///2016 Abstract Details
2016 Abstract Details2019-07-15T10:10:51-05:00

Neighborhood- and Hospital-Level Factors Associated with Severe Peripartum Maternal Morbidity in New York State, 2009-2011.

Abstract Number: BP-02
Abstract Type: Original Research

Jean Guglielminotti MD, PhD1 ; Ruth Landau MD2; Cynthia A. Wong MD3; Catherine Deneux-Tharaux MD, PhD4; Guohua Li MD, Dr PH5

Background: Severe maternal morbidity (SMM) is increasing in the US (1), affecting 1.6% of deliveries with marked variations among hospitals. Explanations include increases in maternal age, comorbidities, high-risk pregnancies, and cesarean deliveries. Research suggests that neighborhood characteristics such as long driving time to hospital or low obstetrician density are associated with adverse maternal outcomes (2). Similar associations are suggested for hospital characteristics such as rural location, low delivery volume, high proportion of minority patients, or low staffing (3). We examined associations of neighborhood- and hospital-level factors with SMM to identify potential targets for maternal safety improvement.

Methods: Discharge records indicating labor and delivery, patient characteristics, and SMM were identified with ICD-9-CM codes in the State Inpatient Database for New York 2009-2011. SMM encompassed organ failures and specific diseases, and included 15 diagnosis: disseminated intravascular coagulation (DIC); acute heart, respiratory, kidney, neurologic, and hepatic failure; postpartum hemorrhage (PPH) associated with blood transfusion or hysterectomy; eclampsia; stroke; pulmonary embolism; status asthmaticus; status epilepticus; severe sepsis; myocardial infarction; diabetic coma. Hospital and neighborhood characteristics were obtained from the AHA and AHRF Files 2010. Driving time to hospital and obstetrician density in the county of residence were calculated. Multilevel modeling was used to examine the association of neighborhood- and hospital-level factors with SMM.

Results: 605,534 discharges in 139 hospitals were analyzed; 7,627 discharges indicated SMM (1.3%) with an increase from 1.1% in 2009 to 1.4% in 2011 (P < 0.0001). The 3 most frequent SMMs were severe PPH (35%), DIC (27%), and heart failure (16%). No neighborhood characteristics were associated with SMM (Table). Four hospital characteristics were associated with decreased rate of SMM: urban location, low proportions of minority patients and high-risk pregnancies, and higher cesarean delivery rate. Between-hospital variation was equally explained by patient- and hospital-level characteristics.

Conclusions: Hospital-, but not neighborhood characteristics were associated with SMM, which may result in interventions to improve maternal safety.


1 Obstet Gynecol 2012;120:1029-36

2 Am J Obstet Gynecol 2005;193:1083-8

3 Am J Obstet Gynecol 2014;211:647.e1-16

SOAP 2016