Research article
Self-reported adherence with the use of a device in a clinical trial as validated by electronic monitors: the VIBES study
1 Hebrew SeniorLife Institute for Aging Research, Musculoskeletal Research Center, Boston, MA, USA
2 Harvard Medical School, Department of Medicine, Boston, MA, USA
3 Boston University School of Public Health, Data Coordinating Center, Boston, MA, USA
4 Cecil G. Sheps Center for Health Services Research Chapel Hill, The University of North Carolina, Chapel Hill, NC, 27412, USA
5 University of Massachusetts Medical School, Worcester, MA, 01655, USA
6 Department of Biomedical Engineering Stony Brook, SUNY Stony Brook, New York, NY, 10027, USA
BMC Medical Research Methodology 2012, 12:171 doi:10.1186/1471-2288-12-171
Published: 14 November 2012Abstract
Background
Adherences to treatments that require a behavioral action often rely on self-reported recall, yet it is vital to determine whether real time self reporting of adherence using a simple logbook accurately captures adherence. The purpose of this study was to determine whether real time self-reported adherence is an accurate measurement of device usage during a clinical trial by comparing it to electronic recording.
Methods
Using data collected from older adult men and women (N=135, mean age 82.3 yrs; range 66 to 98 yrs) participating in a clinical trial evaluating a vibrating platform for the treatment of osteoporosis, daily adherence to platform treatment was monitored using both self-reported written logs and electronically recorded radio-frequency identification card usage, enabling a direct comparison of the two methods over one year. Agreement between methods was also evaluated after stratification by age, gender, time in study, and cognition status.
Results
The two methods were in high agreement (overall intraclass correlation coefficient = 0.96). The agreement between the two methods did not differ between age groups, sex, time in study and cognitive function.
Conclusions
Using a log book to report adherence to a daily intervention requiring a behavioral action in older adults is an accurate and simple approach to use in clinical trials, as evidenced by the high degree of concordance with an electronic monitor.
Trial registration
Clinicaltrials.gov NCT00396994
