Sampling strategies in nested case-control studies.

TL;DR: A stratified version of nested case-control sampling which is called "countermatching" is presented, and asymptotic relative efficiency calculations indicate that a substantial efficiency gain relative to simple random sampling of controls can be expected in these situations.

Abstract: A stratified version of nested case-control sampling which we call "countermatching" is presented This design uses data available for all cohort members to obtain a sample for collecting additional information in a case-control substudy Hitherto the only stratified sampling design for such studies has involved matching of controls to cases with respect to confounding variables However, in some situations, rather than sampling to make controls as similar as possible to cases, we might wish to make them as different as possible This is achieved by the counter-matched design Statistical analysis of counter-matched studies is straightforward using existing computer software We investigate the use of the design when a surrogate measure of exposure is available for the full cohort, but accurate exposure data is to be collected only in a nested case-control study, and when exposure data are available for the whole cohort but data concerning important confounders are not Asymptotic relative efficiency calculations indicate that a substantial efficiency gain relative to simple random sampling of controls can be expected in these situations We also illustrate how the design might be implemented in practice