Diagnostic Test Accuracy of Point-of-Care Ultrasound during Cardiopulmonary Resuscitation to Indicate the Etiology of Cardiac Arrest

profile avatar

ILCOR staff

Commenting on this CoSTR is no longer possible

To read and leave comments, please scroll to the bottom of this page.

This CoSTR is a draft version prepared by ILCOR, with the purpose to allow the public to comment and is labeled “Draft for Public Comment". The comments will be considered by ILCOR. The next version will be labelled “draft" to comply with copyright rules of journals. The final COSTR will be published on this website once a summary article has been published in a scientific Journal and labeled as “final”.

Conflict of Interest Declaration

The ILCOR Continuous Evidence Evaluation process is guided by a rigorous ILCOR Conflict of Interest policy. The following Task Force members and other authors were recused from the discussion as they declared a conflict of interest: none applicable.

The following Task Force members and other authors declared an intellectual conflict of interest and this was acknowledged and managed by the Task Force Chairs and Conflict of Interest committees: none applicable.

CoSTR Citation

Reynolds JC, Nicholson TC, O’Neil BJ, Drennan I, Issa M, Welsford M, on behalf of the ILCOR Advanced Life Support Task Force. Diagnostic Test Accuracy with Point-of-Care Ultrasound During Cardiopulmonary Resuscitation to Indicate the Etiology of Cardiac Arrest: Consensus on Science with Treatment Recommendations [Internet]. Brussels, Belgium: International Liaison Committee on Resuscitation (ILCOR) Advanced Life Support Task Force, 2022 DATE OF POSTING HERE. Available from:

Methodological Preamble (and Link to Published Systematic Review if applicable)

The continuous evidence evaluation process for the production of the Consensus on Science with Treatment Recommendations (CoSTR) started with a systematic review (PROSPERO CRD42020205207) conducted by Joshua Reynolds, Tonia Nicholson, Brian O’Neil, Mahmoud Issa, and Michelle Welsford with involvement of methodologic and clinical content experts. Evidence from adult literature was sought and considered by the Advanced Life Support Adult Task Force.

The index test was point of care ultrasound (POCUS) – any bedside sonographic anatomic or physiologic assessment conducted during CPR. The reference standard was any additional confirmatory test, process, or procedure other than the original POCUS assessment. Duplication of components in the index test and reference standard artificially inflates diagnostic test accuracy and conflates with inter-rater reliability. We considered secondary review of the original POCUS images to be a measure of inter-rater reliability and not a true reference standard. The target condition was a specific etiology or pathophysiologic state that likely led to cardiac arrest.

In addition to standardized risk of bias assessment with the QUADAS-2 tool, we especially considered two sources of bias unique to prognostication during resuscitation of cardiac arrest. Spectrum bias occurs when diagnostic test accuracy is influenced by the case mix of subjects and/or prevalence of the target condition. In the context of cardiac arrest and CPR, these are influenced by the patient population, clinical setting, institutional support for POCUS, and training/skill/availability of the clinician-sonographer. Verification bias can occur when only ‘test positive’ subjects receive the reference standard or the reference standard differs for ‘test positive’ and ‘test negative’ subjects. In the context of cardiac arrest and CPR, short-term clinical outcomes exert their own version of verification bias if only subjects with return of spontaneous circulation or survival to hospital admission receive the reference standard.

Ultimately, we found one observational study enrolling 48 in-hospital cardiac arrest subjects with complete contingency table data on a subset of 31 subjects to estimate sensitivity and specificity. We provide GRADE evidence profile tables for these data in addition to the narrative summary below. Eleven other studies contained data sufficient only to estimate positive predictive value, in that they report the prevalence of a given POCUS finding and then a description of subsequent imaging, procedural success, or post-procedural clinical outcomes that suggest confirmation of the POCUS finding. We do not provide GRADE evidence profile tables for these data and instead provide a narrative summary below.


The PICOST (Population, Intervention, Comparator, Outcome, Study Designs and Timeframe)

Population: Adults in any setting (in-hospital [IHCA] or out-of-hospital [OHCA]) with cardiac arrest.

Intervention: A particular finding on point-of-care ultrasound during CPR

Comparators: A external confirmatory test or process including some component other than point-of-care ultrasound

Outcomes: A specific etiology or pathophysiologic state that led to cardiac arrest

Study Designs: Randomized and non-randomized cohort studies (prospective and retrospective) and case-control studies with data on both point of care ultrasound findings and an external reference standard to contribute to a contingency table (i.e. true positive, false positive, false negative, true negative). Animal studies, ecological studies, case series, case reports, narrative reviews, abstracts, editorials, comments, letters to the editor, or unpublished studies will not be included.

Timeframe: All years and all languages were included as long as there is an English abstract. Literature search updated through October 6, 2021.

PROSPERO CRD42020205207

Consensus on Science

The overall certainty of evidence was rated as very low for diagnosis of all target conditions primarily due to risk of bias, inconsistency, and imprecision. The individual studies were at substantial risk of selection bias, ascertainment bias, and verification bias. Because of this and a high degree of clinical heterogeneity, no meta-analyses could be performed and individual studies are difficult to interpret.

Sensitivity and Specificity

For the target condition of cardiac tamponade, we identified very low-certainty evidence (downgraded for risk of bias, inconsistency, and imprecision) from 1 observational study (van der Wouw 1997 780) enrolling 48 adult IHCA subjects, which estimated a sensitivity of 1.00 (95% CI 0.29-1.00) and a specificity of 1.00 (95% CI 0.88-1.00) among a subset of 31 subjects. The index test was pericardial effusion with collapse of at least one cardiac chamber. The reference standard was autopsy and/or clinical adjudication.

For the target condition of pulmonary embolism, we identified very low-certainty evidence (downgraded for risk of bias, inconsistency, and imprecision) from 1 observational study (van der Wouw 1997 780) enrolling 48 adult IHCA subjects, which estimated a sensitivity of 1.00 (95% CI 0.16-1.00) and a specificity of 0.97 (95% CI 0.82-0.99) among a subset of 31 subjects. The index test was a dilated right ventricle and right atrium with poor filling of the left atrium and left ventricle. The reference standard was autopsy and/or clinical adjudication.

For the target condition of myocardial infarction, we identified very low-certainty evidence (downgraded for risk of bias, indirectness, inconsistency, and imprecision) from 1 observational study (van der Wouw 1997 780) enrolling 48 adult IHCA subjects, which estimated a sensitivity of 0.86 (95% CI 0.57-0.98) and a specificity of 0.94 (95% CI 0.71-0.99) among a subset of 31 subjects. The index test was reduced contractility in a region of myocardium during spontaneous contractions with or without effective cardiac output (i.e. pulseless electrical activity or ‘peri-ROSC’ states). The reference standard was autopsy and/or clinical adjudication.

Positive Predictive Value

For the target conditions of cardiac tamponade, pericardial effusion, pulmonary embolism, myocardial infarction, aortic dissection, and hypovolemia, 11 observational studies (Chua 2017 310, Hilberath 2014 926, Jung 2020 31, Lien 2018 125, Lin 2006 167, Memtsoudis 2006 1653, Shillcutt 2012 362, Tayal 2003 315, Varriale 1997 1717, Zengin 2012 68, Zengin 2016 105) with high risk of bias provided data sufficient only to estimate individual positive predictive values among subsets of one to ten subjects with out-of-hospital cardiac arrest, in-hospital cardiac arrest, or intra-operative cardiac arrest. In each study, a small number of subjects had a notable POCUS finding compared to a reference standard. Most of the individual studies used different definitions of the index test and/or reference standard for a given target condition. Individual estimates of positive predictive value have very wide confidence intervals and are difficult to interpret in the context of very small subsets of subjects.

A complete description of index tests, reference standard, and positive predictive value estimates is available at doi: 10.1016/j.resuscitation.2022.01.006 (Table 3).

Treatment Recommendations

We suggest against the routine use of point of care ultrasound during CPR to diagnose reversible causes of cardiac arrest (weak recommendation, very low-certainty evidence).

We suggest that if point of care ultrasound can be performed by experienced personnel without interrupting CPR, it may be considered as an additional diagnostic tool when clinical suspicion for a specific reversible cause is present (weak recommendation, very low-certainty evidence).

Any deployment of diagnostic point of care ultrasound during CPR should be carefully considered and weighed against the risks of interrupting chest compressions and misinterpreting sonographic findings (good practice statement).

Justification and Evidence to Decision Framework Highlights

This topic was prioritized by the ALS Task Force based on the frequent utilization of point-of-care ultrasound during cardiac arrest without recognizing the potential pitfalls for misinterpretation as a diagnostic tool. A comprehensive and rigorous summary of its intra-arrest diagnostic capabilities provides valuable information to both the resuscitation science community and bedside clinicians.

In making these recommendations, the ALS Task Force considered the following:

  • The inconsistent definitions and terminology used for sonographic evidence of specific causes of cardiac arrest was the primary source of clinical heterogeneity. We strongly encourage the establishment of uniform definitions and terminology to describe sonographic findings of reversible causes of cardiac arrest.
  • The identified studies suffer from high risk of bias related to selection bias and ascertainment bias. Additionally, the logistics of cardiac arrest resuscitation introduce potential for spectrum bias (when diagnostic test accuracy is influenced by the case mix of subjects and/or prevalence of the target condition) and verification bias (when availability or use of the reference standard is influenced by ‘test positive’ or ‘test negative’ status). Verification bias was present in all but one of the included studies, largely restricting contingency tables to positive predictive value. The evidence supporting use of POCUS as a diagnostic tool is uniformly of very low certainty. Clinicians should cautiously interpret sonographic findings during cardiac arrest in light of these limitations. We strongly encourage subsequent investigations of POCUS during cardiac arrest to employ methodology that mitigates these risks of bias. This includes enrolling a consecutive, prospective sample; utilizing clear definitions of the index test, credentials of the sonographer, and testing interval; selecting an objective, uniform reference standard; and blinding appropriately.
  • No included studies reported estimates of inter-rater reliability. The influence of acoustic window, sonographer training/experience, and particular pathology in question on inter-rater reliability is also unknown. As POCUS matures as a field, there are now validated image quality rating scales to promote standardization of assessment. (Gaspari 2021 100097).
  • No POCUS finding had sufficient sensitivity to be used as sole criterion to ‘rule out’ the cause of cardiac arrest, but the certainty of this evidence is very low.
  • POCUS findings had higher point estimates and/or narrower confidence intervals of specificity to ‘rule in’ certain causes of cardiac arrest, but this evidence is from a single study and of very low certainty.
  • The diagnostic utility of POCUS is affected by the clinical context. For example, a post-operative cardiac surgery patient with acute cardiac arrest has given pre-test probabilities for specific causes such as cardiac tamponade, pulmonary embolism, or acute hemorrhage. Conversely, the diagnostic utility of POCUS may be more limited in the context of undifferentiated cardiac arrest in the out-of-hospital setting.
  • Clinicians should be cautious about introducing additional interruptions in chest compressions with a transthoracic approach to point-of-care echocardiography during cardiac arrest. (Huis In’t Veld 2017 95, Clattenburg 2018 65) Several logistical strategies mitigate these concerns, including use of transesophageal echocardiography. (Clattenburg 2018 69; Gaspari 2021 100094; Teran 2019 409).
  • The task force noted several pitfalls and logistical questions around the feasibility of diagnosing a myocardial infarction in the context of pulseless electrical activity or similar low-flow states. In this context, wall motion abnormalities may result from the ischemia of a low-flow state or a pre-existing infarct, as opposed to a de novo myocardial infarction.
  • Not treating a reversible cause of cardiac arrest risks failure of resuscitation or more severe post-cardiac arrest injury. Treating an incorrect diagnosis suggested by POCUS risks iatrogenic injury or delayed identification of the true underlying cause.
  • POCUS is subject to the availability of equipment and skilled operators. Starting a new POCUS program requires material fixed and recurring costs and resources to obtain equipment and train clinicians. An existing POCUS program requires fewer incremental resources to be used in the context of cardiac arrest. In either case, the development and maintenance of the requisite skill sets both obtain and interpret images under the compromised conditions of cardiac arrest presents an additional burden for a POCUS program. The task force expects that most diagnostic applications of POCUS will occur in a hospital-based setting as opposed to the prehospital setting.
  • Given the items listed, many task force members advocated for restriction of diagnostic applications of POCUS to circumstances in which the clinical suspicion for a readily treatable abnormality is high and justifies interruption of CPR. In such instances, the time allotted for imaging should be as brief as possible.
  • The prognostic utility of POCUS to predict clinical outcomes is covered in a separate PICOST ( Gaps

    There are no studies of the diagnostic accuracy of point-of-care ultrasound during cardiac arrest with methodology that sufficiently minimizes risk of bias, especially selection bias, ascertainment bias, and verification bias.

    There are no uniform definitions and terminology to describe sonographic findings of reversible causes of cardiac arrest or the associated reference standards.

    The inter-rater reliability of POCUS diagnostic findings during cardiac arrest is unknown.

    No identified studies provided data on resource requirements, cost-effectiveness, equity, acceptability, or feasibility.

    Some studies reported a ‘change in management’ driven by the diagnostic use of POCUS, but these assertions are not well characterized or quantified. Furthermore, it is unknown whether these ‘changes in management’ led to improved clinical outcomes.




Chua MT, Chan GW, Kuan WS. Reversible Causes in Cardiovascular Collapse at the Emergency Department Using Ultrasonography (REVIVE-US). Ann Acad Med Singap. 2018;46:310-6.

Clattenburg EJ, Wroe P, Brown S, Gardner K, Losonczy L, Singh A, Nagdev A. Point-of-care ultrasound use in patients with cardiac arrest is associated prolonged cardiopulmonary resuscitation pauses: A prospective cohort study. Resuscitation. 2018;122:65-68.

Clattenburg EJ, Wroe PC, Gardner K, Schultz C, Gelber J, Singh A, Nagdev A. Implementation of the Cardiac Arrest Sonographic Assessment (CASA) protocol for patients with cardiac arrest is associated with shorter CPR pulse checks. Resuscitation. 2018 Oct;131:69-73.

Gaspari R, Harvey J, DiCroce C, Nalbandian A, Hill M, Lindsay R, Nordberg A, Graham P, Kamilaris A, Gleeson T. Echocardiographic pre-pause imaging and identifying the acoustic window during CPR reduces CPR pause time during ACLS - A prospective Cohort Study. Resusc Plus. 2021 Mar 6;6:100094.

Gaspari R, Teran F, Kamilaris A, Gleeson T. Development and validation of a novel image quality rating scale for echocardiography during cardiac arrest. Resusc Plus. 2021 Mar 6;6:100097.

Hilberath JN, Burrage PS, Shernan SK, Varelmann DJ, Wilusz K, Fox JA, Eltzschig HK, Epstein LM, Nowak-Machen M. Rescue transoesophageal echocardiography for refractory haemodynamic instability during transvenous lead extraction. Eur Heart J Cardiovasc Imaging. 2014 Aug;15(8):926-32.

Huis In 't Veld MA, Allison MG, Bostick DS, Fisher KR, Goloubeva OG, Witting MD, Winters ME. Ultrasound use during cardiopulmonary resuscitation is associated with delays in chest compressions. Resuscitation. 2017;119:95-98.

Jung WJ, Cha KC, Kim YW, Kim YS, Roh YI, Kim SJ, Kim HS, Hwang SO. Intra-arrest transoesophgeal echocardiographic findings and resuscitations outcomes. Resuscitation. 2020;154:31-7.

Lien WC, Hsu SH, Chong KM, Sim SS, Wu MC, Chang WT, Fang CC, Ma MH, Chen SC, Chen WJ. US-CAB protocol for ultrasonographic evaluation during cardiopulmonary resuscitation: validation and potential impact. Resuscitation. 2018;127:125-31.

Lin T, Chen Y, Lu C, Wang M. Use of transoesophageal echocardiography during cardiac arrest in patients undergoing elective non-cardiac surgery. Br J Anaesth. 2006 Feb;96(2):167-70.

Memtsoudis SG, Rosenberger P, Loffler M, Eltzschig HK, Mizuguchi A, Shernan SK, Fox JA. The usefulness of transesophageal echocardiography during intraoperative cardiac arrest in noncardiac surgery. Anesth Analg. 2006 Jun;102(6):1653-7.

Shillcutt SK, Markin NW, Montzingo CR, Brakke TR. Use of rapid "rescue" perioperative echocardiography to improve outcomes after hemodynamic instability in noncardiac surgical patients. J Cardiothorac Vasc Anesth. 2012 Jun;26(3):362-70.

Tayal VS, Kline JA. Emergency echocardiography to detect pericardial effusion in patients in PEA and near-PEA states. Resuscitation. 2003;59:315-8.

Teran F. Resuscitative Cardiopulmonary Ultrasound and Transesophageal Echocardiography in the Emergency Department. Emerg Med Clin North Am. 2019 Aug;37(3):409-430.

van der Wouw PA, Koster RW, Delemarre BJ, de Vos R, Lampe-Schoenmaeckers AJ, Lie KI. Diagnostic accuracy of transesophageal echocardiography during cardiopulmonary resuscitation. J Am Coll Cardiol. 1997;30:780-3.

Varriale P, Maldonado JM. Echocardiographic observations during inhospital cardiopulmonary resuscitation. Critical Care Medicine. 1997;25:1717-20.

Zengin S, Yildirim C, Al B, Genc S, Kilic H, Dogan M. The effectiveness of ultrasound in patients with non-traumatic cardiopulmonary arrest. Journal of Academic Emergency Medicine. 2012;11:68-72.

Zengin S, Yavuz E, Al B, Cindoruk S, Altunbas G, Gumusboga H, Yildirim C. Benefits of cardiac sonography performed by a non-expert sonographer in patients with non-traumatic cardiopulmonary arrest. Resuscitation. 2016;102:105-9.


Sort by

Time range




Review Type