This analysis suggests that state Medicaid/CHIP policy changes such as reimbursement increases, incentives, and delivery system changes may be associated with greater preventive care receipt among publicly insured children. By using hazard analyses, which have not been used in prior research to study these issues, our work also provides new insights into how policy changes can influence the timing of preventive care receipt following enrollment in Medicaid/CHIP. Given the paucity of research on Medicaid policy changes and the limitations to our analytic approach driven largely by data constraints, more research is needed on this topic. Stronger research designs will require new data investments, because no data are currently available to track the receipt of well-child and preventive dental care for publicly enrolled children in each state consistently over time or to measure Medicaid reimbursement rates for such visits.
In Kentucky, the dental policy changes that were adopted were associated with an increased likelihood of receiving preventive dental care and earlier receipt of such care following enrollment. The combination of the introduction of a second covered preventive dental visit, the 30 percent dental reimbursement rate increase (compared with the 12.5 percent increase for well-child visits), and the longer postperiod observed for dental changes (24 months) compared with well-child changes (12 months) may have contributed to the stronger dental findings. In Idaho, it appears that it might have been slightly easier for families to obtain preventive dental care for their children under managed care relative to the prior fee-for-service system, which is consistent with reports that the number of dentists accepting Medicaid/CHIP increased (Kenney and Pelletier 2010
), although the effects are small and only marginally significant in our main model specification. While the reimbursement rate increases for well-child visits were associated with small increases in the receipt of preventive care in Idaho, no consistent effects were found for Kentucky.
The new wellness incentives introduced in Idaho appeared to increase rates of well-child care by a substantial amount among the children targeted by the incentives. Similar attempts at rewarding beneficiaries for obtaining preventive care in California, Pennsylvania, and Florida have had mixed results, as states have struggled to make beneficiaries aware of and encourage redemption of the rewards (Redmond, Solomon, and Lin 2007
; Coughlin et al. 2008
;). The automatic redemption of rewards (i.e., premium offset) in Idaho's program may be a unique feature which contributed to its success. In fact, the share of children who lost coverage due to failure to pay premiums dropped from 15 to 20 percent before introduction of the incentives to 4 percent in state fiscal year 2008 and less than 1 percent in state fiscal year 2009 (Kenney and Pelletier 2010
). However, despite the apparent success of the incentives in Idaho, the overall increase in well-child receipt was low in the state because such a small percentage of enrolled children (just around 11 percent) were eligible for the benefit. Moreover, children in Idaho's premium-paying CHIP categories are substantially more likely than children in Medicaid to receive well-child care, raising equity concerns. The state currently has no plans to expand the program to Medicaid children due to budget constraints and the lack of a CMS-approved mechanism for rewarding families who are not required to pay premiums. These challenges, along with our findings that suggest demand-side initiatives in Idaho may have produced substantial increases in well-child visit receipt, demonstrate the need for further research on ways that states can incentivize families to seek preventive care for their publicly insured children.
There are a number of limitations to this analysis. First, the reliance on a pre–post design to identify impacts may not adequately control for confounding changes in case mix or service delivery systems that could affect preventive care receipt for children covered by Medicaid/CHIP. Although previous case study work did not identify any major changes that would have confounded the analysis of these policy changes, that possibility cannot be ruled out (Kenney, Pelletier, and Costich 2010
). Second, experiences of children who are served exclusively by community health centers and rural health clinics in Idaho could not be examined due to the lack of claims data on visits to these providers.
Third, coding practices may bias downward the estimates of the extent to which children are receiving preventive care based on claims data (Steinwachs et al. 1998
) and the new incentives may have led to more coding of preventive care during primary care visits in Idaho and not to an increase in the receipt of well-child care. However, we found no evidence in Idaho of offsetting decreases in nonpreventive primary care among the higher-income CHIP category of children in either age group or among the younger children in the lower-income CHIP category, which were the groups where the largest increases in preventive service receipt were observed (data not shown). Therefore, it appears that the increased reported receipt of well-child care in these three premium-paying categories may reflect greater provision of well-child care and not simply changes in coding. Fourth, the impact estimates should be interpreted as early impacts since the postperiod is short for many of the analyses. Though we ran sensitivity analyses that suggested a small lagged effect or a change in trend, we cannot be sure of the long run effect of the reforms. Finally, the analysis provides no information on the content of the preventive care that is being provided or whether the increased receipt of preventive care led to improvements in child health and functioning. More research is needed on that question.
The recently enacted Patient Protection and Affordable Care Act (PL 111–148) mandates an increase in Medicaid reimbursement for primary care services provided by primary care physicians up to Medicare levels in 2013 and 2014, paid entirely by the federal government. This provision, designed to address access problems in Medicaid, should increase rates in Kentucky by about 17 percent on average but will have little or no effect on rates in Idaho since that state already reimburses Medicaid providers at or above Medicare rates (American Academy of Pediatrics 2009
While this analysis suggests that Medicaid programs may be able to improve preventive care receipt through the use of demand and supply side initiatives such as increased reimbursement rates, incentives, and changes in the service delivery system, rates of well-child and preventive dental visits still fall short of recommended guidelines in both states even after implementation of the policy changes. The modest impacts we find for most of the policy changes we examine suggest that behavior change around prevention is difficult even when financial incentives are better aligned. Thus, Medicaid programs may need to address other factors (e.g., training, perceived gaps in cultural competence, negativism about Medicaid) that limit the supply of Medicaid providers (Edelstein 2009
). However, national survey data suggest that Kentucky and Idaho may be performing on par with other public programs in terms of children's preventive care use and that receipt of preventive care may actually be higher in Kentucky and Idaho among children with public insurance than among children with private coverage (National Survey of Children's Health 2007
). Thus, increasing the receipt of preventive care for children may require addressing both demand and supply side barriers, regardless of insurance type.
To meet recommended targets, states may want to consider conducting outreach about the benefits of preventive care and testing alternative approaches for rewarding both families and providers for preventive care provision. For example, states with PCCM models may need to give primary care providers greater incentives to increase preventive care receipt and to make appropriate referrals when follow-up care is needed. Addressing access to specialty care for children in Medicaid/CHIP may also be needed so that providers can make appropriate referrals to ensure that children receive needed follow-up care for problems that are diagnosed during preventive visits. Finally, in order for states to adequately monitor preventive care receipt and to ensure that the services that are provided to individual children are tailored to their specific health needs and risks, Medicaid/CHIP claims and encounter data may need to include additional fields on the child's health status and risk factors and on the procedures and counseling occurring during visits (Schor 2004
; Bergman, Plsek, and Saunders 2006