The CACTUS-D randomized trial purpose was to evaluate if an urinary health education workshop was able to improve urinary incontinence and change beliefs in various countries and languages [1]. Women included who benefited from the urinary health education workshop reported, a year later, more frequently an impression of improvement, an improvement in their quality of life, and a reduction in use of pads.
Our goal in this secondary analysis was to identify which women were most likely to benefit from the urinary health workshop.

Our target population was women aged 65 years or more, leaving in the community, not treated for urinary incontinence despite reporting urinary leakage at least 2 times by week.
Each workshop led by a research assistant lasted approximately one hour and gathered 1 to 10 women. The aim of the urinary health education workshop was to change erroneous beliefs about accepting incontinence as a normal part of ageing, to explain that urinary incontinence is not inevitable and irreversible, and that it is possible to fight against urinary incontinence in everyday life and do pelvic floor contraction exercises regardless of age.
The severity of urinary incontinence symptoms was assessed by the ICIQ-FLUTS symptom questionnaire. The quality of life related to urinary incontinence was assesses by the I-QOL questionnaire. We asked participating women if they thought UI was part of normal aging and whether they practiced Kegel exercises. We collected also by phone information about demographic characteristics and lifestyle habits (age, education level, smoking, coffee or tea consumption, BMI), comorbidities (estimated health status, depression, high blood pressure, diabetes, falls), pelvic floor health (duration of urinary incontinence, faecal incontinence, constipation).
We interviewed again included women by phone at 6 and 12 months after the workshop.
Improvement in urinary symptoms and improvement in quality of life were defined by a favorable change between baseline and 12 months of the ICIQ-FLUTS and I-QOL scores equal or greater than the minimum clinical difference (MCID) [2]. In the absence of the 12-month follow-up, we used, when available, the 6-month follow-up. Impression of improvement was reported by women using a 7-likert scale (Patient Global Impression of Improvement or PGI-I). To explore if women had made change in their believes and behaviors, we asked women again at follow-up if they thought UI was part of normal aging and whether they practiced Kegel exercises.
Analysis: We used three logistic regressions to compare improved women with unimproved women considering the impression of improvement (PGI-I), urinary symptoms (ICIQ-FLUTS), and urinary quality of life (I-QOL). Variables tested focused on demographic characteristics and lifestyle habits, comorbidities, pelvic floor health, and changes observed during follow-up (change in ICIQ-FLUTS and IQOL scores for the regression about improvement defined by the PGI-I, Kegel exercises, beliefs about UI). Variables eligible for the multivariate analysis were the variables with a p <0.10 in univariate analysis (maximum model). The final model kept only the significant variables and the center (mandatory for all models).

552 incontinent women aged 65 or older were included in the intervention group and benefited from the urinary health education workshop; 392 women (71%) were followed at 6 or 12 months after the workshop, completed questionnaires, and constitute our analysis population.
In our sample at baseline, 39.0% of women were 80 years of age or older and 99.5% reported some comorbidities (8 comorbidities in average). Frequency of urinary incontinence was daily for 57.1% of women (“once a day” for 28.0% and “several times a day” for 29.1%). Incontinence types were stress urinary incontinence for 49 (12.5%), urge urinary incontinence for 109 (27.8%), mixed incontinence for 188 (48.0%), and other or non-classifiable incontinence types for 46 women (11.7%). 191 women (48.7%) thought urinary incontinence was normal at their age and 205 (52.3%) did not do Kegel exercises at baseline.
At follow-up, the average change in the ICIQ-FLUTS score was an improvement of -4.2 points and that of I-QOL an improvement of +10.4 points. There was a moderate correlation between the improvement in the ICIQ-FLUTS score and that of the I-QOL (Spearman coefficient -0.41, p <0.001). A significant improvement ("much better" or more according to the PGI-I) was reported by 61 women (15.6%). A significant improvement in urinary symptom score and urinary quality of life was observed in 55.1 and 61.7% of women respectively (Figures 1 and 2); and 39.5% of women reported a clinically significant improvement in both symptoms and urinary quality of life. 
Factors associated with greater chances of improvement (patient reported outcome using PGI-I) were not to believe that incontinence is part of normal aging at baseline, and improvement of urinary symptoms at follow-up. The improvement reported by PGI-I was similar in the case of SUI, UUI, and MUI. Factors associated with a significant improvement in urinary symptoms (MCID or more) were the greater severity of urinary incontinence at baseline and being overweight . Factors associated with a significant improvement in urinary quality of life (MCID or more) were a poor urinary quality of life at baseline and be under 81 years of age.

Our analysis shows that a good improvement was more often reported after the education workshop among older women who declared that they refuse to believe that urinary incontinence is part of aging, and in those who have seen their urinary symptoms improved. Chances of improving urinary symptoms were higher in obese women, in case of severe urinary symptoms at baseline, and in women who started Kegel exercises. Chances of improvement in quality of life were higher for the oldest women and those who reported the poorer quality of life at baseline.
Among the characteristics of the urinary health education workshop evaluated in the CACTUS-D study, we must take into account in interpreting our results, the lack of individual adaptation and clinical expertise. That makes the difference between an education workshop and medical care. The facilitator in charge of the workshop was not a doctor and did not know the severity of the symptoms, the history, and the expectations of the women present. It is interesting to observe a positive effect on urinary symptoms in women who reported the most severe symptoms and a positive effect on quality of life among those with the most degraded urinary quality of life. These results reinforce the rational of the message delivered during the workshop: urinary incontinence is not part of normal aging; it is always possible to do something. One of the obstacles to improvement in our secondary analysis was to believe otherwise, it is likely that in this case the subject does not have the courage to change his habits.

It was possible to change the mind about a quarter of older women who thought that urinary incontinence was normal and to change the habits of more than half of those who did not do Kegel exercises. A short, inexpensive and non-medical intervention can change the mindset and habits of older women suffering from incontinence and not seeking care. A significant improvement is possible even in case of severe symptoms; it is therefore legitimate to offer the workshop to all incontinent elderly women living in the community.

Tannenbaum C, van den Heuvel E, Fritel X, Southall K, Jutai J, Rajabali S, Wagg A. Continence Across Continents To Upend Stigma and Dependency (CACTUS-D): study protocol for a cluster randomized controlled trial. Trials 2015
Halme AS, Fritel X, Benedetti A, Eng K, Tannenbaum C. Implications of the Minimal Clinically Important Difference for Health-Related Quality-of-Life Outcomes: A Comparison of Sample Size Requirements for an Incontinence Treatment Trial. Val Health 2015