Patterns of Health Care Utilization for Voiding Dysfunction in an Ambulatory Population: Implications for Clinical Trial Results

RJ Jeddeloh, M.D., AM Carlson, Ph.D., KA Oleson, DF Milam, M.D. and Sacral Nerve Stimulation Study Group

Urinary incontinence is a highly prevalent condition in the U.S. population with consequences that are devastating to the lives of affected individuals and immediate family member.⁽¹⁾ Current estimates indicate that the condition affects 13 million Americans.⁽²⁾ The actual prevalence may be higher because the condition is widely underreported, underdiagnosed and untreated. Results from a study in the state of Minnesota indicate that care-seeking for voiding problems is very low; about 29% of men and 13% of women with incontinence symptoms actually seek treatment.⁽³⁾ In a 1990 survey of 541 healthy women aged 42-50 years, Burgio et al found that while 58% experienced incontinence at some time and 31% reported incontinence on a regular basis, only 25.5% of the women had sought treatment.⁽⁴⁾ A British study found similar care-seeking patterns in a study of 384 women 20 years of age and older. The overall reported prevalence of incontinence was 53% and, although the 167 women who reported incontinence were offered treatment at a women's clinic, only 13 attended and 10 entered a treatment trial.⁽⁵⁾

Although the economic issues of incontinence in institutionalized elderly has been reported, the economic burden that results from urinary incontinence for an ambulatory population has received less attention. In part this is because of the societal misperception that urinary incontinence represents an inconvenience rather than a health burden.^(6a) Many of the economic studies have been reported by Hu and colleagues.

Their 1987 estimate indicated that the direct costs (initial diagnostic and medical evaluation, costs of treatment, and extended use of disposable medical supplies such as absorbent pads and urinary collection devices) exceeded $10 billion.^(6a) By 1995 the costs had exceeded $26.3 billion for individuals 65 years of age, which translated to $3,565 per individual with urinary incontinence.^{(6b, 6c)}

Urinary incontinence also carries other serious cost consequences like skin irritation, urinary tract infection, sleep deprivation, falls, weight gain and depression. These co-morbid conditions can be costly to treat, significantly contributing to the total health care expenditures of affected individuals.⁽⁷⁾ In some cases, symptoms may be so difficult to manage that individuals can no longer work, further adding to their economic burden. Incontinent individuals also generate greater costs to publicly funded home care programs. One study of 1,468 individuals using home care programs found that incontinent persons generated a 25% increase in average 18-month expenditures for home programs.⁽⁸⁾

Although recent progress has been made in understanding the underlying pathophysiology of voiding dysfunction, patients diagnosed with incontinence have had limited treatment options.⁽⁹⁾ Non-surgical interventions, including diet modification, drug therapies, and behavioral techniques such as timed voiding, pelvic muscle exercises, and biofeedback, have all been employed to treat the condition.^(10,11) Drug therapies are intended to act directly on the bladder to calm or relax detrusor muscle activity while behavioral techniques are intended to condition the pelvic floor and suppress detrusor muscle contraction by modulation of neural reflexes that have an inhibitory effect on the bladder. Cure rates for either technique are low. Drug therapies have reported cure rates of less than 45% of patients and behavioral techniques have rates as low as 12%.⁽²⁾ Partial cure rates (a 50% or greater reduction in incontinence episodes) are higher--up to 77% for drug therapies and 75% for pelvic muscle exercises.⁽²⁾ Unfortunately, patients are still left to manage distressing voiding symptoms that prevent them from realizing improved quality of life. In addition, drug therapies have associated side effects, though newer agents have an improved adverse events profile further contributing to declines in quality of life.⁽¹²⁾ Many patients with voiding dysfunction are caught in a trial and error process that ultimately results in the continued use of less than satisfactory therapies, progression to surgical intervention, or both.

Surgical interventions for voiding dysfunction include cystourethropexy, augmentation cystoplasty, bladder denervation, and bladder removal with urinary diversion. Complication rates following surgery range from 3% to 15% depending on the surgery type,⁽²⁾ and repeat operations are not uncommon. One study of women enrolled in an HMO found repeat operations performed on 29% of women with surgically managed pelvic organ prolapse and urinary incontinence.⁽¹³⁾

Clinical literature addressing voiding dysfunction in general, and urinary incontinence specifically, has identified newer therapeutic measures to treat patients. One approach is the neuromodulation of sacral nerves through electrostimulation of sacral afferent nerve pathways thereby inhibiting detrusor overactivity ^{(10, 14-16)} or facilitating an underactive detrusor.^{(17, 18)} Sacral nerve stimulation (SNS) has demonstrated efficacy and safety ^(19-21), is reversible, and carries with it the advantage of a test period during which an assessment of successful treatment for individual patients can be made. Test stimulation is an office-based procedure designed to assess sacral nerve integrity through percutaneous stimulation of the S3 or S4 sacral nerve, and helps clinicians assess the merits of a surgical implant. A thin lead is percutaneously implanted in cases where favorable responses are obtained during acute test stimulation, and patients can track their urinary symptoms in voiding diaries during a home trial stimulation period of 3-7 days. Patients with a satisfactory response during the test can then have the device surgically implanted for chronic use.^{(22, 23)}

Persons enrolled in clinical trials testing the safety and efficacy of this method have all been patients without satisfactory amelioration of distressing voiding symptoms even after repeated surgical intervention. As such, they may not be representative of a more general population. It has often been suggested that persons enrolled in clinical trials may not be representative of patients found in clinical practice situations, thereby making it difficult for physicians and other interested persons to apply the results of evidence found under clinical trial conditions to office-based practice.⁽²⁴⁾ If the clinical trial populations are similar, if not identical, to a more general population presenting with voiding dysfunction, the short- and long-term outcomes of the clinical trial population may predict the impact of this therapy on a broader population.

The purpose of this paper is two-fold. First, from a large claims database in a Midwest independent practice model (IPA) health plan, we wanted to review the treatment options and quantify the health care expenditures incurred by the population identified with voiding dysfunction. Second, we wanted to compare and contrast the treatment options identified in the claims population with those identified among subjects prior to their enrollment in a clinical trial of an implantable device (InterStimÒ Therapy for Urinary Control; Medtronic, Mpls, MN) for SNS in the treatment of urinary incontinence. The intent was to determine whether or not the clinical trial patients represented the extreme, worst-case patients or were comparable to more mainstream patients being cared for in a large health plan.

Claims Data

The claims database from which the data was drawn includes member and provider information, and physician/health provider, facility, and outpatient pharmacy claims. A member enrollment number assigned by the health plan links all files. Members may be enrolled in one of several insurance product groups--commercial insured, Medicaid certified, and Medicare certified. Extent of benefit coverage can be determined through linked enrollment files to ensure similar benefit coverage for identified members.

Patients with claims evidence of voiding dysfunction (codes for diagnoses, medical procedures, medical supplies and drugs) during a 36-month period beginning 04/01/96 and ending 03/31/99 were identified from the health plan's database. Only individuals without capitation limits were identified and all patients had similar levels of prescripdion drug coverage that included therapeutic agents for the treatment of voiding dysfunction. Exclusion criteria were applied to rule out individuals for whom adequate claims evidence appeared lacking, those less than 20 years of age, and those with non-continuous enrollment. Inadequate claims evidence was a diagnosis of voiding dysfunction associated with a single laboratory or radiology claim in a 12-month period or for which a non-medical provider was identified. The exclusion criteria were designed to parallel the exclusions found in the clinical trial population used as a comparison.

Table 1: Diagnosis and Procedure Codes for Claims Analysis

	ICD-9-CM Codes
Inclusions
Cystitis/other bladder/urethritis	595 -597.8
Prolapse	618.0
Retention of Urine	788.2-788.29
Incontinence of Urine	788.3-788.39
Urinary Frequency	788.4-788.41
Exclusions
Lumbar disk disease	722.10;722.2;722.5722.7; 724.4
Multiple sclerosis	340; 340.19
Spinal cord injury	342-342.99; 344.0-344.09; 344.6-344.69; 344.8-344.99
Spina bifida	741-741.99
Myopathy	358-359.99
Parkinson's Disease	332-332.99
Benign prostatic hypertrophy	600-601.99
Diabetes	250-250.99; 648.0; 648.8
Neurogenic bladder	344.61; 596.54

Patient identification results indicated that about 3% of the insured population within the health plan (approximately 13,600 persons per year) had claims evidence of voiding dysfunction. Overall, the prevalence among women was approximately twice that of men (39/1000 for females compared to 19/1000 for males) but only through the 6^th decade. The highest prevalence rate, 908/1000, was found among men 80+ years of age.

Patients identified with voiding dysfunction were subset into two groups that would be most comparable to those included in a Medtronic sponsored clinical trial of SNS therapy--urinary retention (UR) and urinary urge incontinence and/or urinary (UI/UF). Patients with voiding dysfunction who also had claims evidence of concomitant conditions that affect voiding (lumbar disk disease, multiple sclerosis and related conditions, spinal cord injury, spina bifida, myopathy, Parkinson's Disease, benign prostatic hypertrophy, and/or diabetes) were excluded, as were patients with evidence of neurogenic bladder and those whose only voiding problem was stress incontinence.

A total of 733 patients were classified into the UR group (6% of all voiding dysfunction patients) and 3,443 were classified into the UI/UF group (26% of all voiding dysfunction patients). The overall ratio of UI/UF to UR was 5:1. Compared to males, approximately 2.5 times as many females were identified with retention and 9 times as many females were classified with urinary urge incontinence/frequency.

Annual total health care expenditures for patients identified in the UR group were $6.9 million of which $896,000 was directly associated with voiding problems. This is equivalent to average total health care expenses of $9,514 per member identified with UR and $1222 per member for voiding-specific care. Of the expenditures related specifically to voiding problems, hospitalizations accounted for 82%, outpatient/office visits for 11% and outpatient pharmacy claims accounted for only 1%.

For patients identified in the UI/UF group $16,009,000 was expended on total health care. This is an average of $4,650 per member with UI/UF per year for total health care expenditures. A total of $960,000 was directly associated with voiding problems, an average of $279 per member. Of these voiding-specific costs, hospitalizations accounted for 44%, outpatient/office visits for 42% and outpatient pharmacy for 6% of the dollars.

It is important to note, however, that these costs are from the perspective of the insurer and represent limited direct medical costs, being a summation of the costs for health services available as a covered benefit only. The financial data elements used were as reported following claims adjudication, a process that takes into account benefit restrictions and provider contractual arrangements. The resultant costs are the actual amounts paid by the plan. As such these costs represent the minimum direct costs determined for voiding disorders.

Claims histories for patients with UR and UI/UF were also examined for evidence of surgical procedures related to voiding dysfunction. During a 12-month period approximately 12% of persons with UR and 7% of persons with UI/UF had at least one surgical procedure. To examine the issue of repetitive surgeries, we reviewed the claims histories of patients who had surgery during the middle 12 months of the study time frame and had continuous enrollment across all 36 months. Of the identified patients, 8% had at least one surgery in the previous 12 months and 9% required additional surgeries for voiding dysfunction during the subsequent 12 months.

Clinical Trial

A 581 subject, multicenter, prospective, randomized clinical trial (MDT-103) of an implantable, multiprogrammable neurostimulator (InterStim^Ò, Medtronic, Inc., Minneapolis, MN*) was recently concluded. Study candidates, documented as refractory to standard medical therapies via medical audit, were enrolled from the general urologic population at key global centers under identical protocols approved by bioethics committees. Upon enrollment, patients were categorized into one of three patient groups based upon dysfunctional voiding patterns: urge incontinence (N = 184); retention (N = 177) and urgency/frequency (N = 220). Exclusion criteria included neurologic conditions such as multiple sclerosis, primary stress incontinence, and primary pelvic pain symptoms. The patient population was 79% female. The average age was 43.3 ±12.5 yrs. At the time of enrollment in the clinical trial the patients presented with an average voiding dysfunction duration of 8.0 years.

All enrolled patients completed test stimulation of the sacral nerves. Patients with a favorable response during acute test stimulation tracked their urinary symptoms in voiding diaries during a home test stimulation period of 3-7 days. Patients demonstrating a 50 percent or greater reduction in symptoms during the test were qualified for surgical implantation.

Baseline evaluation included medical history, physical evaluation, complete urodynamic testing and quantification of symptoms in voiding diaries. In addition to the general medical history, a detailed medical record history of all previous treatments employed by the patients to treat their underlying urinary problem was made. Drug therapies were the most widely used treatment. A total of 541 patients (93%) of those enrolled had used drugs to treat their urinary problems some time during the two years prior to enrollment in the clinical trial. 268 of the 581

*The specific components of the surgically implanted system included Model 7424 Itrel^Ò II Implantable Pulse Generator; Model 7495 Extension; and Model 3886 or Model 3080 Lead and were manufactured by Medtronic, Inc., Minneapolis, MN.

patients (46.1%) used non-surgical interventions including external stimulation/TENS, biofeedback, urethral dilatation, pain management, psychologic counselling, vaginal weighted cones, or timed voiding. Urethral dilatation, pain management and biofeedback were the most commonly employed.

A total of 336 persons (58% of enrolled patients) had undergone surgery at some time since diagnosis to treat urinary problems. In all 1295 surgical procedures were verified from the medical histories. The average was 2.2 surgeries (± 4.5). One person reported undergoing 41 different procedures to treat voiding problems. 245 persons reported surgical procedures such as biopsies, laparoscopies, and hysterectomies.

Long-term efficacy of the surgically implanted device was evaluated in all treated patients. All study results were source-verified through complete reviews of institutional medical records and independent audits by the U.S. Food and Drug Administration. Based on the submitted results the device was approved for use in the United States. Through the most recent follow-up (with an average patient post-implant follow-up duration of 28 months), use of concomitant urologic drugs dropped from 93% to 54% in the urge incontinent group, with slight reductions noted for the remaining indications. More importantly, non-InterStim related urologic surgeries dramatically reduced from 58% of patients at baseline down to 4%-7% for the 3 indications.

Discussion

Claims data analysis identified that 3% of an insured population sought medical services for voiding dysfunction during a 12-month period. The rates may be compared to estimated national rates of urinary incontinence presented in a national clinical guideline. Based on an estimated 13-15 million persons in the U.S. population affected with urinary incontinence, a prevalence rate of approximately 5% can be determined (the estimated resident U.S. population for the study time period was 269,623,000). The guideline estimate includes, however, persons who are affected by urinary incontinence that may or may not have accessed the health care system for treatment. The estimated prevalence from claims includes only those persons who actively sought health care for the diagnosis and/or treatment of voiding dysfunction using insurance benefits and suggests that national estimates may be low.

The average annual expenditure per patient with either UR or UI/UF was found to be considerably higher than the estimated $1,600 spent annually on average for a commercially insured member. Patients identified with claims evidence of UR averaged $9,514 in total health care expenses; those identified with UI/UF averaged $4,650. For UR patients this is a 6-fold increase over average plan expense per member and for UI/UF patients a 3-fold increase.

A comparison of the experiences of the clinical trial subjects and patients identified in claims data suggests that, in an effort to control their symptoms, persons identified with voiding problems will experiment with a variety of treatment options. It is generally believed that patients enrolled in clinical trials do not represent patients found in clinical practice situations. In the case of sacral nerve stimulation for urinary incontinence, the patients included in the clinical trial are considered therapeutic failures since, having exhausted all forms of treatment without success they have voluntarily participated in the evaluation of a new treatment option. The comparative results between clinical trial patients and those identified in a claims database indicate that this may not be the case. Patients in general clinical practice have tried multiple therapy options, including repetitive surgery during the course of a 12-month period, in an attempt to alleviate their voiding problems. They may be more similar to patients in clinical trial than different. Clinical trial results for sacral nerve stimulation therapy may be more directly applicable than would otherwise be the case.

For those patients enrolled in the clinical trial all previous therapy attempts clearly failed to alleviate the problem. Voiding problems continued to such an extent that they were prompted into enrollment in a clinical trial, despite even drastic surgical intervention. The same sense of treatment failure cannot be identified using the claims data, although persons with repetitive surgeries could be found. About 8% of patients undergoing surgery had a previous surgery within 12 months and 9% would undergo an additional surgery within 12 months.

Conclusions

Claims analysis indicates that though the number of patients identified with voiding problems appears small, it is a population that utilizes health care to a great extent including multiple surgeries in an effort to alleviate symptoms. The claims based population has care-seeking patterns for voiding dysfunction similar to that described by patients enrolled in a clinical trial. Outcomes of the clinical trial could demonstrate the impact of SNS treatment of voiding dysfunction in a larger health plan population. Treatment of voiding dysfunction with newer treatment options like SNS may lead to reduced health care utilization and costs.

References

1. Diokno AC. Epidemiology and psychosocial aspects of incontinence. Urologic Clinics of North America. 1995; 22(3):481-485.

2. Fantl JA, Newman DK, Colling J, et al. Urinary Incontinence in Adults: Acute and Chronic Management. Clinical Practice Guideline No. 2. 1996 Update. Rockville, Maryland: U.S. Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research. AHCPR Publication No. 96-0682. March, 1996.

3. Roberts RO, Jacobsen SJ, Rhodes T, Reilly WT, et al. 1998. Urinary incontinence in a community-based cohort: prevalence and healthcae-seeking. J Am Geriatr Soc 46: 467-472.

4. Burgio KL, Matthews KA, Engel BT. Prevalence, incidence and correlates of urinary incontinence in healthy, middle-aged women. J Urol. 1991; 146(5): 1255-9.

5. Harrison GL, Memel DS. Urinary incontinence in women: its prevalence and its management in a health promotion clinic. Br J Gen Pract. 1994; 44(381): 149-52.

6a. Hu T. Impact of urinary incontinence on health-care costs. JAGS. 1990; 38: 292-295.

6b. Hu T. Impact of urinary incontinence on health-care costs. JAGS. 1990; 38: 292-295.

6c. Wagner TH, Hu TW. Economic costs of UI in 1995. Urology 1998; 51: 355-361.

7. Zhou Z, Jensen GA. Insurance claim costs for overactive bladder disorder. Drug Benefit Trends April, 2001; 45-58.

8. Baker DI, Bice TW. 1995. The influence of urinary incontinence on publicly financed home care services to low income elderly people. Gerontologist 35: 360-369.

9. Yalla SV. Editorial: management of urinary incontinence-progress and innovative strategies. J Urol. 1998; 159:1520-1522.

10. Shaker HS, Hassouna M. Sacral nerve root neuromodulation: an effective treatment for refractory urge incontinence. J Urol. 1998; 159:1516-19.

11. Diokno AC. Advances in the management of urge incontinence. J Urol Index & Reviews 1 (1): 4-5

12. Appell RA. Clinical efficacy and safety of tolterodine in the treatment of overactive bladder: a pooled analysis. Urology. 1997; 50 (Supp 6A): 90-6.

13. Olsen AL, Smith VJ, Bergstrom JO, et al. Epidemiology of surgically managed pelvic organ prolapse and urinary incontinence. Obstet Gynecol. 1997; 89(4): 501-6.

14. Bosch J, Groen J. Sacral (S3) segmental nerve stimulation as a treatment for urge incontinence in patients with detrusor instability: results of chronic electrical stimulation using an implantable neural prosthesis. J of Urol 1995; 154:504-507.

15. Dijkema H, Ijzerman W, Mijs P, et al. Neuromodulation of sacral nerves for incontinence and voiding dysfunctions. Euro Urol. 1993; 24: 72-77.

16. Thon W, Baskin L, Jonas U, et al. Neuromodulation of voiding dysfunction and pelvic pain. World J of Urol. 1991; 9: 138-141.

17. Fowler CJ, Swinn MJ, Goodwin RJ, et al. Studies of the latency of pelvic floor contraction during peripheral nerve evaluation show that the muscle response is reflexly mediated. J Urol 2000 Mar;163(3):881-3.

18. Schultz-Lampel D, Jiang C, Lindstrom S, Thuroff JW. Experimental results on mechanisms of action of electrical neuromodulation in chronic urinary retention. World J Urol 1998;16(5):301-4.

19. Schmidt RA, Jonas U, Oleson KA, et al. Sacral nerve stimulation for treatment of refractory urinary urge incontinence. Sacral Nerve Stimulation Study Group. J Urol 1999; 162(2):352-7.

20. Hassouna MM, Siegel SW, Nyeholt AAB, et al. Sacral neuromodulation in the treatment of urgency-frequency symptoms: A multicenter study on efficacy and safety. J Urol 2000; 163(6):1849-54.

21. Jonas U, Fowler CJ, Chancellor MB, et al. Efficacy of Sacral Nerve Stimulation (SNS) in Urinary Retention: Results up to 18 months following implantation. J Urol 2000 (in publication).

22. Schmidt RA, Senn E, Tanagho EA. Functional evaluation of sacral root integrity. Urology 1990; 35:388-92.

23. Siegel SW. Management of voiding dysfunctions with an implantable neuroprosthesis. Urol. Clinics of North Am 1992; 19(1):163-70

24. Friedman LM, Turberg CD, Demets DL. Fundamentals of Clinical Trials 3rd Edition. New York: Springer-Verlag, 1998.

Sacral Nerve Stimulation Group

In addition to D. Milam, the principal investigators include:

RA Schmidt, University of Colorado Health Sciences Center, Denver, Colorado, USA

RA Janknegt, Academisch Ziekenhuis Maastricht, Maastricht, the Netherlands

MM Hassouna, Toronto Hospital Western Division, Toronto, Ontario, Canada

PEV van Kerrebroeck and EJ Meuleman, Academisch Ziekenhuis Nijmegen, Nijmegen, the Netherlands

SW Siegel, Metropolitan Urologic Specialists, St. Paul, Minnesota, USA

M Fall, Sahlgrenska Hospital, University of Gothenborg, Gothenborg, Sweden

CJ Fowler, National Hospital for Neurology & Neurosurgery, London, United Kingdom

A.A.B. Lycklama à Nyeholt, Academisch Ziekenhuis Leiden, Leiden, the Netherlands

U Jonas, Medizinische Hochschule Hannover, Hannover, Germany

JB Gajewski, Queen Elizabeth II Health Science Center, Halifax, Nova Scotia, Canada

MM Elhilali, Royal Victoria Hospital, Montreal, Canada

DA Rivas, Thomas Jefferson University, Philadelphia, Pennsylvania, USA

AK Das, Albany Medical College, Albany, New York, USA

F Catanzaro, Ospedale di Desio, Desio, Italy

TB Boone, Baylor College of Medicine, Houston, Texas, USA

R Tutrone, Greater Baltimore Medical Center, Baltimore, Maryland, USA

S Kaplan and A Te, Columbia University, New York, New York, USA

MB Chancellor, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA

S Madersbacher, University of Vienna, Innsbruck, Austria

HE Dijkema, Academisch Ziekenhuis Almelo, the Netherlands

AR Stone, University of California Davis School of Medicine, Sacramento, California, USA

Acknowledgement: Funded by Medtronic Neurological .

Corresponding Author Address:

A.M. Carlson, Ph.D.

Data Intelligence Consultants, LLC

P.O. Box 44993

Eden Prairie, MN 55344