for the Teflon group. However, again there was

one patient in which the day of discharge was not recorded,

and, therefore, only 25 data points were available on this

variable for the Teflon group. All 25 potential data points

were available for the patients making up the pushback group.

The raw data for operative time and length of hospitali-

zation are shown in Appendix D.

Postoperative Complications Data

The operative reports, discharge summaries, progress

notes, and nursing notes pertaining to any hospitalization

for one of the operations under investigation were carefully

read, and any immediate postoperative event which was docu-

mented in any of the above as unfavorable to the postoperative

course was recorded on a given patient's data recording sheet.

This investigator attempted not to impose his own beliefs

and/or opinions on what did or did not constitute an unfavor-

able postoperative event, but, rather, he endeavored to only

record events that had been noted in the patient's medical


In this manner a list of postoperative events was com-

piled and, after eliminating duplication, was numbered con-

secutively. The number of different postoperative events

was twelve. The numbers one through twelve were taken from

a table of random numbers, and the list of postoperative

events was rearranged in random order. Together with a cover

letter the list of postoperative events was distributed to

three board-certified surgeons (two Ear-Nose-Throat, one

Plastic), none of whom had operated on any of the patients

in this study, who were requested to rate the severity of

the postoperative events on a scale from one (not serious)

to seven (very serious). In this way an individual post-

operative event could be quantitatively compared to another.

Also, the average of the severity rating as well as the fre-

quency of postoperative events or complications could be com-

pared between the two groups.

A listing of the postoperative events and their severity

ratings is shown in Appendix E.

Procedure for Obtaining Data
from the Cinefluorographic Films

Since the mid-1960's a library of 16 mm sound cinefluoro-

graphic films has been built at the University of Florida by

both the Department of Communicative Disorders, College of

Health Related Professions, and the Department of Basic Dental

Sciences, College of Dentistry. The procedure for obtaining

the cinefluorographic films has remained remarkable constant

over the years and is briefly summarized here for the con-

venience of the reader.

Radiographic Procedures

All the cinefluorographic films were made utilizing the

facilities of the Department of Radiology, Shands Teaching

Hospital and Clinics, J. Hillis Miller Health Center. For

the past several years a Philips Super 100 fluoroscopy/

radiography unit mounting a Philips six-inch image intensi-

fier has been used. With the patient seated in an upright

position and his or her head held in a stable position by

a lucite cephalostat, 16 mm sound movie films were made of

the fluorographic image of the patient while he or she pro-

duced a predetermined speech sample. All the cinefluoro-

graphic films have been obtained by mounting a modified

Auricon Cine-voice 16 mm Sound-on-Film camera (Model CM-

72A-FM) on the Philips image intensifier. Prior to 1981,

Eastman Double-X Panchromatic Negative film (Type 7222) was

used. Since 1981, Eastman Plus-X Negative film (Type 7231)

has been used. One hundred foot film rolls, magnetic stripped

and with a B winding, were utilized in all cases. In recent

years a 50% barium sulfate suspension has been frequently used

as a contrast material to coat the soft tissues of the naso-

pharynx and oropharynx. Filming speech was 24 frames per sec-

ond which has been demonstrated to be adequate for analyzing

speech movement patterns (Shelton, Brooks, Youngstrom, Died-

rich, & Brooks, 1963). A simultaneous soundtrack was placed

on each film by positioning an Auricon High-Fidelity micro-

phone (Model E-7) approximately six inches from the patient's

mouth and connecting movie camera and microphone via an Auricon

Filmagnetic amplifier (Model MA-11C). All films were processed

automatically using a Combilator Automatic Film Processor

(Model CMB-3).

Each filming procedure resulted in no more than 4.9 Rads

per minute of localized radiation exposure to the patient,

as determined by the Department of Radiation Physics, J. Hillis

Miller Health Center.

Speech Sample

The speech sample, as the filming procedure, has remained

remarkably constant over the years, with only several changes

and/or modifications taking place. Briefly, the sample con-

sisted of prolongations of the vowels /i/, /a/, and /u/;

repetition of three consonant-vowel syllable sets including

/pi/, /pa/, and /pu/; word or phrase repetition of several sen-

tences as well as the patient blowing, swallowing and whistling.

Appendix F lists the formal speech sample in its entirety.

Frequently, the entire speech sample would not have been

elicited from a given patient. However, the repetition of

the three consonant-vowel syllable sets and the sentence,

"In the evening Connie watches T.V. with me," were always

elicited unless, by chance, the operator accidentally over-

looked them or there was some type of equipment malfunction

during the filming procedure. Because these speech acts were

potentially available on every film for every patient, the

first utterance of the syllable /pi/ was used to determine

velar rise rate. The maximum elevation of the velum during

the production of the initial consonant /k/ in the word

"Connie" was used to measure the effective length of the

soft palate as well as to measure the height of the velum

above or below the palatal plane.

Method of Making Single Frame Tracings

The cinefluorographic films were reviewed using a

Moviola Two-Plate 16mm Flatbed Console Editor (Model M-84).

This projector is capable of advancing and rewinding film

at a variable rate from less than one frame per second up

to 240 frames per second. Tracings were needed for later

measurements; therefore, a piece of Unitek eight inch by

ten inch tracing film was taped to the Moviola viewing

screen, and the structures of interest on a particular frame

were traced onto the matte surface of the tracing film. The

actual frames that were traced numbered four on each film.

First, a calibration frame (ruler shot) was traced. The

ruler shot consisted of the filming of a radiopaque stain-

less steel ruler which had holes drilled at centimeter in-

tervals which was placed in the patient's mouth or along-

side his or her cheek. The second tracing was of a frame

with the patient breathing quietly through the nose (rest

shot). The third tracing was of the frame in which the velum

reached its maximum excursion during the production of the

/k/ in "Connie" (/k/ shot). Lastly, the frame just prior

to any velar movement during the first repetition of /pi/

in that syllable set was traced. The film was then advanced

frame-by-frame until the velum reached its point of maximum

excursion during this utterance of /pi/ and, repositioning

the tracing film on the previously drawn boney structures

of the anterior nasal spine, posterior nasal spine, and the

tubercle of the atlas, the outline of the velum was then

traced in an elevated configuration (/pi/ shot). From these

tracings, lines were drawn and measurements of the lines

were taken and were corrected to true size according to the

following descriptions.

Plane of Reference and Definitions of Linear Measurements

Palatal plane. A line drawn from the anterior nasal

spine (ANS) through the posterior nasal spine (PNS) and

projected to the posterior pharyngeal wall (PPW), as shown

in Figure 6, forms the palatal plane.

Depth of the nasopharynx (DNP). The distance along the

palatal plane from the PNS to the PPW with the patient at

rest constituted the depth of the nasopharynx.

Velar length: rest (VLR). The straight line distance

from the PNS to the midpoint of the uvula (U) with the velum

at rest determined the velar length at rest.

Velar length: function (VLF). The combined distance

of the lines drawn through the midline of the velum in its

functional position during the production of /k/ was the

velar length in function. One line began at PNS and was pro-

jected through the midline posteriorly. The second line

began at U and was projected through the midline superiorly.

Velar height (VH). The distance along a line drawn

perpendicular to the palatal plane and either ascending or

descending to the highest point on the nasal surface of the

velum at its maximum point of excursion during the produc-

tion of /k/ constituted the measurement of velar height.

Velar rise (VR). The distance along a line drawn per-

pendicular to the line PNS U and ascending to the highest

point on the nasal surface of the velum at its maximum point

of excursion during the production to /pi/ was the velar

rise. The line PNS U was drawn on a tracing of the velum

just prior to any detectable movement of the velum during

the uttering of the syllable /pi/. Figures 6 and 7 will

aid the reader in a clearer understanding of the above

linear measurements.






Figure 6. Illustration of the anterior nasal spine (ANS),
posterior nasal spine (PNS), uvula (U), pos-
terior pharyngeal wall (PPW), palatal plane (PP),
depth of the nasopharynx (DNP), and the velar
length at rest (VLR).



Figure 7. Illustration of the velar length during function
(VLF), velar height (VH), and velar rise (VR).

Computed Measurements

Velar stretch (VS). The percent increase (decrease) in

velar length during function as compared to velar length at

rest describes velar stretch. It was computed by subtract-

ing the resting velar length (VLR) from the velar length

during function (VLF), dividing by the resting length (VLR)

and then multiplying by 100.


% VS = --------- x 100


Velar rise rate (VRR). This measurement is the speed

(expressed in millimeters per second) of velar ascent during

the production of the syllable /pi/. It was computed by

multiplying the velar rise (VR) by 24 (the number of frames

per second) and dividing by the counted number of frames it

took to actually accomplish the velar rise.

24 x VR

VRR = --------------

No. frames during VR

All measurements were made to the nearest millimeter

and computed measurements are reported to the nearest whole


Tracing and Measurement Reliability

Twenty-five percent (4 subjects, 10 films) of the cine-

fluorographic films had each individual frame of film used

to make the previously described measurements marked with

a small dot of India ink in order that a given frame might

be precisely identified at some future time for the purpose

of establishing reliability.

The reliability of the author's tracing technique and

the measurements made from the tracings (intratracer relia-

bility) were determined by a tracing-retracing procedure

separated by a time interval of three weeks. Retracings

were made from the precise frame of film that was originally

traced. The measurements taken from the retracings were then

compared to the original data, and a Pearson product-moment

correlation coefficient (Pearson r) was computed.

The validities of the author's tracing technique and his

measurements (intertracer reliability) were assessed by hav-

ing another independent and very experienced tracer produce

a set of tracings and the measurements from them. The sec-

ond tracer traced 13% of the cinefluorographic films (2 sub-

jects, 6 films). Measurements from these tracings were

then compared to the author's original data, and a Pearson

product-moment correlation coefficient was computed. The

correlation coefficients for the intratracer reliability

and the intertracer reliability can be found in Table 4.

These correlation coefficients were considered to be

within acceptable limits for the purposes of this study.

Method of Analysis

The data generated by the above described procedures

will first be reduced to an intelligible and interpretable

form using simple descriptive statistics. Reduced data


Table 4
Correlation Coefficients for Intratracer
and Intertracer Reliabilities

Correlations Correlation Coefficient

1. Intratracer reliability 0.97

2. Intertracer reliability 0.94

will be grouped according to which research question they per-

tain. Data will be statistically manipulated so as to com-

pare the obtained results with chance expectations. With

the exception of morphological changes following surgery (re-

search question two) there are no previously reported studies

to suggest testing hypotheses other than null hypotheses,

and, for this reason, specific research hypotheses were not

generated except under the second research question. Hy-

potheses testing will utilize the appropriate parametric or

nonparametric statistical test dependent upon the level of

measurement employed for a given variable. In choosing the

appropriate statistical test the concept of robustness (Marks,

1982) has been adhered to whenever possible. For the pur-

pose of hypothesis testing all tests of difference will be

carried out using a two-tailed statistical test except under

research question two where the direction of change has been

predicted. Hypotheses will be tested at the 1% level of

confidence in order to minimize the probability of rejecting

a null hypothesis when it is, in fact, true.

Finally, the outcome of the statistical manipulations

will be summarized prior to discussion any interpretation

in later chapters of this thesis.


The data and the data analyses presented in this chap-

ter are arranged in the same order as the research questions

outlined in Chapter II. All data reduction and statistical

manipulations were performed by the author either manually

or by utilizing the computer programs of the Advanced

Statistical Analysis package (Radio Shack, 1979) run on a

TRS-80 Model I microcomputer.

Preoperative Group Homogeneity

The first question addresses whether the patients in

the two surgical groups were selected from the same popula-

tion. Six variables were selected to describe and compare

the patients. These variables are age at the time of the

operation, the sex distribution of the two groups, the pre-

sumptive etiology of the VPI, the preoperative indices of

hypernasality and nasal emission and the number of gross

articulation errors demonstrated prior to surgery. The

data describing these variables are presented in Table 5.

Table 6 summarizes the results of the tests for dif-

ferences carried out between the two groups of patients

on each variable. The chi-square tests were corrected for

r0 U

4J -H

H 0

0 :

*H U

4-) (0
tl X

Q)4Jr- -

Cd W)
E-4 > 4J

*a4 Z

ui 0









- 0I




C a
- I




I l

Ln r-i
N m
ust "









O-i 4






A4 P4
* II
4 HH


r-1 ,--






a a







I I1



0 0

00 I



Cd I


- I

cn I
4 I

o I
0 I













O >


-H I
.0 1

0 I






a I

0 I






0 0

0 (
4 .


4-4 0

0 Q)
a) )


E-4 rx4
0 a)

M >1

0 U!
O1 0





1 oo CM
00 e 00 0
CN CN (N4 0 -O r


(N N
-P1 X X Nj NI NI



0 >1 >1 >I

I ) a)
(U (U (U C 0 C

u4 -- w +j 0
C C 4 *H *H *H A
a) r-t (

4-4 -H *H r
I U U ( ( (

> o

m 0 0
< *H +1

0) X *- 0a m O
13 a) >1 Cd
< 02 M Z U

continuity, and the Mann-Whitney U tests were corrected for

ties (Siegel, 1956). No statistically significant differ-

ences were found between the two groups on any of the vari-

ables considered. Appendix C lists the pertinent raw data

for each patient in each group.

Morphological Changes Following Surgery

The second research question was directed toward de-

termining whether the two surgical procedures did what they

were intended to do. That is, did the Teflon injection re-

duce the depth of the nasopharynx and did the combined push-

back with pharyngeal flap increase velar length, as each is

presumed to do? Table 7 summarizes the average preoperative

and postoperative measure of the appropriate variable as-

sociated with each surgical procedure. These data were

generated by the nine patients in the Teflon group and the

seven patients in the pushback group who had undergone both

a preoperative and a postoperative cine analysis. The raw

measurements of each variable were converted into standard

deviations above or below the normal for the age of each

patient in order to correct for any postoperative changes

that might be due solely to growth. The average time be-

tween the preoperative and postoperative cine analyses was

over a year for each group (Teflon = 21 mos., pushback =

17 mos.); however, the difference between the two groups

was not statistically significant (t = .3531, df = 14).

The Walsh test (Siegel, 1956) was used to test for

I >1

(C 0
2Z 0

0 I P




0 0
,- II -
IA 0
a)z 0

















0 0


0 -H

significant differences in the preoperative and postopera-

tive measures of depth of the nasopharynx for the Teflon

group and velar length for the pushback group. This par-

ticular test as well as other nonparametric tests was

chosen to evaluate the data taken from the cine analyses

because strong arguments can be put forth that these sub-

groups were not representative of the larger groups of pa-

tients from which they came (see Chapter V). As the direc-

tion of change might be predicted for the hypotheses under

this research question, the hypotheses were tested using

a one-tailed Walsh test. Table 7 reveals that the change

in the depth of the nasopharynx was statistically signifi-

cant following the Teflon augmentation pharyngoplasty but

that the change in velar length was not statistically sig-

nificant following the combined pushback with pharyngeal


Effectiveness in Eliminating Hypernasality
and Nasal Emission

The third research question was asked in order to de-

termine if the two surgical procedures were equally effec-

tive in correcting the symptoms of VPI. A surgical outcome

was considered as a success only if it resulted in the total

elimination of both the patient's hypernasality and inappro-

priate nasal emission. These criteria for judging a out-

come have been advocated by Furlow et al. (1982) and by

Hogan and Schwartz (1977) and are more stringent than the

criteria for judging success which are typically found in

the literature.

The Teflon augmentation pharyngoplasty resulted in a

successful outcome in 80% (16/20) of the cases. The com-

bined pushback with pharyngeal flap resulted in a success-

ful outcome in 92% (23/25) of the cases. Table 8 shows the

observed outcomes for each operation in a 2 x 2 contingency

table. A chi-square test, corrected for continuity (Siegel,

1956) was used to test both null hypotheses under research

question three. Table 8 reveals that there was no statis-

tically significant difference in the frequency of a success-

ful outcome between the two surgical groups.

Persisting Postoperatively Induced Hyponasality

The fourth research question was designed to determine

whether either operation produced persisting hyponasality.

The Teflon augmentation pharyngoplasty did not result in

any cases of hyponasality persisting past one year post-

operatively. The combined pushback with pharyngeal flap

resulted in one case (4%) of hyponasality persisting past

one year postoperatively. The Fisher exact probability test

(Siegel, 1956) revealed essentially no statistical differ-

ence (p = 1.0) between the two surgical groups in the fre-

quency of postoperatively induced hyponasality.