Impact of Regional Variation and Practitioner Bias on Infant Mortality Rate: A Survey Study [12L

Abstract

INTRODUCTION: Periviable birth is a primary driver of IMR in the US and varies significantly among regions. We measured practice variation in how periviable pregnancy outcomes are categorized and how this impacts reported regional IMRs. METHODS: We distributed a 28-item questionnaire to practitioners associated with US birth hospitals. Five clinical scenarios were included to identify how participants classify birth outcomes from 15-22 weeks. Practice characteristics potentially affecting reporting of birth outcomes were assessed. IRB approval was obtained. RESULTS: Of 1024 questionnaires distributed, 237 were completed (23% response rate). Responses were compared among US census regions; 8.4% from the Northeast, 56.1% Midwest, 13.9% South, 5.4% West; 16% did not disclose. Despite >90% participants classifying a 14-week loss without signs of life as a miscarriage, 19% would provide a certificate of fetal death. 51% who would not initially provide a certificate would give one upon request, yielding a potential fetal death reporting rate of 32% for a 14-week pregnancy loss. Likewise, 32% would provide a certificate of live birth for a 16-week loss with signs of life, with notable regional variation in classifying this as a live birth (41% Northeast, 44% Midwest, versus 13% South, 18% West, p=0.003). CONCLUSION: Practice variation in recording birth outcomes at periviable gestational ages was noted in our study. This may contribute to significant misrepresentation of the frequency of miscarriage, stillbirth and live births and could influence regional variation in reported birth outcomes.