Days at Home After Traumatic Brain Injury: Moving Beyond Mortality to Evaluate Patient-Centered Outcomes Using Population Health Data.

Abstract

Background and objectives: Administrative data are invaluable for assessing outcomes at the population level. However, there are few validated patient-centered outcome measures that capture morbidity following traumatic brain injury (TBI) using these data. We sought to characterize and validate days at home (DAH) as a measure to quantify population-level outcomes after moderate to severe TBI. We additionally assessed the earliest feasible outcome assessment period for patients with TBI using this outcome measure.

Methods: This multicenter retrospective cohort study used linked health administrative data sources to identify adults with moderate to severe TBI presenting to trauma centers in Ontario, Canada, between 2009 and 2021. DAH at 180 days (DAH_{180 days}) reflects the total number of days spent alive and at home excluding the days spent institutionalized across care settings. Construct validity was determined using hierarchical quantile regression to assess the associations between clinical and injury covariates with DAH_{180 days}. Predictive validity was assessed using Spearman rank correlation. We estimated minimally important difference (MID) in DAH_{180 days} to aid with outcome measure interpretability.

Results: There were 6,340 patients who met inclusion criteria. Median DAH_{180 days} was 70 days (interquartile range 0-144). Mortality occurred in 2,162 (34.1%) patients within 90 days following injury. Patients in the lower DAH_{180 days} group were more commonly older (absolute standardized difference [ASD] = 0.68) with higher preinjury health resource utilization (ASD = 0.36) and greater injury severity (ASD = 0.81). Increased baseline health resource utilization (-10.1 days, 95% CI -17.4 to -2.8, p = 0.0041), older age (-4.6 days, 95% CI -5.7 to -3.4, p < 0.001), higher cranial injury severity (-84.6 days, 95% CI -98.3 to -71.0, p < 0.001), and major extracranial injuries (-14.2 days, 95% CI -19.5 to -8.93, p < 0.001) were significantly associated with fewer DAH_{180 days}. DAH_{180 days} was positively correlated with DAH at up to 3 years (r = 0.91, 95% CI 0.90-0.92) and negatively correlated with direct health care expenditure (r_s = -0.89, 95% CI -0.88 to -0.90). The average MID estimated from anchor-based and distribution-based methods was 18 days.

Discussion: We validate DAH_{180 days} as a potentially useful outcome measure with construct, predictive, and face validity in a population with moderate to severe TBI. Given the intensity of acute care requirements for patients with TBI, our work highlights DAH_{180 days} as a feasible and sufficiently responsive outcome measure.