Lymphocyte-monocyte-neutrophil index: a predictor of severity of coronavirus disease 2019 patients produced by sparse principal component analysis

Abstract

Background: It is important to recognize the coronavirus disease 2019 (COVID-19) patients in severe conditions from moderate ones, thus more effective predictors should be developed.

Methods: Clinical indicators of COVID-19 patients from two independent cohorts (Training data: Hefei Cohort, 82 patients; Validation data: Nanchang Cohort, 169 patients) were retrospected. Sparse principal component analysis (SPCA) using Hefei Cohort was performed and prediction models were deduced. Prediction results were evaluated by receiver operator characteristic curve and decision curve analysis (DCA) in above two cohorts.

Results: SPCA using Hefei Cohort revealed that the first 13 principal components (PCs) account for 80.8% of the total variance of original data. The PC1 and PC12 were significantly associated with disease severity with odds ratio of 4.049 and 3.318, respectively. They were used to construct prediction model, named Model-A. In disease severity prediction, Model-A gave the best prediction efficiency with area under curve (AUC) of 0.867 and 0.835 in Hefei and Nanchang Cohort, respectively. Model-A's simplified version, named as LMN index, gave comparable prediction efficiency as classical clinical markers with AUC of 0.837 and 0.800 in training and validation cohort, respectively. According to DCA, Model-A gave slightly better performance than others and LMN index showed similar performance as albumin or neutrophil-to-lymphocyte ratio.

Conclusions: Prediction models produced by SPCA showed robust disease severity prediction efficiency for COVID-19 patients and have the potential for clinical application.

Keywords: COVID-19; Prediction; Principal component analysis; SARS-CoV-2; Severity.

Fig. 1

Results of the sparse principal components analysis using clinical data of Hefei cohort. Sparse principal analysis (SPCA) was performed based on the 44 clinical variables of Hefei cohort and the alpha parameter was adjusted from 0.0001 to 0.002 with stepsize 0.0001. For models of each alpha, the cumulative variance of the first 13 principal components (PCs) were summed and the number of variables selected in the first 13 PCs was counted. Variance of different alpha values in SPCA was plotted (a) and the number of selected clinical variables in the 13 PCs of each SPCA were added upon the point. b Distribution of the coronavirus disease 2019 (COVID-19) patients projected to principal components of SPCA with alpha of 0.0015. Depending on each patient's first (X-axis) and 12th (Y-axis) principal components value, COVID-19 patients were projected on the principal components plot of SPCA. c Scatter plot of the clinical markers selected in the first and 12th principal components of SPCA with alpha being 0.0015. Depending on each variable's first (X-axis) and 12th (Y-axis) principal components loadings, 44 clinical variables were projected on the principal components plot of SPCA. The first (X-axis) and 12^th (Y-axis) principal components accounted for the 17.8% and 2.9% of the total variance of the 44 clinical markers, respectively

Fig. 2

The association between prediction models with clinical characteristics of COVID-19 patients. Using prediction model Model-A and LMN index, in COVID-19 patients, CD8 + T lymphocytes negatively correlated with Model-A probability (a) and LMN index (b), while, patients age always positively correlated with Model-A probability (c) and LMN index (d). COVID-19 patients with comorbidity always have higher Model-A probabilities (e) and LMN index (f). Abbreviations: COVID-19 coronavirus disease 2019, Model-A prediction model based on the first and 12th principal components produced by sparse principal component analysis, LMN index lymphocyte–monocyte–neutrophil index, a simplified version of Model-A

Fig. 3

The dynamics of prediction models of COVID-19 patients from hospital admission to discharge. For all of the patients, results of three time points during hospitalization were collected: the first time point upon hospitalization (After Admission), the medium-term after hospitalization (Middle Stage), and the last time of laboratory test before hospital discharge (Before Discharge). With results of the three time points, the Model-A probability and LMN index were calculated and plotted. a Dynamics of Model-A probability for individual patients (P < 0.05); b dynamics of LMN index of individual patients (P < 0.05). Abbreviations: COVID-19, coronavirus disease 2019; Model-A, prediction model based on the first and 12th principal components produced by sparse principal component analysis; LMN index lymphocyte–monocyte–neutrophil index, a simplified version of Model-A

Fig. 4

Decision curve analysis of prediction models produced by SPCA. Decision curve analysis of prediction models in the training Hefei Cohort (a) and independent validation Nanchang Cohort (b). Model-A showed slightly better net benefit both in Hefei Cohort and Nanchang Cohort. Abbreviations: ALB Albumin, COVID-19 coronavirus disease 2019, Model-A prediction model based on the first and 12th principal components produced by sparse principal component analysis, LMN index lymphocyte–monocyte–neutrophil index, a simplified version of Model-A, NLR Neutrophil-to-lymphocyto ratio