ABSTRACT
Background: COVID-19 pandemic has created many challenges for clinicians. The monitoring trend for laboratory biomarkers is helpful to provide additional information to determine the role of those in the severity status and death outcome. Objective: This article aimed to evaluate the time-varying biomarkers by LOWESS Plot, check the proportional hazard assumption, and use to extended Cox model if it is violated. Methods: In the retrospective study, we evaluated a total of 1641 samples of confirmed patients with COVID-19 from October until March 2021 and referred them to the central hospital of Ayatollah Rohani Hospital affiliated with Babol University of medical sciences, Iran. We measured four biomarkers AST, LDH, NLR, and lymphocyte in over the hospitalization to find out the influence of those on the rate of death of COVID-19 patients. Results: The standard Cox model suggested that all biomarkers were prognostic factors of death (AST: HR=2.89, P<0.001, Lymphocyte: HR=2.60, P=0.004, LDH: HR=2.60, P=0.006, NLR: HR=1.80, P<0.001). The additional evaluation showed that the PH assumption was not met for the NLR biomarker. NLR biomarkers had a significant time-varying effect, and its effect increase over time (HR(t)=exp (0.234+0.261×log(t)), p=0.001). While the main effect of NLR did not show any significant effect on death outcome (HR=1.26, P=0.097). Conclusion: The reversal of results between the Cox PH model and the extended Cox model provides insight into the value of considering time-varying covariates in the analysis, which can lead to misleading results otherwise.
ABSTRACT
OBJECTIVE: Intensive care unit (ICU) admission occurs at different times during hospitalization among patients with COVID-19. We aimed to evaluate the time-dependent receive operating characteristic (ROC) curve and area under the ROC curve, AUC(t), and accuracy of baseline levels of inflammatory markers C-reactive protein (CRP) and neutrophil-to-lymphocyte ratio (NLR) in predicting time to an ICU admission event in patients with severe COVID-19 infection. METHODS: In this observational study, we evaluated 724 patients with confirmed severe COVID-19 referred to Ayatollah Rohani Hospital, affiliated with Babol University of Medical Sciences, Iran. RESULTS: The AUC(t) of CRP and NLR reached 0.741 (95% confidence interval [CI]: 0.661-0.820) and 0.690 (95% CI: 0.607-0.772), respectively, in the first 3 days after hospital admission. The optimal cutoff values of CRP and NLR for stratification of ICU admission outcomes in patients with severe COVID-19 were 78 mg/L and 5.13, respectively. The risk of ICU admission was significantly greater for patients with these cutoff values (CRP hazard ratio = 2.98; 95% CI: 1.58-5.62; NLR hazard ratio = 2.90; 95% CI: 1.45-5.77). CONCLUSIONS: Using time-dependent ROC curves, CRP and NLR values at hospital admission were important predictors of ICU admission. This approach is more efficient than using standard ROC curves.