The first time I ever performed CPR was on my 19th birthday. My official title was “summer employee,” a minimum wage job that encompassed a variety of menial but necessary tasks in the emergency department of a busy rural hospital. One of the accompanying benefits of my job was the chance to learn CPR, and by the summer’s end, by my birthday, I had finally received my CPR certification.
Now, 12 years later, some details are a little hazy; but I remember that the patient was in her 80s and had gone into cardiac arrest at home. The dispatched emergency medical services team had been providing chest compressions for 14 minutes by the time she arrived at the emergency department. I remember how quickly my arms tired, despite an adrenaline surge and my racing heart as I watched my compressions on the telemetry monitor, seeing that the force provided was adequate. But, most of all, I remember feeling sad and defeated when we could not save her. Afterward, the emergency physician tried to comfort me, telling me that our efforts were likely futile from the start– that most patients who come into the hospital after an outside cardiac arrest do not survive.
Since that fateful afternoon, I have performed CPR many more times, usually with similar disappointing outcomes. CPR guidelines have changed twice since my initial certification, and now focus on fast, hard compressions, often suggested, ironically, to be given to the beat of “Stayin’ Alive.” However, despite our widespread CPR training, patients who develop cardiac arrest outside the hospital generally continue to have poor outcomes, spurring debate on the utility and value of CPR training for non-medical professionals.
This debate is the subject of two articles by a group from the Center for Resuscitation Science in Sweden that appear in this week’s NEJM that examine the potential benefits of early out-of-hospital bystander CPR in cardiac arrest. In the first study by Hasselqvist-Ax et al., the authors used the Swedish Cardiac Arrest Registry in a retrospective analysis of outcomes from over 30,000 cases of witnessed cardiac arrest that occurred from 1990 to 2011. About half of all cardiac arrest patients received out-of-hospital CPR by a bystander, while the other half did not.
Although survival overall was poor at 30 days, the survival for those who received out-of-hospital CPR was more than double that of those who did not (10.5% vs 4.0%, p<0.001). These findings remained robust when the authors controlled for multiple potential confounding factors such as sex, age, cardiac etiology, initial heart rhythm, and location of cardiac arrest (in the home versus in a public place– Table 2 in the original article). There was, however, in preplanned subgroup analysis, a positive interaction for male sex and arrest outside the home- these 2 subgroups had a more marked benefit from out-of-hospital CPR compared to other subgroups (in contrast to my first CPR patient).
There are a few characteristics of interest in the out-of-hospital CPR group. First, men outnumbered women in the out-of-hospital group, and were more likely to have ventricular fibrillation as their initial ECG rhythm. The out-of-hospital CPR group also had a shorter collapse-to-call time for EMS (by 1 minute), indicating that people trained in CPR appear to recognize a medical emergency faster and therefore call emergency services sooner. Despite calling emergency services faster, however, there was a delay seen in time to EMS arrival after the call (difference of 2 minutes) as well as a delay in time from collapse to initiation of defibrillation (2 minutes). Thus, early initiation of out-of-hospital CPR leads to improved survival benefits sufficiently robust to outweigh a delay in arrival of emergency medical services and a delay in initiation of defibrillation.
The second study, by Svensson and colleagues, evaluated the utility of a mobile phone positioning system to dispatch CPR-trained lay volunteers to out-of-hospital cardiac arrests in an effort to increase the rate of out-of-hospital CPR actually delivered.
Svensson et al. performed a community-based randomized controlled trial in the county of Stockholm, Sweden from April 2012 to December 2013. 5,989 CPR-trained lay volunteers were recruited at the beginning of the study. When an out-of-hospital cardiac arrest was called into the dispatcher between 6am and 11pm, the mobile phone positioning system was launched to locate any trained lay volunteers within a 500-meters of the incident. Randomization was performed in a 1:1 ratio by the mobile phone positioning system. In cases assigned to the intervention group, a voice call, text message and web link to a map showing the location was then sent out to any volunteers in a 500-meter radius. The dispatcher and all investigators were blinded until the final analysis was completed. Primary outcome was initiation of bystander CPR before arrival of EMS. Secondary outcomes included return of spontaneous circulation and 30-day survival.
There were 667 out-of-hospital cardiac arrests randomized and evaluated (46% in the intervention group and 54% in the control group). Baseline characteristics were similar for the two groups. 61.6% of patients randomized to the intervention group (utilization of mobile phone dispatch system) received bystander CPR compared to 47.8% of those in the control group, a difference of about 14% (p<0.001). However, about 700 out-of-hospital cardiac arrests were not randomized by the dispatcher, which would reduce the effect of the intervention by about half, resulting in only a 7% difference when the whole population of persons having an out-of-hospital cardiac arrest is considered.
Despite the increase in bystander CPR in the intervention group, there was no significant difference in return of spontaneous circulation or 30-day survival between the two groups, which the authors postulate is because the trial was not adequately powered for those outcomes. Additionally, only 65% of lay volunteers contacted by the mobile phone dispatch attempted to reach the scene. And even when lay volunteers reached the scene, only 13% of all cases for whom calls were dispatched had CPR initiated before trained personnel arrived (see Table 3 from study).
The technology to dispatch trained laypersons to a nearby cardiac arrest victim may have potential to increase bystander CPR rates, if implemented on a wider scale; but many obstacles may be encountered, including legal, social and economic factors. One might speculate that greatest benefit may occur in rural communities, where there may be considerable delay until emergency medical services can reach the scene or in senior communities where many cardiac arrests events occur at home where CPR-trained and capable lay persons are not usually around. Maybe my first patient 12 years ago would have benefited from such a system, or at least have had a slightly increased chance in survival once she got to the hospital under the care of the emergency room team, and my 19-year-old hands.
The authors of these studies are available through June 19th to answer your questions on the NEJM Group Open Forum.