28 Effects of Emergency Department Operational Parameters on Left Without Being Seen: Length of Stay is Most Important

Abstract

Emergency departments (EDs) have experienced an increase in volume and wait times. Markers of operations for an ED include left without being seen (LWBS), with individual ED’s setting site-specific goals. In fact, The Joint Commission and the Centers for Medicare & Medicaid Services (CMS) believe the proportion of patients who left without being seen should be a quality indicator for hospitals. Press Ganey scores from 2010 show an increase of 31 minutes in wait times for patients in an ED since 2002. Up to 51% of all patients are willing to wait for a maximum of 2 hours for emergency care. Multiple models to address this increase in wait times and limiting LWBS have been postulated. These include split flow models, up front physician screening and specialized nurse case managers. To assess the effect of covariates on LWBS rates, in order to identify the most important parameters associated with LWBS. At an academic ED, with an annual census of approximately 50,000, a LWBS rate of < 1% was set as the desired goal. Tighter controls on operations performance, including obtaining new ED software/tracking capabilities, were instituted. A prospective analysis of logistics data was performed. The time period was 100 consecutive days, September 1, 2013 through December 9, 2013. The unit of analysis was “day.” The number of LWBS patients and specific time parameters were recorded. Time parameters collected included the time a patient was signed into ED, triaged, placed in an exam room, initially seen by physician, and left the ED. LOS in the ED was determined for admitted and discharged patients. Descriptive statistics were employed with means and standard deviation for normal data and medians and interquartile ranges for non-normal data. Fisher’s exact testing was utilized to assess for univariate significance between categorical variables. Non-normal data was assessed with nonparametric trend testing. Multivariate regression was utilized to adjust for covariates while assessing relationships between dependent and independent variables. Significance was set at the P<0.05 level. The correlation between LWBS and the overall length of stay (LOS) was by far the major predictor of LWBS; this remained the case even after adjusting for covariates such as ED volume and acuity (as measured by triage level and by admission rate). For each minute of decreased LOS the LWBS decreased by 6% (OR 0.94, 95% CI 0.90-0.97, p<.001). LWBS is a major operations marker throughout most EDs. This study identifies LOS as the strongest single predictor of LWBS at the study hospital. This relationship remains independent of other ED operations metrics (eg, other time intervals). The results are being used to drive LWBS-reducing efforts in the study ED, towards focus on reducing LOS.