- Email: [email protected]
The Effect of Study Methodology on Reported Success Rates of the Modified Pereyra Bladder Neck Suspension
Vol. 154, 1732-1735, November 1995 Printed in U.S.A.
THE EFFECT OF STUDY METHODOLOGY ON REPORTED SUCCESS RATES OF THE MODIFIED PEREYRA BLADDER NECK SUSPENSION LARRY T. SIRLS, CHARLES M. KEOLEIAN, HOWARD J. KORMAN
AND
AARON K. KIRKEMO
From the Department of Urology, Henry Ford Hospital, Detroit, Michigan
ABSTRACT
Purpose: We compared surgical results in a cohort of women after modified Pereyra bladder neck suspension using questionnaire based outcomes analysis versus a retrospective chart review. Materials and Methods: Of 151 patients who underwent modified Pereyra bladder neck suspension 102 had complete questionnaire and chart data for review. Mean patient age was 56 years and followup was 25 months. Results: According to outcomes analysis 48 patients (47.1%) were cured and in 65 (64%) stress urinary incontinence improved compared to 74 (72%) cured and 89 (89%)improved by retrospective review. Of the 102 chart review patients 10 (9%)reported daily pad use compared to 55 of the 102 (53%) in the questionnaire study. Conclusions: "his study controls for patient selection, definition of cure and length of followup, and demonstrates that study methodology profoundly affects reported outcomes for the modified Pereyra bladder neck suspension. KEY WORDS:bladder; urinary incontinence, stress; outcome assessment (health care)
The reported success rates of bladder neck suspension procedures for the treatment of stress urinary incontinence in the literature vary. Many factors may account for the differences in reported success rates aRer transvaginal bladder neck suspension procedures, including patient selection, surgical technique, definition of success, physician bias and length of followup. Stamey noted that referral centers may be biased toward decreased success rates secondary to having patients with more complications and severe stress urinary incontinence.' Similarly, patients with a failed Raz modification of the modified Pereyra bladder neck suspension were believed to have had intrinsic sphincter dysfunction initially misdiagnosed as anatomical incontinence.2 Poor surgical technique, such as incorporating inadequate anterior vaginal wall anchoring tissue with the suspension sutures, has also been implicated.2 Another important variable is the investigator definition of cure. Using a strict definition of cure, such as no urine loss regardless of provocative maneuvers, may provide less favorable results than results based on patient improvement. Furthermore, a more global assessment of outcome, such as overall patient satisfaction with surgical intervention, may more accurately reflect patient perspective.3 If the length of followup is inadequate, treatment failures may be missed. It has been suggested that bladder neck suspension procedures fail secondary to suture pull-through from the anchoring tissue of the anterior vaginal wall. Bruskewitz et al demonstrated in an animal model that suture pullthrough was a time-dependent phenomenon related to suture tension.4 Interestingly, in a study of the modified Pereyra bladder neck suspension Kelly et a1 reported that in nearly 25% of the patients with a failed procedure and recurrent hypermobility the failure occurred 2 or more years after ~urgery.~ The literature increasingly suggests that study methodology is a critical factor in reported success rates following transvaginal bladder neck suspension procedures.3 A review of the study methodology in reports of transvaginal bladder neck suspension reveals that studies using questionnaire and interview based outcomes analysis formats have less favorable results than those based on retrospective chart Accepted for publication April 7, 1995
review data. The disparity in surgical success rates reported by retrospective chart review versus patient based outcomes analysis can be dramatic. For example, in 1 retrospective study of the Stamey bladder neck suspension 91%of patients were reportedly cured of stress urinary incontinence.' In contrast, a questionnaire based outcomes study of the Stamey bladder neck suspension reported that 40% of patients were cured of stress urinary incontinence.6 These 2 studies had similar followup. The reported success rates of the modified Pereyra bladder neck suspension have also varied with study methodology. Retrospective chart reviews have reported "cure" or "improved" rates ranging from 77 to 91%.7-l0 However, in an interview and questionnaire based outcomes study using a strict definition of cure Kelly et a1 reported that only 51% of patients were -curedn of stress urinary incontinence and 76% were subjectively "impr~ved."~ We evaluated and reported on a group of women with genuine anatomical stress urinary incontinence treated with the Raz modification of the modified Pereyra bladder neck suspension, using a questionnaire based outcomes analysis format.3 A blinded independent team subsequently performed a retrospective chart review of this patient population. The objective of this study was to determine the influence of study methodology on success rates for modified Pereyra bladder neck suspension. By comparing the results of both study methods in a single population we controlled for the variables of patient selection, surgical technique, definition of success and length of followup. The questionnaire based results from the initial study were compared to the chart review results during the same followup in this single patient population. The combined group and individual patient data for each study method were compared. Our surgical results as assessed by each study method were compared to other questionnaire and chart review studies in the literature. MATERIALS A N D M E T H O D S
Preoperative evaluation and surgical technique. Between 1989 and 1991. 151 patients with urodynamically documented stress urinary incontinence were treated with the Raz modification of the Pereyra bladder neck suspension."
1732
STUDY METHODOLOGY FOR BLADDER NECK SUSPENSION
patients had type 2 (anatomical) genuine stress urinary incontinence based on history, physical examination, and fluoroscopic assessment of the bladder neck and urethra.11 Patients were assessed for imtative voiding symptoms, obstructive voiding symptoms and pad use. Urethral hypermobility and associated pathological conditions (cystocele, rectocele, enterocele or vault prolapse) were noted at physical examination. Routine urinalysis, urine cytology and postvoid residual volume measurements were done. A multichannel video urodynamic evaluation was performed using a 1OF triple lumen catheter and a medium fill rate. The presence or absence of detrusor instability was noted. Surgica2 technique. The modified Pereyra bladder neck suspension was performed using permanent suture incorporating the anterior vaginal wall and urethropelvic ligament.10 All bladder neck suspension procedures were performed by a single surgeon (A.K.K.). Associated vaginal vault pathological conditions were repaired concurrently. Perioperative antibiotics were routinely administered intravenously. Postoperatively patients were treated with a suprapubic tube or clean intermittent catheterization regimen. Postoperative evaluation. Of the 151 patients 106 had previously responded to the questionnaire based outcome analysis study.3 Of the 106 charts 102 were available for review. The charts were reviewed by one of us (C. M. KJ who was blinded to the individual results of the questionnaire based outcome analysis. The chart review was limited to outpatient notes logged by the surgeon or referring physician during the same followup as the questionnaire based outcome study. The questionnaire is shown in the Appendix.3 The questionnaire addressed stress urinary incontinence, pad use, obstructive and imtative symptoms, and overall satisfaction. The responses to questions 9 and 15 were incorporated into the stress urinary incontinence results. Imtative voiding symptoms (questions 6 to 8) and obstructive voiding symptoms (questions 2 to 5)were scored. The number and types of pads used for protection were assessed (questions 10 and 11). Although the questionnaire does not directly address quality of life, factors (such as dyspareunia and pelvic pain) that may affect quality of life were addressed. The retrospective chart review included the preoperative evaluation, and the postoperative history and physical exam-
~
gorized as worse. The quantity and type of pad use were assessed. The types of protection reported were panty liners, minipads, maxipads or diapers. Posttreatment irritative symptoms were defined as frepuency, nocturia, urgency and/or urge incontinence. De novo h t a t i v e symptoms and the persistence of irritative symp hmS were recorded. If irritative symptoms were persistent, they were compared to the preoperative irritative symptom and classified as better, the same or worse.
1733
Statistical analysis. Results were tested for symmetry of agreement between the retrospective chart review and the questionnaire based outcomes analysis. Marginal homogeneity was assessed, determining if the percent reported in each category correlated between retrospective chart review and questionnaire based analysis. Further assessment of each response category was performed by the McNemar test for paired data with significance defined by the chi-square statistic with 3 degrees of freedom. A kappa statistic was generated measuring the amount of agreement between 2 variables beyond that expected by chance alone. RESULTS
Of the 106 patients responding to the initial outcomes analysis study 102 (96%) underwent a retrospective chart review. Mean patient age was 56 years (range 18.8 to 81.5). Mean followup of the questionnaire study was 25 months (range 8.5 to 44)and mean followup calculated from the chart review study was 23 months (range 4 to 44). Retrospective notes were only reviewed if logged during the perid of the questionnaire based study. A total of 41 patients underwent only a m&ed Pereyra bladder neck suspension. Concurrent vaginal vault pathological conditions were corrected in 61 patients (60%), including a c y s t o c e l d d l e in 27,cystocele in 21,cystoceldrectoceldentemlein 7 and rectocele in 6. Stress urimry incontinence. The chart review and questionnaire based results for stress urinary incontinence are shown in the table. Using the same definition of cure (no urine loss regardless of provocative maneuvers) the chart review stress urinary incontinence cure rate of 74 of 102 patients (72%) was significantly m e r e n t from the questionnaire based cure rate of 48 of 102 (47%, chi-square, p <0.001). The results of the 2 study methods for stress urinary incontinence were then compared for each individual patient with each serving as her own control. Of the 74 patients described as "cured" of stress urinary incontinence in the chart review only 42 (57%)reported no incontinence with provocative maneuvers in the questionnaire based outcomes study. Of the 91 patients in whom stress urinary incontinence was classified as "cured" or "better"in the chart review
Improved same
worse
Pad use: None 1 3 Minipads
daily
1-4 Mexipads or
17 (17) 1 (1) 10 (10)
17 (17). 27 (26)* 10 (10).
92 (91) 6 (6)
48 (47) 33 (32)
4 (3) 21 (21) diapers daily * sie;nifieantly differentby chi-square end marginal homogeneiw4- (P
CO.001).
1734
STUDY METHODOLOGY FOR BLADDER NECK SUSPENSION
charting error, physician bias and patient desire to please the physician. In contrast, questionnaires allow assessment of patient satisfaction without confronting the surgeon directly. Questionnaires may also provide a more realistic estimate of overall patient satisfaction with outcome than simply if a patient has urine loss with provocative maneuvers. An example is the importance of assessing the role of irritative symptoms in patient outcome since these symptoms can be debilitating and are often difficult to treat. Persistent or de novo irritative symptoms are known to impact adversely patient sati~faction.~. l2 The limited number of questions detailing irritative symptoms in the questionnaire, specifically if irritative symptoms were persistent or de novo, illustrates a potential design problem with any questionnaire based study. Other potential shortcomings of questionnaire studies include the difficulty in assessing recurrent vaginal vault pathological conditions and postoperative complications. These types of objective data are likely to be better documented in patient charts. It is interesting that our chart review findings are comparable to those of previous chart review studies and our questionnaire based findings are comparable to those of other questionnaire based studies of modified Pereyra bladder neck suspension. The charted stress urinary incontinence cure or improved rate of 89% in our study is comparable to other retrospective chart review results of 77 to 91%.7-’0 Surgical technique is difficult to assess and compare but our rate of technical complications is similar to that reported by Raz et a1 (postoperative vaginal vault prolapse in 8% of patients versus 6% and prolonged urinary retention in 3.99 versus 2.5%h2 Similarly our questionnaire based results of 47% “cured” of stress urinary incontinence and 67% “improved” compares to the physician-blinded questionnaire based study reported by Kelly et a1 of 51% “cured” and 76% “subjectively improved.”6 This correlation with previous studies in the literature supports the critical role of study methodology on modified Pereyra bladder neck suspension outcome. It is clear from our study that the current mechanism of DISCUSSION evaluating patients after bladder neck suspension needs to Previous studies of transvaginal bladder neck suspension procedures have documented disparate results. Differences become standized. The methodology of evaluating outcomes in patient selection, surgical expertise, definition of cure, for treatment modalities has been under intense review and physician bias and length of followup have been implicated has appropriately evolved to emphasize patient perceived as factors contributing to this disparity. Our current study outcome. It is not sufficient to state that patients “cured” of controls for these potential artifacts to isolate the impact of stress incontinence are satisfied with the outcome since they study methodology on reported results specifically.Using the may be dissatisfied with persistent or de novo irritative same definition of cure in the same patient population the symptoms. An emphasis on patient satisfaction with outcome questionnaire based outcomes analysis reported that 47% is appropriate when dealing with treatments that affect qualwere cured of stress urinary incontinence and 64% had im- ity of life.13 Most patients with stress urinary incontinence or provement, while the retrospective chart review reported urgency symptoms seek treatment for quality of life reasons. 89% of patients as “cured” or “improved.” If the 2 study Few will have significant pathological problems, such as a methodologies used were ideal, the results for this single large cystocele with outlet obstruction, elevated post-void patient cohort would have been identical. These data dem- residual or hydronephrosis. Our goal with symptomatic paonstrate that study methodology is a critical determinant of tients should-be to characterize the dominant complaint and reported success after modified Pereyra bladder neck suspen- advise if our treatment modalities are capable of effectively alleviating this complaint. It may be necessary to integrate a sion. Our retrospective chart review data contrast with the standard assessment of the degree of dissatisfaction with the questionnaire based outcomes analysis data when compared voiding problem, which is commonly done with bothersomeas a group and when patients were compared individually. ness scores. Although this type of in-depth information may The retrospective chart data providing surgeon perspective be obtained with a standard interview, it is time-consuming suggest a more favorable outcome for the treatment of stress and inefficient. A prospective validated questionnaire assessurinary incontinence. Of the patients 89% were documented ing voiding symptoms, their impact on quality of life and in the chart as “cured” of stress urinary incontinence or bothersomeness to the patient is an essential means of un‘improved” compared to 64% in the questionnaire based out- derstanding our patients. This global evaluation would be comes analysis.3 Although these data are not urodynamically combined with objective parameters, such as the degree of documented, they are internally consistent with the degree of anatomical defect, urethral function and the presence of depad use supporting the reported incontinence rates. The ret- trusor instability to establish a pretreatment baseline. After rospective chart review reported less pad use supporting a intervention (behavioral, pharmacological or surgical) repeat lower seventy of incontinence (see table). objective evaluation and a followup questionnaire would alRetrospective chart reviews are potentially influenced by low a standard assessment of therapeutic efficacy and mean-
( h e m or minipads) and maximal protection (maxipads or diapers). The results of the 2 study methods for pad use were then compared for each individual patient with each serving as her own control. Of the 92 patients who did not require pads according to the chart review 54 (59%) reported pad use in the questionnaire study. Of these 54 patients 37 used minipads, 11 used maxipads and 6 used diapers for protection. In contrast, the chart review revealed that only 1 patient used pads for protection who did not report pad use by questionnaire. Zrritatiue symptoms. Of the 102 patients 68 (67%) evaluated by chart review reported preoperative irritative symptoms.Of these 68 patients 25 (37%) described complete resolution of the irritative symptoms. Of those with persistent irritative symptoms the symptoms were “better” in 11 (IS%), the “same” in 20 (29%)and “worse” in 12 (17%). In addition, 17 patients (17%) had de novo irritative symptoms after modified Pereyra bladder neck suspension. The results of the 2 study methods for irritative symptoms were then compared for each individual patient with each serving as her own control. Postoperatively 60 of the 102 patients were recorded as having bothersome irritative symptoms by chart review. Of note, 22 of the 42 patients (52%) without documented irritative symptoms by chart review reported postoperative urge incontinence by questionnaire. Complications. Of the 102 patients in the chart review complications included prolonged urinary retention in 4,pelvic hematoma in 2,surgical release of 1 suspension suture to relieve suprapubic discomfort in 2, deep venous thrombosis in 1, recurrent urinary tract infections in 1, suprapubic wound infection in 1 and pseudomembranous colitis requiring hospitalization in 1.In 8 patients (8%)symptomatic vaginal prolapse developed that required treatment postoperatively. Two patients had severe incontinence secondary to documented intrinsic sphincter deficiency.
STUDY METHODOLOGY FOR BLADDER NECK SUSPENSION
bgf~l comparisons between treatment modalities and institutions. CONCLUSIONS
Our study demonstrates that study methodology profoundly affects reported outcomes for the modified Pereyra bladder neck suspension. Other important factors to consider when reviewing published reports include patient selection, length of followup and definitions of cure. Combined with objective findings a standard validated prospectively administered questionnaire is needed for accurate comparison of surgical outcomes. Questionnaires have the advantage of best assessing patient satisfaction with treatment, arguably the most important parameter in clinical medicine. Chart review data remain an important source of information, such as findings on physical examination, urodynamic data and treatment complications. Standard outcomes research is needed to allow accurate comparison of different treatment modalities or modifications, such as changes in surgical technique.
1735
13. If you are having intercourse, is it painful? (Yes, No, Sometimes) 14. Knowing what you know now, would you still have chosen to have surgery for this condition? (Yes, NO, Maybe) 15. All things considered, how is your urine control compared to what it was before the operation? (Better, Same, Worse) 16. What medications are you currently taking? 17. What other surgeries have you had since the bladder surgery? REFERENCES
1. Stamey, T. A.: Endoscopic suspension of the vesical neck for urinary incontinence in females. Report on 203 consecutive patients. Ann. Surg., 192: 466, 1980. 2. Raz,S.,Sussman, E. M., Erickson, D. B., Bregg, K J. and Nitti, V. W.: The Raz bladder neck suspension: results in 206 patients. J. Urol., 148: 845,1992. 3. Korman, H. J.. Sirls, L. T. and Kirkemo, A. El: Success rate of modified Pereyra bladder neck suspension determined by outcomes analysis. J. Urol., 152 1453,1994. APPENDIX: PATIENT ADMINISTERED QUESTIONNAIRE USED 4. Bruskewitz, R.C.,Nielsen, K. T., Graversen, P. H., Saville, W. D. IN OUTCOMES ANALYSIS3 and Gasser, T. C.: Bladder neck suspension material investigated in a rabbit model. J. Urol., 142: 1361,1989. 1. Compared to before surgery, how is your urinary 5. Kelly, M. J., Knielsen, K., Bruskewitz, R, h k a m p , D. and stream? (Better, Same, Worse) Leach, G. E.: Symptom analysis of patients undergoing modi2. Do you have to strain to urinate? (Always, Somefied Pereyra bladder neck suspension for stress urinary incontimes, Never) tinence. Pre- and postoperative findings. Urology, 31: 213, 3. Do you feel that you empty your bladder com1991. pletely? (Always, Sometimes, Never) 6. Walker, G. T.and Texter, J. H., Jr.: Success and patient satis4. Do you have dribbling after you void? (Always, faction following the Stamey procedure for stress urinary inSometimes, Never) continence. J. Urol., 141: 1521,1992. 7. McDougal, E. M., Klutke, C. G. and Tiemann, D. A. Comparison 5. Do you use a catheter in order to empty your bladof two surgical procedures for the management of anatomical der? (Yes, No, Sometimes) female stress urinary incontinence. J. Urol., part 2,141:2794 6. When you get the urge to urinate, can you make it abstract 261,1992. to the bathroom without wetting? (Always, Some8. Golomb, J., Goldwasser, B. and Mashiach, S.: Raz bladder neck times, Never) suspension in women younger than 65 years compared to 7. During the daytime, how often do you urinate? elderly w o m e n 3 years experience. J.Urol., part 2,149:404A, (To hours between voids) abstract 764,1993. How often do you urinate at night? ( TO __ 8. 9. Paty, R., Maidenberg, M. and Badlani, G.: Long term continence rate with Raz urethropexy. J. Urol., part 2,141:279& abstract hours between voids) 262,1992. 9. Do you leak with any of the following maneuvers? (Cough, Strain, Sneeze, Run, Jog, Walk, Change 10. Raz, S.: Modified bladder neck suspension for female stress incontinence. Urology, 11: 82,1981. position, Lying flat, Simply standing, Always) J. G. and Olsson. C. A: Stress incontinence: classifica10. Do you have to wear pads to stay dry? (To __ 11. Blaivas, tion and surgical approach. J. Urol., 139.727,1988. minipads a day, -To -maxipads a day, - 12. Kelly, M. J., Zimmern, P. E. and Leach, G. E.:Complications of To diapers a day) bladder neck suspension procedures. Urol. Clin. N. h e r . , 18: 11. Do you have to change panties during the day? 339. 1991. TO __ times a day) 13. H o l G w e , H. L.:Outcomes research and BPH, new concepts for deciding therapy. AUA Update Series, vol. 10, lesson 20, p. 12. Do you have problems with pelvic pain? (Yes, NO, 154, 1991. Sometimes) ~
~
(z
THE EFFECT OF STUDY METHODOLOGY ON REPORTED SUCCESS RATES OF THE MODIFIED PEREYRA BLADDER NECK SUSPENSION LARRY T. SIRLS, CHARLES M. KEOLEIAN, HOWARD J. KORMAN
AND
AARON K. KIRKEMO
From the Department of Urology, Henry Ford Hospital, Detroit, Michigan
ABSTRACT
Purpose: We compared surgical results in a cohort of women after modified Pereyra bladder neck suspension using questionnaire based outcomes analysis versus a retrospective chart review. Materials and Methods: Of 151 patients who underwent modified Pereyra bladder neck suspension 102 had complete questionnaire and chart data for review. Mean patient age was 56 years and followup was 25 months. Results: According to outcomes analysis 48 patients (47.1%) were cured and in 65 (64%) stress urinary incontinence improved compared to 74 (72%) cured and 89 (89%)improved by retrospective review. Of the 102 chart review patients 10 (9%)reported daily pad use compared to 55 of the 102 (53%) in the questionnaire study. Conclusions: "his study controls for patient selection, definition of cure and length of followup, and demonstrates that study methodology profoundly affects reported outcomes for the modified Pereyra bladder neck suspension. KEY WORDS:bladder; urinary incontinence, stress; outcome assessment (health care)
The reported success rates of bladder neck suspension procedures for the treatment of stress urinary incontinence in the literature vary. Many factors may account for the differences in reported success rates aRer transvaginal bladder neck suspension procedures, including patient selection, surgical technique, definition of success, physician bias and length of followup. Stamey noted that referral centers may be biased toward decreased success rates secondary to having patients with more complications and severe stress urinary incontinence.' Similarly, patients with a failed Raz modification of the modified Pereyra bladder neck suspension were believed to have had intrinsic sphincter dysfunction initially misdiagnosed as anatomical incontinence.2 Poor surgical technique, such as incorporating inadequate anterior vaginal wall anchoring tissue with the suspension sutures, has also been implicated.2 Another important variable is the investigator definition of cure. Using a strict definition of cure, such as no urine loss regardless of provocative maneuvers, may provide less favorable results than results based on patient improvement. Furthermore, a more global assessment of outcome, such as overall patient satisfaction with surgical intervention, may more accurately reflect patient perspective.3 If the length of followup is inadequate, treatment failures may be missed. It has been suggested that bladder neck suspension procedures fail secondary to suture pull-through from the anchoring tissue of the anterior vaginal wall. Bruskewitz et al demonstrated in an animal model that suture pullthrough was a time-dependent phenomenon related to suture tension.4 Interestingly, in a study of the modified Pereyra bladder neck suspension Kelly et a1 reported that in nearly 25% of the patients with a failed procedure and recurrent hypermobility the failure occurred 2 or more years after ~urgery.~ The literature increasingly suggests that study methodology is a critical factor in reported success rates following transvaginal bladder neck suspension procedures.3 A review of the study methodology in reports of transvaginal bladder neck suspension reveals that studies using questionnaire and interview based outcomes analysis formats have less favorable results than those based on retrospective chart Accepted for publication April 7, 1995
review data. The disparity in surgical success rates reported by retrospective chart review versus patient based outcomes analysis can be dramatic. For example, in 1 retrospective study of the Stamey bladder neck suspension 91%of patients were reportedly cured of stress urinary incontinence.' In contrast, a questionnaire based outcomes study of the Stamey bladder neck suspension reported that 40% of patients were cured of stress urinary incontinence.6 These 2 studies had similar followup. The reported success rates of the modified Pereyra bladder neck suspension have also varied with study methodology. Retrospective chart reviews have reported "cure" or "improved" rates ranging from 77 to 91%.7-l0 However, in an interview and questionnaire based outcomes study using a strict definition of cure Kelly et a1 reported that only 51% of patients were -curedn of stress urinary incontinence and 76% were subjectively "impr~ved."~ We evaluated and reported on a group of women with genuine anatomical stress urinary incontinence treated with the Raz modification of the modified Pereyra bladder neck suspension, using a questionnaire based outcomes analysis format.3 A blinded independent team subsequently performed a retrospective chart review of this patient population. The objective of this study was to determine the influence of study methodology on success rates for modified Pereyra bladder neck suspension. By comparing the results of both study methods in a single population we controlled for the variables of patient selection, surgical technique, definition of success and length of followup. The questionnaire based results from the initial study were compared to the chart review results during the same followup in this single patient population. The combined group and individual patient data for each study method were compared. Our surgical results as assessed by each study method were compared to other questionnaire and chart review studies in the literature. MATERIALS A N D M E T H O D S
Preoperative evaluation and surgical technique. Between 1989 and 1991. 151 patients with urodynamically documented stress urinary incontinence were treated with the Raz modification of the Pereyra bladder neck suspension."
1732
STUDY METHODOLOGY FOR BLADDER NECK SUSPENSION
patients had type 2 (anatomical) genuine stress urinary incontinence based on history, physical examination, and fluoroscopic assessment of the bladder neck and urethra.11 Patients were assessed for imtative voiding symptoms, obstructive voiding symptoms and pad use. Urethral hypermobility and associated pathological conditions (cystocele, rectocele, enterocele or vault prolapse) were noted at physical examination. Routine urinalysis, urine cytology and postvoid residual volume measurements were done. A multichannel video urodynamic evaluation was performed using a 1OF triple lumen catheter and a medium fill rate. The presence or absence of detrusor instability was noted. Surgica2 technique. The modified Pereyra bladder neck suspension was performed using permanent suture incorporating the anterior vaginal wall and urethropelvic ligament.10 All bladder neck suspension procedures were performed by a single surgeon (A.K.K.). Associated vaginal vault pathological conditions were repaired concurrently. Perioperative antibiotics were routinely administered intravenously. Postoperatively patients were treated with a suprapubic tube or clean intermittent catheterization regimen. Postoperative evaluation. Of the 151 patients 106 had previously responded to the questionnaire based outcome analysis study.3 Of the 106 charts 102 were available for review. The charts were reviewed by one of us (C. M. KJ who was blinded to the individual results of the questionnaire based outcome analysis. The chart review was limited to outpatient notes logged by the surgeon or referring physician during the same followup as the questionnaire based outcome study. The questionnaire is shown in the Appendix.3 The questionnaire addressed stress urinary incontinence, pad use, obstructive and imtative symptoms, and overall satisfaction. The responses to questions 9 and 15 were incorporated into the stress urinary incontinence results. Imtative voiding symptoms (questions 6 to 8) and obstructive voiding symptoms (questions 2 to 5)were scored. The number and types of pads used for protection were assessed (questions 10 and 11). Although the questionnaire does not directly address quality of life, factors (such as dyspareunia and pelvic pain) that may affect quality of life were addressed. The retrospective chart review included the preoperative evaluation, and the postoperative history and physical exam-
~
gorized as worse. The quantity and type of pad use were assessed. The types of protection reported were panty liners, minipads, maxipads or diapers. Posttreatment irritative symptoms were defined as frepuency, nocturia, urgency and/or urge incontinence. De novo h t a t i v e symptoms and the persistence of irritative symp hmS were recorded. If irritative symptoms were persistent, they were compared to the preoperative irritative symptom and classified as better, the same or worse.
1733
Statistical analysis. Results were tested for symmetry of agreement between the retrospective chart review and the questionnaire based outcomes analysis. Marginal homogeneity was assessed, determining if the percent reported in each category correlated between retrospective chart review and questionnaire based analysis. Further assessment of each response category was performed by the McNemar test for paired data with significance defined by the chi-square statistic with 3 degrees of freedom. A kappa statistic was generated measuring the amount of agreement between 2 variables beyond that expected by chance alone. RESULTS
Of the 106 patients responding to the initial outcomes analysis study 102 (96%) underwent a retrospective chart review. Mean patient age was 56 years (range 18.8 to 81.5). Mean followup of the questionnaire study was 25 months (range 8.5 to 44)and mean followup calculated from the chart review study was 23 months (range 4 to 44). Retrospective notes were only reviewed if logged during the perid of the questionnaire based study. A total of 41 patients underwent only a m&ed Pereyra bladder neck suspension. Concurrent vaginal vault pathological conditions were corrected in 61 patients (60%), including a c y s t o c e l d d l e in 27,cystocele in 21,cystoceldrectoceldentemlein 7 and rectocele in 6. Stress urimry incontinence. The chart review and questionnaire based results for stress urinary incontinence are shown in the table. Using the same definition of cure (no urine loss regardless of provocative maneuvers) the chart review stress urinary incontinence cure rate of 74 of 102 patients (72%) was significantly m e r e n t from the questionnaire based cure rate of 48 of 102 (47%, chi-square, p <0.001). The results of the 2 study methods for stress urinary incontinence were then compared for each individual patient with each serving as her own control. Of the 74 patients described as "cured" of stress urinary incontinence in the chart review only 42 (57%)reported no incontinence with provocative maneuvers in the questionnaire based outcomes study. Of the 91 patients in whom stress urinary incontinence was classified as "cured" or "better"in the chart review
Improved same
worse
Pad use: None 1 3 Minipads
daily
1-4 Mexipads or
17 (17) 1 (1) 10 (10)
17 (17). 27 (26)* 10 (10).
92 (91) 6 (6)
48 (47) 33 (32)
4 (3) 21 (21) diapers daily * sie;nifieantly differentby chi-square end marginal homogeneiw4- (P
CO.001).
1734
STUDY METHODOLOGY FOR BLADDER NECK SUSPENSION
charting error, physician bias and patient desire to please the physician. In contrast, questionnaires allow assessment of patient satisfaction without confronting the surgeon directly. Questionnaires may also provide a more realistic estimate of overall patient satisfaction with outcome than simply if a patient has urine loss with provocative maneuvers. An example is the importance of assessing the role of irritative symptoms in patient outcome since these symptoms can be debilitating and are often difficult to treat. Persistent or de novo irritative symptoms are known to impact adversely patient sati~faction.~. l2 The limited number of questions detailing irritative symptoms in the questionnaire, specifically if irritative symptoms were persistent or de novo, illustrates a potential design problem with any questionnaire based study. Other potential shortcomings of questionnaire studies include the difficulty in assessing recurrent vaginal vault pathological conditions and postoperative complications. These types of objective data are likely to be better documented in patient charts. It is interesting that our chart review findings are comparable to those of previous chart review studies and our questionnaire based findings are comparable to those of other questionnaire based studies of modified Pereyra bladder neck suspension. The charted stress urinary incontinence cure or improved rate of 89% in our study is comparable to other retrospective chart review results of 77 to 91%.7-’0 Surgical technique is difficult to assess and compare but our rate of technical complications is similar to that reported by Raz et a1 (postoperative vaginal vault prolapse in 8% of patients versus 6% and prolonged urinary retention in 3.99 versus 2.5%h2 Similarly our questionnaire based results of 47% “cured” of stress urinary incontinence and 67% “improved” compares to the physician-blinded questionnaire based study reported by Kelly et a1 of 51% “cured” and 76% “subjectively improved.”6 This correlation with previous studies in the literature supports the critical role of study methodology on modified Pereyra bladder neck suspension outcome. It is clear from our study that the current mechanism of DISCUSSION evaluating patients after bladder neck suspension needs to Previous studies of transvaginal bladder neck suspension procedures have documented disparate results. Differences become standized. The methodology of evaluating outcomes in patient selection, surgical expertise, definition of cure, for treatment modalities has been under intense review and physician bias and length of followup have been implicated has appropriately evolved to emphasize patient perceived as factors contributing to this disparity. Our current study outcome. It is not sufficient to state that patients “cured” of controls for these potential artifacts to isolate the impact of stress incontinence are satisfied with the outcome since they study methodology on reported results specifically.Using the may be dissatisfied with persistent or de novo irritative same definition of cure in the same patient population the symptoms. An emphasis on patient satisfaction with outcome questionnaire based outcomes analysis reported that 47% is appropriate when dealing with treatments that affect qualwere cured of stress urinary incontinence and 64% had im- ity of life.13 Most patients with stress urinary incontinence or provement, while the retrospective chart review reported urgency symptoms seek treatment for quality of life reasons. 89% of patients as “cured” or “improved.” If the 2 study Few will have significant pathological problems, such as a methodologies used were ideal, the results for this single large cystocele with outlet obstruction, elevated post-void patient cohort would have been identical. These data dem- residual or hydronephrosis. Our goal with symptomatic paonstrate that study methodology is a critical determinant of tients should-be to characterize the dominant complaint and reported success after modified Pereyra bladder neck suspen- advise if our treatment modalities are capable of effectively alleviating this complaint. It may be necessary to integrate a sion. Our retrospective chart review data contrast with the standard assessment of the degree of dissatisfaction with the questionnaire based outcomes analysis data when compared voiding problem, which is commonly done with bothersomeas a group and when patients were compared individually. ness scores. Although this type of in-depth information may The retrospective chart data providing surgeon perspective be obtained with a standard interview, it is time-consuming suggest a more favorable outcome for the treatment of stress and inefficient. A prospective validated questionnaire assessurinary incontinence. Of the patients 89% were documented ing voiding symptoms, their impact on quality of life and in the chart as “cured” of stress urinary incontinence or bothersomeness to the patient is an essential means of un‘improved” compared to 64% in the questionnaire based out- derstanding our patients. This global evaluation would be comes analysis.3 Although these data are not urodynamically combined with objective parameters, such as the degree of documented, they are internally consistent with the degree of anatomical defect, urethral function and the presence of depad use supporting the reported incontinence rates. The ret- trusor instability to establish a pretreatment baseline. After rospective chart review reported less pad use supporting a intervention (behavioral, pharmacological or surgical) repeat lower seventy of incontinence (see table). objective evaluation and a followup questionnaire would alRetrospective chart reviews are potentially influenced by low a standard assessment of therapeutic efficacy and mean-
( h e m or minipads) and maximal protection (maxipads or diapers). The results of the 2 study methods for pad use were then compared for each individual patient with each serving as her own control. Of the 92 patients who did not require pads according to the chart review 54 (59%) reported pad use in the questionnaire study. Of these 54 patients 37 used minipads, 11 used maxipads and 6 used diapers for protection. In contrast, the chart review revealed that only 1 patient used pads for protection who did not report pad use by questionnaire. Zrritatiue symptoms. Of the 102 patients 68 (67%) evaluated by chart review reported preoperative irritative symptoms.Of these 68 patients 25 (37%) described complete resolution of the irritative symptoms. Of those with persistent irritative symptoms the symptoms were “better” in 11 (IS%), the “same” in 20 (29%)and “worse” in 12 (17%). In addition, 17 patients (17%) had de novo irritative symptoms after modified Pereyra bladder neck suspension. The results of the 2 study methods for irritative symptoms were then compared for each individual patient with each serving as her own control. Postoperatively 60 of the 102 patients were recorded as having bothersome irritative symptoms by chart review. Of note, 22 of the 42 patients (52%) without documented irritative symptoms by chart review reported postoperative urge incontinence by questionnaire. Complications. Of the 102 patients in the chart review complications included prolonged urinary retention in 4,pelvic hematoma in 2,surgical release of 1 suspension suture to relieve suprapubic discomfort in 2, deep venous thrombosis in 1, recurrent urinary tract infections in 1, suprapubic wound infection in 1 and pseudomembranous colitis requiring hospitalization in 1.In 8 patients (8%)symptomatic vaginal prolapse developed that required treatment postoperatively. Two patients had severe incontinence secondary to documented intrinsic sphincter deficiency.
STUDY METHODOLOGY FOR BLADDER NECK SUSPENSION
bgf~l comparisons between treatment modalities and institutions. CONCLUSIONS
Our study demonstrates that study methodology profoundly affects reported outcomes for the modified Pereyra bladder neck suspension. Other important factors to consider when reviewing published reports include patient selection, length of followup and definitions of cure. Combined with objective findings a standard validated prospectively administered questionnaire is needed for accurate comparison of surgical outcomes. Questionnaires have the advantage of best assessing patient satisfaction with treatment, arguably the most important parameter in clinical medicine. Chart review data remain an important source of information, such as findings on physical examination, urodynamic data and treatment complications. Standard outcomes research is needed to allow accurate comparison of different treatment modalities or modifications, such as changes in surgical technique.
1735
13. If you are having intercourse, is it painful? (Yes, No, Sometimes) 14. Knowing what you know now, would you still have chosen to have surgery for this condition? (Yes, NO, Maybe) 15. All things considered, how is your urine control compared to what it was before the operation? (Better, Same, Worse) 16. What medications are you currently taking? 17. What other surgeries have you had since the bladder surgery? REFERENCES
1. Stamey, T. A.: Endoscopic suspension of the vesical neck for urinary incontinence in females. Report on 203 consecutive patients. Ann. Surg., 192: 466, 1980. 2. Raz,S.,Sussman, E. M., Erickson, D. B., Bregg, K J. and Nitti, V. W.: The Raz bladder neck suspension: results in 206 patients. J. Urol., 148: 845,1992. 3. Korman, H. J.. Sirls, L. T. and Kirkemo, A. El: Success rate of modified Pereyra bladder neck suspension determined by outcomes analysis. J. Urol., 152 1453,1994. APPENDIX: PATIENT ADMINISTERED QUESTIONNAIRE USED 4. Bruskewitz, R.C.,Nielsen, K. T., Graversen, P. H., Saville, W. D. IN OUTCOMES ANALYSIS3 and Gasser, T. C.: Bladder neck suspension material investigated in a rabbit model. J. Urol., 142: 1361,1989. 1. Compared to before surgery, how is your urinary 5. Kelly, M. J., Knielsen, K., Bruskewitz, R, h k a m p , D. and stream? (Better, Same, Worse) Leach, G. E.: Symptom analysis of patients undergoing modi2. Do you have to strain to urinate? (Always, Somefied Pereyra bladder neck suspension for stress urinary incontimes, Never) tinence. Pre- and postoperative findings. Urology, 31: 213, 3. Do you feel that you empty your bladder com1991. pletely? (Always, Sometimes, Never) 6. Walker, G. T.and Texter, J. H., Jr.: Success and patient satis4. Do you have dribbling after you void? (Always, faction following the Stamey procedure for stress urinary inSometimes, Never) continence. J. Urol., 141: 1521,1992. 7. McDougal, E. M., Klutke, C. G. and Tiemann, D. A. Comparison 5. Do you use a catheter in order to empty your bladof two surgical procedures for the management of anatomical der? (Yes, No, Sometimes) female stress urinary incontinence. J. Urol., part 2,141:2794 6. When you get the urge to urinate, can you make it abstract 261,1992. to the bathroom without wetting? (Always, Some8. Golomb, J., Goldwasser, B. and Mashiach, S.: Raz bladder neck times, Never) suspension in women younger than 65 years compared to 7. During the daytime, how often do you urinate? elderly w o m e n 3 years experience. J.Urol., part 2,149:404A, (To hours between voids) abstract 764,1993. How often do you urinate at night? ( TO __ 8. 9. Paty, R., Maidenberg, M. and Badlani, G.: Long term continence rate with Raz urethropexy. J. Urol., part 2,141:279& abstract hours between voids) 262,1992. 9. Do you leak with any of the following maneuvers? (Cough, Strain, Sneeze, Run, Jog, Walk, Change 10. Raz, S.: Modified bladder neck suspension for female stress incontinence. Urology, 11: 82,1981. position, Lying flat, Simply standing, Always) J. G. and Olsson. C. A: Stress incontinence: classifica10. Do you have to wear pads to stay dry? (To __ 11. Blaivas, tion and surgical approach. J. Urol., 139.727,1988. minipads a day, -To -maxipads a day, - 12. Kelly, M. J., Zimmern, P. E. and Leach, G. E.:Complications of To diapers a day) bladder neck suspension procedures. Urol. Clin. N. h e r . , 18: 11. Do you have to change panties during the day? 339. 1991. TO __ times a day) 13. H o l G w e , H. L.:Outcomes research and BPH, new concepts for deciding therapy. AUA Update Series, vol. 10, lesson 20, p. 12. Do you have problems with pelvic pain? (Yes, NO, 154, 1991. Sometimes) ~
~
(z