Background: COVID-19 pandemic has led to a catastrophic shortage of ICU beds. This has resulted in the need to identify patients that can be discharged early before full clinical recovery. We designed this study to determine if in changes routine tests like CBCD and CRP can be a useful complement to clinical status when deciding to discharge patients from ICU.
Methods: This retrospective study was conducted in Rafic Hariri University Hospital. Levels of biomarkers measured at admission (T1) and within 3 days of outcome (T2) were collected and ratios (T2/T1) were calculated. The Odds Ratios of association between the changes in these biomarkers and outcome were estimated. Multivariate analysis and AUC for the performance of these biomarkers were also conducted.
Results: We found on multivariate analysis that reduction in counts of lymphocyte and platelets and elevation in counts of neutrophils and level of CRP (T2/T1 ratio > 1) are strongly associated with mortality with respective ORs estimated at 6.74, 3.26, 5.65 and 4.34 [p-values < 0.001]. AUCs were found to lie in a range of 0.68 to 0.81 indicating fair to good performance. Other factors found to impact survival were AKI, AF and ACS [p-values < 0.01]. In contrast to other studies, risk factors didn't show an association with survival when adjusted for effects of complications and changes in biomarker levels.
Conclusions: Our results confirm that inexpensive tests like lymphocyte count and CRP can be reliably used to follow COVID-19 patients in ICU and to support the decision to discharge patients.