Asthma Care Flow Modeling for Children in the Medicaid Program

PI’s: Nicoleta ServanJulie Swann James Bost

In this project, we studied  patient healthcare utilization and expenditure using the Medicaid Analytic Extract (MAX) claims files for pediatric asthma care for Georgia and North Carolina. The objective was to profile patient-level utilization with inferences on provider-to-provider care pathways, potential cost-saving interventions and the effects of geographical access on utilization. Our study was divided into two analyses: an annual study and multi-year study.

In the annual study we profile and characterize utilization and expenditure for pediatric asthma care using one year of patient-level claims data. The objective of this study is to provide approaches for annual reporting of pediatric asthma care across a large number of patients. The output of this study consists of multiple patient-level utilization profiles for multiple utilization profiles. We further used the derived utilization profiles in inferences on care expenditure and determinants of variations in utilization.


In the multi-year study, we examine multi-year longitudinal utilization profiles using the claims files from years 2005 – 2009, allowing for analysis of a larger number of healthcare visits and transitions between care events over several years of care.  In this study, we also considered the timing between different care events.  Similar to the annual study, the outputs consists of clustering of patients according to their longitudinal utilization of the care system.


With similar Medicaid populations but different care coordination systems and effectiveness rankings, we find some striking similarities in the utilization behaviors for pediatric asthma care between GA and NC.

Overall, this study has some important implications for health care providers and policy makers. According to the care practice recommendations, if a child visits the emergency department for asthma care then he/she needs to be referred back to primary care. In both Georgia and North Carolina, the transition from emergency department or from hospitalization to physician’s office varies across utilization profiles, with very low probability of physician’s office follow-up visits for the patients using emergency department and hospitalization regularly. Those follow-up visits vary with the patient’s profile, indicating that different interventions should be considered for each of the profile of patients. More importantly, in both states, patients who are visiting emergency department regularly for asthma care are few, with long periods of time between readmissions.

Asthma-controlled medication uptake is strongly connected with physician’s office visits across three utilization profiles, where in one profile where it is not, patients are regularly taking medication with no significant severe outcomes recorded. From the strength of the relationship between physician’s office and medication (re)fills, and lack of connection of those two event types to the emergency department, those patients who visit a physician’s office on a regular basis while staying on asthma-controlled medication are unlikely to have emergency department visits in both states. This finding provides evidence that asthma can be controlled with regular physician’s office visits and medication, with the potential of eliminating costly emergency department visits.

When comparing expenditure for the two states, there are more dissimilarities than similarities. GA experiences lower expenditures than NC across all types of care in 2009. The total Medicaid payments for emergency department and primary care visits are higher in NC than in GA by 38% and 179%, respectively. While the Medicaid payments due to emergency department and primary care are similar for NC, they are different for GA with the Medicaid payments for emergency departments almost twice those of PC.