Using Machine Learning to Predict Mortality for COVID-19 Patients on Day Zero in the ICU

Abstract

Rationale Given the expanding number of COVID-19 cases and the potential for upcoming waves of infection, there is an urgent need for early prediction of the severity of the disease in intensive care unit (ICU) patients to optimize treatment strategies. Objectives Early prediction of mortality using machine learning based on typical laboratory results and clinical data registered on the day of ICU admission. Methods We studied retrospectively 263 COVID-19 ICU patients. To find parameters with the highest predictive values, Kolmogorov-Smirnov and Pearson chi-squared tests were used. Logistic regression and random forest (RF) algorithms were utilized to build classification models. The impact of each marker on the RF model predictions was studied by implementing the local interpretable model-agnostic explanation technique (LIME-SP). Results Among 66 documented parameters, 15 factors with the highest predictive values were identified as follows: gender, age, blood urea nitrogen (BUN), creatinine, international normalized ratio (INR), albumin, mean corpuscular volume, white blood cell count, segmented neutrophil count, lymphocyte count, red cell distribution width (RDW), and mean cell hemoglobin along with a history of neurological, cardiovascular, and respiratory disorders. Our RF model can predict patients outcomes with a sensitivity of 70% and a specificity of 75%. Conclusions The most decisive variables in our model were increased levels of BUN, lowered albumin levels, increased creatinine, INR, and RDW along with gender and age. Complete blood count parameters were also crucial for some patients. Considering the importance of early triage decisions, this model can be a useful tool in COVID-19 ICU decision-making.

Platelet size as a predictor for severity and mortality in COVID-19 patients: a systematic review and meta-analysis

“…The pooled mean difference of the MPV values was significantly higher in the severe group (SMD = 0.34 [95% CI: 0.14, 0.53], p = 0.0006, n = 1,337); however, there was substantial heterogeneity by the random effect model ( I 2 = 55%) (Supplementary material 5: Figure 1). Sensitivity analysis based on blood tests taken at hospital admission for four studies [19, 24, 25, 38] showed similar results (SMD = 0.40 [95% CI: 0.19, 0.61], p = 0.0002, n = 849) and heterogeneity ( I 2 = 50%). Similar results and heterogeneity were also shown for sensitivity analysis based on the clinical outcome by excluding Barrett et al [25] (SMD = 0.31 [0.10, 0.52], p = 0.004, n = 1237, I 2 = 59%).…”

“…Wu et al [20] was an international multi-centre study, but we extracted data for patients recruited from one centre. Five studies were identified from preprint databases [20-24]. The disease outcomes for 16 studies were severity of COVID-19, and seven studies assessed mortality.…”

“…Values for platelet indices reported by each study are listed in Table 3. Although there was a trend toward lower PLT in the severe and non-survivor groups for most studies, it was only significantly lower in three studies of severe patients [20, 29, 30] and two studies of non-survivors [19, 24]. Only one study reported a mean PLT for the severe COVID-19 patients that was within the mild thrombocytopenia range (<150 ×10 9 ) [20].…”

“…Eight studies reported using MPV in COVID-19 patients who survived and died [19, 24, 25, 37, 38, 40-42]. All studies measured survival and death, except Barrett et al [25] who measured hospitalised patients who had a thrombotic event and death.…”

Symptom Prediction and Mortality Risk Calculation for COVID-19 Using Machine Learning

“…In general, complications are common among elderly patients and those with pre-existing conditions. The intensive care unit (ICU) admission rate is substantially higher for these groups ( Abate et al, 2020 ; Jamshidi et al, 2021a ).…”