Table 2.

Predicted effect sizes from the sample size calculation.

	Control	Treatment	Risk difference
Non‐psychiatric group
Total sample	1000	1000	–
Event	35	61.25	–
No event	965	938.75	–
Risk	0.035	0.061	0.026
Psychiatric group
Total sample	1000	1000	–
Event	70	122.5	–
No event	930	877.5	–
Risk	0.070	0.123	0.053
Overall
Total sample	2000	2000	–
Event	105	183.75	–
No event	1895	1816.25	–
Risk	0.053	0.092	0.039