TY - JOUR
AU - Walker, Alexander M.
AB - LETTERS TO THE EDITOR 921 Yet to measure the intake of vitamin A requires an the less likely the surrogate informant will provide intimate daily knowledge of the decedent's eating accurate information about the decedent, and the more habits. Next-of-kin must then not be those related by likely he will reflect information about his diet which blood who are available to fill out a death certificate, is now ascribed to the decedent. Again, some infor- but someone who shared meals with the study subject. mation about the time lag between death and interview In fact, in some cases this may not even be a next-of- for cases and controls would be helpful. kin but a close friend. Both of these points need to be considered in carry- If there were substantial differences in the study by ing out studies of dietary practices using, so called, Stehr et al. (1) between cases and controls in the "next-of-kin" as informants. As a general principle, number of "next-of-kin" who lived with and shared any study that seeks information of a highly detailed, meals, then the finding of a vitamin A deficiency in personal or complex nature that must be obtained their stomach 
TI - MISCLASSIFIED CONFOUNDERS
JO - American Journal of Epidemiology
DO - 10.1093/oxfordjournals.aje.a114180
DA - 1985-11-01
UR - https://www.deepdyve.com/lp/oxford-university-press/misclassified-confounders-QFdc0C4eUc
SP - 921
EP - 922
VL - 122
IS - 5
DP - DeepDyve
ER -