Table 1. Power and Sample Size for Hypothesis Tests of the 30-Day Inpatient Mortality Difference Between Usual Care and Intervention^a.

Test	Sample size assumptions and results
	Usual care P2	Intervention P1	Difference (P2 − P1)	No. of participants		Statistical power, %
				Per group	Total
1	0.06	0.03	0.03	748	1496	80
2	0.06	0.04	0.02	1863	3726	80
3	0.06	0.03	0.03	1001	2002	90
4	0.06	0.04	0.02	2493	4986	90

^a The usual care group mortality rate (P2) is assumed to be 0.06. To detect a proportion difference (P2 − P1) of 0.03 (or P1 of 0.03) with 90% power, the number of participants needed will be 1001 in the intervention arm and 1001 in the usual care arm.