C A Mazurick1, J R Landis. 1. Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania School of Medicine, Philadelphia 19104-6021, USA.
Abstract
PURPOSE: We address the importance of 3 consecutive daily frequency volume charts relative to a single day in research studies for interstitial cystitis. In addition, differences in voiding patterns between weekend days and weekdays were evaluated. MATERIALS AND METHODS: Longitudinal frequency volume chart data for 305 women enrolled in the National Interstitial Cystitis Data Base Study before December 31, 1995 and meeting protocol specifications were summarized for 3 outcome measures. Longitudinal data methods designed to accommodate different within patient correlation patterns for repeat measures within and among patient contacts were implemented in a generalized estimating equation modeling framework, permitting overall testing of day and weekend effects. RESULTS: Day 2 and day 3 measures were not statistically different (1% level) from day 1 for nocturnal voiding frequency and nocturia. Although statistically significant for 24-hour frequency, the mean difference of -0.36 void daily between days 1 and 3 was not likely to be clinically significant. Also, measures on weekdays were not significantly different from those on weekends for each of the 3 outcome measures. CONCLUSIONS: Our results suggest that the current clinical research requirement of 3 consecutive days for frequency volume charts can be reduced to a single day. Furthermore, day selection need not be affected by weekday versus weekend considerations.
PURPOSE: We address the importance of 3 consecutive daily frequency volume charts relative to a single day in research studies for interstitial cystitis. In addition, differences in voiding patterns between weekend days and weekdays were evaluated. MATERIALS AND METHODS: Longitudinal frequency volume chart data for 305 women enrolled in the National Interstitial Cystitis Data Base Study before December 31, 1995 and meeting protocol specifications were summarized for 3 outcome measures. Longitudinal data methods designed to accommodate different within patient correlation patterns for repeat measures within and among patient contacts were implemented in a generalized estimating equation modeling framework, permitting overall testing of day and weekend effects. RESULTS: Day 2 and day 3 measures were not statistically different (1% level) from day 1 for nocturnal voiding frequency and nocturia. Although statistically significant for 24-hour frequency, the mean difference of -0.36 void daily between days 1 and 3 was not likely to be clinically significant. Also, measures on weekdays were not significantly different from those on weekends for each of the 3 outcome measures. CONCLUSIONS: Our results suggest that the current clinical research requirement of 3 consecutive days for frequency volume charts can be reduced to a single day. Furthermore, day selection need not be affected by weekday versus weekend considerations.