Association of inflammatory biomarker abnormalities with mortality in COVID-19: a meta-analysis

Abstract

Abstract Background COVID-19 outbreak has engulfed different parts of the world, affecting more than 163 million people and causing more than 3 million deaths worldwide due to human transmission. Thus, it has become critical to identify the risk factors and laboratory parameters to identify patients who have high chances of worsening clinical symptoms or poor clinical outcomes. Therefore, the study aims to identify inflammatory markers that can help identify patients at increased risk for progression to critical illness, thus decreasing the risk of any mortality. Our study focussed on the predictive utility of C-reactive protein, Interleukin-6, D-dimer and Procalcitonin in assisting the management of COVID-19 patients with adverse clinical effects. Through literature search in electronic databases, we included the retrospective studies that evaluated the biomarkers among confirmed COVID-19 patients before initiation of treatment and who had a definite outcome (dead or discharged). Biomarkers were expressed in standardized difference in mean value, calculated based on study sizes and mean values between survivors and non-survivors considered the effect size. We carried out a meta-regression analysis to identify the causes of the heterogeneity between the studies. Results Number of studies eligible for C-reactive protein, D-dimer and Interleukin-6 markers were eight, seven and four, respectively. Using random effect model revealed that the overall effect size with 95% confidence interval (CI) for C-reactive protein, D-dimer and Interleukin-6 were 1.45 (0.79–2.12) milligrams/litre, 1.12 (0.64–1.59) micrograms/millilitre Fibrinogen Equivalent Units and 1.34 (0.43–2.24) picograms/millilitre respectively was statistically significant (P < 0.05) inferring that the mean scores of these marker were significantly higher among the non-survivors compared to the survivors. Two studies were eligible for Procalcitonin marker and there was no heterogeniety (I2-statistics = 0) between these studies. Therefore, fixed-effect model revealed that the overall effect size (95% CI) for Procalcitonin was 0.75 (0.30–1.21) Nanograms/millilitre was also high among non-survivors. Conclusions The study found that serum levels of C-reactive protein, Interleukin-6 and D-dimer showed significant elevation in non-survivors compared to survivors. Raised inflammatory markers aid in the risk stratification of COVID-19 patients and their proper management.