Sample size for a case-control study |
Example: We wish to conduct a
case-control study to assess whether bladder cancer may be
associated with past exposure to cigarette smoking. Cases will
be patients with bladder cancer and controls will be patients
hospitalized for injury. It is assumed that 20% of controls will
be smokers or past smokers, and we wish to detect an odds-ratio
of 2 with power 90%. Three controls will be recruited for one
case.
We enter the values 2 in the field "Minimum Odds-ratio to detect", 20 in the field "Percentage exposed among controls", 3 in the field "Number of controls per case" and leave "Alpha risk" and "Power" at their default values. The check box "One-sided test" is left unchecked. Sampsize returns an estimate sample size of 600: 150 cases and 450 controls. An alternative is to conduct a matched case-control study rather than the above unmatched design. One case will be matched to one control. With all other parameters equal to above specified, sampsize returns a sample size of 226 case-control pairs (total sample size 452). |
Power of a case-control study |
Example: Kessler and Clark (1978) studied 365 males with
bladder cancer and an equal number of controls, 35% of whom reported
past use of nonnutritive sweeteners. Using a one-sided test at the
alpha = 5% level, the smallest risk greater than one that can
be detected with 90% power is RR = 1.55: Click the check box to compute minimum OR, enter 35 in the field "percentage exposed among controls", 365 in the field "number of cases", and select a one-sided test. Note: the minimum/maximum detectable OR is only computed for an equal number of cases and controls. |
glaziou@gmail.com