Insurance-based disparities in traumatic brain injury length of stay and discharge dispositions: A retrospective cohort analysis of the national trauma data bank.

Abstract

Insurance-based disparities associated with outcomes in traumatic brain injury (TBI), including acute care discharge dispositions, remain understudied. This study aims to evaluate the associations between Medicaid versus private insurance (PI) on hospital length of stay (LOS) and discharge disposition in patients with TBI across severities. This is a retrospective multicenter cohort study utilizing data from United States trauma centers from the National Trauma Data Bank (NTDB) years 2014-2016. Inclusion criteria were: identified to have TBI based on ICD-9 codes. Exclusion criteria were: missing gender, race, Glasgow Coma Scale (GCS) score, or LOS; not admitted as inpatient to hospital; did not have Medicaid or PI; age < 14 or > 89 years; extracranial AIS= 6. Patients were grouped into those with Medicaid and those with PI. Of 808,120 NTDB patients with TBI, 188,756 met inclusion criteria (Medicaid: N = 58,867, PI: N = 129,889). Overall, PI patients were older and more frequently female, with differences in racial distribution and TBI severity compared to Medicaid patients. On multivariable analysis, Medicaid patients demonstrated longer inpatient LOS across all TBI severities, corresponding to a 7.3% increase after mild TBI (95% CI: 6.4%-8.2%, p < 0.001), a 7.4% increase after moderate TBI (95% CI: 2.8%-12.2%, p = 0.001), and an 8.7% increase after severe TBI (95% CI: 6.2%-11.3%, p < 0.001) compared to PI patients. Across TBI severities, Medicaid patients had higher adjusted odds of routine discharge to home/self-care (mild: 15.9% increase; moderate: 28.2% increase; severe: 42.3% increase; all p < 0.001) and to skilled nursing facilities (SNF) (mild: 63.6% increase; moderate: 54.6% increase; severe: 54.9% increase; all p < 0.001). Conversely, Medicaid patients had significantly lower odds of discharge to inpatient rehabilitation (mild: 33.7% lower; moderate: 37.9% lower; severe: 42.6% lower; all p < 0.001). For discharge to home health services, Medicaid patients had lower odds after mild TBI (13.0% lower, p < 0.001), higher odds after severe TBI (17.5% higher, p = 0.017), and no significant difference after moderate TBI. Sensitivity analysis using highest in-hospital GCS for severity classification demonstrated consistent direction and magnitude of associations, supporting the robustness of these findings. Medicaid patients had longer LOS compared to PI patients, which was more pronounced within increasing TBI severity. Medicaid patients were more likely to be discharged to home/self-care and SNF, and less likely to be discharged to acute rehabilitation compared to those with PI. Our study confirms and extends prior studies on Medicaid as a risk factor for increased hospital LOS and lower likelihood for discharge to acute rehabilitation after TBI, which should be studied in near-term healthcare policy interventions.