Predictors and adverse pregnancy outcomes associated with antepartum discharge against medical advice

Brownsyne Tucker Edmonds, Corinne Ahlberg, Katie McPherson, Sindhu Srinivas, Scott Lorch

Research output: Contribution to journalArticle

3 Scopus citations


To determine predictors and pregnancy outcomes associated with antepartum discharge against medical advice (AMA D/C). Retrospective cohort study of state-level maternal and infant hospital discharge data linked to vital statistics data for antepartum admissions in California from 1995 to 2005. (N = 203,250). After adjusting for comorbid conditions, the odds of AMA D/C for Black women were twice that of white women (OR = 2.00, 95 % CI 1.70-2.35). Publicly insured women had 3.5 times the odds of AMA D/C compared to privately insured women (OR = 3.54, 95 % CI 3.02-4.15). AMA D/C was also higher among substance abusers and women with mental illness (OR = 1.96, 95 % CI 1.43-2.67 and OR = 4.45, 95 % CI 3.81-5.21 respectively). Most notably, AMA D/C tripled the odds of fetal death in patients admitted for pregnancy-induced hypertension (OR = 3.08, 95 % CI 1.36-6.98) and increased the odds of neonatal morbidity (respiratory distress syndrome OR = 1.35, 95 % CI 1.07-1.70 and small-for-gestational-age OR = 1.47, 95 % CI 1.15-1.89) in patients admitted with preterm premature rupture of membranes. Vulnerable populations and patients with comorbid medical and mental illnesses are at increased risk for AMA D/C and its associated adverse pregnancy outcomes. Targeted interventions and resources to support at-risk populations are needed.

Original languageEnglish (US)
Pages (from-to)640-647
Number of pages8
JournalMaternal and Child Health Journal
Issue number3
StatePublished - Apr 2014



  • Antepartum admission
  • Discharge against medical advice
  • Pregnancy outcomes

ASJC Scopus subject areas

  • Public Health, Environmental and Occupational Health
  • Epidemiology
  • Obstetrics and Gynecology
  • Pediatrics, Perinatology, and Child Health
  • Medicine(all)

Cite this