Hypothesis Testing in Case-Control Studies
本文扩展了Prentice和Pyke的经典结果，证明在病例对照数据中拟合前瞻性逻辑回归模型得到的似然比检验，在仅涉及比值比参数时渐近精确，涉及分层效应时渐近保守。
Prentice & Pyke (1979) have shown that valid estimators of the odds-ratio parameters in a logistic regression model may be obtained from case-control data by fitting the model as if the data had been obtained in a prospective study. The resulting estimates of standard errors are also valid asymptotically. We extend the Prentice-Pyke result to likelihood ratio tests obtained from fitting a prospective model. These tests are asymptotically exact if the hypothesis involves only the odds-ratio parameters and are asymptotically conservative if stratum effects are involved.