Validation of bladder health scales and function indices for women's research

Abstract

Background: Existing bladder-specific measures lack the ability to assess the full range of bladder health, from poor to optimal health.

Objective: This study aimed to report evidence of validity of the self-administered, multidimensional bladder health scales and function indices for research in adult women.

Study design: A cross-sectional population-based validation study with random assignment to paper or electronic administration was conducted using national address-based probability sampling supplemented by purposive sampling of women with lower urinary tract symptoms in 7 clinical research centers. Construct validity of the bladder health scales and function indices was guided by a multitrait-multimethod approach using health and condition-specific questionnaires, bladder diaries, expert ratings of bladder health, and noninvasive bladder function testing. Internal dimensional validity was evaluated using factor analysis; internal reliability was assessed using paired t-tests and 2-way mixed-effects intraclass correlation coefficient models. Chi-square, Fisher exact, or t-tests were used for mode comparisons. Convergent validity was evaluated using Pearson correlations with the external construct measures, and known-group validity was established with comparison of women known and unknown to be symptomatic of urinary conditions.

Results: The sample included 1072 participants. Factor analysis identified 10 scales, with Cronbach's alpha ranging from 0.74 to 0.94. Intraclass correlation coefficients of scales ranged from 0.55 to 0.94. Convergent validity of the 10 scales and 6 indices ranged from 0.52 to 0.83. Known-group validity was confirmed for all scales and indices. Item distribution was similar by mode of administration.

Conclusion: The paper and electronic forms of the bladder health scales and function indices are reliable and valid measures of bladder health for use in women's health research.

Trial registration: ClinicalTrials.gov NCT04016298.

Keywords: adaptive behavior; construct validity; instrument; lower urinary tract symptoms; multitrait-multimethod.

FIGURE 1. Study sampling and disposition

Asterisk denotes 2 did not complete diaries. CASI, computer-assisted self-interview; CRC, clinical research coordinator; CSS, clinical research site sample; LUTS, lower urinary tract symptoms; NS, national sample; PAPI, paper-and-pencil interview; PPBC, Patient Perception of Bladder Condition; RUCC, Rural-Urban Continuum Codes.

FIGURE 2. Validation framework

MTMM validation framework of external construct measures from different data sources and different data collection methods. The (gray arrows) indicate measures a priori excluded as valid external constructs. LUTS, lower urinary tract symptoms; MOS, Medical Outcomes Study; MTMM, multitrait-multimethod; PFDI, Pelvic Floor Distress Inventory; QOL, quality of life.

FIGURE 3. Adaptive behavior–adjusted and unadjusted BHS distributions for the NS

The distribution of individual BHS scores with and without adaptive behavior adjustment applied. All scores run from 0 (absence of bladder health) to 100 (maximum bladder health). Vertical axis depicts percentage of response frequency to scale levels, with the maximum response frequency varying by scale. BHS, bladder health scale; NS, national sample.