Self-report measurement tools are used to characterize patients with lower urinary tract symptoms (LUTS). In clinical settings, bladder diaries are often used to assess voiding frequency, volume, urgency and incontinence episodes. Patients are instructed to complete bladder diaries in real time (though they may or may not), in contrast to self-report questionnaires, which usually ask respondents to describe their experiences over a recall period (e.g., “in the past 7 days”). In this study, we assessed correlations between bladder diary derived variables and recalled survey responses for 3- and 7-day recall items.

Participants with LUTS were recruited from six US tertiary referral centers and completed a modified version of the International Consultation on Incontinence Questionnaire (ICIQ) 3-day paper bladder diary. Participants recoded voiding frequency, bladder sensation at voiding, incontinence episodes, and pad changes. No information on voided volumes or intake data was collected. Participants also completed a 15 item LUTS questionnaire electronically consisting of items from the LURN Comprehensive Assessment of Self-reported Urinary Symptoms (CASUS). Items were asked using a 3- and 7-day recall period, covering the same time period as the bladder diary. The full study duration was 30 days, and participants were assigned to one of two study groups in order to assess study order effects: Group A completed the bladder diary, 3-day, and 7-day recall surveys in their first week in the study followed by a week of daily questionnaires, and Group B completed the bladder diary, 3-day, and 7-day recall surveys in the second week of the study after a week of daily questionnaires. We mapped items from the questionnaires to measurements captured on the bladder diary and calculated biserial or polyserial correlation coefficients to assess the linear relationship between the bladder diary and self-report items. Fisher’s z-transformation was used to construct 95% confidence intervals for each correlation and z-transformed correlations were compared using two-sample independent t-tests. Correlation coefficients above 0.7 were considered strong, between 0.5-0.7 were considered moderate, between 0.3-0.5 were considered low, and less than 0.3 were considered negligible. A difference in correlation coefficient of 0.2 was considered a meaningful difference.

Of 290 participants instructed to complete a bladder diary and the recalled reports, 175 participants completed bladder diaries and at least one corresponding questionnaire (i.e., 3- and/or 7-day recall). These participants had an average age of 58.7±14.8 years, were 53% female and predominantly white (87%). Correlations with bladder diary measurements ranged from 0.18-0.75 for 3-day recall items and from 0.04-0.74 for 7-day recall items. For both 3- and 7-day recall, we observed strong correlations between the diary and recalled symptoms for frequency of daytime and nighttime voids. Diary and recalled frequency of voids with urgency sensations were moderately correlated. There were low correlations between diary and recalled symptoms of urinary incontinence across all types of incontinence (i.e., reported and for both 3- and 7-day recall periods, with the exception of “leaking or wetting or pad without feeling it” which had negligible correlation for both recall periods. Correlation coefficients were higher for the 3-day recall period for seven of the fifteen items. That said, the 3- and 7-day correlations were generally similar to each other, and the differences for any pair did not exceed 0.15 (maximum difference was 0.13), and the confidence intervals overlapped. We also calculated correlations between bladder diary measurements and 3-day recall survey items by gender. Correlation coefficients ranged from -0.01-0.73 for men and 0.29-0.76 for women; correlations were similar between both genders for each item. Likewise, correlations between study groups were similar, correlation coefficients for the items ranged between 0.06-0.79 for participants in Group A and 0.30-0.78 for participants in Group B.

Missing bladder diary data were common, highlighting the particular patient burden associated with this method of data collection. We found strong correlations between paper bladder diaries and electronically collected 3- and 7-day recalled reports for daytime and nighttime voiding frequency, and weak to moderate correlations for urgency and incontinence.

Previous studies have suggested poor concordance between voiding diaries and recalled reports of particular symptoms, but in most of these studies, patients recalled experience was collected as part of their medical history, before completing a voiding diary. This study included a heterogeneous sample of men and women with diverse LUTS who completed both paper bladder diaries and recalled reports over the same time period. We found that missing or unusable clinical bladder diary data were common, whereas there were fewer missing data for electronically collected recalled reports of LUTS. Correlations between clinical bladder diaries and 3- and 7-day recalled reports were strong for voiding frequency but weak to moderate for urgency and incontinence. Additional analyses will investigate the mismatch to determine whether participants were over- or underestimating their LUTS on the bladder diaries.