Talk to the Veterans Crisis Line now
An official website of the United States government
Search | Search by Center | Search by Source | Keywords in Title
Rinne S, Wong ES, Perkins M, Bryson CL, Liu C. Non-responders to a patient satisfaction survey incurred higher medical utilization and lower medication adherence. Poster session presented at: AcademyHealth Annual Research Meeting; 2014 Jun 10; San Diego, CA.
Patient satisfaction surveys are commonly used to assess quality improvement efforts -Little is known about the effect of non-response bias on patient satisfaction surveys -Identify risk factors for survey non-response -Compare medical costs, health care utilization, and medication adherence among responders and non-responders of a patient satisfaction survey -Study sample: 40,766 veterans -diabetics on oral antihyperglycemic agents -seen in VA outpatient clinics and subsequently sent a patient satisfaction survey (SHEP) -during period of 9/30/04 - 10/1/07 -SHEP: Survey of Healthcare Experiences of Patients -Stratified random sampling of each ambulatory care clinic in VA -26,051 (64%) patients in our sample responded to SHEP and 14,715 (36%) did not -Outcomes: -Medical costs in the year following SHEP -Clinic visits and hospitalizations -Medication adherence (yes = MPR > 0.8) -Analysis: -Logistic regression to examine risk factors for survey non-response -Generalized linear models to estimate incremental cost for survey non-response -Negative binomial regression to assess impact of survey non-response on health care utilization -Logistic regression to assess the effect of -Only analyzed information on VA costs, health care utilization, and medication use -Limited generalizability to non-VA settings -Risk factors for non-response included younger, single, non-white race, and receiving free care -Patients with depression and schizophrenia were less likely to respond to surveys -Non-responders had significantly higher health care costs and utilization, but lower medication adherence
