A randomized controlled trial comparing traditional training in cardiopulmonary resuscitation (CPR) to self-directed CPR learning in first year medical students: The two-person CPR study

Study Aim

The primary purpose of this study was to compare two, shorter, self-directed methods of cardiopulmonary resuscitation (CPR) education for healthcare professionals (HCP) to traditional training with a focus on the trainee's ability to perform two-person CPR.

Methods

First-year medical students with either no prior CPR for HCP experience or prior training greater than 5 years were randomized to complete one of three courses: 1) HeartCode BLS System, 2) BLS Anytime, or 3) Traditional training. Only data from the adult CPR skills testing station was reviewed via video recording by certified CPR instructors and the Laerdal PC Skill Reporter software program (Laerdal Medical, Stavanger, Norway).

Results

There were 180 first-year medical students who met inclusion criteria: 68 were HeartCode BLS System, 53 BLS Anytime group, and 59 traditional group Regarding two-person CPR, 57 (84%) of Heartcode BLS students and 43 (81%) of BLS Anytime students were able to initiate the switch compared to 39 (66%) of traditional course students (p = 0.04). There were no significant differences in the quality of chest compressions or ventilations between the three groups. There was a trend for a much higher CPR skills testing pass rate for the traditional course students. However, failure to “clear to analyze or shock” while using the AED was the most common reason for failure in all groups.

Conclusion

The self-directed learning groups not only had a high level of success in initiating the “switch” to two-person CPR, but were not significantly different from students who completed traditional training.     

Poor quality teaching in lay person CPR courses

Recent studies have found that poor cardiopulmonary resuscitation (CPR) is commonly performed in resuscitation attempts, both by health professionals and lay people. One of the contributing factors to poor performance of CPR may be poor initial teaching. This study was conducted to investigate the quality of 14 CPR courses complying with New Zealand Qualifications Authority standards, which includes formal assessment of CPR. While courses taught by the large first aid training organisations in New Zealand had a student to manikin ratio of around 3:1, courses taught by smaller providers had a ratio of over 4:1. During the 4h course, only 20+/-2 min were spent demonstrating CPR, and 26+/-4 min were spent with students practising CPR. The assessment of adult, child and infant CPR took on average less than 2.5 min in total. Importantly, in the majority of courses (71%), certification was granted when the CPR technique was performed incorrectly, with both compression depth and compression place being corrected only 57% of the time. Courses only discussed the importance of early defibrillation 57% of the time, and provided limited information on symptoms of acute coronary syndromes. In light of these observations it is suggested that the current style of teaching is unlikely to result in students being able to perform adequate CPR if required in the community.     

Public access resuscitation program including defibrillator training for laypersons: a randomized trial to evaluate the impact of training course duration

BACKGROUND: Time to cardiopulmonary resuscitation (CPR) is a main determinant of survival after out-of-hospital cardiac arrest. Only widespread implementation of training courses for laypersons can decrease response time. METHODS AND RESULTS: In this prospective randomized trial, we evaluated how laypersons retained CPR skills and skills in using the automated external defibrillator (AED). A total of 1095 volunteers were randomly assigned to receive CPR/AED-training courses of 2h (375 persons), 4h (378 persons) or 7h (342 persons) duration. Courses were held in accordance with the guidelines for CPR. All trainees were tested immediately after the initial class in a standardized test scenario using an AED and a manikin. Either at 6 or at 12 months, retests were given to 164 and 206 volunteers, respectively. In 479 volunteers, retesting was completed at both 6- and 12-month intervals. At the immediate tests, the 7-h training group showed a slightly higher rate of correct responses (7h: 96%, 4h: 94%, 2h: 92%) (p<0.001). Skill retention decreased significantly in the three groups and was lowest after 12 months if no 6-month retests were done. In trainees who did undergo retesting at 6 months, skills did not deteriorate at 12 months. There were no significant differences between the three groups (overall correct responses: 2h: 72%, 4h: 73%, 7h: 74%) (ns). CONCLUSIONS: A 2-h class is sufficient to acquire and retain CPR and AED skills for an extended time period provided that a brief re-evaluation is performed after 6 months.     

Performance of supraglottic airway devices and 12 month skill retention: a randomized controlled study with manikins

Purpose

Airway management for successful ventilation by laypersons and inexperienced healthcare providers is difficult to achieve. Bag-valve mask (BVM) ventilation requires extensive training and is performed poorly. Supraglottic airway devices (SADs) have been successfully introduced to clinical resuscitation practice as an alternative. We evaluated recently introduced (i-gel™ and LMA-Supreme™) and established SADs (LMA-Unique™, LMA-ProSeal™) and BVM used by laypeople in training sessions on manikins.

Methods

In this randomized controlled study, 267 third-year medical students participated with informed consent and IRB approval. After brief standardized training, each participant applied all devices in a randomized order. Success of device application and ventilation was recorded. Without further training, skill retention was assessed in the same manner 12 months later. Outcome parameters were the number of application attempts, application time, tidal volume and gastric inflation rate recorded at successful attempts, and subjective ease-of-use rating by the participants.

Results

i-gel™ and LMA-Supreme™ were the most successful in the first attempt at both assessments and in the subjective ease-of-use rating. The shortest application time was found with BVM (8 ± 5 s in 2008 vs. 9 ± 5 s in 2009) and i-gel (10 ± 3 s vs. 12 ± 5 s). Tidal volumes were disappointing with no device reaching 50% volume within the recommended range (0.4-0.6 L). Gastric inflation rate was highest with BVM (18% vs. 20%) but significantly lower with all SADs (0.4-6%; p < 0.001 for 2008 and 2009).

Conclusion

SADs showed clear advantages over BVM. Compared with LMA-Unique™ and LMA-ProSeal™, i-gel™ and LMA-Supreme™ led to higher first-attempt success rates and a shorter application time.     

Training to deeper compression depth reduces shallow compressions after six months in a manikin model

Introduction

Studies show that students, trained to perform compressions between 40 and 50 mm deep, often do not achieve sufficient depth at retention testing. We hypothesized that training to achieve depths >50 mm would decrease the proportion of students with depth <40 mm after 6 months, compared to students trained to a depth interval of 40–50 mm.

Methods

A basic life support (BLS) self-learning station was attended by 190 third year medicine students. They were first offered the possibility to refresh their skills, following the instructions of a 15 min abbreviated Mini Anne™ video (Laerdal, Norway) using a full size torso and a face shield. This was followed by further training using Resusci Anne Skills Station™ software (Laerdal, Norway). Voice feedback was provided according to randomisation to a standard group (SG) 40–50 mm and a deeper group (DG) >50 mm. Quality of compressions was tested after 6 months.

Results

The SG and DG groups consisted of 90 (67% female) and 100 (58% female) participants respectively. At the end of training, all students reached the target depth without overlap between groups. After 6 months, the proportion of students achieving a depth <40 mm was 26/89 (29%) in the SG vs. 12/89 (14%) in the DG (P = 0.01). The proportion of students with a depth >50 mm was 5/89 (6%) for the SG and 44/89 (49%) in the DG (P < 0.001).

Conclusions

The educational strategy to train students to a deeper depth, reduced shallow compressions 6 months after training.     

Retraining basic life support skills using video,voice feedback or both: A randomised controlled trial

Introduction

The optimal strategy to retrain basic life support (BLS) skills on a manikin is unknown. We analysed the differential impact of a video (video group, VG), voice feedback (VFG), or a serial combination of both (combined group, CG) on BLS skills in a self-learning (SL) environment.

Methods

Two hundred and thirteen medicine students were randomly assigned to a VG, a VFG and a CG. The VG refreshed the skills with a practice-while-watching video (abbreviated Mini Anne™ video, Laerdal, Norway) and a manikin, the VFG with a computer-guided manikin (Resusci Anne Skills Station™, Laerdal, Norway) and the CG with a serial combination of both. Each student performed two sequences of 60 compressions, 12 ventilations and three complete cycles of BLS (30:2). The proportions of students achieving adequate skills were analysed using generalised estimating equations analysis, taking into account pre-test results and training strategy.

Results

Complete datasets were obtained from 192 students (60 VG, 69 VFG and 63 CG). Before and after training, ≥70% of compressions with depth ≥50 mm were achieved by 14/60 (23%) vs. 16/60 (27%) VG, 24/69 (35%) vs. 50/69 (73%) VFG and 19/63 (30%) vs. 41/63 (65%) CG (P < 0.001). Compression rate 100–120/min was present in 27/60 (45%) vs. 52/60 (87%) VG, 28/69 (41%) vs. 44/69 (64%) VFG and 27/63 (43%) vs. 42/63 (67%) CG (P = 0.05). Achievement of ≥70% ventilations with a volume 400–1000 ml was present in 29/60 (49%) vs. 32/60 (53%) VG, 32/69 (46%) vs. 52/69 (75%) VFG and 25/63 (40%) vs. 51/63 (81%) CG (P = 0.001). There was no between-groups difference for complete release.

Conclusions

Voice feedback and a sequential combination of video and voice feedback are both effective strategies to refresh BLS skills in a SL station. Video training alone only improved compression rate. None of the three strategies resulted in an improvement of complete release.     

Effects of monthly practice on nursing students' CPR psychomotor skill performance

Aims

The study examined the effects of brief monthly practice on nursing students’ CPR psychomotor skill performance at 3, 6, 9, and 12 months compared to a control group with no practice, and of repeating the initial BLS course at 12 months.

Methods

Nursing students (n = 606) completed either HeartCode™ BLS or an instructor-led course and were then randomly assigned to an intervention group practice schedule, consisting of experimental (6 min of monthly practice on a voice advisory manikin) or control (no practice) and test out month. Every 3 months, a subset of students was randomly selected from both groups for reassessment of their CPR psychomotor skills. Outcome measures were compression rate and depth, percent of compressions performed with adequate depth, percent performed with correct hand placement, ventilation rate and volume, and percent of ventilations with adequate volume.

Results

At 3 months, there were no differences between the groups in mean ventilation volume (p = 0.71), but with practice by 6 months students were able to ventilate with an adequate volume; this skill continued to improve with monthly practice. In the control group, the mean ventilation volumes were less than the recommended minimum throughout the 12 months. The control group had a significant loss of ability to compress with adequate depth between 9 and 12 months (p = 0.004). By practicing only 6 min a month, students maintained or improved their CPR skills over the 12-month period.

Conclusion

The findings confirmed the importance of practicing CPR psychomotor skills to retain them and also revealed that short monthly practices could improve skills over baseline.     

A manually powered mechanical resuscitation device used by a single rescuer: a randomised controlled manikin study

Purpose

The goal of this randomized, open, controlled crossover manikin study was to compare the performance of “Animax”, a manually operated hand-powered mechanical resuscitation device (MRD) to standard single rescuer basic life support (BLS).

Methods

Following training, 80 medical students performed either standard BLS or used an MRD for 12 min in random order. We compared the quality of chest compressions (effective compressions, compression depth and rate, absolute hands-off time, hand position, decompression), and of ventilation including the number of gastric inflations. An effective compression was defined as a compression performed with correct depth, hand position and decompression.

Results

The use of the MRD resulted in a significantly higher number of effective compressions compared to standard BLS (67 ± 34 vs. 41 ± 34%, p < 0.001). In a comparison with standard BLS, the use of the MRD resulted in less absolute hands-off time (264 ± 57 vs. 79 ± 40 s, p < 0.001) and in a higher minute-volume (1.86 ± 0.7 vs. 1.62 ± 0.7 l, p = 0.020). However, ventilation volumes were below the 2005 ERC guidelines for both methods. Gastric inflations occurred only in 0 ± 0.1% with the MRD compared to 3 ± 7% during standard BLS (p < 0.001).

Conclusion

Single rescuer cardio-pulmonary resuscitation with the manually operated MRD was superior to standard BLS regarding chest compressions in this simulation study. The MRD delivered a higher minute-volume but did not achieve the recommended minimal volume. Further clinical studies are needed to test the MRD's safety and efficacy in patients.     

心脏骤停后心肺复苏和心肺脑复苏成功病例的对比分析

目的 探讨影响心脏骤停患者成功脑复苏的相关因素.方法 回顾对比分析心脏骤停后成功心肺脑复苏(A组,n=38)和仅心肺复苏成功(B组,n=42)患者之间的相关指标,包括性别、年龄、原发疾病、心脏骤停原因、心脏骤停环境、心脏骤停相关时间和心肺复苏后相关治疗持续时间.结果 两组性别比和平均年龄比较差异无统计学意义(P＞0.05).原发疾病:A组以外科为主(78.9%),B组以内科为主(61.9%),两组比较差异有统计学意义(P＜0.005).心脏骤停原因:A组31例(81.6%)为急性缺氧、低血压、内脏神经反射和单纯心脏疾患, B组30例(71.4%)为慢性缺氧和慢性心脏病,两组比较差异有统计学意义(P＜0.005).心脏骤停环境:A组24例(63.2%)发生在手术室和ICU,B组22例(52.4%)发生在普通病房,两组比较差异有统计学意义(P＜0.005).心脏骤停相关时间:A组心脏骤停持续时间(8.2±8.7)min,自主心跳恢复时间(6.7±8.4)min,脑缺血缺氧时间(1.5±1.3)min,均明显短于B组[分别为(30.8±26.2)min、(27.7±24.9)min和(3.1±3.1)min,P＜0.001或P＜0.005].心肺复苏后相关治疗持续时间:A组亚低温持续时间(4.0±2.6)d,呼吸机持续时间(11.1±19.7)d,与B组[(5.9±3.8)d和(15.4±29.3)d]比较差异无统计学意义(P＞0.05).Logistic多因素回归分析显示,原发疾病(OR=6.22,95%CI 1.64～23.46)、心脏骤停持续时间(OR=1.11,95%CI 1.04～1.19)和心脏骤停发生环境(OR=4.51,95%CI 1.22～16.61)与成功脑复苏的关系更密切,成为三个独立影响因素.结论 没有明显慢性疾病,在手术室和ICU以急性缺氧、低血压和单纯心脏原因发生的心脏骤停,抢救及时有效,复苏后处理恰当、合理,尽早实施全面脑保护是成功脑复苏的有利因素.     

Nationwide survey of resuscitation education in Finland

Aims

Good-quality cardiopulmonary resuscitation (CPR) is highlighted in the International Resuscitation Guidelines, but clinically the quality of CPR is often poor. Education of CPR has a major role in the primary skills imparted to students. Different methods can be used to teach CPR quality. We evaluated the current status of their usage in Finland institutes teaching students of emergency medicine at different levels.

Methods

The following institutes were included in an anonymous survey: medical schools (teaching future physicians), universities of applied sciences (paramedics), colleges (emergency medical technicians) and emergency services college (fire-fighters). Hours of teaching theory lessons of CPR and hours of small group training were evaluated. In particular, we focussed on the teaching methods for adequate chest compression rate and depth.

Results

Twenty-one of 30 institutes responded to the questionnaire. The median for hours of theory lessons of CPR was 8 h (range: 2–28 h). The median for hours of small group training was 10 (range: 3–40 h). The methods of teaching adequate chest compression rate were instructors’ visual estimation in 28.5% of the institutions, watch in 33.3%, metronome in 9.5% and manikins’ graphic in 28.5% of institutions. The methods of teaching adequate chest compression depth were instructors’ visual estimation in 33.3%, in manikins light indicators in 23.8% and manikins’ graphics in 52.3% of institutions.

Conclusion

The hours of theoretic lessons and small group training vary widely among different institutes. In one-third of institutions, the instructor's visual estimation was a sole method used to teach adequate chest compression rate and depth. Different technical methods were surprisingly seldom used.     

The retention of cardiopulmonary resuscitation skills

Although a great emphasis has recently been placed on training both the medical profession and the general public in cardiopulmonary resuscitation (CPR), studies have demonstrated that retention of resuscitation skills is poor. Although CPR certification is generally valid for a 1- to 2-year period, evaluation of trainees at all levels has demonstrated a marked lack of proficiency over this course of time. This paper reviews the studies that have disclosed this lack of skills retention, as well as proposed solutions and reinforcement techniques. CPR course content and certification criteria must be appropriate to maximize retention as well as learning. To this end a simplification of basic life-support training curricula is recommended.     

Changes in internal medicine residents' attitudes about resuscitation after cardiac arrest over a decade

Background

Resident physicians' beliefs about cardiopulmonary resuscitation (CPR) may impact their communication with patients about end-of-life care. We sought to understand how these perceptions and experiences have changed in the past decade because both medical education and American society have focused more on this domain.

Method

We surveyed 2 internal medicine resident cohorts at a large academic medical center in 1995 and 2005. Residents were asked of their beliefs about survival after CPR, perceived patient understanding, and regret after attempted resuscitation. Residents in 2005 reported more numerical experience with CPR. Current internal medicine residents are more optimistic than the 1995 cohort about survival after an inpatient cardiac arrest. They believe that far fewer patients and families understand resuscitation but report less regret about attempting to resuscitate patients.

Conclusions

These pilot data reveal potential changes in the attitudes of resident physicians toward CPR. The perceived poor understanding among decision makers calls into question the standard of informed consent. Despite this, residents report less regret leading one to ask what factors may underlie this response.     

甲泼尼龙对心肺复苏后患者血清细胞因子的影响

目的 探讨甲泼尼龙对心肺复苏后患者血清细胞因子表达的影响.方法 将2005年5月至2007年5月上海市闸北区中心医院心肺复苏恢复自主循环(ROSC)30例患者,经心电监护或心电图证实院内心跳呼吸骤停,即刻心肺复苏恢复自主循环、生存≥48 h、年龄≥18岁患者;入选病例排除因各种疾病终末期、晚期肿瘤、自然死亡因素所致心跳呼吸停止者,及发病前一周内伴有各种感染、休克、严重创伤者.随机分为A组(甲泼尼龙组,n=14):复苏后加用甲泼尼龙每日3 mg/kg,分两次静脉滴注,连续3 d;B组(对照组,n=16):采用常规心肺复苏治疗.两组患者原发病因基本相仿.ELISA法检测两组心肺复苏即刻、ROSC后24 h、48 h、72 h、7 d的血清肿瘤坏死因子-α(TNF-α)、白细胞介素-1β(IL-1β)、白细胞介素-6(IL-6)、白细胞介素-8(IL-8)、白细胞介素-10(IL-10)水平.资料数据采用SPSS11.5版统计软件进行分析处理,计量资料以均数±标准差(-x±s)表示,两组各不同时间点计量资料两两比较采用成组t检验,两组心肺复苏后SIRS患病率和病死率采用Chi-square test,以P＜0.05为差异有统计学意义.结果 两组心跳骤停至心肺复苏恢复自主循环平均间期[(8.9±4.6)min,(9.6±5.0)min]及心肺复苏即刻血清TNF-α、IL-1β、IL-6、IL-8、IL-10水平差异无统计学意义(P＞0.05).与B组比较,A组ROSC后24 h、48 h血清TNF-α、IL-1β、IL-6、IL-8水平明显降低(P＜0.05～0.01),ROSC后72 h A组血清IL-8水平(114.33±149.72)仍低于B组(332.09±277.45)(P＜0.05),ROSC后7 d两组血清各细胞因子水平差异无统计学意义(P＞0.05).血清IL-10水平两组在不同时间点差异均无统计学意义(P＞0.05).结论 早期应用甲泼尼龙可降低心肺复苏后患者血清TNF-α、IL-1β、IL-6、IL-8释放,对复苏患者有保护作用.     

Training rates and willingness to perform CPR in King County, Washington: a community survey

Background

It has been hypothesized that high rates of cardiopulmonary resuscitation (CPR) training in a community will lead to improved survival for out-of-hospital cardiac arrest. However, factors to consider when designing a far-reaching community CPR training program are not well defined. We explored factors associated with receiving CPR training in the survey community and characteristics contributing to willingness to perform CPR in an emergency.

Methods

A telephone survey was administered to 1001 randomly selected residents in September 2008 assessing CPR training history, demographics, and willingness to perform CPR. Characteristics of survey respondents were compared to examine factors that may be associated with reports of being trained compared to reports of never being trained. A stratified analysis compared characteristics of respondents who reported a high level of willingness to perform CPR in those trained compared to those never trained.

Results

The survey response rate was 39%. Seventy-nine percent of survey respondents reported ever attending a CPR training class. A majority of people (53%) attended their most recent class more than five years ago. People who had never been trained in CPR were older, were more likely to be men and were less likely to have at least a 2-year college degree than those who had ever been trained. Among those who had been trained, younger age, male gender, time of last training and number of times trained were all significantly associated with willingness to perform CPR and none of these factors were associated with willingness in those who had not been trained.

Conclusions

Retraining rates, methods for reaching underserved populations and measures that will improve the likelihood that bystanders will perform CPR in an emergency should be considered when designing a community CPR education program.     

Neonatal resuscitation--an analysis of the transatlantic divide

AIM: To highlight the main differences between the current editions of the Newborn Life Support (NLS; Resuscitation Council, UK) and the Neonatal Resuscitation Program (NRP; American Academy of Pediatrics and American Heart Association), and to analyse differences between the evidence underlying NLS and NRP. MATERIAL AND METHODS: We undertook a detailed comparison of recommendations and references, based on the NLS and the NRP provider course manuals issued in 2006. Literature on neonatal resuscitation, published in 2005 and thereafter, was searched, focusing on controversies between NLS and NRP. RESULTS: A multitude of important differences between NLS and NRP have been reaffirmed in their current editions, leading to conflicting messages regarding many aspects of resuscitation. An incongruent selection of evidence appears to be a major factor accounting for this divergence. CONCLUSION: To avoid confusion among health care providers and to support the credibility of both NLS and NRP, an intensified dialogue and a more congruent evidence base between NRP and NLS is required. Mutual recognition of equivalency appears unrealistic until substantial progress in this direction has been achieved.     

Comparison of ventilation and chest compression performance by bystanders using the Impact Model 730 ventilator and a conventional bag valve with mask in a model of adult cardiopulmonary arrest

"Bystanders" or lay persons are typically the first caregivers to attend to a victim of out-of-hospital cardiopulmonary arrest. Astronaut crew medical officers (CMO) play a similar role to bystanders aboard the International Space Station (ISS). Studies have demonstrated the importance of bystander cardiopulmonary resuscitation (BCPR) for patient survival before the arrival of emergency medical care. Recent apprehension from bystanders about the threat of contracting communicable diseases during BCPR, however, has led to the consideration of other ventilation systems such as the bag-valve mask (BVM) and automatic transport ventilators (ATV). BVM use is called for during CPR aboard the ISS. This study evaluated the ventilation and compression performance of 40 basic CPR-trained bystanders using either a BVM (adult-sized self-inflating bag with face mask) or an ATV (Model 730 ventilator (M730), Impact Instrumentation, Inc., West Caldwell, NJ). Each two-bystander team gave BCPR to a simulated cardiopulmonary arrest victim using the 2-breath/15-compression cycle for 4 min and then switched roles for another 4-min interval. Compared to BVM use, the M730 led to significantly (p<0.05) lower number of breaths, smaller tidal volumes, airway flows, airway pressures, volume of gas entering the stomach per breath and chest compressions for the 4-min period. The M730 also enabled a bystander to meet the recommendation of 4-breath and compression cycles per minute as per Guidelines 2000. Lastly, ease-of-use scores were significantly higher for the M730 compared to the BVM. Overall, the data suggest that the M730 improves the quality of performance for a bystander performing BCPR.     

Epidemiology, trends, and outcome of out-of-hospital cardiac arrest of non-cardiac origin

INTRODUCTION: The majority of victims who experience out-of-hospital cardiac arrest (OHCA) have ventricular fibrillation (VF) as the presenting rhythm and are thought to have a cardiac etiology for their arrest. Over the past decade, the incidence of VF OHCA has declined. The aims of this study were to describe the epidemiology of OHCA of non-cardiac origin in Olmsted County MN and to determine the trends that have occurred over time. METHODS: All residents with a traumatic OHCA from 1995 to 2005 were included for analysis. OHCA data were collected prospectively according to the Utstein method. Cardiac arrests were classified as cardiac or non-cardiac in origin and the etiology determined based on autopsy reports, electronic medical records, and/or emergency medical services reports. RESULTS: During the study period, 414 OHCAs were identified, 90 (21.7%) of which were classified as non-cardiac. Mean age was 61.5+/-19.7 years. Response time was 7.73+/-2.9 min, and 40 (44.4%) were bystander-witnessed. Sixty-eight (75.6%) arrests occurred at home, 13 (14.4%) in a public place, and 9 (10%) in other locations. Bystander CPR was performed in 17 (18.9%) cases. The presenting rhythm was VF in 2 (2.2%) cases, PEA in 54 (60%), and asystole in 34 (37.8%). Eight (8.9%) patients survived to hospital discharge. Respiratory failure (35.6%), unknown (15.6%), and pulmonary embolism (13.3%) were the most common etiologies. The mean percentage of arrests due to a non-cardiac cause in three sequential time-periods (1995-1999, 2000-2002, 2003-2005) was 9.4%, 20.1% and 37.7%, respectively. CONCLUSIONS: Over the study period, 21.7% of OHCAs were non-cardiac in origin. PEA was the most common presenting rhythm and respiratory failure the most common etiology. 8.9% of patients survived. The decreasing number of VF arrests may be a contributing factor to the increasing proportion of OHCAs of non-cardiac etiology observed in the out-of-hospital setting.