Death to Discharge - Surviving the System: Coronary Reperfusion for Out-of-Hospital Cardiac Arrest

Survival from out-of-hospital cardiac arrest (OHCA) requires the interaction of complex diversified systems and the implementation of evidence-based interventions associated with improved outcomes in a coordinated and seamless manner. Complex system interactions cannot be quantified, but individual impediments to overall system function can be identified, as well as important “devil’s in the details.” Peberdy and colleagues¹ in the above-referenced AHA Scientific Statement address an important and evolving individual, institutional, and system bias potentially resulting in the exclusion of the highest risk acute coronary syndrome (ACS) patients^{2, 3} from optimal clinical treatment and outcomes.

One intervention associated with improved outcome in patients resuscitated from OHCA is coronary angiography⁴ and emergent coronary revascularization.⁵ Intermediate and long-term mortality in patients undergoing successful percutaneous coronary intervention (PCI) for ST-segment elevation myocardial infarction (STEMI) with OHCA appear similar to those without arrest. Resuscitation from cardiac arrest following PCI may not be an independent predictor of mortality.^6-8

The importance and rationale for a strategy of triage to PCI centers for coronary angiography following resuscitation from OHCA have evolved over time since Davies first carefully demonstrated that an unstable but not necessarily completely occlusive coronary lesion is associated with the majority of out-of-hospital adult sudden cardiac deaths.^{9, 10} As coronary intervention matured, Spaulding in Paris developed a pragmatic but sound pathophysiological strategy by triaging resuscitated victims of cardiac arrest without an obvious non-cardiac etiology to interventional centers for coronary angiography and PCI.¹¹ Coronary occlusion was a frequent finding and angioplasty appeared to improve survival. Although largely observational and non-randomized, subsequent data have confirmed this association with improved outcome. So what barriers and impediments exist that continue to stall the rapid access of patients resuscitated from cardiac arrest to STEMI systems of care, timely reperfusion, and interdisciplinary critical care in appropriate specialty units and resuscitation centers?

To help answer this question, Peberdy and colleagues expertly target and discuss an important impediment to access and intervention: public reporting of PCI outcome. But the title is deceiving since it reviews much more than the impact of public reporting methodology and consequences of PCI outcome on patients resuscitated from OHCA in large registries. Expanding the scope of the statement, the authors do much more and delineate the premise and current data supporting the inclusion of a reperfusion strategy for appropriate patients. Compelling observational data also suggest that patients resuscitated from OHCA require separate analysis and consideration of confounding parameters and systems of care.

Tracking and public reporting of health-related outcomes can be an important quality tool but the intended outcome of changing individual and system behavior is not convincingly established.^{12, 13} In reality, public (and probably confidential) reporting of performance and outcome can have unintended consequences that may initiate and sustain adverse outcomes and less than optimal care.^{14, 15} Two harmful consequences are the limitation of system access (specifically STEMI systems of care and bundled interdisciplinary specialty critical care) and clinician decision making in the treatment of high-risk patients.^{13, 16} Performance reporting shines an altruistic quality light on individual performance but also sends a probing and indiscriminate beacon on system performance with multiple individual, institutional, and stakeholder biases. These biases can act in concert to promote the exclusion of high-risk patients from access to potentially beneficial therapy. In fact, a recent analysis of the Massachusetts PCI Registry demonstrated that PCI centers that were publicly reported as outliers shifted their risk profile (decreasing the number of high-risk patients) with an associated improvement in reported PCI mortality.¹⁴

Most adults are risk averse by nature.¹⁷ Raise the stakes and this inherent tendency is subliminally stimulated or manifestly magnified. Individual biases are then easily recruited to minimize this risk, perceived or real. Clinicians attempt to minimize risk for their patients and “do no harm.” In this light, the harm to patients may be the assumption that an intervention is futile and unnecessary with attendant misuse of costly resources. Poor data collection and interpretation will reinforce this behavior and perpetuate an individual or institutional conformational bias. Poor system function and the failure to appreciate system inadequacies and design provide an outcome bias to sustain it. Bad data generate an anchoring bias and provide false benchmarks and a confounded reference point, distorting assessment and future outcome analysis. For example, in the National Cardiovascular Data Registry (NCDR®) CathPCI, the only current metric employed to account for OHCA is “cardiac arrest within 24 hours.” As a result, a brief episode of ventricular fibrillation (VF) in the catheterization laboratory or emergency department is weighted similar to a resuscitated OHCA presenting with VF. In the latter, additional variables such as witnessed arrest, time to defibrillation, bystander CPR, quality of CPR, and EMS response times may be additional independent variables.

So what can be done as a temporary fix to prevent the development and perpetuation of inadequate, contradictory, and confounding data? What prognostic factors and treatment goals require incorporation into system access and redesign? Excluding inadequate and substantially confounded data would be an immediate first step. Patients resuscitated from OHCA should be excluded from mortality and quality of care models. The playing field would be leveled while a uniform reporting and appropriate risk assessment model is developed. Then, improved and uniform risk prediction models could be developed and refined to address the unique and additional clinical parameters and treatment, for example the use of induced hypothermia and bundled goal driven critical care.

Additionally, one major purpose of public reporting is to encourage quality assessment and improvements in care. However, process of care requires validation, and validation is premised on successful system design, implementation, and ongoing assessment. A system redesign will be required to permit a cycle of uniform outcome assessment, a setting for continuing evaluation and research, and development, implementation, and testing of new strategies. For example, Mylotte recently reported that a strategy of primary PCI with multivessel revascularization following cardiac arrest appears to improve survival in resuscitated patients presenting with cardiogenic shock.¹⁸ This strategy is not supported by current guideline recommendations and appropriate use criteria. In these patients at highest risk, a risk-averse physician and system could easily invoke futility pending “clinical outcome.” Beyond targeting recommendations for public reporting and PCI, organizations and systems need to recognize, provide, and nurture a culture and environment that encourages comprehensive, clinically relevant data collection, research, and clinical guideline development.

My ”Top Ten” take-away messages from this AHA Scientific Statement applied to patients following out-of-hospital resuscitation from cardiac arrest are as follows:

1. Most out-of-hospital sudden adult cardiac arrests occur in the setting of ACS and an unstable plaque.
2. The majority of in-hospital deaths following a reperfusion strategy after PCI for OHCA are due to multi-organ dysfunction and related clinical and ethical decisions. Deaths are not primarily cardiovascular or related to PCI performance.
3. ACS patients resuscitated from cardiac arrest have the highest mortality but potentially the greatest absolute benefit. Their long-term neurological function is good, and survival may be comparable to survival from cardiogenic shock complicating STEMI.
4. The electrocardiogram is a useful triage tool following cardiac arrest, but STEMI criteria miss one in five patients with a completely occluded artery. A unique reperfusion and hemodynamic support strategy may deserve consideration in the presence of multi-organ reperfusion injury and with multi-vessel disease and cardiogenic shock.
5. Use of coronary angiography and appropriate intervention are associated with improved mortality and neurological function.
6. Coma and neurological impairment adversely affect prognosis but do not independently preclude triage to coronary angiography. Patients with acute myocardial infarction undergoing PCI have one of the best prognoses for return of good neurological function.
7. Following transfer from the catheterization laboratory these patients require multidisciplinary assessment, interventions, and comprehensive critical care. A checklist post-PCI ACS mentality will not optimize or improve outcome.
8. Real-world risk aversion occurs at an individual, institutional, and system level, impeding access and optimal care due to public (and probably confidential) reporting of mortality.
9. Current risk prediction models do not adequately account for the unique circumstances and multiple confounding clinical and system factors impacting patient and outcome prediction. The development of heterogeneous models and varied quality interventions will confound a quality process and strategy development.
10. Organizations and quality care processes should not publicly report outcomes for patients undergoing a reperfusion strategy and PCI following cardiac arrest. They should develop a nested registry concept (similar to the Syntax Registry)¹⁹ including indicators of outcome²⁰ appropriate for quality care and regional system development, assessment, and process improvement.
11. One Power Ball number for Good Measure- National organizations and registries should focus on quality outcome processes and strategies for system re-design by hypothesis-generating data analysis and research mindful of heterogeneous and complex system interactions. Regional and local systems should re-design or implement strategies of care that meld successful STEMI systems of care best practices and concepts of resuscitation centers.