T H Wagner1, J H Hibbard, M R Greenlick, L Kunkel. 1. VA HSR&D Health Economics Resource Center, Center for Primary Care and Outcomes Research, and Center for Health Policy, Stanford University, CA, USA. twagner@stanford.edu
Abstract
OBJECTIVE: To determine whether providing health information to residents of Boise ID had an effect on their self-reported medical utilization. RESEARCH DESIGN: The Healthwise Communities Project (HCP) evaluation followed a quasi-experimental design. SUBJECTS: Random households in metropolitan zip codes were mailed questionnaires before and after the HCP. A total of 5,909 surveys were returned. MEASURES: The dependent variable was self-reported number of visits to the doctor in the past year. A difference-in-differences estimator was used to assess the intervention's community-level effect. We also assessed the intervention's effect on the variance of self-report utilization. RESULTS: Boise residents had a higher adjusted odds of entering care (OR = 1.27, 95% CI 0.88, 1.85) and 0.1 more doctor visits compared with residents in the control cities; however, for both outcomes, the effects were small and not significant. Although the means changed little, the data suggest that the variance of utilization in Boise decreased. CONCLUSIONS: The HCP had a small effect on overall self-reported utilization. Although the findings were not statistically significant, a posthoc power analysis revealed that the study was underpowered to detect effects of this magnitude. It may be possible to achieve larger effects by enrolling motivated people into a clinical trial. However, these data suggest that population-based efforts to provide health information have a small effect on self-reported utilization.
OBJECTIVE: To determine whether providing health information to residents of Boise ID had an effect on their self-reported medical utilization. RESEARCH DESIGN: The Healthwise Communities Project (HCP) evaluation followed a quasi-experimental design. SUBJECTS: Random households in metropolitan zip codes were mailed questionnaires before and after the HCP. A total of 5,909 surveys were returned. MEASURES: The dependent variable was self-reported number of visits to the doctor in the past year. A difference-in-differences estimator was used to assess the intervention's community-level effect. We also assessed the intervention's effect on the variance of self-report utilization. RESULTS: Boise residents had a higher adjusted odds of entering care (OR = 1.27, 95% CI 0.88, 1.85) and 0.1 more doctor visits compared with residents in the control cities; however, for both outcomes, the effects were small and not significant. Although the means changed little, the data suggest that the variance of utilization in Boise decreased. CONCLUSIONS: The HCP had a small effect on overall self-reported utilization. Although the findings were not statistically significant, a posthoc power analysis revealed that the study was underpowered to detect effects of this magnitude. It may be possible to achieve larger effects by enrolling motivated people into a clinical trial. However, these data suggest that population-based efforts to provide health information have a small effect on self-reported utilization.
Authors: Anne C van der Gugten; Cuno S P M Uiterwaal; Theo J M Verheij; Cornelis K van der Ent Journal: Br J Gen Pract Date: 2015-02 Impact factor: 5.386
Authors: Wouter A Spoelman; Tobias N Bonten; Margot W M de Waal; Ton Drenthen; Ivo J M Smeele; Markus M J Nielen; Niels H Chavannes Journal: BMJ Open Date: 2016-11-09 Impact factor: 2.692
Authors: Anne C van der Gugten; Rob J R J de Leeuw; Theo J M Verheij; Cornelis K van der Ent; Marijke C Kars Journal: Scand J Prim Health Care Date: 2016-04-11 Impact factor: 2.581