Machine-learning-based COVID-19 mortality prediction model and identification of patients at low and high risk of dying

Table 4 The conditional probabilities for each cluster are shown for each response category of 20 variables in the analysis

Variable	Category	Cluster 1	Cluster 2	Cluster 3
Age > 65	No	0.8791	0.3844	0.0429
Age > 65	Yes	0.1209	0.6156	0.9571
Nursing home	No	0.9976	0.9509	0.3191
Nursing home	Yes	0.0024	0.0491	0.6809
Smoking	No	0.9157	0.9411	0.9255
Smoking	Yes	0.0843	0.0589	0.0745
Alcohol	No	0.5509	0.7871	0.9453
Alcohol	Yes	0.4491	0.2129	0.0547
Chest pain	No	0.8947	0.7753	0.996
Chest pain	Yes	0.1053	0.2247	0.004
Dyspnea	No	0.4492	0.2709	0.5011
Dyspnea	Yes	0.5508	0.7291	0.4989
Headache	No	0.8459	0.9365	0.9793
Headache	Yes	0.1541	0.0635	0.0207
AMS	No	0.9954	0.9214	0.5785
AMS	Yes	0.0046	0.0786	0.4215
Consolidation	No	0.8887	0.7795	0.727
Consolidation	Yes	0.1113	0.2205	0.273
O2 saturation < 88	No	0.9979	0.8809	0.855
O2 saturation < 88	Yes	0.0021	0.1191	0.145
yno2	No	0.8128	0.5695	0.5008
yno2	Yes	0.1872	0.4305	0.4992
CAD	No	0.9978	0.9125	0.6985
CAD	Yes	0.0022	0.0875	0.3015
Atrial fibrillation	No	0.9911	0.9019	0.7769
Atrial fibrillation	Yes	0.0089	0.0981	0.2231
Hypertension	No	0.8206	0.173	0.1622
Hypertension	Yes	0.1794	0.827	0.8378
PVD	No	0.9977	0.9447	0.8524
PVD	Yes	0.0023	0.0553	0.1476
Stroke	No	0.9892	0.976	0.6911
Stroke	Yes	0.0108	0.024	0.3089
Dementia	No	0.9981	0.9984	0.4778
Dementia	Yes	0.0019	0.0016	0.5222
Diabetes	No	0.9772	0.5565	0.4624
Diabetes	Yes	0.0228	0.4435	0.5376

No and yes values are considered as the absence and presence, respectively, for the clinical variables

ISSN: 1364-8535