Main factors predicting somatic, psychological, and cognitive patient outcomes after significant injury: a pilot study of a simple prognostic tool

Abstract

Abstract Background There are still insufficient data on the complexity and predictability of patient-related outcomes following trauma. The aim of this study was to assess longer-term outcomes in patients with significant injury and to develop a simple scoring method to identify patients at high risk of subsequent deficits 1–2 years after injury. Methods We conducted a prospective cohort study of survivors of significant injury (New Injury Severity Score, NISS greater than or equal to 8), with analysis of patients’ 1- to 2-year health-related quality of life (HRQoL) and their functional outcomes based on Short Form-36 (SF-36), Trauma Outcome Profile (TOP), and Quality Of Life after Brain Injury (QOLIBRI). Documented variables suspected or known from the literature to be possible factors associated with outcome were first analysed by univariate analysis, and significant variables were entered into a stepwise logistic regression analysis. Scores predicting longer-term impaired outcome were constructed from risk factors resulting from multivariate analysis. Results Depending on the patient-reported outcome measure (PROM) used, up to 30 per cent of 1052 study patients (mean NISS 18.6) indicated somatic, 27 per cent psychological, and 54 per cent cognitive deficits. The investigated sociodemographic, injury-related, treatment, and early hospital outcome variables demonstrated only low associations with longer-term outcome in univariate analysis that were highest for preinjury pain or function (R = 0.4) and outcome at hospital discharge (R = 0.3). After logistic regression, the study variables explained a maximum variance of 23 per cent for somatic, 11 per cent for psychological, and 14 per cent for cognitive longer-term outcomes. The resulting Aarau trauma prognostic longer-term outcome scoring (ATPLOS) system, developed by checking eight risk factors, had a specificity of up to 80 per cent, and importantly may facilitate early detection of patients at risk of a poorer longer-term outcome. Conclusion Despite the high rate of deficits recorded for survivors of significant injury, particularly in loss of cognitive function, the multiple variables analysed only led to a limited characterization of patient-related longer-term outcomes. Until more is known about additional individual influencing factors, the proposed scoring system may serve well for clinical evaluation. Registration number NCT 02165137 (http://www.clinicaltrials.gov)