Abstract
OBJECTIVE: This study aimed to determine whether the circadian rhythm in heart rate independently associated with in-hospital mortality among patients with TBI. METHODS: Data were extracted from the eICU-CRD. Heart rate circadian rhythm was characterized using three parameters: mesor, amplitude, and peak time. The prognostic association and incremental discriminatory value of these circadian variables, alongside the APACHE IV score, was assessed in relation to in-hospital mortality. RESULTS: Among 3202 patients showing a circadian rhythm in heart rate, each 10-beat/min increase in mesor was associated with a 1.18-fold higher odd of in-hospital mortality (95% CI: 1.08-1.30; P < 0.001), while each 5-beat/min increase in amplitude corresponded to a 1.14-fold increase (95% CI: 1.03-1.25; P < 0.001). The link between mesor and in-hospital mortality varied with comorbidities such as heart failure (P = 0.039), atrial fibrillation (P = 0.014), infection (P = 0.003), and fever (P = 0.027). Adding circadian rhythm metrics to the APACHE IV score improved model discrimination (c-index 0.795; 95% CI: 0.747-0.844) compared to APACHE IV alone (c-index 0.719; 95% CI: 0.665-0.773; P < 0.001). CONCLUSION: Circadian rhythm features of heart rate are independent prognostic factors associated with in-hospital mortality in TBI patients. Integrating these variables with conventional scoring systems may improve prognostic accuracy.