P3B079: Health Care Utilization for Children with Medical Complexity in Foster Care

Background: While children with medical complexity (CMC) represent a small percentage of US children, they account for up to one-third of pediatric healthcare spending and over half of pediatric inpatient charges. Children in foster care (FC) also account for disproportionate healthcare utilization and spending. The objective of the study was to describe the population of CMC in FC and their healthcare use, compared to Medicaid-enrolled CMC not in FC.

Methods: Secondary analysis of a 40% random sample of Medicaid claims data from 28 states over 2010-2012. We used a sample of children enrolled in all three years and applied the Pediatric Medical Complexity Algorithm (PMCA) to inpatient and ED claims data to identify CMC based on having a complex, chronic disease. Reported summary statistics describe child demographic characteristics along with health utilization. Based on child eligibility over the 3-year period, we compare children “ever” enrolled in FC; “ever” enrolled because of a disability and never in FC; or “always” enrolled as low-income child (never FC, never disabled). All analyses conducted using STATA17.

Results: Among our sample, 8.2% (17,590/195,969) of children in FC group were CMC, compared to 22.1% (88,298/311,453) CMC in disabled group and 3.0% (155,577/4,962,742) CMC in low-income group (Table). There was significant variation across states, e.g. 4.8% CMC in FC in NC and 11.7% in NY (Figure). Child demographics and health utilization for CMC vs non-CMC by Medicaid eligibility category are presented in Table. CMC in FC were more likely to be non-Hispanic Black and older than other categories. Over the 3-year period, nearly all CMC had at least 1 ED visit. Percent of CMC with at least 1 hospital encounter was similar between FC (43.3%) and disability (43.8%) groups, though less in low-income group (35.6%). Among CMC with at least 1 admission during study period, the average number of hospitalizations per year varied by group: 0.86 for FC, 0.98 for disabled, and 0.61 for low-income. CMC in all groups had higher ED and inpatient use than non-CMC. Top primary admission ICD-9 codes for CMC varied by category, with mental health diagnoses accounting for 3/10 top codes for CMC in FC (Table). Data analysis ongoing, including addition of outpatient claims, describing use of durable medical equipment and top services, and conducting statistical tests of significance.

Conclusion: There are nearly 3 times more CMC in FC group than low-income group never in FC. CMC in FC are higher utilizers of costly inpatient care than CMC in low-income group, but use less on average than CMC in the disabled group. There is need to explore comorbid medical complexity with mental health disorders, as mental health diagnoses were among the top primary admission diagnoses, particularly for CMC in FC.