Predictors of successful prompted voiding among incontinent nursing home residents

Abstract

Objective: To develop a simple, noninvasive assessment strategy that will enable nursing home staff to identify incontinent residents who respond well to prompted voiding.

Design: Incontinent nursing home residents underwent an extensive clinical and functional assessment and then 7 days of prompted voiding. Data from the assessment and the first 3 days of prompted voiding were used to predict responsiveness to the intervention.

Setting: Seven nursing homes.

Patients: A cohort of 191 incontinent, long-stay nursing home residents who passed a simple behavioral screen (able to state their name or reliably point to one of two named objects).

Intervention: Prompted voiding was carried out by trained research nurse's aides from 7 AM to 7 PM for 7 days. The intervention was maintained in responsive residents 5 days per week for an additional 9 weeks.

Main outcome measures: Physical checks for wetness were done by research staff hourly from 7 AM to 7 PM for 3 days in a baseline condition, during days 5 through 7 of the 7-day prompted voiding intervention, and for 3 days at the end of 9 weeks of prompted voiding in the responsive group. Outcome measures were percentage of checks wet and response to prompted voiding, with "responders" defined as residents with an average of one or fewer wet episode per day on days 5 through 7 of prompted voiding.

Results: Seventy-eight (41%) of the residents were responders. Their wet percentage went from 26.7% to 6.4% at the end of 1 week and was maintained at 9.6% after 9 weeks of prompted voiding. The best predictors of responsiveness were the wet percentage and the appropriate toileting percentage during the first 3 days of prompted voiding, the self-care subscale score of the Multidimensional Observational Scale for the Elderly, and the ability to ambulate without human assistance. The best sensitivity and specificity in identifying responders was achieved when either the wet percentage was lower than 20% or the appropriate toileting percentage was higher than 66% during the first 3 days of prompted voiding (sensitivity, 87%; specificity, 69%). Those residents falsely identified as responders by these criteria still had a 46% relative reduction in wetness.

Conclusions: A substantial proportion of nursing home residents respond well to prompted voiding. The most responsive residents can be easily identified using data collected during a 3-day trial of the intervention. The assessment strategy is consistent with federal guidelines and could be used to facilitate quality control by assessing changes in percentage of wetness from the expected norm.