Are reductions in emergency department length of stay associated with improvements in quality of care? A difference-in-differences analysis
Are reductions in emergency department length of stay associated with improvements in quality of care? A difference-in-differences analysis
复制标题
DOI：
10.1136/bmjqs-2015-004189
复制
发表时间：
2016-07-01
影响因子：
5.4
通讯作者：
Schull, Michael
中科院分区：
文献类型：
--
作者：
Vermeulen, Marian J.;Guttmann, Astrid;Schull, Michael
背景 我们试图确定在引入安大略省急诊室等待时间策略后的 2 年内，在急诊科 (ED) 整体住院时间 (LOS) 减少的医院就诊的患者是否更有可能在三种重要情况下的急诊科护理质量的其他衡量指标中获得改善。方法 使用双重差分分析进行回顾性医疗记录审查，以比较 11 家安大略省医院在 3 年期间的质量指标绩效变化，这些医院的 ED LOS 中位数较 2008 财年有所改善到 2010 年，13 个匹配站点的 ED LOS 没有变化或恶化。对 2008 年和 2010 年这些医院的急性心肌梗死 (AMI)、哮喘以及儿童和成人上肢骨折患者的 18 项质量指标进行了评估，这些指标反映了急诊室护理的及时性和安全性/有效性。在二次分析中，我们检查了患者就诊时的轮班急诊拥挤情况以及质量指标的表现。 结果 从 2008 年到 2010 年，改善医院的中位 ED LOS 改善了高达 26%（63 分钟），而在未改善的医院则恶化了高达 47%（91 分钟）。我们分别从改善和未改善的医院中提取了 4319 个和 4498 个图表。 2008 年至 2010 年期间，医院整体 ED LOS 中位数的改善与同期任何其他 ED 质量指标的变化无关。在我们的二次分析中，轮班拥挤仅与反映护理及时性的指标相关。在不太拥挤的轮班期间，AMI 患者更有可能在目标间隔内再灌注（比率 1.59，95% CI 1.03 至 2.45），哮喘患者更经常及时接受类固醇给药（比率 1.88，95% CI 1.59 至 2.24）和 β 受体激动剂（比率 1.47，95% CI 1.25 至 2.24）。 1.74）和成人（但不是儿童）骨折患者更有可能在一小时内接受镇痛或夹板固定（比率 1.66，95% CI 1.22 至 2.26）。 结论 这些结果表明，仅针对减少 ED LOS 的政策方法与所选质量指标的更广泛改进无关。与此同时，没有证据表明解决拥挤问题的努力会对护理质量产生不利影响。
Background We sought to determine whether patients seen in hospitals who had reduced overall emergency department (ED) length of stay (LOS) in the 2 years following the introduction of the Ontario Emergency Room Wait Time Strategy were more likely to experience improvements in other measures of ED quality of care for three important conditions.Methods Retrospective medical record review using difference-in-differences analysis to compare changes in performance on quality indicators over the 3-year period between 11 Ontario hospitals where the median ED LOS had improved from fiscal year 2008 to 2010 and 13 matched sites where ED LOS was unchanged or worsened. Patients with acute myocardial infarction (AMI), asthma and paediatric and adult upper limb fractures in these hospitals in 2008 and 2010 were evaluated with respect to 18 quality indicators reflecting timeliness and safety/effectiveness of care in the ED. In a secondary analysis, we examined shift-level ED crowding at the time of the patient visit and performance on the quality indicators.Results Median ED LOS improved by up to 26% (63 min) from 2008 to 2010 in the improved hospitals, and worsened by up to 47% (91 min) in the unimproved sites. We abstracted 4319 and 4498 charts from improved and unimproved hospitals, respectively. Improvement in a hospital's overall median ED LOS from 2008 to 2010 was not associated with a change in any of the other ED quality indicators over the same time period. In our secondary analysis, shift-level crowding was associated only with indicators that reflected timeliness of care. During less crowded shifts, patients with AMI were more likely to be reperfused within target intervals (rate ratio 1.59, 95% CI 1.03 to 2.45), patients with asthma more often received timely administration of steroids (rate ratio 1.88, 95% CI 1.59 to 2.24) and beta-agonists (rate ratio 1.47, 95% CI 1.25 to 1.74), and adult (but not paediatric) patients with fracture were more likely to receive analgesia or splinting within an hour (rate ratio 1.66, 95% CI 1.22 to 2.26).Conclusions These results suggest that a policy approach that targets only reductions in ED LOS is not associated with broader improvements in selected quality measures. At the same time, there is no evidence that efforts to address crowding have a detrimental effect on quality of care.