Maternal levels of care and association with severe maternal morbidity during birth hospitalizations

BACKGROUND: Maternal level of care guidelines were published to facilitate risk-appropriate care and improve outcomes, yet evidence of benefit is lacking. All birth hospitals should be equipped to provide routine childbirth care and be prepared to manage common, treatable childbirth complications. Thus, our objective was to examine the association of maternal level of care with severe maternal morbidity (SMM) without transfusion among 1) all obstetric patients and 2) obstetric patients with common, treatable childbirth complications (e.g., hemorrhage and infection). METHODS: Using a retrospective observational cohort of hospital-based births, we analyzed linked vital statistic and hospital administrative data from Michigan (2010-2020), Oregon (2010-2020), Pennsylvania (2010-2018), and South Carolina (2010-2020). Maternal level of care (I [basic], II [specialty], III [subspecialty], and IV [regional perinatal health centers]) were empirically assigned using administrative data based on the care provided to patients who gave birth in each hospital annually. The primary outcome was SMM without blood product transfusion. We used modified Poisson regression to estimate adjusted risk ratios (aRRs), accounting for covariates present prior to admission for childbirth and associated with SMM. RESULTS: Among 3,296,520 births, the SMM rate rose with increasing level (I: 0.47%, II: 0.56%, III: 0.74%, IV: 0.94%). Compared to level IV care, the SMM aRR was lower in hospitals with level I care (aRR 0.87, 95% confidence interval [CI] 0.76-0.999). Whereas, among 245,604 patients with common, treatable childbirth complications the SMM aRR was higher in hospitals with level III care (aRR 1.21, 95% CI 1.04-1.40, reference level IV). CONCLUSION: We found differential associations between maternal levels of care and SMM among all obstetric patients and those with common, treatable complications that arise during childbirth. While unmeasured differences in patient triage, referral, and transfer patterns may influence associations, these findings suggest differences in how hospitals of varying levels of care assess patient risk, manage complications, and perceive their capabilities, illustrating the ongoing need to improve risk-appropriate obstetric care and optimize patient outcomes.