A comparison of gestational age reporting methods based on physician estimate and date of last normal menses from fetal death reports.

Utilizing 10,587 cases from the 1980 National Center for Health Statistics Fetal Death Statistics File, we examined the comparability of two methods of determining the gestational age of a fetal death, the calculated interval from date of last normal menses (DLNM) and the physician's estimate. The physician estimated gestational age distribution exhibits even number digit preference and a distinct clustering at the 40-week value. The DLNM distribution appears more smoothly distributed but with a more pronounced post-term tail. An exact agreement between the two methods is observed in only 27.9 per cent of the cases. A 1.7 week mean difference between the methods indicates a systematic underestimation by physician reported gestational age when compared to that calculated from the DLNM, potentially biasing gestational age distributions when the physician estimate is substituted for cases with a missing DLNM. Over 8 per cent of cases 20+ weeks by DLNM are estimated as less than 20 weeks by the physician. This underestimation has important implications for the completeness of reporting of fetal deaths on vital records and the comparability of fetal death rates. Further, it may limit investigations of the completeness of reporting of less than 500 gram live births.