Does the ‘halo effect’ of trauma center verification extend to severe postpartum hemorrhage? A four-year retrospective review of level 1 trauma centers in the United States

Abstract

Introduction

Designated high-quality trauma services have been shown to improve outcomes of trauma patients by virtue of access to specialized personnel and resources. It remains unclear if a ‘halo effect’ extends these benefits more generally to non-trauma populations. Obstetric patients who develop severe postpartum hemorrhage often require close attention in intensive care units and utilize more resources. Given the overlapping needs between trauma and obstetric patients, we hypothesize that the ‘halo effect’ might extend to patients with severe postpartum hemorrhage.

Methods

The Nationwide Inpatient Sample for years 2008 to 2011 was queried. Patients with severe postpartum hemorrhage were identified as those requiring transfusion, hysterectomy, or uterine repair. After stratifying by level 1 trauma center versus non-level 1 trauma center status, unadjusted univariate comparisons were made. Adjusted odds ratio of end-organ failure and death were analyzed using multivariable logistic regression.

Results

A total of 11,135 patients were identified with severe postpartum hemorrhage. The majority were hospitalized at non-level 1 trauma centers rather than level 1 trauma centers (71.4% vs. 28.6%). Patients at non-level 1 trauma centers were younger, more likely to be white, admitted electively, insured, and healthier with a lower comorbidity index. There was no significant difference in rates of mortality or organ failure. However, after adjustment for differences in comorbidity index, race, and emergency admission, patients at non-level 1 trauma centers had a significantly higher risk of respiratory failure (odds ratio, 1.27; 95% confidence interval, 1.01–1.59).

Conclusions

These findings suggest that the outcomes of obstetric patients with severe postpartum hemorrhage admitted in level 1 trauma centers are not overall significantly different when compared to those in non-level 1 trauma centers. However, after adjusting for baseline characteristics, there was a reduced risk of respiratory failure in patients admitted to level 1 trauma centers.