Hypothesis / aims of study
Supervised pelvic floor muscle (PFM) physiotherapy is effective and recommended by international clinical practice guidelines as first-line treatment for stress or mixed urinary incontinence (UI) in women (Evidence Level 1; Recommendation Grade A). In many countries, PFM physiotherapy is routinely provided in individual one-on-one sessions, but not widely covered by the publicly funded health system, resulting in an important economic burden for patients, particularly older women. Evidence from randomized controlled trials (RCT) suggests that group-based PFM physiotherapy may also be effective to treat younger women with UI, and help overcome financial and human resource barriers. However, economic evaluations of PFM physiotherapy programs are scarce, and evidence of the cost-effectiveness of group-based vs. individual physiotherapy for UI in older women is lacking to date. Therefore, this study aims to present an economic evaluation of group-based and individualized PFM physiotherapy alongside the GROUP trial, focusing on participant and provider perspectives.
Summary of the GROUP trial.
GROUP (Group Rehabilitation Or IndividUal Physiotherapy) was a multicenter RCT aiming to determine whether or not group-based physiotherapy is inferior to individual pelvic floor physiotherapy in older women with stress or mixed UI . 362 women aged 60 years and over received an individual session with a trained pelvic floor physiotherapist to learn how to effectively contract their PFM. Participants were then randomized to a 12-week PFM training program, either in one-on-one sessions or as part of a group of up to eight women. Both study arms received the same intervention program, including education and progressive PFM training. Participants in the group-based intervention arm who reported difficulty with their exercises were offered (up to three) 20-minute one-on-one assessment sessions with their physiotherapist to confirm correct PFM contractions. Participants in the individual arm used intra-vaginal electromyography biofeedback as per usual practice. Women in both intervention groups were also assigned home PFM exercises during the entire study period. Treating physiotherapists monitored attendance to training sessions. Adherence to the home PFM exercise program was assessed through the participant exercise diary and follow-up contact calls at six, nine and 12 months. Participant- and provider-related costs (i.e. from a participant or provider perspective) were collected at baseline, post-treatment and at one year, using a modified version of the Dowell Bryant Incontinence Cost Index. Cure or improvement of UI was based on the percent reduction in the total number of UI episodes per week at one year. Cure or improvement of UI was assessed via the 7-day bladder diary relative to the number of UI episodes at baseline. For the cost-effectiveness analysis, we considered the 50% reduction cut-off in leakage episodes as a clinically important measure of effectiveness, with a non-inferiority threshold of 10%. 319 (153/178 from the group-based and 165/184 from the individual arm) participants completed the one-year follow-up and were included in a prespecified per protocol analysis. No imbalance on any potential confounders was found between intervention groups at baseline or follow-up. Twelve (7%) participants in the group-based arm requested short one-on-one PFM assessment sessions with their physiotherapists to confirm their capacity to contract PFM. Mean costs related to incontinence products were similar between groups at baseline, post-treatment or at one year (P=0.92, P=0.52 and P=0.91 respectively; details on Figure 2). Use of other treatments was low and similar between groups: 3% vs. 5% (P=0.49; details on Figure 2). 113 (74%) participants completing the group-based physiotherapy and 118 (72%) participants completing the individual physiotherapy presented clinically important improvement of their symptoms at one year. There was no evidence of meaningful difference between groups (risk ratio 1.03, 95% CI 0.90 to 1.18, P=0.64), with the lower CI bound within the 10% for non-inferiority margin.
Study design, materials and methods
Cost-analysis. Costs were estimated from both participant and provider perspectives and based on the current (2019) costs related to UI treatment over one year in the study area. All costs are reported in Figure 1 in Canadian dollars. Both group and individual physiotherapy costs were based on treatment session attendance and either the current private clinic practice (participant perspective), or the current mean labor hour of the specialized physiotherapists (provider perspective) participating in the GROUP trial, in addition to room cost (including staff time, consumables and all equipment necessary for clinical practice and specific to each treatment arm). Specifically for the group-based physiotherapy arm, costs for participants were considered as a package (initial assessment plus 12 treatment sessions), while provider costs were based on the number of participants’ attending per session. One-on-one assessments of a mean duration of 20 minutes were estimated to cost 1/3 of a one-hour individual assessment (details in Figure 1).
Additional participant costs were estimated based on reported use of incontinence products (various washable or disposable brands/sizes) and other treatments undertaken during the study period (any medication or additional treatment sessions) as reported in the Dowell Bryant Incontinence Cost Index. From these reports, we estimated the annual cost per participant based on the frequency and mean unit cost commonly applied to the study area’s population. The additional cost of the provider follow-up calls was estimated to be 1/6 of the physiotherapist labor hour, based on their mean duration of 10 minutes. The additional cost of repeated attempts to contact participants was also included. Statistics. Given non-normality of cost data, we used a Mann-Whitney test to compare the costs between the two independent samples. Bootstrapping (using 4,000 iterations) was used to estimate the 95% confidence interval (CI) around the difference in mean costs between groups. Group-based physiotherapy was expected to be less costly than individual physiotherapy. A cost-effectiveness analysis was planned if GROUP study results showed that group-based physiotherapy was inferior to individual-based treatments.
Interpretation of results
This is the first economical evaluation comparing group-based and individual physiotherapy interventions to treat UI in older women. For the same efficacy, the main cost driver from both perspectives was the mean primary treatment cost, which was significantly less costly in the group-based physiotherapy arm. Overall, group-based physiotherapy was at least as clinically effective as individual physiotherapy and at least 60% less costly, regardless of who pays for treatment. Group-based physiotherapy can be considered a more affordable first-line conservative treatment of UI.