Subak, Leslee L. MD; Marinilli Pinto, Angela PhD; Wing, Rena R. PhD; Nakagawa, Sanae; Kusek, John W. PhD; Herman, William H. MD; Kuppermann, Miriam PhD
Urinary incontinence is highly prevalent among middle aged and older women and is costly, exceeding $20 billion per year in direct costs in the United States.1,2 Approximately one half to three quarters of the costs of incontinence are attributed to resources used for incontinence management or “routine care” such as absorbent pads, protection, and laundry. Despite the burden of urinary incontinence (UI), there are few primary data on UI management costs, and cost estimates vary widely between studies ($50 to $1,000 per person per year) and by method of cost determination.1–8 Whereas increased frequency of UI is associated with higher incontinence management costs, it is not known whether successful incontinence treatment can decrease costs for UI management.
We previously conducted a randomized clinical trial of a weight loss and maintenance program among obese and overweight women with UI. Briefly, women in the weight loss group had greater reduction in body weight compared with the control group at 6 months (−8.0% compared with −1.6%; P<.01) and 18 months (−5.5% compared with −1.5%; P<.001) and greater decrease in weekly incontinence episode frequency compared with the control group at 6 months (−47% compared with −28%; P=.04) but not at 18 months (−61% compared with −54%; P=.31).9–11 We estimated the effect of change in UI frequency on UI management costs and consider factors that may predict change in cost in these women.
PATIENTS AND METHODS
From July 2004 through April 2006, 338 overweight and obese community-dwelling adult women with at least 10 UI episodes per week were enrolled in a randomized clinical trial to determine whether an 18-month weight loss and maintenance intervention would significantly reduce the frequency of incontinent episodes at 6- and 18-month follow-up compared with a structured education program control group. Details of the design and results of the clinical trial have been previously reported.9–11 As a “minimal treatment” for incontinence, all participants were given at time of randomization a self-help behavioral treatment booklet with instructions for improving bladder control.9,12 The study was approved by the institutional review board at each site and written consent was obtained from all participants before enrollment.
Data were collected at baseline (before randomization) and at 6 and 18 months after randomization. Resources used for UI management or “routine care” were assessed by the question “During a typical week, how many of each of the supplies listed below do you use specifically for your urine leakage?” Supplies included panty liners or minipads, maxipads, incontinence pads, urethral inserts, toilet paper (number of changes), and paper towels (number of sheets). Laundry was assessed by the question “During a typical week, how many loads of wash do you do because of your urine leakage?” and dry cleaning by “During a typical week, how many items of clothing do you dry clean because of your urine leakage?” Women recorded the number of each type of clothing dry cleaned per week, including pants, skirts, dresses, suits, and blouses.
National unit costs were estimated for each type of supply by a survey of 14 stores in 6 states and one national Internet source accessed in 2006.5 The cost of a load of laundry and dry cleaning was estimated by a survey of five laundromats and five dry cleaners in each of four geographically distant areas in the United States.5 To approximate home laundering cost, we used 80% of the mean laundromat cost ($2.24 per load) to adjust for laundromat-related expenses such as rent and profit. The cost of dry cleaning was estimated as pants, $5.03; skirt, $5.35; blouse, $5.45; dress, $9.45; and two-piece suit, $9.64. Because routine care costs are paid out-of-pocket by women in the United States, the analysis is from the patient's perspective.
Demographic characteristics, medical history, and other factors potentially associated with costs were assessed by self-reported questionnaires. Incontinence symptom bother was measured by the 19-item Urogenital Distress Inventory and UI-specific quality of life was measured using the 30-item Incontinence Impact Questionnaire.13 Both the Urogenital Distress Inventory and Incontinence Impact Questionnaire are scored from 0 to 400, with higher scores indicating more severe UI symptoms or life effect. A difference of 16 on the Incontinence Impact Questionnaire and 11 on the Urogenital Distress Inventory are considered clinically significant.14 Frequency of UI was assessed by a participant-completed 7-day voiding diary that included time of each void and each incontinence episode, which was identified by the participant as stress (involuntary loss of urine with coughing, sneezing, straining, or exercise), urgency (loss of urine associated with a strong need or urge to void), or other. The amount of urine lost involuntarily was quantified using a 24-hour pad test using methods standardized by the International Continence Society.15
Mean costs for UI management (“cost”) were calculated by multiplying units of resources used by the mean cost per unit and presented in 2006 U.S. dollars. Because the distribution of the cost data were skewed, we calculated the median and 25th and 75th percentiles as well as mean and standard deviation for total routine care costs. We report both mean and median cost estimates because the aggregate costs to society are best summarized by mean cost times the number of people affected whereas the cost for individuals is best summarized by median cost.6
We compared baseline demographic and clinical characteristics by treatment group using linear mixed models for continuous variables and generalized estimating equations multinomial models for categorical variables to account for recruitment clusters (a recruitment wave consisted of one control group and two weight-loss groups). At baseline, separate generalized linear gamma models were used to assess the association of weekly incontinence episode frequency, 24-hour involuntary urine loss, Incontinence Impact Questionnaire score, and Urogenital Distress Inventory score with UI management cost. Multivariable models adjusted for age, race (white or nonwhite), clinical site, and the baseline characteristics associated with P<.2 in both univariable and multivariable models (marital status, weight, hysterectomy, ever smoking more than 100 cigarettes, and SF-36 Physical Component Summary and Mental Component Summary scores).
Because no significant differences were observed between the two intervention groups for cost at baseline or change in cost from baseline to 6 or 18 months (all P>.05), the groups were combined to examine the effects of change in incontinence cost from baseline to 18 months. To estimate the effect of treatment on cost at 6 and 18 months, we used generalized estimating equations gamma repeated measures models, adjusting for clinical site, with robust standard errors to account for intervention clusters.16 The gamma model is useful for badly right-skewed outcomes like costs.17 Analyses were implemented in SAS 9.1 and STATA 10.0.
At baseline among all study participants, the mean age±standard deviation was 53±10 years, weight was 97±17 kg, total number of weekly UI episodes was 24±18, and 19% were African American. Annual household income was less than $40,000 for 27%, $40,000–$99,999 for 53%, and $100,000 or more for 20% of participants. No differences were observed between the weight loss and control group participants on baseline characteristics (Table 1).
Nearly 92% (n=311) of women reported some costs associated with incontinence management at baseline and more than 75% of women reported using absorbent pads, the most commonly used resource category. The mean estimated cost for UI management was $7.76±$14.40 (median $3.79; 25th, 75th percentile $0.87, %7.66) per week (Table 2) at baseline or a mean annual cost of $403 (median $197).
In univariable analysis, baseline mean weekly incontinence management cost increased by 15% for each increase of seven total incontinence episodes per week (one episode per day) and 18% for each increase of seven urgency incontinence episodes per week (both P<.01; Table 3). Increased costs were also associated with greater 24-hour involuntary urine loss (pad weight) and poorer (higher) Incontinence Impact Questionnaire and Urogenital Distress Inventory scores. In multivariable adjusted models, cost increased by 11% with each increase of seven incontinence episodes per week (P<.01). Using pad weight instead of episode frequency as a measure of incontinence severity, we observed a 5% increase in cost for every 10-g increase in 24-hour pad weight (P<.01). Costs also increased with poorer incontinence-specific quality of life (12% increase in costs for each 16-point increase in Incontinence Impact Questionnaire score [P<.001] and 4% increase in costs for each 11-point increase in Urogenital Distress Inventory score [P=.049]). There was no association between age, ethnicity, annual household income, or type of incontinence and costs in the adjusted models.
Changes in weight and UI frequency by treatment group at 6-month and 18-month follow-up have been reported previously.9 Cost data were available for 318 (94%) women at 6 months and 292 (86%) women at 18 months. Change in mean weekly costs for incontinence management were similar in the weight loss group and the control group at 6 months (−$3.28 compared with −$2.88; P=.27) and at 18 months (−$3.93 compared with −$5.16; P=.79). Change in median weekly costs and the proportion of women reporting any costs at 6 and 18 months (data not presented) were also similar in the two groups. Therefore, all subsequent analyses were performed on the combined randomized groups.
Among all women, we observed significant decreases in weight, weekly incontinence episode frequency, and pad weight from baseline to 6 and 18 months (all P<.001; Table 4). The proportion of women reporting any costs associated with incontinence decreased from 92% (311 of 338) at baseline to 79% (252 of 318) at 6 months and 68% (200 of 292) at 18 months (P<.001 for both). Reductions of mean costs were observed in all cost categories (Table 4). Total mean incontinence management costs, adjusted for clinical site, reduced by an average of 54% at 6 months and 81% at 18 months (P<.001 for both). For a woman with a mean baseline cost of $7.76 per week, this would amount to a reduction of $4.19 per week at 6 months and $6.29 per week at 18 months. Median costs also decreased from $3.79 per week at baseline to $1.97 per week at 6 months and $0.69 per week at 18 months (Wilcoxon signed rank sum test P value effect of time on cost was less than .001 for both comparisons).
In multivariable analyses, decreases in total, stress, and urgency UI frequency were each strongly and independently associated with reduction in costs at 6 and 18 months (Table 5). For example, for every decrease of seven total incontinence episodes per week, we observed an 18% decrease in mean weekly cost at 6 months and a 23% decrease at 18 months (both P<.001). Decreased 24-hour pad weight was associated with reductions in cost only at 18 months (P<.006). Greater improvement in Incontinence Impact Questionnaire and Urogenital Distress Inventory scores were also associated with decreased cost at 6 and 18 months (P<.001). In addition, decreases in costs were independently associated with weight loss at 6 and 18 months: for every 5-kg decrease in weight, we observed a 22% decrease in mean weekly cost at 6 months and a 21% decrease at 18 months (both P<.001). However, randomization group was not a significant predictor of change in cost at either 6 or 18 months. Decreases in cost were also associated with nonwhite race at 6 months (45% [95% CI 13%, 65%] greater decrease in weekly cost compared with white race).
Among overweight and obese women with UI who participated in a clinical weight loss trial, we observed a significant decrease in incontinence resource use and out-of-pocket costs for UI management after improvement in incontinence frequency. On average, at 18 months after randomization, women reported a 60% decrease in weekly incontinence episodes and an 81% decrease in incontinence management costs. This represents an annual savings of $327 per participant.
The reported costs of UI at baseline were substantial. More than 90% of women reported some costs associated with incontinence management, spending a mean of $7.76 per week or more than $400 per year out of pocket for UI management. This expenditure is also 0.5–1.0% of the median annual household income of women in the study ($40,000–$59,999), which is comparable to the mean annual out-of-pocket spending on prescription drugs for workers with health insurance benefits.18 Costs reported at study entry by this cohort were similar to costs observed in cross-sectional studies of women with incontinence.1–8,19–21 As with previous studies, we observed that patient costs for UI management were strongly associated with UI frequency but not with annual household income.5–7
Marked decreases in costs associated with incontinence of greater than 50% were observed in this trial. The strongest independent predictor of a reduction in cost was a decrease in total incontinence episode frequency, with mean cost decreasing by 18% ($1.40 per week) for each decrease of seven UI episodes per week. This suggests that women use fewer resources to manage UI as the frequency of incontinence decreases. Other significant predictors of reduction in cost were improvement in Urogenital Distress Inventory and Incontinence Impact Questionnaire scores. Interestingly, there was an association between change in 24-hour weight and change in cost only at 18 months, possibly suggesting a lag in time between decreased frequency of incontinence or volume of loss and a reduction in resource use or the inherent lack of reproducibility of pad testing, or both. The significance of the association between weight loss with decreased incontinence management costs independent of change in incontinence frequency remains uncertain.
Previous cross-sectional studies have observed that African American women report higher costs for incontinence management compared with white women.5,6 At baseline in this study, we observed a trend toward nonwhite women (84% of nonwhite women were African American) reporting higher costs associated with incontinence compared with white women (53% higher costs [−7%, 152%]; P=.09 in multivariable analyses). Independent of baseline incontinence frequency, baseline costs, and change in incontinence frequency, nonwhite women had greater reduction in cost compared with white women (P=.007) at 6 months, an association that was not observed at 18 months. These data suggest that UI management costs for African American women may be more responsive to changes in incontinence frequency than for white women.
In an earlier report,9 we observed a greater decrease in incontinence frequency in the weight loss group compared with the control group at 6 months. However, we found no difference in change in incontinence management cost over the same time period when the behavioral weight reduction group was compared with the control group. Thus, decreased incontinence frequency predicted decreased incontinence management cost regardless of randomized group assignment. Our inability to show a difference in cost reduction when randomized groups were compared may be due to no difference between groups in volume of urine lost estimated by the 24-hour pad test or insufficient power to detect a difference or both.
In contrast to incontinence costs for diagnosis, treatment, and institutionalization, a majority of routine care costs for incontinence incurred by community-dwelling people are paid out of pocket without reimbursement by third-party payers. This places a large cost burden directly on individuals, most often the elderly and women, who are adversely affected by disease as well as the cost of care. As we observed in this study, this cost burden may be reduced by other approaches to effective treatment for incontinence.
We studied overweight and obese community-dwelling women with UI enrolled in a randomized clinical trial of weight loss. Therefore, these results may not be generalizable to other women with incontinence or to those who undergo other incontinence treatments. Pads, laundry, and dry cleaning may also have been used for reasons other than UI, resulting in overestimating resource use for UI. Recall of resource use during a typical week was used to minimize participant burden. Although earlier reports have found both over- and underreporting of utilization of resources for management of UI,5,20 the fact that costs were reduced similarly in the randomized groups in this nonblinded trial suggests that women accurately reported resource use.
In summary, we found that managing UI was associated with substantial costs and that a reduction in incontinence frequency was associated with a significant reduction in incontinence management costs among obese and overweight women with UI. A decrease in the frequency of episodes was the strongest predictor of a reduction in incontinence management costs through 18 months. The trade-off between the costs of incontinence treatment and benefits in both patient-focused outcomes and decreased incontinence management costs should be explored with cost-utility analyses. The substantial incontinence-related costs incurred by women with UI and the reduction of costs we observed with decreased incontinence frequency suggests that other effective incontinence treatments may be cost-effective.
1. Wilson L, Brown JS, Park GE, Luc KO, Subak LL. Annual costs of urinary incontinence. Obstet Gynecol 2001;98:398–406.
2. Wagner TH, Hu TW. Economic costs of urinary incontinence in 1995. Urology 1998;51:355–61.
3. Dowell CJ, Bryant CM, Moore KH, Simons AM. Calculating the direct costs of urinary incontinence: a new test instrument. BJU Int 1999;83:596–606.
4. McClish DK, Wyman JF, Sale PG, Camp J, Earle B. Use and costs of incontinence pads in female study volunteers. Continence Program for Women Research Group [see comments]. J Wound Ostomy Contin Nurs 1999;26:207–8–210–3, –.
5. Subak LL, Brown JS, Kraus SR, Brubaker L, Lin F, Richter HE, et al.; Diagnostic Aspects of Incontinence Study Group. The “costs” of urinary incontinence for women. Obstet Gynecol 2006;107:908–16.
6. Subak L, Van Den Eeden S, Thom D, Creasman JM, Brown JS; Reproductive Risks for Incontinence Study at Kaiser Research Group. Urinary incontinence in women: direct costs of routine care. Am J Obstet Gynecol 2007;197:596.e1–9.
7. Subak L, Brubaker L, Chai TC, Creasman JM, Diokno AC, Goode PS, et al.; Urinary Incontinence Treatment Network. High costs of urinary incontinence among women electing surgery to treat stress incontinence. Obstet Gynecol 2008;111:899–907.
8. Tediosi F, Parazzini F, Bortolotti A, Garattini L. The cost of urinary incontinence in Italian women. A cross-sectional study. Gruppo di Studio Incontinenza. Pharmacoeconomics 2000;17:71–6.
9. Subak LL, Wing R, West DS, Franklin F, Vittinghoff E, Creasman JM, et al.. Weight loss to treat urinary incontinence in overweight and obese women. N Engl J Med 2009;360:481–90.
10. West DS, Gorin AA, Subak LL, Foster G, Bragg C, Hecht J, et al.; Program to Reduce Incontinence by Diet and Exercise (PRIDE) Research Group. A motivation-focused weight loss maintenance program is an effective alternative to a skill-based approach. Int J Obes (Lond) 2011;35:259–69.
11. Wing RR, West DS, Grady D, Creasman JM, Richter HE, Myers D, et al.; Program to Reduce Incontinence by Diet and Exercise Group. Effect of weight loss on urinary incontinence in overweight and obese women: results at 12 and 18 months. J Urol 2010;184:1005–10.
12. Burgio KL, Pearce LK, Lucco AJ. Staying dry: a practical guide to bladder control. Baltimore (MD): The Johns Hopkins University Press; 1989.
13. Shumaker SA, Wyman JF, Uebersax JS, McClish D, Fantl JA. Health-related quality of life measures for women with urinary incontinence: the Incontinence Impact Questionnaire and the Urogenital Distress Inventory. Continence Program in Women (CPW) Research Group. Qual Life Res 1994;3:291–306.
14. Barber MD, Spino C, Janz NK, Brubaker L, Nygaard I, Nager CW, et al.; Pelvic Floor Disorders Network. The minimum important differences for the urinary scales of the Pelvic Floor Distress Inventory and Pelvic Floor Impact Questionnaire. Am J Obstet Gynecol 2009;200:580.e1–7.
15. Abrams P, Blaivas JG, Stanton SL, Andersen JT. The standardisation of terminology of lower urinary tract function. The International Continence Society Committee on Standardisation of Terminology. Scand J Urol Nephrol Suppl 1988;114:5–19.
16. Zeger SL, Liang KY. Longitudianl data analysis for discrete and continuous outcomes. Biometrics 1986;42:121–30.
17. McCullagh P, Nelder J. Generalized linear models. 2nd ed. London: Chapman and Hall; 1989.
18. Joyce GF, Escarce JJ, Solomon MD, Goldman DP. Employer drug benefit plans and spending on prescription drugs. JAMA 2002;288:1733–9.
19. Samuelsson E, Mansson L, Milsom I. Incontinence aids in Sweden: users and costs. BJU Int 2001;88:893–8.
20. Sandvik H, Hunskaar S. The epidemiology of pad consumption among community-dwelling incontinent women. J Aging Health 1995;7:417–26.
21. Wyman J. The ‘costs’ of urinary incontinence. Eur Urol 1997;32(suppl 2):13–9.