Female Stress Urinary Incontinence Clinical Guidelines Panel Summary Report on Surgical Management of Female Stress Urinary Incontinence

00~~-534~~~~~i583-n~~5$03.no~o THEJOURNAL OF UROLOGY Copyright 0 1997 by AMERICAN UROI.OGICAL ASSOCIATION, INC.

Vol. 158. 875-880.September 1997 Printed in V.S.A.

-

Special Communications FEMALE STRESS URINARY INCONTINENCE CLINICAL GUIDELINES PANEL SUMMARY REPORT ON SURGICAL MANAGEMENT OF FEMALE STRESS URINARY INCONTINENCE GARY E. LEACH, ROGER R. DMOCHOWSKI, RODNEY A. APPELL, JERRY G. BLAIVAS, H. ROGER HADLEY, KARL M. LUBER, JACEK L. MOSTWIN, PAT D. O'DONNELL AND CLAUS G. ROEHRBORN

ABSTRACT

Purpose: The American Urological Association convened the Female Stress Urinary Incontinence Clinical Guidelines Panel to analyze t h e literature regarding surgical procedures for treating stress urinary incontinence i n the otherwise healthy female subject and to make practice recommendations based on t h e treatment outcomes data. Materials a n d Methods: The panel searched t h e MEDLINE data base for all articles through 1993 on surgical treatment of female stress urinary incontinence. Outcomes data were extracted from articles accepted after panel review. The data were t h e n meta-analyzed to produce outcome estimates for alternative surgical procedures. Results: The data indicate t h a t after 48 months retropubic suspensions a n d slings appear to be more efficacious t h a n transvaginal suspensions, a n d also more efficacious t h a n anterior repairs. The literature suggests higher complication rates when synthetic materials are used for slings. Conclusions: The panel found sufficient acceptable long-term outcomes d a t a (longer than 48 months) to conclude that surgical treatment of female stress urinary incontinence is effective, offering a long-term cure i n a significant percentage of women. The evidence supports surgery as initial therapy and as a secondary form of therapy after failure of other treatments €or stress urinary incontinence. Retropubic suspensions a n d slings a r e the most efficacious procedures for long-term success (based on cureldry rates). However, i n the panel's opinion retropubic suspensions a n d sling procedures a r e associated with slightly higher complication rates, including longer convalescence a n d postoperative voiding dysfunction. KEY WORDS:urinary incontinence; urinary incontinence, stress To provide detailed information specific to surgical management of stress urinary incontinence in women, the American Urological Association (AUA) convened the Female Stress Urinary Incontinence Clinical Guidelines Panel and charged this panel with 2 principal tasks: 1) analyze published outcomes data on surgical procedures to treat female stress urinary incontinence and 2) produce practice recommendations based primarily on outcomes evidence from the literature. When adequate outcomes evidence was lacking, the panel based its recommendations on expert opinion according to commonly accepted clinical practices and principles. The type of patient to whom the recommendations apply, termed the index patient, is an otherwise healthy woman with stress urinary incontinence, either previously untreated or treated (surgically or nonsurgically), without significant Pelvic organ prolapse, who has decided to seek surgical treatment. The index patient may have urethral hypermobility and/or intrinsic sphincteric deficiency. The purpose of the Panel is to distinguish between alternative surgical procedures and provide estimates of the outcomes that can be expected with each. These outcome estimates are intended to help the patient and physician decide which surgical procedure is best for her. The complete report, which contains all Accepted for publication March 7, 1997.

results of the study, is available from the AUA.1 The major sections of that report are summarized. METHODOLOGY

To extract outcomes data for surgical interventions to treat stress urinary incontinence, a literature search was performed using the MEDLINE data base. The data base was searched several times up to January 1994, using the MESH subject headings, "urinary incontinence" and "urinary incontinence, stress." In addition, the panel hand searched bibliographies from appropriate references. The panel reviewed abstracts and citations (author, title, date) of 5,322 articles, of which 457 were retrieved for detailed article review and data extraction. Upon panel review of these 457 articles 175 were rejected. The net result was 282 articles with some type of acceptable outcomes data. For accepting or rejecting an article, the panel established specific review criteria. Articles were rejected if not published in a peer reviewed publication in English, the article was in abstract from only, the data reported were irrelevant to surgical management of stress incontinence in the otherwise healthy woman, there was concomitant prolapse surgery, original data were absent (as in a review article) or data were updated in a more recent article. Articles were also rejected for various miscellaneous reasons. For example, the

875

876

FEMALE STRESS URINARY INCONTINENCE CLINICAL GUIDELINES PANEL

panel voted to reject articles on polytetrafluorethylene because it has not been approved by the Food and Drug Administration in the United States for urological applications, although a s of 1997 it is undergoing clinical trials for use in such applications. A major reason for article rejection was insufficient patient followup in reporting outcomes data. Whether patients were cured andor dry (referred to a s “cure/dry”) was the primary outcome the panel evaluated. For length of followup, the panel established a 12-month minimum as an exclusion criterion. Any cure/dry data with reported followup of less than 12 months were excluded from analysis. For articles that reported a range of followup, the minimum of the range had to be 12 months or longer. For articles reporting followup only as an average or mean, that average or mean had to be a t least 24 months or the data were excluded. The panel used each author’s objective or subjective definition of cure/dry for inclusion of data to compute values for outcome estimates. Also, in defining “cure“ the panel recognized that “cure“ does not always equate to “dry.” “Cure” means being cured of stress incontinence. A patient who was reported as cured of stress incontinence but who experienced incontinence due to urgency or urge related incontinence was considered cured for purposes of data analysis. A comprehensive data abstraction form was devised to capture as much relevant information as possible from each selected article. The basic task of extracting outcomes data and transcribing the data onto the form was performed by a paid staff of reviewers selected by the panel chairman and panel facilitator. All reviewers were residents in urology a t Kaiser Permanente Medical Center in Los Angeles, California. They were trained by the facilitator coordinator for AUA clinical practice guidelines, aided by other AUA consultants

and staff. The panel chairman and panel facilitator closely monitored the entire data extraction process. The data resulting from this process were combined to yield the outcome estimates displayed in a comparative outcomes table (table 1).To combine the data the AUA used the Confidence Profile Method, which allows analysis of data from studies that are not randomized controlled trials.2 :$ The FAST*PRO computer software was used in the analysis.4 The results of the Confidence Profile Method incorporated in FAST*PRO are probability distributions, which can be described reporting a median probability estimate with a 95% confidence interval. The 95% confidence interval is that interval such that the probability (Bayesian) of the true value being outside the interval is 5%. Table 1 displays probability estimates for the 4 major procedure categories of retropubic suspensions, transvaginal suspensions, anterior repairs and pubovaginal sling procedures. All outcome probability estimates were developed by combining outcomes data for the various individual procedures that make up each of the 4 major procedure categories. For example, in the first cell of table 1 under the retropubic suspensions, column the median probability estimate for cureldry a t 12 to 23 months is 84%. This estimate was derived from combined cureldry data for Burch, MarshallMarchetti-Krantz and various other individual retropubic suspension procedures. The other probability estimates in this column, including estimates of potential complications, were also derived from combined outcomes data for the various individual procedures that make up the retropubic suspension grouping. Similarly, the estimates in the transvaginal suspensions column were derived from combined outcomes data for Pereyrdmodified Pereyra, Stamey, Gittes and other types of individual transvaginal suspension proce-

TABLE1. Comparative outcomes for procedure categories Retropubic Suspensions

Suspensions

Outcomes

Sling Procedures ~~

Gm ~

-__

Anterior Repairs

~~

-

Median CI

Median CI

Median C1

Median CI

(2.5-97.5)

(2.5-97.51’3

(2.5-97.5)%

12.5-97.5lq

~

Curddry: 12-23 Mos. 2 4 4 7 mos. 48 Mos. and longer Curddry1improved: 12-23 Mos. 24-47 Mos. 48 Mos. and longer Postop. urgency: For pts. with urgency and detrusor instability preop. For pts. with urgency and no detrusor instability preop. For pts. with no urgency and detrusor instability preop. For pts. with no urgency and no detrusor instability preop. Retention: Longer than 4 wks. Permanent Days in hospital (panel survey) Resumption of normal activities Death

151943 2311,870 1712.196

84 (77-89) 84 (80-88) 84 (79-88)

131700 81424 4/292

79 171-861 65 150-771 67 (53-791

61310 31113 511,088

68 155-80) 85 169-951 61 147-72)

51135 71344 71473

82 173-89 I RZ 173-89 I 63 175-88)

161961 24/1,941 1812,204

86 (80-90) 88 (85-91) 90 (87-921

131700 81424 41292

82 ( 7 4 4 7 ) 78 (71-83) 82 (73-891

61310 31113 611,101

78 (65-88) 95 (89-98) 73 (70-761

51135 71344 71473

9 1 184-96) 85 177-911 87 180-92)

6/78

66 (50-791

6133

54 135-73)

4145

46 (24-681

61319

36 (22-52)

51110

34113-61) 20 (5-451

116 81241

4 10-33)

11 (8-16)

113

7 (0-54)

4136

6/150

5 13-10)

71140

7 13-11)

(44) 71578 8 16-11) Less than 5% From 0 to 5 days Typically 6 weeks for all treatment modalities Death rate for all prwedures presumed to be no different than for other types of elective vaginallabdominal surgery, approximately 5110,000 911,131 5 13-81 1011,083 3 (14) 31857 3 11-91 61279 4 (2-71 5/340

Transfusion General medical complications: Significant 2113,136 Not significant 7/1,549 Intraop. complications: 1311,992 Significant 16/2,284 Not significant Periop. complications: 4013.598 Significant 6416.044 Not significant Subjective complications 1311,001 Complications requiring surgery 1512,718 ~ _ Note: G P : &No. Groupnreatment Arms ExtractedP-No.

5

(3-7)

61479

5

2 2

(2-3) (1-4)

101805 71646

2 4

11-3) (1-91

511,005 31650

2 11-3) 8 (2-17)

14/1,127 21256

4 12-51 6 (3-10)

2 3

(1-3) 11-41

51532 2511,835

2 5

(1-5)

11294 11313

1 0

10-2)

6/326 1911,077

3

(44)

4 (3-51 14114-15) 9 15-15] 2 (1-3)

4012.814 5413.330 2411,412 1511.575

Pts. in Groups.

7 (591 12 19-15) 11 (8-15) 2 11-4)

51970 1312,322 31341 W1.074

10-11

2 11-51 16110-231 2 (141 0 10-1)

2011,723 2611.916 41301 1111.119

H

11-6) (5-121

7 15-10) 12 (8-171 6 12-131 3 12-51

877

FEMALE STRESS URINARY INCONTINENCE CLINICAL GUIDELINES PANEL

dures. The panel found no significant differences between individual procedures within each of the 4 groupings. OUTCOMES ANALYSIS RESULTS

After analyzing cureldry data variously reported in the literature a t time points 12 to 120 months, the panel defined the 3 discrete time frames in table 1 as short term-12 to 23 months of followup, intermediate-24 to 47 months and long term-48 months and longer. The cureldry rates indicate that after 48 months retropubic suspensions and slings appear to be more efficacious than transvaginal suspensions and anterior repairs. However, for anterior repairs the rates after 24 months are anomalous, with a sharp increase in cure/dry status at 24 to 47 months (from 68 to 85%) and then a sharp decline after 48 months (to 61%). These rates are believed to represent a n aberrancy of data reporting rather than true outcomes. Such anomalous results may occur because not all studies report outcomes at all time points. For intrinsic sphincteric deficiency therapy cureldry results are incompletely reported in the literature because recognition of intrinsic sphincteric deficiency as a diagnostic entity is relatively new. In the panel’s opinion sling procedures are the most effective type of treatment for patients with intrinsic sphincteric deficiency. However, the available data are insufficient to determine whether other types of procedures might be equally effective. The panel also attempted to determine from the literature the degree to which factors, such as patient age and prior surgery, had an impact on cure/dry outcomes. These factors were prior incontinence surgery, subjective urgency, urodynamically proved detrusor instability, menopausal status, parity, prior hysterectomy and age. Because of incomplete data and varying definitions of data, the panel could make no determinations regarding the effects of these variables on cure/dry rates. Cure/dry rates alone do not fully represent all potential treatment benefits. Many patients, even if not completely cured of stress urinary incontinence, report substantial improvement and consider the surgery successful. For this reason, the panel included a cure/dry/improved category in table 1. As would be expected, cure/dry/improved rates are higher than cure/dry rates alone for all treatment types across all 3 time frames. Probability estimates for treatment complications, beginning with estimates of postoperative urgency, are displayed in table 1 relative to patient preoperative status regarding urgency and detrusor instability. Analysis of the 4 patient subgroups revealed no major statistical differences between transvaginal suspensions, sling procedures and retropubic suspensions with regard to postoperative urgency. Insufficient data exist to evaluate anterior repairs for this outcome. In general, if urgency and detrusor instability are present preoperatively, there is significant risk of these symptoms persisting postoperatively. If either is absent, the risk of postoperative urgency is reduced. However, no accurate statistical statements can be made regarding these tendencies because of small total sample sizes. The panel determined that the 2 periods to be reported for urinary retention should be greater than 4 weeks and permanent. For temporary retention the panel recognized that the data for less than 4 weeks reflect varying practice patterns, and changes in hospitalization practices and catheter management during the time frame of the literature reviewed. The data cannot be meaningfully combined because of differing lengths of hospital stay and different studies measuring retention a t different points in time. The estimated probability of temporary urinary retention lasting longer than 4 weeks is 5% for retropubic and transvaginal suspensions, and 8% for sling procedures. For anterior rePairs the retention data were insufficient to generate a prob-

ability estimate. For permanent retention there are no accurate data to date. In the panel’s opinion the risk is somewhat higher for sling procedures than for other procedures. However, for all procedures panel opinion is that the risk of permanent retention generally does not exceed 5%. Days in the hospital as reported in the literature again reflect practice patterns that in the panel’s opinion are no longer current. Assuming an uncomplicated stay and based on current practice, the panel determined that length of hospital stay for each treatment type ranges from 0 to 5 days, with the shortest length of stay for transvaginal suspensions and anterior repairs. The panel attempted to analyze obstruction and other possible complications of surgery. The obstruction data were variable and incomplete, with differing definitions and a lack of objective data. Urodynamic diagnosis of obstruction was noted to be poorly standardized and often dependent on subjective symptoms rather than defined pressure-flow criteria. For urinary tract infection, the diagnoses were also variable, and reflected different definitions and reporting biases in the urological as contrasted to the gynecological literature. In general, urinary tract infection was more commonly reported in the gynecological literature. Death rate as an outcome was a difficult variable to assess based on explicit evaluation of the available literature. Because death as a complication of surgery for stress urinary incontinence is so rare, it is seldom mentioned. Therefore, when considering death rate values, the panel attempted to estimate a reasonable rate for all procedures and agreed upon a death rate of approximately 5/10,000procedures. This estimate is consistent with rates for hysterectomy, a pelvic procedure for which defined mortality rates are available.5 For transfusion the panel noted no important differences in probability estimates among the 4 procedure categories in table 1. The rates range from 3 to 5% (median values) across all 4 categories. For considering other complications, 5 general groupings were developed, including general medical complications, intraoperative complications, perioperative complications, subjective complications and complications requiring surgery. Under these general headings, significant and nonsignificant complications were determined for summary purposes by panel opinion. In evaluating the significant and nonsignificant complications under each of the 5 general headings, the panel again noted no major differences in complication rates among the 4 procedure categories in table 1. With regard to synthetic materials for slings, the literature suggests higher complication rates when using such materials (table 2). PANEL CONCLUSIONS AND TREATMENT RECOMMENDATIONS

Overall, after thorough review of the literature and rigorous analysis of the reported data, the panel found sufficient acceptable long-term outcomes data (48months and longer) to conclude that surgical treatment of female stress urinary incontinence is effective, offering a long-term cure in a significant percentage of women. The evidence supports surgery as an initial therapy and as a secondary form of therapy after failure of other treatments for stress urinary incontinence. TABLE2 . Synthetic compared with autologous (rectus fascia and fascia lata) and with homologous materials (nonsynthetic, non-self grafts, porcine xenografts) for specific complications No. Autologous No. Homologous No. Synthetic Complications Vagmal erosion Urethral erosion Fistula Wound sinus Wound infection Seroma

Materials 11,715 pts.i 1 l0.0001~ 5 (0.003I 6 10.003) 3 t0.0021 11 (0.006) 6 (0.003)

Materials 1414 pts.1

Materials 11.515 pts.i

0

10 (0.007)

0

27

0 0 9 (0.02)

4 (0.002) 11 (0.007) 15 (0.009) l(0.0007)

0

(0.02)

878

FEMALE STRESS URINARY INCONTINENCE CLINICAL GUIDELINES PANEL

Specific panel conclusions are embodied in practice recom- long-term success (based on cureldry rates). However, in the mendations for treatment of stress urinary incontinence in panel’s opinion retropubic suspensions and sling procedures the index patient as previously defined. The recommenda- are associated with slightly higher complication rates, intions were produced mainly from analysis of outcomes data cluding postoperative voiding dysfunction and longer convaand partially by expert opinion of panel members. The panel lescence. In patients who are willing to accept a slightly graded its treatment recommendations by 3 levels of flexibil- higher complication rate for the sake of long-term cure, retity, standard, guideline and 0ption,6.~which have been de- ropubic suspensions and slings are appropriate choices. In fined previously, based primarily on the strength of the sci- the patient for whom decreased hospital stay, less likelihood entific evidence for estimating outcomes of interventions. A of morbidity and/or earlier return to work are paramount standard is the least flexible of the 3, a guideline is more transvaginal suspensions are appropriate procedures. In the flexible and an option is most flexible. None of the panel’s panel’s opinion anterior repairs are the least likely of the 4 recommendations fits the definition of a guideline, 2 are major procedure categories to be efficacious in the long term. The panel made a further observation regarding choice of labeled as standards and 1 is an option. The labels indicate the strength of the recommendations. Recommendations procedure. Not only is the range of surgical options for treatwere labeled standards if the panel concluded that they ing stress urinary incontinence already extremely diverse should be followed by virtually all health care providers for with more than 100 procedures, but new techniques and virtually all patients. The recommendation labeled a n option procedure modifications continue to emerge. Some of these, had considerable implementation flexibility because out- as in the past, will not stand the test of experience with time. comes evidence in the literature was insufficient for a stron- Moreover, different surgeons may perform the same proceger recommendation and because of significant trade-offs dure quite differently. The panel believes that surgeons (such as a higher long-term cure/dry rate but also higher should take these considerations into account, together with complication rates for a particular procedure) that could re- their own areas and levels of expertise and their own previous treatment results, when counseling patients regarding sult in variable patient preferences. Standard. Based on the panel’s opinion, the preoperative choice of procedure. evaluation of women with symptoms of stress urinary incontinence should comprise the 4 major components of history LIMITATIONS (impact of symptoms on life-style), physical examination (obLimitations to the process of developing evidence based jective demonstration of stress incontinence) urinalysis and other appropriate diagnostic studies designed to assess treatment recommendations became apparent during our resymptom causes (particularly consideration of the relative view of the literature, and are the reason the panel was contributions of urethral hypermobility, intrinsic sphincter unable to develop a stronger recommendation regarding deficiency and detrusor dysfunction), frequency and severity choices of surgical procedure. Prospective, randomized, conof incontinent episodes and patient expectations from treat- trolled trials, although optimal for providing reliable data, ment. Although components of diagnostic evaluation are not are almost never undertaken. Standardized diagnostic critethe main focus of this document, the panel created the afore- ria are lacking. The panel unanimously agreed that it is mentioned standard because an accurate preoperative eval- conceptually important to classify the cause(s) of stress inuation is obviously indispensable for selection of a treatment continence as urethral hypermobility and/or intrinsic sphincoption. As indicated in the list of recommended evaluation teric deficiency. Yet, there are no universally accepted diagcomponents the essential goals are to ascertain the impact nostic parameters for making these distinctions, and only a of incontinence on quality of life, to demonstrate stress uri- small number of studies reviewed by the panel actually clasnary incontinence objectively, to identify the etiological fac- sified the patients in this manner. (As previously noted, tors contributing to incontinence and to identify prognostic recognition of intrinsic sphincteric deficiency as a diagnostic indicators. entity is relatively new.) The panel also agreed unanimously Standard. The index patient should be informed of the that it is important to record the presence or absence of available surgical alternatives. Patients should also be in- urinary urgency, urge incontinence and detrusor instability formed for each procedure of estimated benefits and risks. before and after treatment, which was rarely done in the The choice of treatment should be made between surgeon and studies the panel reviewed. There are no standardized methpatient, taking into consideration patient preferences and ods of recording outcomes. Moreover, the vast majority of the experience and judgment of the surgeon. In weighing the studies under panel review did not even define the methods potential benefits and harms of each surgical alternative, the by which incontinence was assessed or how the results were surgeon and patient should discuss not only cure/dry and obtained. cure/dry/improvement rates but such issues as length of opThere is a lack of specific intervals over which results were eration, length of hospital stay and possible complications. reported. Although many studies claimed to have long folPatients should be informed especially of the possible occur- lowup periods, the majority did not record the total number rence of common complications, with information on how of patients treated, the number actually evaluated and the they would be treated if they occurred. number lost to followup. Most studies included statements If a surgical treatment is a relatively short procedure with such as, “100 women were followed for 1 t o 10 years with a low risk of short-term morbidity, a patient may choose this mean followup of 5 years.” This kind of statement does not treatment although it has a lower estimated long-term cure/ allow the drawing of meaningful conclusions about followup dry rate than an alternative treatment that requires more beyond the first year. time and poses a higher risk of complications. On the other There is a lack of uniformity in the terminology describing hand, a patient may be willing to accept a longer operation, surgical procedures. Many authors described significant with greater likelihood of complications and short-term mor- modifications of surgical technique under 1 name but the bidity, if there is also greater likelihood of long-term success. actual technique they used was closer to a technique known Option. Although the 4 major types of procedure groupings, by another name. For example, 1 author used the name retropubic suspensions, transvaginal suspensions, anterior “Stamey bladder neck suspension” in the title of the paper repairs and sling procedures, have important differences in but the text indicated a technique that was almost identical their outcomes, all are options for the index patient. Based on to the Raz procedure. evidence from the literature as indicated by the probability Authors fail to indicate whether their study population estimates in table 1 and on expert opinion, retropubic sus- was a new or updated series. A number of authors published pensions and slings are the most efficacious procedures for their results as often as every few years, and in most in-

FEMALE STRESS URINARY INCONTINENCE CLINICAL GUIDELINES PANEL

stances it was not possible to determine how many patients represented followup data and how many were new. Complications and morbidities are not described in the literature consistently, often making i t impossible to determine their true incidence.

REC0MMENL)ED STANDARDS O F EFFICACY FOR CLINICAL TRIALS

The following recommendations were adapted from the ”Standards of Efficacy for Evaluation of Treatment Outcomes in Urinary Incontinence,” developed by a multidisciplinary committee of the Urodynamics Society and approved by the AUA and Urodynamics Society.# These are minimal standards to be used in clinical trials for assessing the efficacy of urinary incontinence therapies. General recommendations. All clinical trials should consist of pretreatment and posttreatment evaluations adhering to specific standards recommended. Protocols should further specify the criteria by which treatment success or failure is determined. Evaluation of surgical, prosthetic and implantation therapies should be conducted no less often than 1 , 6 and 12 months after treatment. Ideally, evaluation should be done a t yearly intervals thereafter. With regard t o surgical therapies, minimum 5-year followup is highly desirable. At each posttreatment interval the number of patients treated during that interval, number of patients actually evaluated and number of patients lost to followup and why should be clearly described. Some therapies, including most of those involving pharmacological agents, exert their effects only during active treatment. They should be evaluated a t intervals dictated by the expected outcome but no less often than monthly. After a successful pilot study, tests of all drug therapies should be double-blinded and placebo controlled. Whenever repeated therapies are contemplated (such as periurethral injections), the indications for re-treatment, number of re-treatments and interval since the last re-treatment should be specified. The efficacy assessments should be conducted a t a specified interval after the last treatment. Pretreatment evaluation. The pretreatment evaluation should consist of a structured micturition history a n d o r questionnaire, structured physical examination, micturition diary, a quantitative measure of urinary loss such as a pad test, urodynamics and a quality of life assessment. The history should specifically include patient age, previous operations that might affect the results, such as pelvic a n d o r incontinence surgery and abdominal-perineal resection of the rectum, pelvic irradiation, neurological conditions known to affect micturition, menstrual and obstetric history, and concurrent medications known to affect micturition. The micturition history a n d o r questionnaire should include a t least number of micturitions day and night, number of incontinent episodes day and night, type of incontinence symptoms (stress, urge, mixed, unconscious, continuous leakage) voiding symptoms, subjective evaluation of severity of symptoms, such as a “bother” index and type and number of pads. The structured physical examination should be conducted with a comfortably full bladder and include a neurourological and vaginal examination. The neurourological examination should assess perineal sensation, the bulbocavernosus reflex, and anal sphincter tone and control. A brief, focused neurologwal examination should be performed to differentiate normal, paraplegc, quadriplegic, hemiplegic, dementia and so forth. The vaginal examination should include demonstration of urinary leakage a s spontaneoudcontinuous, synchronous with stress and after stress, and it should detect the Presence and degree of cystocele, urethral hypermobility, uterine prolapse/vaginal vault prolapse, enterocele and rectocele. An estimate of urethral hypermobility (such as a cotton swab test) should be done.

879

The micturition diary, self-reported by the patient, should include time of micturition, time and type of incontinence, and voided volume. The pad test should be quantitative or semiquantitative. A quantitative pad test estimates the volume of urine loss (for example, by weighing pads). A semiquantitative pad test describes the amount of loss (such as large or small), and urinary dyes such as phenozpyridine hydrochloride may be useful. The specific technique of the pad test should be delineated. The minimum evaluation should consist of cystometry (liquid) with simultaneous measurement of vesical and abdominal pressure for determination of detrusor pressure, detrusor pressure-uroflow study and simple uroflow. A measure or estimate of sphincteric strength and urethral hypermobility (for example, leak point pressure and cotton swab test) should be performed to assess for the relative contribution of urethral hypermobility and intrinsic sphincteric deficiency. The cause of incontinence should be diagnosed as urethral hypermobility a n d o r intrinsic sphincteric deficiency. How the diagnosis was made should be stated (for example examination, cotton swab test, cystourethrogram, ultrasound, leak point pressure). In addition, the presence or absence of urgency a n d o r urge incontinence should be noted a s well as the presence of detrusor instability. If detrusor instability is present a t cystometry, it should be categorized as symptomatic or asymptomatic. Posttreatment evaluation. The posttreatment evaluation should consist of a structured micturition history a n d o r questionnaire at each followup, a structured physical examination with full bladder a t appropriate intervals, micturition diary at each followup, pad test a t each followup, and measurement of uroflow and post-void residual at least once and a t appropriate intervals thereafter. The exact method by which the data for analysis was obtained should be specified, for example retrospective or prospective chart review, patient questionnaire or analysis by independent researcher. The latter is the preferred method. Specific complications and morbidities should be noted, including postoperative death, urinary retention (incidence, duration and whether it was a planned or unexpected outcome), and incidence of postoperative urinary frequency, urgency and urge incontinence (each as a separate data point), and whether persistent or a de novo occurrence.

FUTURE RESEARCH RECOMMENDATIONS

Research recommendations include the development of a standardized lexicon describing the types of incontinence, a standardized diagnostic methodology to define the types of incontinence and standardized outcome measures. Also recommended is the development of methods to identify intrinsic sphincteric deficiency, particularly in patients with concurrent hypermobility, better techniques for management of patients with combined intrinsic sphincteric deficiency and hypermobility, less obstructive techniques for the correction of patients with hypermobility associated stress incontinence and a better understanding of the pathophysiology underlying stress urinary incontinence. Drs. Jeffrey Csiszar, Jill Gerspach, Steven Kurtz, Susan Martins-Levy, Hate1 Patel and Lisa Stout performed data extraction for this clinical guideline report.

REFERENCES

1. Leach, G. E., Dmochowski, R. R., Appell, R. A., Blaivas, J. G., Hadley, H. R., Luber, K. M., Mostwin, J. L. and ODonnell, P. D.: Female Stress Urinary Incontinence Clinical Guidelines Panel: Report on the Surgical Management of Female Stress Urinary Incontinence. Baltimore: American Urological Association, Inc., 1997.

880

FEMALE STRESS URINARY INCONTINENCE CLINICAL GUIDELINES PANEL

2. Eddy, D. M.: The confidence profile method: a bayesian method

for assessing health technologies. Oper. Res., 37: 210, 1989. 3. Eddy, D. M.,Hasselblad. V. and Shachter, R.: A bayesian method for synthesizing evidence. The Confidence Profile Method. Int. J. Technol. Assess. Health Care, 6 31, 1990. 4. Eddy, D. M. and Hasselblad, V.: FAST*PRO. Software for metaanalysis by the confidence profile method. San Diego: Academic Press. Harcourt Brace Jovanovich, 1992. 5. Wingo. P. A.. Huezo, C . M.. Rubin, G . L., Ory, H. W. and Peterson, H. B.: The mortality risk associated with hysterectomy.

h e r . J. Obst. Gynec., 152: 803, 1985. 6. Eddy, D. M.: A Manual for Assessing Health Practices and Designing Practice Policies: The Explicit Approach. Philadelphia: American College of Physicians, 1992. 7. AAFP Positions on the Clinical Aspects of Medical Practice. Kansas City: American Academy of Family Physicians, 1996. 8. The Urodynamics Society: Standards of efficacy for evaluation of treatment outcomes in urinary incontinence: a report from the Urodynamics Society. Neurourol. Urodyn., 1 6 145, 1997.