Circadian vital sign characteristics in postoperative patients with and without subsequent complications

Abstract

Abstract Background Although vital signs have a well-described 24-hour cyclic variation (circadian rhythm), this variation is often not considered in clinical vital sign monitoring despite being influenced by physiologic stress from complications. We aimed to analyse if circadian vital signs variations could identify patients at risk for Serious Adverse Events (SAE). Methods This was an analysis of prospectively collected, continuously monitored vital sign data (heart rate (HR), respiration rate, oxygen saturation and blood pressure) for up to 5 days after major surgery. The primary outcome was any SAE that occurred up to 24 hours after a circadian value calculation. The primary analysis was the lowest (nadir) HR at night (NN-HR). We also analysed manually collected vital signs according to National Early Warning Score (NEWS) thresholds. Results Of 590 patients, 461 (78%) were eligible for analysis. Ninety-nine (21.5%) patients developed SAE. The NN-HR was significantly higher in patients with impending SAE (median 70.8 min− 1 [IQR 65.9–78.1] vs 65.4 [IQR 58-73.5], p < 0.01). Discrimination based on NN-HR of > 65.8 min− 1 had sensitivity / specificity of 75.7% / 51.0%, with a resulting risk ratio of 3.04 (99% confidence interval (CI) 1.51–6.13) for subsequent SAE. Early warning scores of ≥ 6 at night yielded a RR of 4.02 (99% CI 1.87–8.61). Conclusions Circadian analyses of vital signs can help to identify patients at risk for SAE after major surgery. The clinical utility of circadian vital sign monitoring in SAE prevention should be explored prospectively.