Predicting patient satisfaction: A study of two Emergency Departments

Paul R. Yarnold*, Edward A. Michelson, David A. Thompson, Stephen L. Adams

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

73 Scopus citations


To identify perceptions that predict overall patient (dis)satisfaction with Emergency Department (ED) care, we studied responses to a survey mailed to all discharged patients over a 6-month period (Academic Hospital), and to a telephone interview of a random sample of discharged patients over a 1- year period (Community Hospital). The survey and interview both assessed overall satisfaction, as well as satisfaction with perceived waiting times, information delivery, and expressive quality of physicians, nurses, and staff. Data for 1176 patients (training sample) and 1101 patients (holdout sample) who rated overall satisfaction as either 'very good' or 'very poor' (Academic Hospital), and for 856 patients (training sample) and 431 patients (holdout sample) who rated overall satisfaction as either 'excellent' or 'poor' (Community Hospital), were retained for analysis. For both hospitals, nonlinear tree models efficiently achieved overall classification accuracy exceeding 98% in training analysis and 95% in holdout analysis (all p < .0001). The findings suggest that overall patient (dis)satisfaction with care received in the ED is nearly perfectly predictable on the basis of patient- rated expressive qualities of ED staff, particularly physicians and nurses. Interventions designed to reinforce positive (and extinguish negative) expressive health-care provider behaviors may cut the number of extremely dissatisfied patients in half.

Original languageEnglish (US)
Pages (from-to)545-563
Number of pages19
JournalJournal of Behavioral Medicine
Issue number6
StatePublished - 1998


  • Academic vs. community hospital
  • Emergency service
  • Nonlinear statistical model
  • Optimal data analysis
  • Patient satisfaction
  • Physician behavior

ASJC Scopus subject areas

  • Psychology(all)
  • Psychiatry and Mental health


Dive into the research topics of 'Predicting patient satisfaction: A study of two Emergency Departments'. Together they form a unique fingerprint.

Cite this