Readmission after major surgery: effect of the postdischarge environment

Marcelo Cerullo, Faiz Gani, Sophia Y. Chen, Joseph K. Canner, Timothy M. Pawlik

Research output: Contribution to journalArticlepeer-review


Background Although uncoordinated postdischarge care has been associated with poor clinical outcomes, the effect of discharge to a low healthcare resource area (LHRA) on readmission remains undetermined. We sought to assess how the quality of discharge area health resources impact readmission following major surgery. Methods This cross-sectional study was performed by linking Maryland state data for 2012-2015 to the Agency for Healthcare Research and Quality Area Health-Resource File. Patients undergoing one of 11 common surgical procedures were identified. Multivariable logistic regression was performed to assess the effect of discharge area health resource quality on readmission. Results A total of 76,747 patients were identified of which 9.4% were discharged to a high healthcare resource area (HHRA), whereas 81.9% of patients were discharged to an LHRA. Perioperative morbidity and length of stay were comparable between HHRA versus LHRA patients (both P > 0.05). Among all patients, 30-d and 90-d readmission was 6.5% and 12.4%, respectively. On multivariable analysis, discharge to LHRA was independently associated with a 19% (odds ratio = 1.19; 95% CI, 1.01-1.41; P = 0.043) and 18% (odds ratio = 1.18; 95% CI, 1.04-1.33; P = 0.010) greater odds of 30-d and 90-day readmission, respectively. Conclusions Patients discharged to an area characterized by LHRA were more likely to be readmitted at 30 d and 90 d following index discharge.

Original languageEnglish (US)
Pages (from-to)318-326
Number of pages9
JournalJournal of Surgical Research
Issue number2
StatePublished - Oct 1 2016


  • Postdischarge care
  • Readmission
  • Surgical complications

ASJC Scopus subject areas

  • Surgery


Dive into the research topics of 'Readmission after major surgery: effect of the postdischarge environment'. Together they form a unique fingerprint.

Cite this