RECOVER evidence and knowledge gap analysis on veterinary CPR. Part 2: Preparedness and prevention



This article is corrected by:

  1. Errata: Erratum Volume 23, Issue 5, 571, Article first published online: 9 September 2013

  • The authors and collaborators declare no conflicts of interests.

Address correspondence and reprint requests to Dr. Maureen McMichael, College of Veterinary Medicine, University of Illinois, 1008 West Hazelwood St, Urbana, IL 61802, USA. E-mail:



To systematically examine the evidence on the effect of prevention and preparedness measures on outcomes in veterinary cardiopulmonary resuscitation and to determine knowledge gaps.


Standardized, systematic evaluation of the literature, categorization of relevant articles according to level of evidence and quality, and development of consensus on conclusions for application of the concepts to clinical practice. Relevant questions were answered on a worksheet template and reviewed by the Reassessment Campaign on Veterinary Resuscitation (RECOVER) prevention and preparedness domain members, by the RECOVER committee, and opened for comments by veterinary professionals for 3 months.


Academia, referral practice, and general practice.


Nine worksheets were prepared to determine the extent to which preparation of the environment (charts, visual aids, etc) and personnel (training, debriefing, etc) are beneficial in improving return of spontaneous circulation.


Of the questions evaluated, only the association between anesthesia-related cardiopulmonary arrest and better outcomes was supported by strong evidence. There is some evidence from the human literature that the use of cognitive aids, standardized didactic, and hands-on training with high-fidelity simulators, team and leadership training, and post-cardiac arrest debriefing improve adherence to cardiopulmonary resuscitation guidelines and, in some cases, patient outcomes. Veterinary studies investigating these issues are lacking, and development of initial guidelines is a crucial first step.


advanced life support




basic life support


cardiopulmonary arrest


level of evidence


population, intervention, control group, outcome


Reassessment Campaign on Veterinary Resuscitation


return of spontaneous circulation


The outcome of cardiopulmonary arrest (CPA) is a function of numerous disparate factors that affect the clinical results. Resuscitation programs that are organized and cohesive and that are led by a well-functioning, knowledgeable team should improve survival from CPA. Ideally, strengthening the links in the chain of survival (the integrated set of time-sensitive, coordinated actions necessary to maximize survival from cardiac arrest) will lead to improved outcomes. As it relates to “Preparedness and Prevention” in veterinary cardiopulmonary resuscitation (CPR), the chain of survival includes both environmental and personnel factors.

Environmental factors that may impact performance include well-designed, straightforward checklists; algorithm charts; cognitive aids; and well-stocked, easy to access, organized crash carts. Personnel factors include high-level team member training, specific leadership training, and appropriate intervals of retraining and debriefing. Optimizing outcome requires consistent improvement in resuscitation education and the implementation of support systems to allow the streamlined delivery of CPR in the clinic setting. Systematic evaluations of instructional delivery along with matched evaluations of clinical performance should allow for the development of evidence-based guidelines for course development and instruction. Course evaluation should include the effectiveness of both the acquisition and retention of the knowledge and skills by the participants. More robust forms of programmatic evaluation include the degree of clinical integration of the knowledge and its association with patient outcome.

This article will review clinical and research evidence from nine population, intervention, control group, outcome (PICO) questions, two that concern environment and etiology and seven that address personnel and training.

The key Preparedness and Prevention recommendations made in this consensus statement for canine and feline CPR are as follows:

  • Organized, pre-stocked arrest stations improve CPR performance, and should be located in areas where animals are routinely anesthetized.
  • Post-CPR debriefing is safe, easy, and improves future performance.
  • Standardized training programs have improved adherence to guidelines in human medicine and are needed in veterinary medicine.
  • Leadership and team communication training increase effectiveness of CPR teams.
  • High fidelity manikins for teaching CPR psychomotor skills are highly effective in human medicine and would be valuable in veterinary medicine.

Evidence Summary

Preparation of the environment

An essential aspect of readiness to respond optimally to an event of CPA is the presence of a resuscitation environment that complements the resuscitators during the CPR effort. This section examines core elements of such an environment.

Crash cart and cognitive aids (PRE01)

PICO Question

In dogs and cats with cardiac arrest (P), does the use of a pre-stocked arrest station with checklists/charts/aids (I), compared with not using these methods (C), improve outcome (O) (eg, return of spontaneous circulation [ROSC]; discharge alive)?


Multiple high quality and high level of evidence (LOE) studies in human medicine suggest that the use of both pre-stocked arrest stations and cognitive aids improves compliance with CPR protocols. However, there are no studies investigating the utility of these measures in veterinary medicine.

Summary of the evidence

The use of a pre-stocked, organized and functional arrest station is a key element in the efficient operation of CPR. A recent internet survey in veterinary medicine (LOE 4, poor/supporting) found that of the CPR preparedness measures evaluated, 80% of general practitioners and 98% of specialists in anesthesia or emergency and critical care have a regularly maintained crash cart consistently available to them for use in an emergency setting.[1] Checklists and crash carts have also been the subject of multiple human studies in an effort to improve the outcome from CPA. Deficiencies of the arrest station center around lack of standardization and systematic layout and these affect team organization during a crisis event (LOE 6, fair/supporting).[2] Most problematic is missing equipment due to lack of return to the cart or incomplete stocking of appropriate materials, inability to identify or locate needed medications, and failure to have drugs and syringes in a quick, usable form (LOE 6, fair/supporting).[3] Deficiencies and defects in resuscitation equipment include basic device failure (intrinsic design faults, manufacturing errors, or component failures), external factors (electric power or gas supply failure), and human error (which is the most common reason for failure and includes inadequate knowledge, training or supervision, and lack of experience with the use of the equipment) (LOE 6, good/supporting).[4] Inadequate preventative maintenance and checklists lead to inappropriate equipment components, device wear and tear, defects and incorrect assembly of equipment (LOE 6, good/supporting).[4] In fact, in a retrospective study evaluating cardiac arrest in a district human hospital (LOE 6, fair/supporting), delay in CPR was found to be related to equipment failure in 18% of the cases.[5] This study also found that 9% of the carts had significant deficiencies, which included such problems as unacceptable cart location for immediate use and infrequent assessment of cart content. Solutions for these problems center on detail planning and education. Emphasis should be placed on having identical equipment for every response to minimize the need to relearn what equipment is present, where it is located, and how to operate it (LOE 6, fair/supporting).[6] The code carts should have standardization of contents and location, determination of medications and supplies needed for stocking, sealing of the carts to ensure that supplies are present when needed, regular maintenance checks and restocking of used materials, and providing a method for personnel to familiarize themselves with the cart and have appropriate training on its contents (LOE 6, fair/supporting)[3] (LOE 6, fair/supporting)[2] (LOE 6, fair/neutral).[7] Williams (LOE 6, poor/supporting) devised a mobile educational crash cart, fully stocked with equipment, supplies, and educational tools that was identical to those used in the hospital setting for procedural training on the CPR equipment and materials.[8]

Checklists and cognitive aids are often inconsistently used in CPR settings and reasons for poor reference to these materials included lack of clarity and specifics, time pressure, unfamiliarity, difficulty of use in the middle of a scenario, disagreement with the content, or not having a need for it (LOE 6, good/neutral).[9] Other studies have found an association between formal training in the use of cognitive aids and whether the aids were used in an arrest situation (LOE 6, good/supporting).[10, 11] Two additional studies have evaluated CPR skills performance and the use of checklists, one (LOE 6, good/supporting) evaluating a short versus a long version of a checklist for success in simulated CPR, and the other (LOE 6, good/neutral) examining the influence of adherence to an established protocol on outcome.[12, 13] The former study found that a long and more detailed checklist improved outcome over a shortened version or no checklist; the latter study found that outcome was affected by the number of critical tasks that the CPR team performed but unaffected by the sequence of activities recommended in the protocol. There is, however, support for having a resuscitation chart for drug dosing and calculations readily available (LOE 6, fair/supporting), and based on the veterinary survey mentioned, the majority of practitioners have these charts available for use.[14]

Knowledge gaps

Although there is convincing evidence in human medicine that the use of pre-stocked arrest stations with checklists, charts, and aids improves compliance with standard CPR protocols, there is no evidence in the veterinary literature on the utility of these measures. Further studies are needed prospectively in human and veterinary medicine to evaluate if checklists and cognitive aids contribute to improved CPR outcome.

Arrest etiology and outcome (PRE02)

PICO Question

In dogs and cats with cardiac arrest (P), does the specific etiology (anesthetic arrest, ICU arrest) (I), compared with all arrests (C), predict outcome in CPR (O) (eg, ROSC, survival to discharge)?


Although the overall survival rate in dogs and cats following CPA is poor, there is high level and high quality supporting evidence in veterinary medicine that anesthesia-related CPAs are associated with increased survival compared to arrests from other causes.

Summary of the evidence

The overall survival rate in dogs and cats with cardiac arrest, regardless of etiology (anesthetic arrest, ICU arrest), ranges from 4% to 9.6% (LOE 2, good/supporting; LOE 6, good/supporting; LOE 4, fair/supporting; LOE 4, fair/supporting, respectively)[15-18] compared to in-hospital survival rates of 10% to 20% in humans (LOE 6, good/supporting, 22 only).[10], [19-22] A recent veterinary prospective observational study (LOE 2, poor/supporting) found that of 204 dogs and cats with cardiac arrest, those anesthetized at the time of CPA were significantly more likely to have ROSC and survive to discharge (9/19, 47%) compared to patients not anesthetized at the time of CPA (3/185, 2%).[15] For the 12 patients in this study that survived and were discharged from the hospital, 9 (75%) had anesthetic arrests. This proportion is higher than that found in an earlier veterinary retrospective study (LOE 4, fair/supporting) evaluating 18 CPA survivors to discharge, of which 10 (55.6%) had anesthesia-related arrests.[23] Another clinical prospective observational study (LOE 2, good/supporting) in 11 cats reported a survival rate of 36.4% (4/11) for anesthesia-associated CPA, compared to 0% (0/7) patients with ICU-related arrests.[24] In a retrospective study, Kass and Haskins (LOE 4, fair/supporting) reviewed the records of 135 dogs and 43 cats post-CPR for factors leading to their cardiac arrest and for survival predictability following the procedure.[17] All animals (4 dogs and a single cat) discharged alive from the hospital had cardiac arrests associated with drug and/or anesthetic reactions.

A case report of an inadvertent overdose of ketamine in a cat with prompt CPR and short-term mechanical ventilation (LOE 5, good/supporting) describes a full clinical recovery and subsequent discharge of the animal.[25] A retrospective study evaluating CPR in 15 hospitalized rabbits (LOE 6, good/supporting) found that 3 of 5 rabbits achieving prolonged ROSC, including the single rabbit that survived to discharge, were anesthetized at the time of CPA.[16] In several large studies of cardiac arrest while under anesthesia, the mortality rate in humans was significantly lower than the reported mortality rate for non-anesthesia-related arrests. In one large study comparing cardiac arrest due to all etiologies compared with cardiac arrest attributable to anesthesia (LOE 6, good/supporting), the rate of death in the operating room was 34.9% for all etiologies and 5.5% for anesthesia-related arrests.[26] In a large study in humans (LOE 6, good/supporting), the immediate survival after CPA was 46.6% and hospital survival was 34.5% for anesthetic-related CPA.[27] Another study (LOE 6, good/supporting) evaluating fatal and nonfatal CPA related to anesthesia in humans founds that human error was noted in 91% of arrests reported.[28]

Knowledge gaps

Only a small number of quality veterinary studies are available, and larger prospective clinical studies needed. Prospective studies investigating outcomes after CPA due to anesthesia compared with CPA occurring during anesthesia but due to underlying disease would help to elucidate some of the factors leading to the lower mortality rate in this group of patients. Also, immediate post-CPR debriefing to detect (and eventually correct) human error would be beneficial, since 91% of anesthetic arrests had an element of human error in one study.[28] Further studies are needed to determine risk factors for CPA related to anesthesia regimen, anesthetized patient characteristics, and specific anesthesia-related CPR measures that may impact outcomes such as ROSC and discharge from the hospital.

Preparation of the team

Team readiness to respond optimally to a CPA depends on advanced preparation of the personnel that will be conducting the resuscitation as well as specific leadership training. This section examines core elements of training and preparation of the team.

Realistic, high-fidelity training (PRE03)

PICO Question

In veterinary CPR providers (P), does training with realistic techniques (eg, high-fidelity manikins equipped with pulse, chest movement, etc; in situ training) (I), compared with nonrealistic techniques (low-fidelity manikins; class-room training) (C) improve outcome (eg, skill acquisition, skill retention, confidence, ROSC, survival) (O)?


High-fidelity manikins and other feedback devices appear to have some benefit in the initial learning of psychomotor skills and in preventing the decay in the skills needed for prolonged performance of CPR. However, these skill sets begin to deteriorate within weeks of learning regardless of which manikin is used. It is unclear whether the improvement is due to the high fidelity nature of the manikin or the feedback itself.

Summary of the evidence

The AHA CPR Guidelines of 2005 and 2010 have stressed the provision of “high quality” CPR in an attempt to improve survival. High-quality CPR involves both the cognitive performance of resuscitation (performing the steps of CPR in an orderly and rapid fashion) as well as the psychomotor skills associated with CPR. The psychomotor skills associated with high-quality CPR consist of providing chest compressions with proper hand position along with adequate rate and depth, allowing complete chest recoil after each compression, minimizing interruptions in compressions, and avoiding excessive ventilation in both rate and volume. The preponderance of studies suggests that rescuer competency in executing basic and advanced CPR is poor.[29-32] Several studies have documented poor performance of these skills during basic and advanced CPR in the field.[30, 31] Multiple training methods have been assessed in an effort to improve the performance of these psychomotor skills both in the classroom and in the field, targeting lay people, paraprofessional staff, or professional medical staff.

The development of immediate feedback devices, both auditory and visual, have, in general, shown improved initial acquisition of some of the skills involved in high-quality basic and advanced CPR but no single study has shown improvement in all of the psychomotor skills involved in high-quality basic or advanced CPR.[29], [33-37] These immediate feedback devices include metronomic sounds to time compressions,[38] auditory and visual heart rhythm feedback,[33, 39] and programmable auditory feedback devices[40, 41] that provide negative reinforcement if any psychomotor skill is performed outside of preset parameters and positive reinforcement if the programmed skills are performed correctly. These devices may be within the manikin itself, placed between rescuer hands and the victim's chest, or as modules on automated electronic defibrillators (AEDs). These automated devices provide more consistent feedback compared to instructor provided feedback during psychomotor skill training for chest compressions (LOE 6, good/supporting),[42] but may not provide benefit over instructor feedback in ventilation skills (LOE 6, good/opposing).[37] In fact, in some of the studies evaluated, a subset of psychomotor skills was executed with decreased competency after training on high-fidelity manikins (equipped with immediate feedback such as palpable pulses, chest movements, etc) versus low-fidelity manikins lacking this type of interaction using instructor feedback instead.[37, 39] The one psychomotor skill that manikins (both high- and low-fidelity type) consistently fail to teach is proper hand position for the most effective chest compressions.[43-45]

In a swine CPR model that evaluated audio guidance for improvement of compression rate to within acceptable range (LOE 6, fair/supporting), such guidance lead to an increase EtCO2, which has been associated with a higher rate of ROSC.[46]

Student satisfaction was mixed regarding immediate feedback devices. When the devices were incorporated into the manikin, course satisfaction was significantly higher for studies with crossover design. However, in studies that used control groups (LOE 6, fair/opposing), student satisfaction was not significantly different between the instructor-led low-fidelity course and the high-fidelity training.[39] In studies using the CPREzy device (LOE 6, fair/supporting), its design diminished student satisfaction, as the device caused discomfort to the rescuers and increased fatigue.[34, 47]

No studies evaluated whether either high-fidelity or low-fidelity training translated to improved rates of success (improved psychomotor skills, ROSC, survival) in actual CPR for laypersons, paraprofessionals, medical students, or medical professionals.

High-fidelity manikins and feedback devices have some benefit in the initial learning of some of the psychomotor skills involved in basic life support (BLS) over low-fidelity manikins (LOE 6 good/supporting),[31, 33], [48, 49] (LOE 6, fair/supporting).[40, 50, 51] Training on these manikins and devices has been shown (LOE 6, fair/supporting) to prevent the decay in quality of CPR performed during prolonged periods of compressions and ventilations (times greater than 3 min).[35] However, competency in these skills starts to deteriorate within weeks of learning and does so at the same rate whether the student learned on a high-fidelity or low-fidelity manikin. By 1–2 years after training (LOE 6, good/supporting), the student's psychomotor skill set returns to pretraining levels.[29] This degradation in skills is lessened if the follow-up training is also performed with immediate feedback (high-fidelity) (LOE 6, good/supporting)[52] (LOE 6, fair/supporting).[53] Evidence (LOE 6, fair/supporting) suggests that booster training (continued brief practice sessions after the initial training), especially if administered on the high-fidelity manikin, helps to prolong the retention period of psychomotor skills.[54]

Knowledge gaps

The development of veterinary specific feedback devices, including high-fidelity manikins for learning psychomotor skills and a standardized, veterinary-specific didactic CPR curriculum should be evaluated to ensure that this model of training is most effective for teaching high-quality CPR for animals to veterinarians, veterinary staff, and lay persons. If proven effective, studies evaluating the effect of these teaching methods on outcomes in veterinary CPR will be warranted.

Leadership experience (PRE04)

PICO Question

In veterinary CPR teams (P), does a more experienced team leader (board certified, advanced training) (I), compared to a less experienced team leader (house officer, non-boarded clinician) (C), improve outcome (ROSC, survival to discharge) (O)?


There are no randomized, controlled clinical trials exploring this question in either the human or veterinary medical fields. Based on the existing, conflicting human evidence, there is no clear benefit or harm from a more experienced team leader in CPR.

Summary of evidence

The effect of more experienced personnel acting as team leaders in CPR has been investigated in several human studies. In contrast, there is no published evidence in veterinary medicine on whether the presence of a more experienced team leader improves outcome.

Conventional wisdom suggests that a more experienced leader would enable more effective CPR with fewer errors. Support of this assumption in the human literature, however, is inconsistent and there is not a clear consensus on whether code leader experience is important. The majority of evidence originates from studies investigating the benefit of physicians manning ambulances, and comparing CPR quality metrics in the presence and absence of a physician. Studies indicating a positive impact of physicians involved in CPR (LOE 6, good/supporting, LOE 6, fair/supporting) have mostly focused on quality of CPR execution.[55, 56] Two additional studies (LOE 6, poor/supporting) reported a modest improvement in survival with physicians present.[57, 58]

The preponderance of evidence finds no difference with physicians present in either survival of the event (LOE 6, good/supporting; LOE 6, poor/supporting)[55, 57] or survival to discharge (LOE 6, good/supporting),[55] (LOE 6, fair/neutral),[56], [59-61] (LOE 6, good/neutral),[62, 63] (LOE 6, poor/neutral).[64] There are, in fact, a few studies that report worse outcome when physicians are present (LOE 6, poor/neutral),[65] (LOE 6, poor/opposing),[66] (LOE 6, fair/opposing, respectively).[61]

Knowledge gaps

The majority of the research in human CPR shows no impact of experience level of first responder in out-of-hospital cardiac arrest on survival from CPA, but there have been no investigations of this question for in-hospital cardiac arrest and in-hospital team leadership. There are no studies evaluating experience level and CPR outcome in veterinary medicine.

Leadership training and performance (PRE05)

PICO Question

In veterinary CPR teams (P), does leadership training (I), compared with no leadership training (C), improve performance in simulation scenarios (O)?


There are high-quality and high-level human medical studies supporting the beneficial role of leadership training in improving CPR performance in simulation scenarios. However, there are no veterinary simulation studies addressing leadership training and performance during CPR.

Summary of the evidence

Rapid, appropriate intervention by first responders and advanced life support (ALS) providers is an essential component to improve survival rates from CPA. To maximize CPR performance, frequent and repeated training using simulated scenarios or interactive computer instruction is recommended by the AHA to retain technical skills.[67] Despite these recommendations, CPR outcomes remain unsatisfactory. Therefore, in human medicine, teamwork and leadership training have also been investigated. Both may affect adherence to recommended CPR strategies and positively impact performance during and survival from an arrest.

A recent randomized controlled study in humans (LOE 6, good/supporting) assessed the influence and sustained effect of technical versus leadership instructions on the performance of medical students in high-fidelity simulated CPR scenarios.[68] Leadership instructions involve coordinating activities during group interactions and are more managerial and team focused. Technical instructions focus on concrete tasks and on individual performance. Leadership training resulted in better team performance and superior overall results for time until initiation of CPR, hands-on time, and chest compression rate. Technically instructed groups showed more correct arm positions. Therefore, in addition to teaching technical algorithms, rescuers may benefit from CPR team and leadership training. Teams prepared with leadership instruction (LOE 6, good/supporting) demonstrated better communication such as leadership utterances, and started CPR earlier, resulting in a significantly longer uninterrupted hands-on time during the first 3 minutes of the cardiac arrest.[68]

Numerous studies in human medicine have established the association between clear leadership roles and training, and more efficient cooperation and task performance in the team. A prospective study by Marsch et al in 2004 (LOE 6, good/supporting) examined whether human behavior affected the quality of CPR by comparing the response to simulated CPA among groups of health-care workers.[69] During the simulated arrests, almost two-thirds of the teams failed to provide BLS or defibrillation within an appropriate time window. Moreover, the absence of leadership behavior and explicit task distribution were associated with poor team performance, and failure to translate theoretical knowledge into effective team activity was identified as a major problem. Another prospective human simulation study (LOE 6, fair/supporting) evaluated the influence of directive leadership and structuring inquiry on group performance in ad hoc medical teams performing CPR.[70] This study found that with teams of changing composition (nurse to resident to senior doctor), performance was influenced by differences in leadership skills among the participants. In addition, when leaders adopted a coordinating role by monitoring CPR and communicating effectively instead of participating hands-on in the emergency, they were more likely to be efficient leaders, and team performance improved. The effect of leadership training on resuscitation success was not specifically evaluated in this study. However, the results do underscore the importance of role-specific behavior and supports the significance of integrating human factor aspects in team training. The effects of ad hoc team building on team performance and hands-on time during CPR were studied in a simulator based trial (LOE 6, fair/supporting) involving human physicians.[71] Compared to team building prior to the onset of cardiac arrest, ad hoc team formation during an arrest negatively affected hands-on time and time to defibrillation due to several shortcomings in the process of ad hoc team building, particularly deficits in leadership. Early team structuring and leadership competence are prerequisites for timely and effective execution of CPR.

Several additional human studies have assessed the success of leadership programs during CPR. In one (LOE 6, good/supporting), researchers evaluated the effectiveness of a leadership development seminar introduced into the Resuscitation Council (United Kingdom) Advanced Life Support Provider course by making observational assessments of leadership performance during cardiac arrest scenarios before and after the seminar.[72] The leadership training program significantly improved leadership performance in the arrest situation.

A prospective study (LOE 6, good/supporting) comparing two groups of nurses using cardiopulmonary resuscitation-defibrillation (CPR-D) simulators in a ventricular fibrillation arrest model showed that defining and teaching leadership leads to improved resuscitation performance.[73] Finally, a prospective study (LOE 6, poor/supporting) attempted to address the deficiencies in coordination of team resources in Advanced Cardiac Life Support (ACLS) training in order to improve medical emergency team performance.[74] It concluded that training and organizing multidisciplinary teams using simulation technology is effective at improving successful task completion. However, this study did not specifically compare this method of training to leadership training.

The American Heart Association ALS training guidelines suggest that providers acquire and maintain the necessary team behavior to maximize resuscitation outcome.[67] Patient simulation has been suggested to be the ideal tool for teaching these skills. Trainees can engage actively in their learning process while doing no harm to their “patients.” High-fidelity human simulators allow in-depth investigation of complex human interactions using precise, reproducible methods. These remove variability in the clinical parameters of resuscitation scenarios, allowing studies of the impact of human factors and team interactions in the absence of clinical confounders. It remains to be seen, however, if practice-using simulation will consistently translate into improved, real-life CPR success in veterinary medicine.

Knowledge gaps

Studies investigating the benefits of leadership training in veterinary CPR are lacking, but evidence in human medicine supports the benefits of team and leadership training to advance CPR skills and improve outcomes. High-fidelity simulators, needed for accurate assessment of CPR skills during training, coupled with team training and advancing technical skills should be developed and tested in veterinary studies.

Minimum team size (PRE06)

PICO Question

In veterinary CPR providers (BLS and ALS) (P), does a minimum team size (I), compared with no minimum team size (C), improve outcome (O) (eg, ROSC, survival to discharge)?


There is no existing consensus in human or veterinary medicine on optimal team size for resuscitation. There are few reports in the literature, and all are on simulated models or are retrospective studies. Retrospective studies are difficult to evaluate because of differing team size and varied personnel. Both in-hospital and out-of-hospital settings have been evaluated. A consistent observation from these studies is that with growing team size, the principles of BLS are increasingly underemphasized in favor of ALS techniques. This may be detrimental, as several studies in humans have shown that early implementation of high quality BLS is an independent predictor of improved survival in patients with CPA.[75-79]

Summary of the evidence

Team size is an important component in helicopter rescue crew CPR, as these teams are limited due to space constraints and often consist of only two rescuers and a pilot. A high-fidelity manikin study (LOE 6, fair/supporting) was performed on 20 teams of 2 people to assess CPR effectiveness in two-rescuer scenarios.[80] The study concluded that two-rescuer CPR is feasible but that standardized training is essential if team size is small. Configurations of 2-, 3-, or 4-member crews were compared in a study of paramedics (LOE 6, fair/supporting) using a high-fidelity human simulator.[81] No flow fraction and assessment of effectiveness and timeliness of performing BLS and ALS skills (eg, time to defibrillation, endotracheal intubation, establishment of intravenous access, and medication administration) were the primary outcome measures. There was no significant difference in any of the parameters with different crew sizes. The authors surmised that larger groups may become hindered by distractions related to performance of ALS, minimizing the potential improvements that could be achieved in BLS skills. In another simulator study (LOE 6, fair/supporting), teams of two firefighters were compared with teams of three firefighters with manikin generated strip charts used to assess quality of BLS.[82] Teams composed of three firefighters provided significantly greater minute ventilation and a greater depth of chest compressions compared with teams of only two. A retrospective study (LOE 6, fair/neutral) evaluated 4,229 patients treated by 2 paramedics (9% survival to discharge), 4,459 patients treated by 3 paramedics (9% survival to discharge), and 1,369 patients treated by 4 or more paramedics (8% survival to discharge).[83] Multivariate analyses showed the groups treated by crews with more than 2 paramedics had reduced survival to discharge.

Knowledge gaps

In human medicine the effect of team size is unclear and there are no studies evaluating the value of team size in veterinary CPR.

Standardized versus ad hoc training and improved ROSC (PRE07)

PICO Question

In dogs and cats with cardiac arrest (P), does standardized training of veterinarians and technicians in pulseless arrest algorithms (I), compared to ad-hoc training (C), improve ROSC (O)?


High quality, high LOE studies in human medicine support standardized training methods to improve CPR outcome, but no veterinary research has specifically evaluated this question. Comparison of standardized and ad hoc training in human medicine has documented improvement in the rate of initiation of CPR and performance, despite persistently poor overall outcome.

Summary of the evidence

Training in CPR has been recommended for healthcare professionals for more than 3 decades. The ultimate goal of a broad educational strategy is to improve survival from CPA. A survey of academic veterinarians evaluating their clinical practices in CPR (LOE 4, fair/supporting) revealed numerous differences based on institution, gender, specialty, and position.[15] Veterinarians were found to differ significantly in many aspects of their approach to CPA and resuscitation. Trained individuals reported both higher success rates and greater feelings of competence, highlighting the importance of formal training. However, formal ALS training as is available in human medicine is not widely pursued by veterinarians. Instead, didactic, practical, apprenticeship, or a combination of these training methods are most often employed to teach veterinary CPR.[15]

Several human studies have addressed standardized and ad hoc training as it relates to CPR outcome. A retrospective study (LOE 6, poor/supporting) evaluated the efficacy of an ACLS training program for resuscitation from cardiac arrest in a rural hospital.[84] It was concluded from this study that the institution of an ACLS-provider course in a rural community hospital was associated with improved initial resuscitation for patients with out-of-hospital arrest due to ventricular fibrillation. A second retrospective review (LOE 6, poor/supporting) of CPA in a rural hospital setting was performed before, during, and after organization of an ACLS teaching program to evaluate the effect of such training on resuscitation efforts and survival.[85] Widespread ACLS training and code team organization was associated with a significant increase in resuscitation efforts and ROSC despite a slight decline in the percentage of patients surviving resuscitation attempts.

BLS skills were improved significantly in a study (LOE 6, fair/supporting) of human emergency medicine residents undergoing a combination of standardized theoretical and practical training sessions.[86] A randomized controlled trial (LOE 6, good/supporting) showed improved retention of ACLS guidelines for providers using a case-based computer simulation program compared to textbook review alone.[87]

Despite widespread training, adherence to CPR guidelines remains poor. A prospective study of emergency medical technicians (EMTs) (LOE 6, fair/supporting) found that accurate knowledge of guidelines is associated with correct performance of some but not all aspects of CPR (better performance of chest compressions and compression to ventilation ratio).[88]

Knowledge gaps

Several different types of training have been described in the human medical literature, with textbook, simulator, interactive multimedia, and case-based examples. Standardized and ad hoc training studies are few in human medicine and lacking in veterinary medicine. Further prospective clinical trials in human and veterinary medicine are needed to fully assess the impact of standardized training and specifically of ACLS algorithm training, on CPR performance and patient outcome.

Debriefing (PRE09)

PICO Question

In veterinary CPR providers (P), does debriefing after CPR (I), compared with no debriefing (C), improve outcome (O) (eg, CPR performance, ROSC, survival to discharge)?


Overall, the evidence suggests that debriefing is a practical tool that can provide significant improvement in CPR performance if routinely utilized. Debriefing is an easy to apply method that can be used even without the immediate feedback manikin devices that are utilized in human medicine.

Summary of the evidence

The ability to perform high-quality CPR is an important skill that requires training. Competent CPR performance requires skills that can be enhanced by debriefing at no additional risk to the patient. Debriefing involves an assessment of the event, gathering of pertinent information, discussions of team performance, and identification of safety equipment or drug issues immediately or shortly after CPR is concluded. It is difficult to completely isolate the effect of debriefing from other CPR teaching techniques utilized, but, the majority of studies that examine debriefing as a teaching tool support its use. Since all trainees are human regardless of the target species for CPR, these studies may be directly applicable to veterinary providers. Andreatta et al (LOE 6, good) found significant improvement in survival in human pediatric patients (33–50%) over time with the use of mock codes and debriefing integrated into a resident teaching program.[89] The effect of debriefing on medical emergency team performance (LOE 6, good/supporting) was also evaluated using a high-fidelity human patient simulator.[89] A group of ACLS trained nurses, physicians, and respiratory therapists were trained with a web-based presentation and pretest before the course, a brief reinforcing didactic session on the day of the course, and were then tested with simulated CPA scenarios. Each scenario was followed by debriefing and analysis with the team. Patient survival in the simulations increased from 0% to 89%. A comparison study (LOE 6, good/supporting) evaluated CPR simulation training using immediate audiovisual (AV) feedback or debriefing with a combination of both techniques in training of nurses.[90] Significant improvement in compression depth was seen in both single intervention groups. In the AV feedback and debriefing groups, delivery of adequate compression depth increased from 19% of the participants to 58%, and from 38% to 68%, respectively. The combination of AV feedback and debriefing resulted in the largest increase in the percentage of nurses using adequate depth of compressions (45% increased to 84%). This study also found improved compression rates associated with both training methods.

Morgan et al (LOE 6, good) found significant and sustained improvement in anesthetists’ CPR performance during simulation following debriefing.[91] Savoldelli et al (LOE 6, good) reported improved CPR performance after debriefing with or without the addition of video feedback.[92] There was no positive effect in the absence of debriefing. Overall there is good evidence that simulation training with debriefing improves CPR performance and possibly patient outcome.

The findings of several studies were neutral to the specific clinical question of the effect of debriefing on clinical outcome in CPR. Real-time feedback with an investigational monitor/defibrillator with CPR-sensing and feedback capabilities (LOE 6, fair/supporting) improved consistency in chest compression rates and ventilation rates during CPR.[92] However, no change in ROSC or survival to discharge was seen. In a study of weekly post-event debriefing of CPR attempts (LOE 6, fair/neutral) for internal medicine residents in a human hospital, an increased rate of ROSC was achieved, but no improvement was found in survival to discharge.[93] In a prospective study (LOE 6, fair/supporting), 63 lay persons performed CPR on a manikin, received a debriefing on BLS techniques, and then performed two consecutive 3-minute trials of hands-only CPR on a manikin that provided audio and visual feedback.[94] The participants improved to 73% correct compression depth on the manikin during the 3 minutes. In a case report (LOE 6, fair/supporting) of successful resuscitation in a bupivacaine-induced asystolic cardiac arrest, training was thought to have favorably impacted outcome.[51] Two anesthesia providers involved in the resuscitation had recently completed simulation training involving the management of local anesthetic cardiotoxicity. It was determined that the previous simulation training influenced execution of the following steps: rapid problem recognition, prompt initiation of specific ACLS therapy, and coordinated team efforts. In another study (LOE 6, fair/supporting), the investigators sought to determine whether standardized computer-based multimedia instruction is effective for learning, whether the learning is retained 5 weeks later and to compare multimedia instruction to personalized video-assisted oral debriefing with an expert.[95] No difference was found between the groups. Computer-based multimedia instruction is an effective method of teaching nontechnical skills in simulated crisis scenarios and may be as effective as personalized oral debriefing.

Debriefing can take many forms including verbal debriefing, immediate audiovisual feedback, or discussion of video or defibrillator recording. Evidence suggests that debriefing is effective in improving CPR performance (increased compression depth, optimized ventilation, and increased ROSC) and can be applied at no risk to the patient. It is a simple, easily implemented tool that can improve resuscitation team performance. There is no evidence that utilization of a debriefing strategy leads to worsened outcome.

A potential issue in many of these studies is whether automated feedback is equivalent to debriefing. The two are very similar with the exception that the CPR responder can instantly adjust to automated feedback, while the results of debriefing cannot be applied until the next CPR intervention, when adjustments in technique can be implemented. It seems that the two could be treated equally despite this difference in timing of application.

Knowledge gaps

There are no studies evaluating the effect of debriefing on CPR performance or outcomes in veterinary medicine. Debriefing appears to be safe and easily implemented in several different variations, and veterinary studies should be considered.

Post-training skill assessment and learning retention (PRE10)

PICO Question

In veterinary CPR providers (P), does assessment of skills post-training (I), compared with no assessment (C), improve learning retention after training (O)?


A small number of high-quality human studies favored the use of assessment post-training in improving CPR learning retention but no veterinary studies are available.

Summary of the evidence

Few studies have examined the effect of assessment post-training on learning retention of CPR skills. Two studies (LOE 6, good/supporting)[96] found a positive effect on learning retention 2 weeks after instruction and training, and 1 study (LOE 6, good/neutral)[97],[98] found a trend for improved learning retention 6 months after instruction. The model used in these studies was similar and consisted of CPR training at a 4-hour in-hospital resuscitation course for medical students. The students were randomized into a control and intervention group. The control group received 4 hours of instruction and training, whereas the intervention group received 3.5 hours of instruction and training followed by 30 minutes of testing. For the first study, the participant's learning outcome was assessed 2 weeks after the course in a simulated scenario using a checklist.[99] Learning outcomes were superior in the intervention group compared with the control group. This study concluded that testing as a final element of resuscitation skills training courses increases learning outcome compared to skill practice alone.

A later study by the same research group (LOE 6, good/supporting) repeated the same design but also measured salivary cortisol levels pre-instruction course, half an hour before the end of the course and post-instruction course.[97] The investigators compared the learning outcomes and cortisol responses between groups and genders. They found that a rise in cortisol was noted in men but not in women, which was associated with better retention of CPR skills by the men. The third study by this group evaluated learning outcomes 6 months after the instruction and training course and found a small positive effect in the testing group, although it was not statistically significant.[98]

There are only a small number of studies investigating this question, and all of the research was performed by the same investigators, limiting diversity in study design and execution. However, the relevant human studies favor assessment of post-training skills as part of CPR learning retention.

Knowledge gaps

There are too few human and no veterinary studies on the question of post-training skills assessment and quality and duration of CPR knowledge and skill retention. Further studies are needed by different research groups and in the setting of veterinary medicine.


Although there is a dearth of quality resuscitation research in veterinary medicine regarding prevention and preparedness for CPR, it seems prudent to adapt some practices that are backed by high-quality research from humans or animal models. Because the interventions examined by these PICO questions are focused on improving the performance of the people performing CPR rather than on the physiologic response of the patient, it could be argued that the studies from human medicine are more directly applicable to the development of veterinary CPR guidelines for this domain than for the others in the RECOVER initiative.

The use of pre-stocked arrest stations and cognitive aids may improve adherence to established CPR protocols. In light of the higher overall survival rate in dogs and cats following CPA while under anesthesia, particular consideration should be given to placement of arrest stations and algorithms in areas where animals are routinely anesthetized (eg, anesthesia induction and recovery area, surgical suites, surgical teaching laboratories, etc)

There is strong evidence supporting improvements in CPR performance and adherence to guidelines in human medicine associated with standardized training and the use of high-fidelity manikins for development of psychomotor skills. This suggests that development of standardized veterinary resuscitation education and veterinary specific high-fidelity manikins for teaching psychomotor skills are worthwhile goals. Feedback mechanisms, debriefing, and retraining at appropriate intervals are all areas in need of veterinary specific research and development.

Borrowing from the extensive human literature on CPR prevention and preparedness, we as veterinary health professionals can work toward ensuring optimal translation of the science of resuscitation to veterinary medicine. Through the development of evidence-based CPR guidelines, standardized training methods documented to be effective at teaching the principles and psychomotor skills of CPR and cognitive aids to improve adherence to those guidelines during an arrest, we have the potential to improve clinical outcomes for our patients. Ultimately, meticulously designed studies will be needed to evaluate the potential benefits of these approaches and to refine these techniques to ensure we are providing the best possible care.


The authors would like to thank the American College of Veterinary Emergency and Critical Care (ACVECC) and the Veterinary Emergency and Critical Care Society for their financial and scientific support, as well as Armelle deLaforcade, the ACVECC Executive Secretary and Kathleen Liard, ACVECC Staff Assistant for their administrative and organizational support. Also, we would like to thank the RECOVER Advisory Board for their guidance and invaluable input during the planning and execution of this initiative: Dennis Burkett, ACVECC Past-President; Gary Stamp, VECCS Executive Director; Daniel Chan, JVECC Liaison; Elisa Mazzaferro, Private Practice Liaison; Vinay Nadkarni, ILCOR Liaison; Erika Pratt, Industry Liaison; Andrea Steele, AVECCT Liaison; Janet Olson, Animal Rescue Liaison; Joris Robben, EVECCS Liaison; Kenneth Drobatz, ACVECC Expert; William W. Muir, ACVECC and ACVA Expert; Erik Hofmeister, ACVA Expert. Finally, we would like to thank the many members of the veterinary community who provided input on the RECOVER guidelines at the IVECCS 2011 session and during the open comment period via the RECOVER website.


RECOVER Preparedness and Prevention Domain Worksheet Authors

Collaborators listed in alphabetical order:

Barr, J; Corsi, R; Holahan, M; Levinson, J; Mensack, S; O'Brien, M; Rudloff, E; Walters, P.