ABSTRACT
Improving Medicaid program effectiveness for underserved populations is hampered by low survey response rates. This study determined how to maximize Medicaid consumer satisfaction survey response rates to the Consumer Assessment of Health Plans Study (CAHPS) survey. In a public immunization clinic, 8 focus groups and 15 extended interviews were used to assess consumer-preferred survey design features and incentives. To test hypotheses, we conducted the following trial. Out of 10,733 total participants in a Kansas Medicaid managed care plan, 3,685 eligible for CAHPS were unduplicated by household. After randomization of the 968 households with valid addresses to one of three groups, a controlled trial was conducted to assess response rates to CAHPS survey formats and incentives. Response rates were 35% for a standard mailing, 44% for a user-friendly low-literacy mailing, and 64% for a user-friendly low-literacy mailing with a $10 contingent incentive. Both experimental arms significantly improved response compared with the control; the response rate of the mailing group with the incentive was higher than the response rate of the group receiving that mailing without any incentive (p<0.0001). Using consumer-based preferences significantly increased response rates to this Medicaid satisfaction survey. Raising CAHPS response rates may increase validity of Medicaid consumer satisfaction information.
Subject(s)
Consumer Behavior/statistics & numerical data , Health Care Surveys/methods , Managed Care Programs/standards , Medicaid/standards , Poverty , Social Class , Adolescent , Adult , Centers for Medicare and Medicaid Services, U.S. , Female , Humans , Kansas , Male , Motivation , Postal Service , Research Design , Socioeconomic Factors , Surveys and Questionnaires , United StatesABSTRACT
BACKGROUND: Trauma surgeons have observed an increased rate of penetrating trauma during periods of increased unemployment. METHODS: During a 10-year period, the rate of unemployment in a metropolitan area was compared with the rate of intentional penetrating trauma at two Level I trauma centers. The total number of trauma cases was recorded. Assaults, derived from police records, were examined as an additional indicator of violent behavior. Pearson correlation coefficients were calculated to identify significant correlation between study variables. Stepwise maximum-likelihood estimation was used to derive a model predicting percent penetrating trauma. RESULTS: The rate of unemployment and percent penetrating trauma of individuals presenting to the emergency department were significantly correlated (P = 0.014). After stepwise estimation, a model was derived (r2 = 0.846, P = 0.014) that estimated percent penetrating trauma on the basis of percent unemployment and total number of trauma admissions. CONCLUSION: This long-term ecological study confirms that rates of penetrating trauma increase with increasing unemployment rates. This should inspire further research to identify areas of greatest need to improve delivery of resources and current public policy with the ultimate goal of decreasing the incidence of penetrating trauma.
