“Head Up CPR” Is Not Ready for Widespread Adoption

Neuroprotective CPR (NPCPR) is a term coined for a bundle of care for patients in out-of-hospital cardiac arrest (OHCA) that features a mechanical chest compression device, impedance threshold device, and gradual delivery of head-up CPR (HUCPR) accomplished by an automated head and thorax elevation device. A recent position statement recommended “the widespread adoption of NPCPR as a best practice standard of care for all firefighters involved in first response to OHCA cases” (1, 2). We disagree that either HUCPR or NPCPR are the standard of care and do not feel these practices should be widely adopted without additional, high-quality evidence. Additionally, we are concerned that use of the phrase “standard of care” is a legal term that carries potential liability for agencies not adhering to it. While the original position statement was subsequently rescinded, we believe it is important to highlight the limited evidence for HUCPR, describe the necessary steps for translating knowledge from the animal lab to the ambulance, and discuss implications for EMS practice (3).

The Current Evidence for Head Up CPR Is Limited and Does Not Support Widespread Adoption

Most of the evidence supporting HUCPR is based upon animal studies, some with conflicting findings (4–10). While preclinical animal studies are essential for evaluating new treatments and aiding translation to clinical studies, they do not provide sufficient information to support the routine use of new technologies in clinical practice.

There have been four observational, non-randomized, reports of HUCPR in humans (11–14). These studies have significant limitations and cannot, by themselves, be used to support human application. In a before-after analysis from a single EMS agency, the first study reported increased ROSC rates at ED arrival following adoption of a resuscitation strategy of initial passive oxygenation, LUCAS mechanical CPR, and delayed elevation of the head by placing the stretcher in reverse Trendelenburg position (11). Two subsequent studies relied on the same set of 227 patients from a registry (ACE-CPR) of patients treated using HUCPR facilitated by proprietary technology (Elegard, AdvancedCPR Solutions, Minneapolis, MN) (12, 13).

To briefly review, the first of these ACE-CPR registry studies found that earlier compared with later application of the bundle was associated with increased ROSC (12). The second study compared survival to hospital discharge between the 227 patients from the ACE-CPR registry and 5,196 non-concurrent controls from three distinct historic clinical trials (PRIMED, ALPS, and ResQTrial) claiming improved survival among the subset of patients with device application in under ten minutes (13, 15–18). While combining patient populations from different studies may be reasonable in a meta-analysis following assessment of heterogeneity, the use of different studies as comparison groups is highly vulnerable to selection bias. Additionally, the authors excluded over 60% of patients enrolled in the ROC trials and 40% of agencies participating in the ACE-CPR registry because of missing data, further raising concerns about selection bias. The PRIMED, ALPS, and ResQTrial studies used different pharmacological and device interventions, had different goals, and took place in different populations at different time periods (2005 to 2015) than those in the ACE-CPR registry.

The final study used the same methods as Moore et al., but included additional patients more recently recruited into the ACE-CPR registry, with 56% of agencies enrolled in the registry now being excluded (13, 14). This paper was limited to only those patients with a non-shockable presenting rhythm and found improved survival to discharge and improved neurologically intact survival but not improved ROSC.

While the authors of these final two papers used advanced analytic techniques such as propensity matching, these approaches cannot control for the many unmeasured differences that likely exist between such disparate datasets, including whether patients had the opportunity to receive the intervention at all. Findings of benefit using such methods should be considered exploratory or hypothesis generating only and should not be the basis for a definitive claim of improved outcomes (19). Definitive claims of benefit should be made only after more rigorous confirmation of these initial claims. Well executed, randomized controlled trials are the best method to control for both known and unknown differences in patient groups. We completely disagree with any suggestion that randomized controlled trials are not needed to confirm this limited evidence. Results of observational studies can produce variable results and, as a result, scientists should maintain humility in the confidence of their results and confirm their hypotheses with rigorous controlled trials in the face of scientific uncertainty (20).

The Translation of Scientific Studies to Clinical Practice

While there are many pathways by which new ideas are adopted in medical practice, any clinical recommendations should be based upon a rigorous and systematic assessment of existing literature. For example, the Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) is a commonly used framework when critically appraising scientific evidence for integration into clinical practice (21–24). Optimally, systematic reviews that include replicated, hypothesis-confirming randomized trials should support these recommendations. To date, this rigorous process has not occurred. There is a single systematic review, but it is only of animal studies of HUCPR and reports improvement in vascular pressure measurements but not clinical outcomes such as return of spontaneous circulation (ROSC) (25). In 2021 and 2023 the International Liaison Committee on Resuscitation (ILCOR) used a structured GRADE process to assess NPCPR, HUCPR, and suggested against their routine use in clinical practice (26, 27).

Important Lessons for the EMS Community

In light of the recent interest in HUCPR, it is important to reiterate a simple but important lesson: non-randomized observational studies offer important perspectives, but they have a higher risk of bias than other, more rigorous study designs. The history of EMS is filled with examples of interventions that were widely adopted based upon results of observational studies but later found to be ineffective, or even harmful, when tested in randomized trials. Examples include anti-shock trousers and large-volume crystalloid resuscitation in trauma and high-dose epinephrine, empiric calcium chloride, continuous chest compressions, and mechanical chest compression devices in OHCA (28–34). This tendency to put the “cart-before-the horse” is not just a poor use of resources but is also potentially harmful, exposing patients to unproven interventions and distracting rescuers from other essential resuscitation care (35).

We are all eager for innovations that save additional lives. HUCPR is an exciting concept with early promise, but it requires confirmation with independent higher quality supporting evidence. HUCPR is not currently supported by this high level of evidence and therefore, at least for the time being, should not be considered a standard part of EMS practice. We do, however, encourage and support high quality research, including clinical trials, to determine the value and optimal deployment of HUCPR.

Jeffrey L. Jarvis
Office of the Medical Director, Metropolitan Area EMS Authority, Fort Worth, Texas
Department of Emergency Medicine, Baylor Scott & White, Fort Worth, Texas
[email protected]

Michael R. Sayre
Department of Emergency Medicine, University of Washington, Seattle, Washington
Seattle Fire Department, Seattle, Washington

Remle P. Crowe
ESO, Austin, Texas
James J. Menegazzi
Department of Emergency Medicine, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania

Henry E. Wang
Department of Emergency Medicine, The Ohio State University, Columbus, Ohio

Disclosure Statement

The authors report there are no competing interests to declare with the exception that Dr Crowe acknowledges she was recently elected to the NAEMSP Board of Directors.