I would like to report an erratum to the paper entitled ‘Cutoff Designs for Community-Based Intervention Studies’ by Pennell, Hade, Murray, and Rhoda, Statistics in Medicine. In the proof given in the Appendix, the expression for E(Y | Z,X) should be as follows:
which simplifies to
because W is independent of Z conditional on X. The validity of the proof is unaffected by the correction.