Lengths of Stay Based on Safety-Net Status of EDs

Results of a retrospective study [JAMA. 2012;307(5):476-482] found no difference between safety- and nonsafety-net emergency departments (EDs) in compliance with proposed measures for ED lengths of stay for admitted, transferred, discharged, and observed patients. With the increased use in all areas of healthcare of applying performance measures and pay-for-performance schemes to improve quality of care, there is concern that such measures may have unintended consequences of reducing funding to already chronically underfunded areas of healthcare that serve vulnerable populations. EDs that provide a safety net for vulnerable populations may be particularly at risk. In this study, the investigators hypothesized that safety-net EDs perform worse on proposed ED length-of-stay measures than nonsafety-net EDs. To test this hypothesis, they used proposed length-of-stay measures for admissions (8 hours or 480 minutes) and discharges, transfers, and observations (4 hours or 240 minutes) to compare compliance to these measures between safety- and nonsafety-net EDs. Data on EDs were obtained from the 2008 National Hospital Ambulatory Medical Care Survey (NHAMCS) and stratified by safety-net status and disposition (admission, discharge, observation, and transfer). Of a total of 396 hospitals and 34,134 patients sampled, 24,791 (72.1%) were included in the analysis. Patients were excluded from the study if they were <18 years of age, if data on length of stay or disposition were missing, had they left the ED against medical device, or were dead on arrival. Of the 24,791 ED visits included in the analysis, 42.3% were to safety-net EDs and 57.7% to nonsafety-net EDs. Calculating the median and 90th percentile ED lengths of stay for each disposition and admission/discharge subcategories (critical care, routine, psychiatric), the study found no difference between safety- and nonsafety-net EDs. The median length of stay for admission was 269 minutes (interquartile range [IQR],178-397 minutes) versus 281 minutes (IQR, 178-401 minutes) for safety- and nonsafety-net EDs, respectively; 156 minutes (IQR, 95-239 minutes) versus 148 minutes (IQR, 88-238 minutes) for discharge, respectively; 355 minutes (IQR, 221-675 minutes) versus 298 minutes (IQR, 195-440 minutes) for observations, respectively; and 235 minutes (IQR, 155-378) versus 239 minutes (IQR, 142-368 minutes) for transfers, respectively. Multivariate analysis showed no independent association between safety-net status and compliance with ED length-of-stay measures, with an odds ratio (OR) of 0.83 (95% confidence interval [CI], 0.52-1.34) for admissions, 1.03 (95% CI, 0.83-1.27) for discharge, 1.05 (95% CI, 0.57-1.95) for observations, and 1.30 (95% CI, 0.70-2.45) for transfers. This lack of independent association between safety-net status and compliance with ED length of stay also held true for all subcategories tested, except for psychiatric discharges (OR, 1.67 [95% CI, 1.02-2.74]). According to the authors, these findings suggest that safety-net institutions will not be penalized because of ED lengths of stay. Based on median ED lengths of stay, the study found that both safety- and nonsafety-net EDs stayed well under 8 hours for admissions and 4 hours for discharges. However, the authors emphasized that using median ED lengths of stay alone provide only limited information. When looking at the 90th percentile results, which they say are more revealing, the study found that both safety- and nonsafety-net hospitals demonstrated poor performance with the ED length-of-stay measures. Among routine and critical care admissions at safety-net EDs, the lengths of stay had a 90th percentile of nearly 10 hours each, and for psychiatric admissions it was >15 hours. These lengths of stay at nonsafety-net EDs were only slightly better. Of particular concern with this finding, say the authors, is that the 90th percentile measure is often seen as a surrogate marker for crowding. Limitations of the study include the potential for errors in the data from NHAMCS, inclusion of 2008 data that may not reflect current ED visits in the current economic recession, and the potential impact that the healthcare reform law will have on EDs.