Presentation Authors: Kathryn E. Flynn*, Milwaukee, WI, Sarah A. Mansfield, Abigail R. Smith, Brenda W. Gillespie, Ann Arbor, MI, Catherine S. Bradley, Iowa City, IA, J. Quentin Clemens, Margaret E. Helmuth, Ann Arbor, MI, Pooja Talaty, Glenview, IL, H. Henry Lai, St. Louis, MO, Ziya Kirkali, Bethesda, MD, Kevin P. Weinfurt, and the LURN Study Group, Durham, NC
Introduction: Self-reported measurement tools often refer to a time period over which respondents are asked to recall their experience, e.g., &[Prime]In the past 7 daysâ€¦&[Prime] However, there is limited evidence for how accurately people recall their lower urinary tract symptoms (LUTS). Our aim was to empirically evaluate the correspondence between 1) average daily recall over 7 days and weekly recall and 2) average daily recall over 30 days and monthly recall.
Methods: Participants (261 women, 254 men) were recruited from the 6 sites of the Symptoms of Lower Urinary Tract Dysfunction Research Network (LURN). We evaluated 18 items representing 7 types of symptoms covering storage, voiding, and post-micturition among participants that endorsed the symptom at any time during the study period. Item responses on the 24-hour recall forms were averaged over a 7- or 30-day period and compared to the corresponding 7-day or 30-day recall version of the item. Between-person correlations were assessed using Pearson&[prime]s correlation, and bias was calculated as the difference between the recalled item and the average daily item score. Positive bias is systematic over reporting of 7- or 30-day recall compared to averaged daily recall, negative bias is systematic under reporting.
Results: Accuracy of recall varied by sex and depending on the type of symptom. Among women, the correlations between average 24-hour reports and recalled reports ranged from 0.72-0.89 (7-day) and 0.71-0.91 (30-day), while correlations for men ranged from 0.68-0.90 (7-day) and 0.68-0.95 (30-day). Most items did not show systematic bias (Figure). Exceptions were over-reporting of urgency (7- and 30-day), slow/weak stream (7-day), incomplete emptying (men, 7-day) nighttime frequency (women, 7-day) and post-micturition dribble (women, 7 day). Urgency incontinence was underreported, especially among women. Recall was best (low bias and high correlations) for daytime frequency and stress incontinence.
Conclusions: In self-reported measurement of lower urinary tract symptoms (LUTS), the recall period should balance accurate measurement and patient burden. In general, recalled reports tracked well with average daily reports for men and women and systematic bias was minimal, supporting the use of 7- and 30-day recall periods.
Source of Funding: Funding provided by grants from NIH/NIDDK