From: Methods of sample size calculation in descriptive retrospective burden of illness studies
Expected number of individuals receiving treatment; (Probability of observing treatment at least once) ± Expected 95% confidence interval width for proportion receiving treatment | ||||||
---|---|---|---|---|---|---|
n = 50 | n = 100 | n = 200 | n = 300 | n = 500 | n = 1000 | |
p = 0.01 | 1 (0.39); ±0.03 | 1 (0.63); ±0.02 | 2 (0.87); ±0.01 | 3 (0.95); ±0.01 | 5 (0.99); ±0.01 | 10 (1.00); ±0.01 |
p = 0.05 | 3 (0.92) ±0.06 | 5 (0.99) ±0.04 | 10 (1.00) ±0.03 | 15 (1.00) ±0.02 | 25 (1.00) ±0.02 | 50 (1.00) ±0.01 |
p = 0.10 | 5 (0.99) ±0.08 | 10 (1.00) ±0.06 | 20 (1.00) ±0.04 | 30 (1.00) ±0.03 | 50 (1.00) ±0.03 | 100 (1.00) ±0.02 |
p = 0.25 | 13 (1.00) ±0.12 | 25 (1.00) ±0.08 | 50 (1.00) ±0.06 | 75 (1.00) ±0.05 | 125 (1.00) ±0.04 | 250 (1.00) ±0.03 |
p = 0.50 | 25 (1.00) ±0.14 | 50 (1.00) ±0.10 | 100 (1.00) ±0.07 | 150 (1.00) ±0.06 | 250 (1.00) ±0.04 | 500 (1.00) ±0.03 |
p = 0.75 | 28 (1.00) ±0.12 | 75 (1.00) ±0.08 | 150 (1.00) ±0.06 | 225 (1.00) ±0.05 | 375 (1.00) ±0.04 | 750 (1.00) ±0.03 |