The value of traditional vital signs, shock index, and age-based markers in predicting trauma mortality.

摘要

Systolic blood pressure (SBP), heart rate (HR), and respiratory rate are poor predictors of trauma outcome. We postulate that HR/SBP (shock index [SI]) and novel new markers SI × age (SIA), SBP / age (BPAI), maximum HR (220 - age) - HR (minpulse [MP]), and HR / maximum HR (pulse max index [PMI]) are better predictors of 48-hour mortality compared with traditional vital signs.Data were extracted from the Trauma Audit and Research Network database. Exclusions included any head or spine injury and prehospital intubation or cardiac arrest. Area under receiver operator characteristic curve (AUROC) was determined for 48-hour mortality for all variables and age. A threshold for each marker was derived using the specificity (rule-in) cutoffs at both 90% and 95% from the receiver operator characteristic curve. Positive likelihood ratios were described for each marker's derived threshold.Vital signs, markers, and age were all significantly associated with 48-hour mortality (p < 0.001). HR, SBP, and respiratory rate fared worst overall (AUROC = 0.69, 0.66, and 0.66, respectively). SIA, MP, PMI, BPAI, and SI were significantly (p < 0.05) better than age at predicting 48-hour mortality (AUROC = 0.79, 0.77, 0.77, 0.74, 0.73, and 0.68, respectively; AUROC for age = 0.68). Thresholds derived for these five markers were values 55 or greater, 44 or less, 70% or greater, 1.5 or less, and 0.9 or greater, respectively, each with a specificity of 95% for 48-hour mortality (positive likelihood ratios were 8.4, 6.1, 6.7, 6.6, and 7.5, respectively). The likelihood of death in 48 hours was 8.4 times more likely if SIA was greater than 55 than if it was lower.Older age seems to be significantly associated with early mortality. Newer markers, especially those combining traditional vital signs with age (SIA, BPAI, MP, and PMI), may contribute to better trauma triage of patients with blunt injuries than traditional vital signs.Prognostic/epidemiologic study, level III.