Risk Factors for Maternal Readmission with Sepsis

Megan E. Foeller, Lillian Sie, Timothy M. Foeller, Anna I. Girsen, Suzan L. Carmichael, Deirdre J. Lyell, Henry C. Lee, Ronald S. Gibbs

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


Objective Our primary objective was to identify risk factors for maternal readmission with sepsis. Our secondary objectives were to (1) assess diagnoses and infecting organisms at readmission and (2) compare early (<6 weeks) and late (6 weeks to 9 months postpartum) maternal readmission with sepsis. Study Design We identified our cohort using linked hospital discharge data and birth certificates for California deliveries from 2008 to 2011. Consistent with the 2016 sepsis classification, we defined sepsis as septicemia plus acute organ dysfunction. We compared women with early or late readmission with sepsis to women without readmission with sepsis. Results Among 1,880,264 women, 494 (0.03%) were readmitted with sepsis, 61% after 6 weeks. Risk factors for readmission with sepsis included preterm birth, hemorrhage, obesity, government-provided insurance, and primary cesarean. For both early and late sepsis readmissions, the most common diagnoses were urinary tract infection and pyelonephritis, and the most frequently identified infecting organism was gram-negative bacteria. Women with early compared with late readmission with sepsis shared similar obstetric characteristics. Conclusion Maternal risk factors for both early and late readmission with sepsis included demographic characteristics, cesarean, hemorrhage, and preterm birth. Risks for sepsis after delivery persist beyond the traditional postpartum period of 6 weeks.

Original languageEnglish
Pages (from-to)453-460
Number of pages8
JournalAmerican journal of perinatology
Issue number5
StatePublished - Apr 1 2020


  • infection
  • postpartum
  • pregnancy
  • readmission
  • risk factors
  • sepsis


Dive into the research topics of 'Risk Factors for Maternal Readmission with Sepsis'. Together they form a unique fingerprint.

Cite this