: There is scant research integrating information from state Consumer Assessment of Health Plans Surveys (CAHPS), state Medicaid eligibility files, and state HMO encounter data files in order to analyze caregiver evaluations and assessments of care at the sub-population level. Such efforts could play an important role in allowing states to prioritize discrete populations for targeted quality improvement initiatives. The primary goal of this dissertation is to examine differences in caregiver reported evaluations and experiences with health care between SCHIP and Medicaid children enrolled in HealthChoice, Maryland's combined Medicaid/SCHIP managed care program. Differences will be explored in both a random sample of children enrolled in state Medicaid HMOs and in a cohort of children with special health care needs. Non-response bias will be assessed by comparing differences between respondent and non-respondent children with respect to Medicaid eligibility category (i.e., SCHIP versus Medicaid status), demographic characteristics, and health status characteristics, the latter using information derived from Medicaid encounter data files. This is a cross-sectional study using data from 4,680 respondent caregivers of enrolled children collected for Maryland's 2001 Medicaid Child CAHPS; information from the State's Medicaid/SCHIP eligibility and HealthChoice encounter data files was appended to CAHPS sample and respondent information. The survey, which over-sampled probable children with special health care needs, piloted a special screener and module providing more in-depth indicators of child health status and assessments of access to care, physician communication and coordination of care. The dissertation analysis will involve definition and testing of appropriate statistical models to examine: (1) differences in evaluations and experiences with care between SCHIP and Medicaid children; (2) such differences with care between """"""""carve-out"""""""" and non-""""""""carve-out"""""""" users; and (3) differences in geo-demographic and health status characteristics between respondent and non-respondent children. This study may provide some indication of how HealthChoice is performing among different segments of a growing and diverse low-income, pediatric population. Ultimately, this may allow Maryland to design quality improvement efforts that are more targeted, efficient and effective in ameliorating disparities in the quality of care, if found, for different groups of HealthChoice enrolled children.
