INCIDENCE OF FECAL AND URINARY INCONTINENCE FOLLOWING RADICAL PERINEAL AND RETROPUBIC PROSTATECTOMY IN A NATIONAL POPULATION

00225347~1198/1602-0454$03.00/0

Vol. 160, 454-458, August 1998 Printed in U S A .

TtlE JOURNAL OF UROLOGY Copyright 8 1998 by h m c m U R O ~ ~ I C ASSOCU~ON, AL. hc.

INCIDENCE OF FECAL AND URINARY INCONTINENCE FOLLOWING RADICAL PERINEAL AND RETROPUBIC PROSTATECTOMY IN A NATIONAL POPULATION JAY T. BISHOFF,* GARRICK MOTLEY, SCOTT A. OPTENBERG, CATHERINE R. STEIN, KATHLEEN A. MOON, SCOTT M. BROWNING, EDMUND SABANEGH, JOHN P. FOLEY AND IAN M. THOMPSON* From the Departments of Urology, Brooke Army Medical Center and Wilford Hall Medical Center, and the Center for Healthcare Education and Studies, United States Army Medical Department Center and School, Sun Antonio, Texas

ABSTRACT

Purpose: Since 1991 we have performed more than 300 anatomical radical perineal prostatectomies at Brooke Army and Wilford Hall Medical Centers, and were initially aware of 8 patients who presented with unsolicited postoperative fecal incontinence. We determined the incidence of fecal and urinary incontinence following radical prostatectomy, defined parameters to identify patients at risk for fecal complaints following radical prostatectomy, and estimated the impact of fecal incontinence on lifestyle and activities. Materials and Methods: Initially a validated 26-question telephone survey was used to evaluate 227 patients who had previously undergone radical prostatectomy a t 1of our 2 institutions. Based on results of the telephone survey a national survey was mailed to 1,200 radical prostatectomy patients randomly selected from a nationwide database of Department of Defense health care system beneficiaries. All patients had undergone radical perineal or retropubic prostatectomy at least 12 months before being contacted for the survey. Results: Responses to the telephone survey from 227 patients revealed that fecal incontinence was a problem after radical retropubic (5%)and perineal(18%) prostatectomy and less than 50% of those with fecal incontinence had told the physician. Our mail survey (response rate 80% and 78% usable for analysis, 784 radical perineal and 123 perineal) strongly indicated that fecal incontinence after radical prostatectomy is a problem nationwide. Frequency of fecal incontinence (daily, weekly, monthly or less than monthly occurrences) was significantly higher among radical perineal(3,9,3 and 16%)compared to retropubic prostatectomy (2,5,3, and 8%) patients (p = 0.002). Fecal incontinence had a significant negative effect on patient social or entertainment activities (p = 0.0291, and travel and vacation plans (p = 0.043). Radical perineal compared to retropubic prostatectomy patients were more likely to wear a pad for stool leakage (p = 0.013), experienced more accidents (p = 0.001),had larger amounts of stool leakage (p = 0.002) and had less formed stools (p = 0.001). Of radical perineal prostatectomy patients only 14% and of retropubic only 7% with fecal incontinence had ever told a health care provider about it, even when the incontinence was severe. Responses to our survey concerning urinary incontinence showed that radical perineal prostatectomy patients had a lower rate of urinary incontinence immediately after prostatectomy compared to retropubic (79 versus 85%, p = 0.043). A higher proportion of perineal patients reported that all urinary leakage had ceased, that is full continence had returned (perineal70%, retropubic 53%, p = 0.001). A smaller proportion of perineal patients found it necessary to wear a pad to protect from urinary incontinence (perineal 39%, retropubic 56%, p = 0.004). Conclusions: Fecal incontinence following radical prostatectomy occurs more frequently than previously recognized. In general fecal incontinence among radical perineal and retropubic prostatectomy patients surpasses the expected incidence rate of 4% for this age group (60 to 70 years) but incidence is significantly higher for radical perineal prostatectomy patients. However, radical perineal prostatectomy patients have a significantly lower incidence of urinary incontinence than those treated with retropubic prostatectomy. Surgeons who perform radical retropubic and perineal prostatectomy should be aware of the possibility of fecal and/or urinary incontinence and associated symptoms. KEY WORDS: prostatectomy,fecal incontinence, urinary incontinence, prostatic neoplasms During the last decade the combination of prostate specific antigen (PSA) screening programs and heightened public Accepted for publication February 6, 1998. The opinions expressed herein are those of the authors and do not necessarily reflect those of the Departments of the Army, Air Force or Defense.

awareness has resulted in an increase in the detection of prostate cancer, and a dramatic shift to a lower stage of disease at presentation. Today more than 90%of patients have

* Re uests for reprints: Department of Urology, Wilford H d Medic3 Center, 2200 Berquist Drive, S m Antonio, Texas 782365300.

454

INCONTINENCE FOLLOWING RADICAL PERINEAL AND RETROPUBIC PROSTATECTOMY

localized disease at diagnosis and less than 8% have pelvic lymph node meta~tases.l-~ This stage migration has led to a renewed interest in radical anatomical perineal prostatectomy. This surgical approach, compared to radical retropubic prostatectomy, has been reported to be associated with less blood loss, a more rapid postconvalescence, equivalent or more rapid return of urinary continence and low rates of positive margins.4-6 However, in addition to the usual morbidity associated with radical prostatectomy, the perineal approach has been associated with a higher incidence of rectal complications. The rate of rectal perforation has been reported to be 2 to 11% among patients who have undergone this procedure.7-8Since 1991 we have performed 302 radical perineal prostatectomies using the supra-sphincteric or Young approach.3.lo We became aware of 8 patients who presented with unsolicited complaints of varying degrees of fecal incontinence following surgery, which led to our study. We determined the incidence of fecal incontinence following radical perineal and retropubic prostatectomy performed at our institutions, and compared it to the expected incidence for patients in this age group. For example, in community based surveys of patients 60 to 70 years old baseline fecal incontinence rates of 2 to 4% have been documented.11.12 We identified patients at risk for fecal complaints following radical prostatectomy due to preexisting medical conditions or prior rectal surgery. Finally, we estimated the impact of fecal incontinence on patient lifestyles and activities. MATERIALS AND METHODS

Fecal incontinence was defined as the involuntary loss of liquid or solid stool. A 26-question survey about fecal and urinary complaints following radical prostatectomy was designed using sections from a validated colorectal survey instrument.13 This validated instrument was used to survey patients randomly selected from the population of men who had undergone radical prostatectomy between March 1991 and December 1995 at Brooke Army or Wilford Hall Medical Centers. Radical perineal or retropubic prostatectomy was performed by 1 of 13 staff surgeons. All patients underwent radical prostatectomy a t least 12 months before being invited to participate in the survey. All information was obtained directly from the patient. Patients who had fecal incontinence before prostate surgery, or complained of fecal urgency, frequency or gas incontinence without loss of liquid or solid stool aRer surgery, were not included in our analysis. To determine whether the findings at our institutions were representative of the nationwide experience a broader based group of patients were sampled using a mail survey based on an expanded version of the original telephone survey. Statistical power analysis was done using the telephone survey results to determine the sample size needed to detect at least a 20%difference in fecal incontinence rates with a statistical power of 95% or greater. This second sample was taken randomly from radical prostatectomy cases documented in a national database of Civilian Health and Medical Program of the Uniformed Services beneficiaries. The database comprises data on patient episodes derived from processed claims for health care provided to these beneficiaries (military retirees and dependents) and is maintained by the Center for Healthcare Education and Studies at the United States Army Medical Department Center and School at Fort Sam Houston, San Antonio, Texas. Based on CFT billing codes a sample of patients who underwent radical prostatectomy between March 1991 and January 1996 was randomly selected for participation in the mailed survey. All patients underwent prostatectomy a t least 12 months before being contacted for the survey. A questionnaire was mailed to each patient with a letter of introduction and a self-addressed envelope. No identifying data were included on the survey

455

forms to ensure patient anonymity. In addition to questions about fecal incontinence, several validated questions about urinary incontinence and an expanded evaluation of effects on lifestyle were included in the mail survey that were not used in the telephone interviews. Based on an initial review of returned surveys, patients were excluded from analysis if the surgical approach could not be determined, incisions on the abdomen and perineum were performed (pelvic lymph node dissection followed by perineal prostatectomy) or radiation therapy had been given postoperatively. Statistical differences between perineal and retropubic approaches were analyzed using the chi-square test. RESULTS

Preliminary telephone survey. The telephone survey was completed by 227 patients of whom 100 underwent radical retropubic and 127 perineal prostatectomy. Average patient age was 65 years, and there was no difference in age between the retropubic and perineal groups. Of the patients who underwent radical retropubic prostatectomy 5 (5%)described new onset of fecal incontinence following surgery. However, among the patients who underwent radical perineal prostatectomy 23 (18%)had new onset of fecal incontinence following surgery. Of radical perineal prostatectomy patients 4 had daily, 11 weekly and 8 monthly episodes of fecal incontinence. An analysis of possible risk factors for fecal incontinence showed no significant difference between those with or without incontinence comparing surgeon, operative time, blood loss, complications, prior medical conditions or prior rectal surgery. Of note less than half of our patients with fecal incontinence, including 2 of the 4 with daily incontinence, had told the urologist or another physician about it. Nationwide mail survey. Due to the clear and convincing evidence based on the telephone survey results and power analysis, 1,200 questionnaires were mailed to Civilian Health and Medical Program of the Uniformed Services patients throughout the United States. Of the surveys 31 were returned undelivered due to incorrect address or death of the patient. Of the 1,169 delivered surveys 932 (80%)were completed and returned. Based on an initial review of the data, 25 surveys were excluded from analysis because the surgical approach (perineal or retropubic) could not be determined. Thus, a total of 907 (usable response rate 78%)surveys were used in our analysis of 784 radical retropubic and 123 perineal prostatectomy patients. Average patients age was 63 and there was no difference in age between the 2 surgical groups. Analysis of responses verified that all patients had undergone prostatectomy at least 12 months previously. Moreover, there was no difference in year of surgery between the 2 surgical approaches. An assessment of preoperative risk factors of fecal incontinence, such as diabetes mellitus, disorders of the nervous system or prior anorectal surgery, showed no difference between radical perineal and retropubic prostatectomy patients, and no difference between those with or without fecal incontinence following radical prostatectomy. The responses to questions about preoperative fecal incontinence indicated no statistical difference between radical perineal and retropubic prostatectomy patients in regard t o preoperative stool complaints. When asked to describe the occurrence of stool leakage before radical prostatectomy 95.3% of patients reported never having it, 4.5% reported sometimes (less than 25% of the time) and only 0.2% reported often (more than 25% of the time). No patient complained of leakage on a usual basis (more than 75% of the time). However, following radical prostatectomy more patients experienced stool leakage and the occurrence (never, sometimes, often or usually) differed significantly between radical perineal (69.1, 23.6, 4.1 and 3.2%) and retropubic

456

INCONTINENCE FOLLOWING RADICAL PER1NEAL AND RETROPUBIC PROSTATECTOMY

TABLE1. Nationwide survey of patients who reported fecal prostatectomy (82.9,15.8,1.0and 0.3%) patients (p = 0.001). incontinence following radical prostatectomy The frequency of stool leakage was stratified by episodes that B Retmpubic 9%Perineal Value occurred daily, weekly, monthly or less than monthly. Of (784 pts.) (123 pts.) radical retropubic prostatectomy patients who reported fecal incontinence the incidence rates were 2,5,3and 8%,respec- Wore pad to protect undergarments 4 9 0.013 17 32 0.001 tively (see figure). The frequency of fecal incontinence was Had accidents of leakage Had moderate to large amounts of 4 10 0.002 significantly higher in radical perineal prostatec.tomy paleakage tients (p = 0.002)who reported incidence rates of 3,9,3and Leakage not formed or solid 7 16 0.001 16%, respectively. Radical perineal patients were more likely than retropubic prostatectomy patients to wear a pad to protect undergarments from stool leakage (p = 0.013)and TABLE2. Nationwlde survey of patients who reported urinary less likely to recognize when stool leakage was about to occur incontinence following radical prostatectomy (p = 0.001) (table 1). Moreover, radical perineal prostatec% Retrouubic % Perineal D Value tomy patients experienced larger amounts of leakage (p = 86 79 0.043 0.001)and significantly less formed (solid) stool than retro- Immediate postop. urinary incontinence pubic patients when leakage occurred (p = 0.001). Radical Failure to return to full 47 30 0.001 perineal prostatectomy patients were also more likely to take continence over-the-counter medications in an attempt to prevent stool Need for long-term use of 56 39 0.004 protective pads leakage than retropubic patients (p = 0.057). Patients with fecal incontinence were asked to estimate the effect, if any, that incontinence had on sexual activity, family relations, social and entertainment activities, work DISCUSSION around the house, and sports and recreation. Fecal incontiWithin the last 10 years different factors have combined to nence had a significant negative effect on social or entertainment activities (p = 0.029),and the travel and vacation plans cause a stage migration in prostate cancer with a large frac(p = 0.043). Despite responses indicating a significant tion of patients now presenting with localized disease, and change in certain lifestyle aspects, the need to wear a pad for the need for pelvic lymph node dissection has come into stool protection and the use of over-the-counter medications question. With reports of decreased morbidity associated to prevent stool leakage, only 14% of radical perineal and 7% with radical perineal prostatectomy this approach has genof retropubic prostatectomy patients with fecal incontinence erated a great amount of interest and investigation. We have had ever told a physician or any other health care provider been performing radical perineal prostatectomy at our 2 institutions since 1991 and were initially aware of only 8 of 302 about it. Questions about urinary incontinence before prostatec- patients with varying degrees of fecal incontinence. With tomy showed that radical penned prostatectomy patients reports in the literature of rectal perforation but a lack of were not statistically different from retropubic patients with information about fecal incontinence, we were prompted to respect to preoperative urinary complaints. However, post- investigate the true incidence of fecal incontinence in our operatively radical perineal prostatectomy patients had a radical retropubic and perineal prostatectomy patients. Conlower rate of immediate urinary incontinence compared to cern for the surprisingly high incidence found among our retropubic patients (79versus 85%, p = 0.043)(table 2). Of patients compelled us to examine the incidence of fecal inpatients with postoperative urinary incontinence, radical continence in a national population of patients and surgeons perineal patients reported a lower failure of return to full from all geographic regions of the United States. continence than retropubic prostatectomy patients (30 verWe present evidence of fecal incontinence, to our knowlsus 47%, p = 0.001).h s s postoperative urinary incontinence edge, a previously undescribed complication of radical prosin radical perineal prostatectomy patients was further sup- tatectomy. In our nationwide survey we found that, among ported by the finding that only 39% of them found it neces- patients who underwent radical retropubic or perineal prossary to wear a pad to protect from urinary incontinence tatectomy, the incidence of fecal incontinence far surpassed compared to 56% of retropubic prostatectomy patients (p = the expected incidence of 4% for patients in the same age 0.004). group (60to 70 years).11.12If we only consider patients in the

w 20%

0

32 F 2

15%

0 0

w

10%

(3

33

5%

0

%n

0%

DAILY

WEEKLY

MONTHLY

LESS THAN MONTHLY

FREQUENCY OF FECAL INCONTINENCE Nationwide survey of 784 retropubic and 123 perineal patients who reported fecal incontinence after radical prostatectomy

INCONTINENCE FOLLOWING RADICAL PERINEAL AND RETROPUBIC PROSTATECTOMY

survey who had fecal incontinence episodes of at least monthly frequency then 10%of radical retropubic and 15%of perineal prostatectomy patients were affected by this problem. These percentages are consistent with those from the telephone survey (5%radical retropubic, 18%perineal prostatectomy), which did not examine less than monthly occurrences of fecal incontinence. For the majority these findings represent the new onset of fecal incontinence in patients who had no fecal complaints before prostate surgery. In addition, we found that preoperative anorectal surgery and medical conditions, including diabetes mellitus and neurological disorders, were not risk factors of fecal incontinence after surgery. Fecal incontinence among radical retropubic and perineal prostatectomy patients had a significant effect on several important aspects of patient lifestyles and well-being. These patients reported that social, entertainment, travel and vacation activities were affected by fecal incontinence. For some patients in this age group (60 to 70) these activities are considered the most important events of their retirement years. Fecal incontinence has gained considerable prominence in the general medical literature during the last decade. It is slowly being recognized that this problem represents a financial burden to society and is far more common than previously recognized.14 In our survey we found that patients are not likely to report fecal incontinence to a health care provider, even when it is severe and detrimental to their lifestyle. While this finding is surprising it is consistent with reporting patterns found in a large community based study, which shows that fewer than 30% of patients with fecal incontinence had ever reported it to a health care provider.14 There are many reasons why patients do not readily report fecal incontinence to physicians and surgeons. It has been shown that fecal incontinence is often underestimated during routine physician history evaluation.14-16 Patients may be more likely to tell a surgeon about urinary incontinence that occurs as a result of prostatectomy because the surgeon asks pertinent questions and they have been warned about the possibility of urinary complaints during preoperative counseling. Still, we are well aware of the poor reporting of urinary incontinence and impotence among patients who have undergone prostatectomy. Many patients will not immediately associate prostate surgery, especially a retropubic approach, with any type of rectal or stool change. Furthermore, there is a social stigma attached to fecal incontinence, and a fear that it represents aging and may threaten independent living. The result is that patients are reluctant to discuss these concerns with others, and there is a general lack of appreciation of the prevalence of fecal incontinence among the general and prostatectomy populations. Many surgeons who perform radical perineal prostatectomy have been impressed with the rapid return to urinary continence following surgery compared to radical retropubic prostatectomy. We are not aware of previous reports that have documented such a difference in urinary continence rates. We found that radical perineal prostatectomy patients are more likely to be continent of urine immediately after catheter removal, have a faster return of urinary continence and are less likely to wear protective pads to guard against urinary leakage compared to retropubic patients. Improved urinary continence results are likely due to the superb exposure of the prostatic apex and the facilitated urethral anastomosis under direct vision. Until recently even the colorectal literature lacked a validated self-reporting instrument to assess fecal and urinary incontinence, and associated symptoms related to pelvic floor dysfunction. Reilly et a1 developed an extensive questionnaire that was administered to a large number of patients, and retested by mail s w e y and telephone interviews.13 They

457

found the instrument to be easily understood, readily completed, reliable and valid. We reproduced 25 questions from their questionnaire to use for assessment of fecal and urinary complaints in our 2 patient populations. The exact mechanism of injury leading to fecal incontinence is not yet known. The rectal sphincter is composed of 3 external muscles (subcutaneous, superficial and deep) and the internal anal sphincter. All of these muscles are subject to damage during the perineal approach to prostatectomy, which may be a direct result during dissection of the external anal sphincter, levator ani muscle or the rectourethralis. Prolonged retraction may lead to neurological compromise. Reconstruction of the levator ani muscles may further damage or prevent proper function of the continence mechanism. Additional research to determine the exact nature of this injury is needed, including anorectal electromyography, manometry and anorectal ultrasound. Anorectal electromyography, which measures interference patterns represented by the number of turns per second and the amplitude of each turn, may provide a quantitative method to evaluate the anal sphincter and correlate with symptoms of fecal incontinence. Combining anorectal manometry with new electromyography techniques may allow us to identify preoperatively patients who may be at increased risk for fecal incontinence. As more knowledge is acquired about this problem we may be able to modify current techniques to prevent fecal incontinence while reaping the benefits of improved urinary continence. CONCLUSIONS

Fecal incontinence following radical prostatectomy occurs more frequently than previously recognized. Surgeons who perform radical retropubic and perineal prostatectomy should be aware of the possibility of fecal and/or urinary incontinence and associated symptoms. Patients should be appropriately counseled and informed about the potential risks of these complications. REFERENCES

1. Catalona, W. J.. Smith, D. S., Ratliff. T. L. and Basler. J. W.: Detection of .organ-confined prostate cancer is increased through prostate-specific antigen-based screening. J.A.M.A., 270 948,1993. Bishoff, J. T., Reyes, A., Thompson, I. M., Harris, M. J., St Clair, S. R., Gomella, L. and Butzin, C. A,: Pelvic lymphadenectomy can be omitted in selected patients with carcinoma of the prostate: development of a system of patient selection. Urology, 45:270, 1995. Petros, J. A. and Catalona, W. J.: Lower incidence of unsuspected lymph node metastases in 521 consecutive patients with clinically localized prostate cancer. J. Urol., 141: 1574, 1992. Gibbons, R. P.: Radical perineal prostatectomy:definitive treatment for patients with localized prostate cancer. AUA Update Series. 1 3 1. 1994. 5. Parra, R. O.,Boullier, J. A., Rauscher, J. A. and Cummings, J. M.: The value of laparoscopic lymphadenectomy in conjunction with radical perineal or retropubic prostatectomy. J. Urol., 161: 1599, 1994. 6. kazier, H. A,, Robertson, J. E. and Paulson, D. F.: Radical prostatectomy:the pros and cons of the perineal versus retropubic approach. J. Urol., 147: 888, 1992. 7. Lassen, P. M. and Kearse, W. S., Jr.: Rectal injuries during radical perineal prostatectomy.Urology, 4 5 266, 1995. 8. Levy, D.A. and Resnick, M. I.: Laparoscopic pelvic lymphadenectomy and radical perineal prostatectomy:a viable alternative to radical retropubic prostatectomy. J. Urol., 161: 905, 1994. 9. Harris, M. J. and Thompson, I. M., Jr.: The anatomic radical perineal prostatectomy: an individualized approach to the treatment of men with clinically localized prostate cancer. In: Monogr. Urol., 1 6 p. 84, 1995.

458

INCONTINENCE FOLLOWING RADICAL PERINEAL AND RETROPUBIC PROSTATECTOMY

A. R.: Validation of a questionnaire to assess fecal inconti. 10. Harris,M.J . and Thompson, I. M.: The anatomic radical perinence and associated risk factors. Unpublished data. neal prostatectomy: a contemporary and anatomic approach. 14. Johanson, J. F.and M e r t y , J.:Epidemiologyoffecal incontinence: Urology, 48:762, 1996. the silent affliction. Amer. J. Gastroenterol., 91: 33,1996. 11. OKeefe, E. A,, Talley, N. J., Tangalos, E. G. and Zinsmeister, A. R.: A bowel symptom questionnaire for the elderly. J. Ger- 15. Leigh, R. J. and Turnberg, L. A.: Faecal incontinence: the unvoiced symptom. Lancet, 1: 1349,1982. ontol., 47: M116, 1992. 12. Nelson, R.,Norton, N., Cautley, E. and h e r , S.: Community- 16. Enck, P., Bielefeldt, K, Rathmann, W., Purrmann, J., Tschope, D. and Erckenbrecht, J. F.: Epidemiology of faecal incontibased prevalence of anal incontinence. J.A.M.A., 274: 559, nence in selected patients groups. Int. J. Colorectal Dis., 6: 1995. 143,1991. 13. b i l l y , T. W.,Talley, N. J., Pemberton, J . H. and Zinsmeister,