Validating a model of patient satisfaction with emergency care

Study objective: We sought to validate a previously developed model of emergency department patient satisfaction in a general population using a standard mailed format. The study aims to export the findings of a comprehensive ED quality-of-care study to an easily measured patient population. Methods: A double-sided, single-page survey was mailed to all patients discharged home from 4 teaching hospital EDs during a 1-month period. Determinants of patient satisfaction were analyzed with a previously developed multivariate, ordinal logistic-regression model. Results: The mail survey response rate was 22.9% (2,373/10,381). The survey validates the importance of previously identified determinants of patient satisfaction, including age, help not received when needed, poor explanation of problem, not told about wait time, not told when to resume normal activity, poor explanation of test results, and not told when to return to the ED (P<.01). Greater age predicted higher patient satisfaction, whereas all other variables correlated with lower patient satisfaction. In contrast with prior findings, black race was not a significant predictor of satisfaction in the mail survey population. Low ratings of overall care are strongly correlated with reduced willingness to return (P<.0001). Conclusion: A patient satisfaction model was previously developed from a comprehensive research survey of ED care. We demonstrate the generalizability of this model to a mail survey population and replicate the finding that satisfaction strongly predicts willingness to return. The response rate of this study is typical of commercial patient-satisfaction surveys. The validated model suggests that ED patient satisfaction improvement efforts should focus on a limited number of modifiable and easily measured factors.