Title: PDB44 Differences in Utilization of and Expenditures on Office-Based Health Care Between Uninsured and Insured Children 0-17 Years of Age From 2004-2008: Results From the Medical Expenditure Panel Survey
Abstract: To explain differences in number of and expenditures on office-based health care visits between insured and uninsured children. Using four two-year panels from the Medical Expenditure Panel Survey (MEPS) from 2004-2008, a 1:1 nearest-neighbor-with-replacement case-control design (matching on age category, race, region, health status, change in health status, specific MEPS panel, and chronic condition(s)) was used to match a continuously (24-months) uninsured (case) with a continuously insured (control) child. The Wilcoxon matched-pairs test was used to compare the mean number of office visits and mean expenditures between cases and controls. Then, ordinary-least-squares regression models to explain the size of the difference between case and control were estimated, accounting for survey design. Out of the approximately 257 million (weighted total) children aged 0-17 across 4 MEPS panels, 4.13% had no health-insurance coverage over the 2-year panel, while 76.95% experienced continuous coverage. The average number of office visits for uninsured children was 0.07 (p-value < 0.01) less per month than that for insured children; mean expenditures on these visits per month was $11.57 (p-value < 0.01) less for uninsured children. Regression predictors with statistically significant coefficients included average healthcare status, change in health care status over the two-year panel, age category, race, income category, and region of the country. Children whose health status improved over the panel had a drop (p-value < 0.01) in the difference between case and control in office-based health care expenditure, relative to children whose health status remained steady. Utilization of and spending on office-based health care are significantly higher for children covered by health insurance. Moreover, the penalty for lack of insurance can be explained by important predictors including income, age category, and health status. The lack-of-insurance barrier for office-based visits may potentially also be a barrier for patient-centered medical homes.