Table 3.

Logistic regression results for predicting discharge

Predictor	Sample	Odds Ratio	95% CI	P Value
Initial AM-PAC Mobility	All (n=1246)	0.867	0.841-0.893	<.001
	COVID-19 as primary/admitting
	Yes (n=604)	0.858	0.817-0.901	<.001
	No (n=641)	0.869	0.837-0.903	<.001
	ICU
	Yes (n=419)	0.861	0.818-0.906	<.001
	No (n=823)	0.867	0.835-0.901	<.001
	Cohort
	First (n=696)	0.883	0.848-0.919	<.001
	Second (n=550)	0.844	0.806-0.884	<.001
Initial AM-PAC activity	All (n=1029)	0.892	0.862-0.923	<.001
	COVID-19 as primary/admitting
	Yes (n=494)	0.868	0.817-0.922	<.001
	No (n=516)	0.896	0.857-0.937	<.001
	ICU
	Yes (n=365)	0.889	0.840-0.940	<.001
	No (n=643)	0.891	0.852-0.935	<.001
	Cohort
	First (n=560)	0.904	0.864-0.946	<.001
	Second (n=469)	0.855	0.810-0.901	<.001
Final AM-PAC mobility	All (n=1242)	0.837	0.812-0.86	<.001
	COVID-19 as primary/admitting
	Yes (n=600)	0.852	0.745-0.836	<.001
	No (n=641)	0.806	0.770-0.843	<.001
	ICU
	Yes (n=392)	0.789	0.745-0.836	<.001
	No (n=823)	0.865	0.832-0.90	<.001
	Cohort
	First (n=685)	0.842	0.808-0.876	<.001
	Second (n=512)	0.818	0.778-0.860	<.001
Final AM-PAC activity	All (n=999)	0.862	0.832-0.893	<.001
	COVID-19 as primary/admitting
	Yes (n=483)	0.873	0.824-0.925	<.001
	No (n=516)	0.848	0.806-0.892	<.001
	ICU
	Yes (n=356)	0.813	0.759-0.870	<.001
	No (n=643)	0.884	0.844-0.927	<.001
	Cohort
	First (n=557)	0.852	0.813-0.893	<.001
	Second (n=442)	0.829	0.782-0.879	<.001

NOTE. Results of the main (sample=all) and subanalyses on different subsets of the cohort (eg, only those with COVID-19 as a primary or admitting diagnosis, etc). Model P<.001 for all analyses. Nagelkerke R² ranged from 0.482-0.552.