Regional and practitioner variations in reporting infant mortality.

Assess regional differences in categorization of preterm delivery outcomes and impact on variation in reported infant mortality rates. A 27-item questionnaire was distributed to 1072 practitioners associated with U.S. birth hospitals. Five clinical scenarios were created to identify how participants classify delivery outcomes. Statistical analysis included Chi-square analysis and multinomial logistic regression. 234 questionnaires were completed (response rate 22%). While >90% respondents classified a 14-week pregnancy loss with no sign of life as a miscarriage, only 22% would provide a fetal death certificate. Likewise, 37% would provide a certificate of live birth for a loss at 16 weeks with signs of life. There was notable regional variation in classifying these as live births (Northeast: 41%, Midwest: 44%, South: 13%, and West: 18%, Regional practice variation in recording both live births and stillbirths was noted. Greater standardization in reporting practices may be warranted to improve the accuracy of reported birth outcomes in the U.S.