Effect of Season on Longitudinal Analyses of Health Status

• Activity Number: 199
• Type: Topic Contributed
• Date/Time: Tuesday, August 13, 2002 : 10:30 AM to 12:20 PM
• Sponsor: Section on Health Policy Statistics*
• Author(s): Ann Derleth*+ and Paula Diehr and Mary McDonell and Stephan Fihn
• Affiliation(s): University of Washington and University of Washington and VA Puget Sound Health Care System and VA Puget Sound Health Care System
• Keywords: SF-PCS ; SF-MCS ; Seasonality ; Longitudinal ; Health Status ; Health Status Summary Scores

Abstract

OBJECTIVE: Studies of treatment effects often use self-report measures of quality of life as one of their outcomes. We sought to determine whether there is seasonal variation in SF-36 summary scores (PCS and MCS) collected longitudinally. METHODS: We examined the distributions of the SF-PCS and SF-MCS collected longitudinally over 30 months from patients enrolled in primary care clinics at seven Veterans Affairs medical centers in the Ambulatory Care Quality Improvement Project (ACQUIP). We examined scores by the calendar month, adjusted for demographic, socio-economic and comorbid conditions. FINDINGS: There was a "late spring to early summer" peak in SF-MCS scores and to a lesser extent, a "spring" and "fall" peak in SF-PCS scores. These differences are consistent across all sites. CONCLUSION: There appears to be seasonal variation in the SF-36 summary scores. Researchers using the SF-36 measures might take the possibility of such variation into account when designing studies using health status measures. Timing of study enrollment, length of follow-up, and interpretation of change scores are important considerations in light of these findings.