In a case-control study of low birthweight deliveries, comparison of hospital records with interviews conducted soon after delivery fails to confirm the widely-held view that concurrently recorded medical notes are superior to retrospective interviewing. The overall agreement was very good, with slight under-reporting of past abortions by mothers, and marked relative under-recording in the hospital records of vaginal bleeding and especially vaginal discharge. Generally, the rates of disagreement between interviews and hospital notes were similar for those women who had low birthweight babies and those who had normal birthweight babies, except for one single item. Uncertainty or lack of information occurred in the hospital notes as often as in the interviews.

When different accounts of labour were present within a single set of medical records, the level of internal consistency was examined. There were more definite contradictions within the hospital notes than had been observed between the notes and the interviews. Moreover, a very marked association of within-hospital contradiction with birthweight was found, suggesting a possibility of bias when hospital records are used for the investigation of conditions associated with low birthweight. One should not, however, conclude that hospital notes are necessarily more error-prone than interviews, as many of the within-hospital contradictions reflected diagnostic uncertainty in complicated cases.