Incidence of Severe Maternal Morbidity and Other Obstetric Complications in Illinois 2016-2018 Using International Classification of Diseases-10

Description

Severe maternal morbidity underestimates the incidence of other, route-specific International Classification of Diseases Version 10 coded delivery complications.

Abstract

OBJECTIVE: Although maternal morbidity has been increasing in recent years, quality measures for maternal outcomes of childbirth hospitalizations remain controversial. Previous studies have relied upon International Classification of Diseases (ICD-9) coding with virtually no literature on how the 2015 transition to ICD-10 has affected delivery complication rates. This study describes the incidence of maternal obstetric complications in Illinois using ICD10 coding algorithms. METHODS: In a retrospective cohort study of all deliveries at 127 Illinois hospitals from January 2016 to December 2018, ICD-10 codes were used to characterize severe maternal morbidity (SMM) and other delivery route-specific complications of vaginal and cesarean delivery. Poisson regression models were used to estimate the association of maternal sociodemographic and clinical characteristics as well as hospital characteristics with the likelihood of complications.RESULTS: Among 421,426 deliveries, the SMM rate was 0.76% overall, 0.44% for vaginal and 1.50% for cesarean delivery. Other maternal complications occurred in 8.62% of vaginal and 9.96% of cesarean deliveries. Patient age, race and ethnicity, and pre-existing and pregnancy-related clinical conditions were significantly associated with each outcome. Hospital-annual delivery volume, weekend admission and medical record coding intensity, measured by the mean number of ICD-10 codes for uncomplicated deliveries, were also significantly associated with complication incidence.CONCLUSION: SMM significantly underestimates the incidence of other potentially preventable complications. Complication incidence is higher for women from racial and ethnic minority groups and high poverty neighborhoods. However, the association of the likelihood of complications with hospital coding intensity and post-hoc analyses of length of stay differences between complicated and uncomplicated deliveries raise questions about the reliability of ICD-10 coding, which is incommensurate with previous ICD-9 research. New risk adjusted outcome measures based on detailed clinical data and linked antepartum and postpartum care will be necessary to improve obstetric quality of care.

Files

Additional details