Supplemental Table 4.

Results of covariate logistic regression model for prediction of Discharge Destination for excluded and included subjects.

	Excluded (n=641)			Included (n=1242)
Variable	B	SE	p	B	SE	p
Age	0.841	0.362	0.020	0.542	0.190	0.004
Sex	0.189	0.332	0.569	0.084	0.162	0.605
Race			0.595			0.078
1	-0.511	0.708	0.470	-0.420	0.302	0.164
2	-0.098	0.707	0.889	0.013	0.303	0.996
Ethnicity	-0.664	0.945	0.482	-1.683	0.473	0.000
Hospital			0.000			0.000
1	-2.008	0.534	0.000	-1.345	0.299	0.000
2	-1.687	0.516	0.001	-0.444	0.284	0.119
3	-1.499	0.527	0.004	-0.444	0.298	0.136
4	-2.876	1.110	0.010	-0.457	0.323	0.157
Length of Stay	0.072	0.028	0.009	0.075	0.010	0.000
Cohort	0.467	0.346	0.177	0.098	0.165	0.553
COVID Dx	-0.171	0.334	0.609	-0.353	0.160	0.028
Ventilator	3.309	1.335	0.013	0.703	0.307	0.022
ICU	-2.528	0.865	0.003	0.139	0.210	0.509
Insurance			0.016			0.000
1	17.522	8072	0.998	2.373	1.868	0.204
2	19.263	8072	0.998	3.470	1.865	0.063
3	15.128	8072	0.998	2.267	1.863	0.224
4	18.796	8072	0.998	2.391	1.926	0.214
BMI	-0.005	0.020	0.801	0.005	0.009	0.564
CCI	0.149	0.057	0.009	0.067	0.026	0.010
Constant	-20.25	8072	0.998	-4.589	1.971	0.17

For EXCLUDED, Model p <0.001; Nagelkerke R² = 0.565; Sp = 96.4; Sn = 54.2. For INCLUDED, Model p<0.001; Nagelkerke R² = 0.419; Sp = 92.1; Sn = 50.0. CODING: Discharge destination (0=home, 1=facility), Age (1=>75); Sex (1=female); Race (0=other, 1=Black, 2=White); Ethnicity (0=non-hispanic, 1=Hispanic); Length of stay (continuous); Cohort (1=second); COVID Dx (1=COVID as admitting or primary diagnosis); Vent (1=yes); ICU (1=yes); Insurance (0=none, 1=Medicaid, 2=Medicare, 3=Private, 4=Other), BMI (continuous), CCI (continuous).