See Article by Wasfy et al

At a time when a global pandemic has upended our entire health system, debates and discussions that have lasted for years seem less relevant. However, creative quality improvement interventions, and strategies to evaluate their effectiveness, remain key for health systems that need to optimize their care delivery during this time of crisis. The study by Wasfy et al¹ in this issue of Circulation: Cardiovascular Quality and Outcomes evaluates a multifaceted quality improvement intervention to reduce readmissions for patients following hospitalization for acute myocardial infarction. While their intervention was designed and delivered in a time of more certainty, the relevance of their approach extends beyond their specific intervention.

Improving the quality and efficiency of healthcare delivery has been a key focus of health policy in the United States for the past decade.² An important facet of such policies has been to reduce how frequently patients are readmitted after a hospitalization through high-quality postdischarge care that enables successful transition to self-care. This focus on readmissions was codified into law in 2010 in form of the Hospital Readmissions Reduction Program, which introduced financial disincentives for hospitals with higher-than-expected readmission rates.² An important, albeit controversial, aspect of the policy is that it does not prescribe readmission targets and does not give credit for achieving compliance with intermediate process measures.³ Therefore, hospitals do not have a set playbook and are judged by the best outcomes they can achieve. This focus on outcomes also ensures that hospitals can pursue strategies suited to their unique demographic milieu.

While there have been questions about the magnitude of changes in the frequency of readmissions,⁴ the number of patients readmitted following hospitalizations for acute myocardial infarction have decreased since the policy was introduced.^5,6 In recent years, temporal trends of readmission rates have begun to level off nationally, raising questions about hospitals approaching a state of optimal care where a majority of readmissions are appropriate.^5,7 This is relevant for policy as a focus on readmissions is most meaningful if they are preventable through improvements in care.

The study by Wasfy et al challenges the assertion that substantial reduction in readmissions is not achievable in the current era, thereby highlighting the value of site-specific interventions and readmission targets. They report a quality improvement intervention at the Massachusetts General Hospital, where patients discharged after a hospitalization for acute myocardial infarction had mandatory scheduled follow-up in an ambulatory cardiology practice within 2 weeks and received care from the cardiology consultation service in the emergency department (ED) for hospital revisits within 30 days of discharge. The effects of the intervention were tested in an interrupted time series model wherein the study recorded a ≈5% lower average postintervention readmission rates relative to the preintervention period (10.7% versus 15.5%) without a regression to the mean over the subsequent 24-month period.

There are some key features of the study and its implications that merit discussion. First, the study was designed as a pre-post study without a control group, limiting its ability to account for concurrent changes in hospital practices that may have affected readmission rates independent of the quality improvement intervention. However, the study tested individually the immediate changes in readmission rates with the intervention, as well as the gradual changes in readmissions rates over time. The abrupt decrease in readmissions that was concurrent with the introduction of their intervention and did not regress to preintervention values suggests a causal role of the intervention on changes in readmission rates. In the absence of randomization, which may be difficult to operationalize at a single center and may not provide meaningful insights given the small number of events, such careful appraisal of clinical interventions is key.

Second, the authors did not find any increase in mortality with their intervention. Mortality is an important balancing measure for any intervention focusing on reducing readmissions. Based on all available evidence, a decrease in readmissions for acute myocardial infarction has consistently been associated with a concurrent decrease in mortality.^5,6 Nevertheless, their assessment provides helpful information in a prospective intervention that changed triage decisions in the ED wherein patients were discharged home with input from the cardiology team. This experience of the authors is consistent with patterns observed nationally where a temporal increase in the utilization of the ED and observation units as avenues for care in the 30-day postdischarge period has not been associated with a concurrent increase in mortality.⁸

Third, there are questions regarding generalizability of such interventions to other hospitals. Stated broadly, if the study’s intervention is tested in a randomized clinical trial in the future and fails to identify a measurable change across a group of hospitals, is the intervention ineffective? Certainly, care transition randomized trials have revealed mixed results.⁹ However, the observations by Wasfy et al support a causal inference for the effect of their intervention on readmissions, which declined quickly and specifically among patients seen in clinics and triaged home from the ED, the sites of their intervention. Their success is an argument in favor of site-specific interventions. The disconnect in the experience of Wasfy and results of randomized controlled trials that did not find an association of care transition interventions with reduced readmissions highlights the challenges with studying complex interventions in randomized trials, even though pursuing them is ideal.

Fourth, the study has broad policy implications. Optimizing outpatient follow-up in the early postdischarge care addresses a recognizable care gap. However, diverting patients from the ED to reduce readmissions may prevent prolonged hospital-based care but does not necessarily achieve the goal of reducing the need for postdischarge care. The benefit from such an approach would be the appropriate allocation of our limited inpatient healthcare resources. A focus on reducing readmissions has also resulted in lower costs for the Centers for Medicare and Medicaid Services, with $2.28 billion in savings in 2016 relative to 2010 attributable to fewer readmissions and only 240 million in excess spending on postdischarge ED visits or observation stays.²

Therefore, the work from Wasfy et al supports a focus on reducing readmissions as their interventions reduced the utilization of inpatient beds and potentially saved healthcare costs from fewer inpatient visits. Moreover, they achieved this without any observable increase in mortality. If hospitals can support such interventions using the revenue generated in follow-up visits, these changes will also likely be sustainable.

However, their study does not address a key effect of any intentional changes to care delivery—the experience of patients. Before such interventions become the norm, it is critical to know whether the experience of discharge from the ED is physically, emotionally, and financially better for patients than being readmitted. Until we can measure patient experiences reliably, there may be value in complementing a focus on readmissions with measures cataloguing postdischarge care experience of patients in its entirety. Measures such as excess days in acute care that assess the number of days patients spend in inpatient, observation unit, or the ED can, therefore, complement the readmission measure.¹⁰ Further, building patient-reported measures to capture other aspects of their care experience is vital to comprehending the needs of our patients and creating a personalized postdischarge care plan.

In conclusion, the study offers an example to healthcare institutions as they deploy quality improvement interventions, particularly when their ability to generate randomized evidence is limited. The value of their intervention lies in its clear definition, astute implementation, but most importantly, in its ability to achieve measurable improvement in outcomes. Yoda’s words of wisdom, “Do or do not, there is no try” are universal and apply to evaluating quality improvement interventions as well. The success of a quality improvement intervention is not determined by the processes it tries to modify but whether it actually measurably brings about measurable improvement in patient outcomes.

None.

The opinions expressed in this article are not necessarily those of the editors or of the American Heart Association.

参见Wasfy等人的文章

在全球大流行摧毁了我们整个卫生系统的时候，历时多年的辩论和讨论似乎没有什么意义。但是，创造性的质量改进干预措施以及评估其有效性的策略仍然是需要在危机时期优化其医疗服务的卫生系统的关键。Wasfy等人¹在本期《循环：心血管质量与结果》中的研究评估了一项多方面的质量改善干预措施，以减少急性心肌梗塞住院后患者的再入院率。尽管他们的干预措施是在更加确定的时间内设计和交付的，但其方法的相关性超出了他们的具体干预措施。

在过去的十年中，提高医疗保健的质量和效率一直是美国卫生政策的重点。²这些政策的一个重要方面是通过高质量的出院后护理减少住院后再次入院的频率，从而使患者成功过渡到自我护理。对再入院的关注在2010年以减少医院再入院计划的形式编入法律，该计划引入了对再入院率高于预期的医院的经济刺激措施。²尽管有争议，但该政策的一个重要方面是，该政策未规定再入院目标，也未因实现中间流程措施而受到赞誉。³因此，医院没有固定的剧本，只能根据其可以达到的最佳结果来判断。对结果的关注也确保了医院可以采取适合其独特的人口环境的策略。

虽然已经有关于变化的再入院的频率幅度的问题，⁴的患者再次入院人数住院之后急性心肌梗死有所下降，因为政策出台。^5,6近年来，全国再入院率的时间趋势已开始趋于平稳，这引发了有关医院在大多数再入院合适的情况下达到最佳护理状态的质疑。^5,7这与政策有关，因为如果可以通过改善护理来预防再入院，那么对重新入院的关注最为有意义。

Wasfy等人的研究对以下说法提出了挑战：在当前时代，再入院率无法大幅降低，从而突显了针对特定地点的干预措施和再入院目标的价值。他们在马萨诸塞州总医院报告了一项质量改善干预措施，该患者因急性心肌梗塞住院治疗后出院的患者必须在2周内对非门诊心脏病学进行强制性定期随访，并接受急诊科心脏病咨询服务的护理），以便在出院后30天内再次前往医院。在中断的时间序列模型中测试了干预措施的效果，其中该研究记录的干预后平均再入院率相对于干预前期间降低了约5％（10.7％对15）。

该研究有一些关键特征及其影响值得讨论。首先，该研究被设计为没有对照组的事前研究，从而限制了其解释医院实践同时发生的变化的能力，而这些变化可能影响再入院率，而与质量改善干预措施无关。但是，该研究分别测试了干预后再入院率的立即变化以及再入院率随时间的逐渐变化。再入院率的突然下降与引入他们的干预措施同时发生，并且没有退回到干预前的值，这表明了干预对再入院率变化的因果作用。在没有随机分组的情况下，

其次，作者没有发现他们的干预使死亡率增加。对于任何旨在减少再入院的干预措施，死亡率都是重要的平衡措施。根据所有现有证据，急性心肌梗死再入院率的降低一直与死亡率的降低同时相关。^5,6然而，他们的评估在前瞻性干预中提供了有用的信息，该干预改变了ED中的分诊决定，在该决策中，患者因心脏病小组的帮助而出院了。作者的经验与全国观察到的模式一致，在这种模式下，出院后30天期间使用急诊室和观察室作为护理途径的时间增加并未与死亡率的同时增加相关。⁸

第三，对于将此类干预措施推广到其他医院存在疑问。概括地说，如果该研究的干预措施将来会在随机临床试验中进行测试，并且未能在一组医院中确定可测量的变化，那么该干预措施是否无效？当然，护理过渡随机试验显示了不同的结果。⁹然而，Wasfy等人的观察结果支持了他们干预对再入院的影响的因果推论，这种再入迅速下降，尤其是在诊所和从ED进行分诊的患者中（他们的干预地点）。他们的成功是支持特定地点干预的一个论据。Wasfy经验的脱节和随机对照试验的结果均未发现护理转移干预措施与再入院率降低之间的关联，这凸显了在随机试验中研究复杂干预措施所面临的挑战，即使追求理想状态也是如此。

第四，该研究具有广泛的政策含义。在出院后早期护理中优化门诊随访可解决可识别的护理差距。但是，将患者转离急诊室以减少再次入院可能会阻止长期的医院护理，但未必能达到减少出院后护理需求的目标。这种方法的好处是可以适当分配我们有限的住院医疗资源。对减少再入院率的关注还导致医疗保险和医疗补助服务中心的成本降低，与2010年相比，2016年节省了22.8亿美元，这归因于再入院率的减少以及出院后急诊就诊或观察住院的超额支出只有2.4亿美元。²

因此，Wasfy等人的工作支持着重于减少再入院，因为他们的干预措施减少了住院床位的使用，并通过减少住院次数减少了医疗成本。而且，他们实现了这一目标，而死亡率没有任何明显的增加。如果医院可以利用随访中获得的收入来支持此类干预措施，这些变化也将是可持续的。

但是，他们的研究并未解决任何有意改变护理服务（患者的经历）的主要影响。在此类干预成为常态之前，至关重要的是要知道从急诊室出院的经历是否比重新入院对患者的身体，情感和经济状况都更好。在我们能够可靠地衡量患者经历之前，可能有必要对重新入院的关注与对患者出院后整体护理经历进行分类的措施相辅相成。因此，评估患者在住院，观察单位或急诊室花费的天数的措施，例如急诊中的超长工作天数，可以补充再入院措施。¹⁰此外，制定患者报告的措施以捕捉其护理经验的其他方面，对于理解患者的需求以及制定个性化的出院后护理计划至关重要。

总之，该研究为医疗机构部署质量改进措施提供了一个范例，特别是当他们产生随机证据的能力受到限制时。他们的干预措施的价值在于其明确的定义，明智的实施，但最重要的是，其具有取得可衡量的成果改善的能力。尤达（Yoda）的智慧言语“做或不做，没有尝试”是普遍的，也适用于评估质量改进措施。质量改进干预措施的成功与否并不取决于其尝试修改的过程，而取决于它是否实际上可衡量地带来了患者结果的可衡量的改善。

没有。

本文表达的观点不一定是编辑者或美国心脏协会的观点。