Recent health-related quality of life, but not change, predicted mortality and healthcare utilization

Objective To assess whether change in HRQOL over a seven-year period was associated with subsequent mortality and hospitalization, after adjusting for important covariates, in a cohort of U.S. Veterans. Methods We used data from a cohort of Veterans who completed a HRQOL survey in 1998 (Short-Form 36 for Veterans [SF-36V]) and a 2006 follow-up (SF-12V) HRQOL survey and SF12V scores were calculated at both time-points. We used logistic regression analyses to model the relationship between changes in the SF12-V physical component (PCS) and mental health component (MCS) scores and 1-year hospitalization and 1-year and 3.3-year mortality after the 2006 follow-up survey. Results 13,900 participants provided data for both the initial and follow-up surveys. We found no significant associations between changes in PCS or MCS and one-year hospitalization after adjusting for follow-up HRQOL and other covariates. We found significant but relatively weak associations between changes in MCS and our mortality outcomes. Conclusion Given the follow-up MCS and PCS, change in PCS over the previous 7 years added little information for predicting mortality and hospitalization Although the change in MCS added some information for predicting mortality. Therefore, knowledge of patient's current HRQOL generally provides meaningful information for predicting subsequent mortality and hospitalization.