Table 1.

Number of random error units in hypothetical studies with dichotomous exposures and outcomes using odds ratios

N of exposed	N of outcome	Total N	Exposure	Outcome		OR (95% CI)	REU
				Yes	No
500,000	500,000	1 million	Yes	250,000	250,000	1 (0.99–1.01)	1
			No	250,000	250,000
50	50	100	Yes	25	25	1 (0.46–2.19)	10 000
			No	25	25
500,000	1,000	1 million	Yes	500	499,500	1 (0.88–1.13)	250
			No	500	499,500
1,000	500,000	1 million	Yes	500	500	1 (0.88–1.13)	250
			No	499,500	499,500
1,000	1,000	1 million	Yes	1	999	1 (0.14–7.11)	62 625
			No	999	998,001
1,000	1,000	10,000	Yes	100	900	1 (0.80–1.24)	772
			No	900	8,100
1,000	1,000	10,000	Yes	8,169	831	2 (1.67–2.39)	528
			No	831	169
9,000	1,000	10,000	Yes	831	169	0.5 (0.42–0.60)	528
			No	8,169	831
1,000	1,000	10,000	Yes	55	945	0.5 (0.38–0.66)	1,276
			No	945	8,055

REU number of random error units