Waiting Time Correlates with Patient Walk-out Rates But Not with Patient Satisfaction Scores

OBJECTIVES: 1) To correlate patient satisfaction scores by Press-Ganey survey (PSS) and the rate of patients leaving without being seen (LWBS) with waiting times. 2) To ascertain if notification of probable waiting time could modify patient ratings. METHODS: Total registrations between 10am to 7pm and the number LWBS were collected daily for three weeks. In week 1, patients were given no information at triage about probable waits. Expected overall ED waiting times were posted outside the triage window and updated hourly the second week. During the third week, patients were individually notified of the expected waiting time for their triage class and ED section. Press-Ganey surveys were tracked by visit day and reviewed for the second and third weeks to score patient satisfaction with the ED visit and with physician care. Models of LWBS and PSS were generated using week (WK), average waiting time (WT), number of registrations (REG), wait estimates (EST) and estimate accuracy (EA) as predictors. RESULTS: LWBS was correlated with WT, REG, WE. Multiple stepwise forward linear regression of correlated predictors gave a relation of LWBS = 1.8 + 2.5*REG (r(2) =.47, p =.007). PSS was not related to the predictors except for a weak trend with EA (r(2) =.08, p =.054). CONCLUSIONS: Prolonged WT explained almost (1/2) of the LWBS rate but did not adversely affect PSS. Attempts to estimate WT at triage were generally accurate but did not strongly affect PSS.