Dell Kristine C, Grossner Emily C, Staph Jason, Schatz Philip, Hillary Frank G
Department of Psychology, The Pennsylvania State University, University Park, Pennsylvania, USA.
Social and Life and Engineering Sciences Imaging Center, The Pennsylvania State University, University Park, Pennsylvania, USA.
Neurotrauma Rep. 2021 Jun 9;2(1):255-269. doi: 10.1089/neur.2020.0065. eCollection 2021.
Health factors impacting both the occurrence of, and recovery from traumatic brain injury (TBI) vary in complexity, and present genuine challenges to researchers and healthcare professionals seeking to characterize injury consequences and determine prognosis. However, attempts to clarify causal links between injury characteristics and clinical outcomes (including mortality) often compel researchers to exclude pre-existing health conditions (PECs) in their samples, including psychiatric history, medication usage, and other comorbid conditions. In this pre-registered population-based study (total starting = 939,123 patients), we examined trends in PEC incidence over 22 years in the state of Pennsylvania (1997-2019) in individuals sustaining TBI ( = 169,452) and individuals with orthopedic injury ( = 87,637). The goal was to determine how PECs interact with age and injury severity to influence short-term outcomes. A further goal was to determine whether number of PECs, or specific PEC clusters contributed to worse outcomes within the TBI cohort, compared with orthopedic injury alone. Primary findings indicate that PECs significantly influenced mortality within the TBI cohort; patients having four or more PECs were associated with approximately a two times greater likelihood of dying in acute care (odds ratio [OR] 1.9). Additionally, cluster analyses revealed four distinct PEC clusters that are age and TBI severity dependent. Overall, the likelihood of zero PECs hovers at ∼25%, which is critical to consider in TBI outcomes work and could potentially contribute to the challenges facing intervention science with regard to reproducibility of findings.
影响创伤性脑损伤（TBI）发生及恢复的健康因素复杂多样，给试图描述损伤后果并确定预后的研究人员和医疗保健专业人员带来了真正的挑战。然而，试图阐明损伤特征与临床结果（包括死亡率）之间的因果关系时，研究人员往往不得不排除样本中预先存在的健康状况（PECs），包括精神病史、药物使用情况和其他合并症。在这项预先注册的基于人群的研究中（初始样本总数 = 939,123名患者），我们研究了宾夕法尼亚州（1997 - 2019年）22年间TBI患者（n = 169,452）和骨科损伤患者（n = 87,637）中PECs的发病趋势。目的是确定PECs如何与年龄和损伤严重程度相互作用以影响短期结果。另一个目的是确定与单纯骨科损伤相比，PECs的数量或特定的PECs集群是否会导致TBI队列中的结果更差。主要研究结果表明，PECs对TBI队列中的死亡率有显著影响；有四种或更多PECs的患者在急性护理中死亡的可能性大约高出两倍（优势比[OR] 1.9）。此外，聚类分析揭示了四个不同的PECs集群，它们与年龄和TBI严重程度相关。总体而言，无PECs的可能性徘徊在~25%，这在TBI结果研究中至关重要，并且可能导致干预科学在研究结果可重复性方面面临挑战。
