In the United States, state Medicaid programs pay for medical and dental care for children from low-income families and support nondental primary care providers delivering preventive oral health services (POHS) to young children in medical offices (“medical POHS”). Despite the potential of these policies to expand access to care, there is concern that they may replace dental visits with medical POHS. Using Medicaid claims from 38 states from 2006 to 2014, we conducted a repeated cross-sectional study and used linear probability regression to estimate the association between the annual proportion of children in a county receiving medical POHS and the probability that a child received 1) dental POHS and 2) a dental visit in a given year. Models included county and year fixed effects and controlled for child- and county-level factors, and standard errors were clustered at the state level. In a weighted population of 45.1 million child-years (age, 6 mo to <6 y), we found no significant nor substantively important association between the proportion of children in a county receiving medical POHS and the probability that a child received dental POHS or a dental visit. Additionally, we found an almost zero probability (<0.001) that the reduction in dental POHS was at least as large as the expansion in medical POHS (full substitution) and a 0.50 probability that increased medical POHS was associated with an increase in dental POHS of at least 6.6% of the expansion of medical POHS. Results were similar when receipt of dental visits was examined. This study failed to find evidence that medical POHS replaced dental visits for young children enrolled in Medicaid and, in fact, offers evidence that increased medical POHS was associated with increased utilization of dental care. Given lower-than-desired rates of dental visits for this population, delivery of medical POHS should be expanded.
The RAND Corporation is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND’s publications do not necessarily reflect the opinions of its research clients and sponsors.