Dispatcher Instruction in CPR (pediatrics)
Citation
Tijssen JA, Aickin RP, Atkins D, Bingham R, Couto TB, de Caen AR, Guerguerian A-M, Hazinski MF, Meaney PA, Nadkarni VM, Ng KC, Nuthall GA, Ong GYK, Reis AG, Schexnayder SM, Shimizu NS, Van de Voorde P, Nikolaou N, Dainty KN, Couper K, Morrison L, Maconochie IK.: Dispatcher instruction in CPR (pediatrics). Consensus on Science and Treatment Recommendations [Internet] Brussels, Belgium: International Liaison Committee on Resuscitation (ILCOR), Pediatrics Task Force, 2018 October, 2019.
Available from: http://ilcor.org.
Metho... Preamble and Link to Published Systematic Review
The continuous evidence evaluation process for the production of Consensus on Science with Treatment Recommendations (CoSTR) began with a systematic review of the population in the context of the intervention and the comparators (Nikolaou, 2018, registered at PROSPERO - CRD42018091427. This systematic review was led by Nikolaus Nikolaou and Katie Dainty. Evidence for adult and pediatric literature was sought and considered by the Basic Life Support Task Force and the Pediatric Life Support Task Force groups, respectively. The current consensus on science is limited to pediatrics with the corresponding adult counterpart published by the Adult Life Support Task Force.
Systematic Review
Nikolaou N, Dainty K, Couper K, Morley P, Tijssen J, Vaillancourt C on behalf of the International Liaison Committee on Resuscitation Basic Life Support and Pediatric Life Support Task Forces. A Systematic Review and Meta-Analysis of the Effect of Dispatcher-Assisted CPR on Outcome from Sudden Cardiac Arrest in Adults and Children. Accepted xxx, 2019. PUBMED Reference.
Dispatcher Instruction in CPR PICOST
The PICOST (Population, Intervention, Control, Outcomes, Study design and Timeframe)
Population: Infants and children with presumed cardiac arrest in out-of-hospital settings.
Intervention: Patients/cases or EMS systems where dispatch assisted CPR is offered.
Comparators: Studies with comparators where either Systems or specific cardiac arrest cases are not offered dispatch-assisted CPR are included.
Outcomes: Critical outcomes included: Survival with good neurological function (at hospital discharge, 1 month or 6 months), survival (hospital discharge, 1 month or 1 year survival), short term survival (Return of Spontaneous Circulation – ROSC, hospital admission), provision of Bystander CPR (BCPR), whilst important outcomes were initial shockable rhythm, time to CPR.
Study Designs: Randomized controlled trials (RCTs) and non-randomized studies (non-randomized controlled trials, interrupted time series, controlled before-and-after studies, cohort studies) were eligible for inclusion.
Timeframe: All years and all languages are included with the last search performed July 1st 2018. Ongoing or unpublished studies were identified through a search of www.clinicaltrials.gov online registry.
PROSPERO Registration - CRD42018091427
Consensus on Science
Cardiac arrest outcomes in EMS systems with and without dispatch-assisted CPR
Three studies were included in the systematic review comparing the outcomes for patients who were offered dispatch-assisted CPR (Goto 2014 e000499, Akahane 2012 1410, Ro 2016 1331). This study did not analyze the patients by whether or not they actually received the CPR. In one of the three studies, 25% of bystanders who were offered dispatcher CPR assistance did not provide dispatch-assisted CPR (Goto 2014 e000499).
Only the Goto study included adjusted analysis. Where overlapping populations from the same source (e.g. registry) were reported for the same outcome, the larger population of the two studies was used in the analysis (Goto 2014 e000499, Akahane 2012 1410).
Survival with good neurological outcome at 1 month
For this critical outcome (O), we identified very low certainty evidence downgraded for serious risk of bias and imprecision, from 1 study (Goto 2014 e000499). The analysis included 4,306 out-of-hospital pediatric cardiac arrests and did not show benefit in systems which offered dispatch-assisted CPR in both the unadjusted and adjusted analyses [OR 1.03 (0.72-1.48); RR 1.03 (0.73-1.47); p=0.87), 1 survivor with good neurological outcome per 1,000 (95% CI: 8 fewer to 14 more), and OR 1.45 (0.98-2.15); RR not applicable; p=0.06), respectively].
Survival at 1 month
For this critical outcome (O), we identified very low certainty evidence, downgraded for serious risk of bias, from 1 study (Goto 2014 e000499). The analysis included 4,306 out-of-hospital pediatric cardiac arrests and did not show benefit in systems which offered dispatch-assisted CPR in the unadjusted analysis OR 1.17 [0.95-1.45]; RR 1.15 [0.95-1.40]; p=0.14; 14 more survivors per 1,000 (CI 95%: 4 fewer to 35 more survivors)]. In the adjusted analysis, there was benefit in systems which offered dispatch-assisted CPR (OR 1.46 [1.05-2.03]; RR not applicable; p=0.02).
Delivery of bystander CPR
For this critical outcome (O), we identified three studies. The unadjusted analysis (Akahane 2012 1410, Ro 2016 1331) included 3,309 out-of-hospital pediatric cardiac arrests and showed an increase in bystander CPR in systems which offered dispatch-assisted CPR (OR 4.05 [2.43-6.75]; RR 2.25 [2.05-2.47]; p=0.001; 315 more survivors per 1,000 (95% CI: 188 more to 437 more)). The certainty of the evidence was low, downgraded for serious risk of bias. The adjusted analysis (Goto 2014 e000499) also showed benefit in systems which offered dispatch-assisted CPR (OR 7.51 [6.58-8.57]; RR not applicable; p<0.0001), with a moderate certainty of evidence.
Shockable initial rhythm
For this important outcome (O), we identified very low certainty evidence, downgraded for serious risk of bias, from one study (Goto 2014 e000499). The unadjusted analysis included 4,306 out-of-hospital pediatric cardiac arrests and showed no benefit in systems which offered dispatch-assisted CPR (OR 0.81 [0.60-1.11]; RR 0.82 [0.61-1.10]; p=0.19; 8 fewer survivors per 1,000 (CI 95%: 5 more to 18 fewer)).
Arrest to CPR-initiation time
For this important outcome (O), we identified very low certainty evidence, downgraded for serious risk of bias, from one study (Goto 2014 e000499) which enrolled 4,306 out-of-hospital cardiac arrests that showed shorter times to CPR in systems that offered dispatch-assisted CPR (median 4 min (inter quartile range 1-9); vs. 11 min (inter quartile range 7-16), p<0.0001) when compared with systems without dispatch-assisted CPR.
Cardiac arrest outcomes among patients who received Dispatch-assisted CPR (DACPR) compared with patients who did not receive DACPR
Three studies were included comparing outcomes among patients who received dispatch-assisted CPR compared with those who did not receive dispatch-assisted CPR (Goto 2014 e000499, Chang 2018 120, Ro 2016 1331). There were 2 groups of analyses: 1- patients who received dispatch-assisted CPR vs patients who did not receive any bystander CPR, and 2- patients who received dispatch-assisted CPR vs patients who received unassisted bystander CPR. Only Chang and Goto studies provided adjusted analyses for comparison of children receiving dispatch-assisted CPR versus no bystander CPR (group 1). Although we present the results of the adjusted (1st group only) and unadjusted analyses (both groups), it is important to note that the patients differed significantly within the Chang and Ro studies (those who received unassisted bystander CPR generally had more favorable prognostic factors [shockable initial rhythm and witnessed arrest] than those who did not receive dispatcher-assisted CPR) and weren’t compared in the Goto study. As in the previous section, where there were overlapping populations for the same outcome (Chang 2018 120 and Ro 2016 1331), the larger population of the two studies was used in the analysis.
Survival with good neurological outcome at 1 month
1- Dispatch-assisted CPR vs no bystander CPR
We identified very low certainty evidence downgraded for serious risk of bias and imprecision from 1 study (Goto 2014 e000499). The analysis included 4,306 out-of-hospital pediatric cardiac arrests and showed benefit for patients who received dispatch-assisted CPR in both the unadjusted and adjusted analyses [OR 1.49 (1.05-2.11); RR 1.47 (1.05-2.07); p=0.03, 12 more per 1,000 (CI 95%: 1 survivor with good neurologic outcome to 26 more)) and OR 1.81 (1.23-2.67); p=0.003), respectively].
2- Dispatch-assisted CPR vs unassisted bystander CPR
We identified very low certainty evidence downgraded for serious risk of bias, from 1 study (Goto 2014 e000499). The unadjusted analysis included 2,722 out-of-hospital pediatric cardiac arrests and patients who received dispatch-assisted CPR compared with patients who received unassisted bystander CPR had lower survival at 1 month with good neurological outcome (OR 0.57 [0.39-0.84]; RR 0.59 [0.41-0.84]; p=0.004, 26 fewer per 1,000 (CI 95%: 9 fewer to 37 fewer,).
Survival with good neurological outcome at hospital discharge
1- Dispatch-assisted CPR vs no bystander CPR
We identified low certainty evidence downgraded for serious risk of bias, from 1 study (Chang 2018 120). The analysis included 1,661 out-of-hospital pediatric cardiac arrests and showed benefit for patients who received dispatch-assisted CPR in both the unadjusted (OR 3.63 [2.18-6.03]; RR 3.43 [2.10-5.59]; p<0.0001, 54 survivors with good neurologic outcome more per 1,000 (CI 95%: 25 more to 99 more)) and adjusted analyses (OR 2.22 [1.27-3.88]; p=0.005).
2- Dispatch-Assisted CPR vs unassisted bystander CPR
We identified very low certainty evidence, downgraded for serious risk of bias and very serious imprecision from 1 study (Chang 2018 120). The unadjusted analysis included 970 out-of-hospital pediatric cardiac arrests and showed no benefit for patients who received dispatch-assisted compared with patients who received unassisted bystander CPR (OR 0.97 [0.58-1.62]; RR 0.97 [0.61-1.56]; p=0.91, 2 fewer survivors with good neurologic outcome per 1,000 (CI 95%: 32 fewer to 43 more)).
Survival at 1 month
1- Dispatch-assisted CPR vs no bystander CPR
We identified very low certainty evidence downgraded for serious risk of bias and imprecision, from 1 study (Goto 2014 e000499). The analysis included 4,306 out-of-hospital pediatric cardiac arrests and showed benefit for patients who received dispatch-assisted CPR in both the unadjusted (OR 1.42 [1.16-1.74]; RR 1.38 [1.15-1.65]; p=0.006, 31 more survivors per 1,000 (CI 95%: 12 more survivors to 53 more survivors)) and adjusted analyses (OR 1.63 [1.32-2.01]; p<0.0001).
2- Dispatch-assisted CPR vs unassisted bystander CPR
We identified very low certainty evidence downgraded for serious risk of bias from 1 study (Goto 2014 e000499). The unadjusted analysis included 2,722 out-of-hospital pediatric cardiac arrests and patients who received dispatch-assisted CPR compared with those who received unassisted bystander CPR had lower survival at 1 month (OR 0.74 [0.58-0.95]; RR 0.77 [0.62-0.95]; p=0.02), 34 fewer survivors at one month per 1,000 (from 6 fewer survivors to 57 fewer survivors)).
Survival at hospital discharge
1- Dispatch-assisted CPR vs no bystander CPR
For the critical outcome (O) of survival at hospital discharge, we identified 1 study (Chang 2018 120). The analysis included 1,661 out-of-hospital pediatric cardiac arrests and showed benefit for patients who received dispatch-assisted CPR in both the unadjusted (OR 3.14 [2.16-4.58]; RR 2.87 [2.02-4.06]; p<0.0001, 84 more survivors per 1,000 (CI 95%: 47 more survivors to 132 more survivors) and adjusted analyses (OR 2.23 [1.47-3.38]; p=0.002). The unadjusted analyses were deemed of moderate certainty and the adjusted were low certainty, downgraded for serious risk of bias.
2- Dispatch-assisted CPR vs unassisted bystander CPR
We identified one study (Chang 2018 120) with very low certainty evidence, downgraded for serious risk of bias and very serious imprecision. The unadjusted analysis included 1,661 out-of-hospital pediatric cardiac arrests and showed no benefit for patients who received dispatch-assisted CPR (OR 0.98 [0.65-1.48]; RR 0.99 [0.69-1.41]; p=0.94, 2 fewer survivors at hospital discharge per 1,000 (CI 95%: 42 fewer survivors at hospital to 51 more survivors at hospital discharge).
Sustained ROSC
1- Dispatch-assisted CPR vs no bystander CPR
For this critical outcome, we identified low certainty evidence, downgraded for serious risk of bias, from one study (Chang 2018 120). The unadjusted analysis included 1,661 out-of-hospital pediatric cardiac arrests and showed benefit for patients who received dispatch-assisted CPR (OR 2.95 [2.07-4.20]; RR 2.68 [1.94-3.70]; p<0.0001, 89 more survivors per 1,000 (CI 95%: 51 more survivors to 137 more survivors)).
2- Dispatch-assisted CPR vs unassisted bystander CPR
We identified one study (Chang 2018 120) with very low certainty evidence, downgraded for serious risk of bias and very serious imprecision. The unadjusted analysis included 1,661 out-of-hospital pediatric cardiac arrests and showed no benefit for patients who received dispatch-assisted CPR (OR 0.82 [0.56-1.19]; RR 0.84 [0.62-1.16]; p=0.29, 26 fewer sustained ROSC per 1,000 (CI 95%: 26 more achieving ROSC per 1,000 to 66 fewer achieving ROSC)).
Shockable initial rhythm
For this important outcome (O), we identified very low certainty evidence, downgraded for serious risk of bias and very serious imprecision, from two studies (Goto 2014 e000499 and Chang 2018 120) .
1- Dispatch-assisted CPR vs no bystander CPR, i.e. pooled results from both these papers.
The unadjusted analysis included 5,967 out-of-hospital pediatric cardiac arrests and showed no benefit in systems which provided dispatch-assisted CPR (OR 1.59 [0.78-3.21]); RR 1.52 [0.81-2.86]; p=0.20, 26 survivors with a shockable rhythm per 1,000 (CI 95%: 10 fewer survivors to 89 more survivors)).
2- Dispatch-assisted CPR vs unassisted bystander CPR
The unadjusted analysis included 3,692 out-of-hospital pediatric cardiac arrests and showed an inferior outcome in patients who received dispatch-assisted CPR compared with those who received unassisted bystander CPR (OR 0.50 [0.33-0.74]; RR 0.54 [0.35-0.82]; p=0.0007), 61 fewer survivors with an initially shockable rhythm per 1,000 (CI 95%; 31 fewer survivors with an initially shockable rhythm per 1,000 to 83 fewer survivors with an initially shockable rhythm per 1,000)).
Arrest to CPR-initiation time
For this important outcome, (O) we identified very low certainty evidence, downgraded for serious risk of bias, from two studies (Goto 2014 e000499, Ro) for both groups of analyses.
1- Dispatch-assisted CPR vs no bystander CPR
Goto analyzed 4,306 out-of-hospital cardiac arrest patients and showed shorter times to CPR for patients with dispatch-assisted CPR (median 1 min [inter quartile range 0-5 minutes]; vs. 11 min [inter quartile range 7-15]).
Ro enrolled 1,265 out-of-hospital cardiac arrest patients and showed shorter times to CPR for patients with dispatch-assisted CPR (median 4 min [inter quartile range 0-13 minutes]; vs. 10 min [inter quartile range 6-18; p=0.01]).
2- Dispatch-assisted CPR vs unassisted bystander CPR
Goto analyzed 2,722 out-of-hospital cardiac arrest patients and showed longer times to CPR for patients with dispatch-assisted CPR compared with those who received unassisted bystander CPR (median 4 min [inter quartile range 0-13 minutes]); vs. 1 min (inter quartile range 0-5). Ro enrolled 766 out-of-hospital pediatric cardiac arrest patients and showed longer times to CPR for patients with dispatch-assisted CPR compared with patients who received unassisted bystander CPR (median 4 min [inter quartile range 0-13 minutes]); vs. 2 min [inter quartile range 0-10]).
Treatment recommendations
We recommend emergency medical dispatch centers offer dispatch-assisted CPR instructions for presumed pediatric cardiac arrest (strong recommendation, very low certainty evidence).
We recommend emergency dispatchers provide CPR instructions for pediatric cardiac arrest when no bystander CPR is in progress (strong recommendation, low certainty evidence).
We cannot make a recommendation for or against emergency dispatchers to provide CPR instructions for pediatric cardiac arrest when bystander CPR is already initiated (no recommendation, very low certainty evidence).
Justification and Evidence to Decision Highlights
This topic was prioritized by the Pediatric Life Support Task Force following publication of several new studies since the previous systematic review was published in 2011. The 2011 review found limited evidence to support dispatch-assisted CPR (Bohm, 2011 1490). In considering the importance of this topic, the Pediatric Life Support Task Force noted that bystander CPR significantly improves the likelihood of survival from OHCA but bystander CPR rates remain very low. In developing the consensus on science and treatment recommendations, the Pediatric Life Support Task Force agreed that consideration of both unadjusted and adjusted analyses was essential to provide a full picture of the evidence. We recognize that unadjusted analysis might be confounded by temporal changes, systematic and patient care differences between and within EMS systems.
In making a strong recommendation for dispatch centers to offer dispatch-assisted CPR despite very low certainty evidence, the Pediatric Life Support Task Force considered the benefit for the critical outcome of survival in the adjusted analyses, as well as the large positive effect for the increase in delivery of bystander CPR and reduced time to initiation of CPR when dispatch assistance was offered. Implementation of dispatch-assisted CPR appears to be acceptable and feasible as many EMS systems have done so successfully. However, its cost effectiveness and impact on health equity have also not been evaluated and until documented, may present barriers to implementation in under-resourced regions. Also, successful implementation of any program of dispatch-assisted CPR requires a process of continuous quality improvement to ensure that dispatchers can quickly identify a likely cardiac arrest and assist the bystander in starting CPR in a very short time (Lerner 2012 648).
In making a strong recommendation despite low certainty evidence for emergency dispatchers to provide CPR instructions to callers for pediatric patients in cardiac arrest when no bystander CPR is in progress, the Task Force valued the consistency of results indicating benefit for all critical and important outcomes, with the exception of shockable rhythm (no benefit). This failure to demonstrate contribution of dispatch-assisted CPR to improvement in outcomes for shockable initial rhythm aligns with the adult meta-analysis (link to adult CoSTR, SR), except for the first analysis where the adult literature found higher rates of shockable initial rhythm for centers offering dispatch-assisted CPR. This may be explained by the significant difference in etiology of OHCA in pediatrics compared with adults (i.e., the majority of episodes of pediatric OHCA have a respiratory etiology, while most adult sudden cardiac arrest is associated with myocardial infarction and a sudden arrhythmia).
In abstaining from recommending for or against dispatch- assisted CPR when bystander CPR is already in progress in the face of very low certainty evidence, the Task Force balanced the consistency of inferior and neutral results for all of the critical outcomes with the lack of any adjusted analyses for this group. The negative results may have several potential explanations: 1- bystander CPR was initiated earlier than dispatch-assisted CPR because the bystander did not experience the delay required in calling a dispatcher, the dispatcher identifying OHCA, and then the dispatcher providing CPR instructions, and 2- a bystander who performed CPR and refused dispatch assistance was likely to be trained in CPR and may have provided a higher quality of CPR than that provided by the untrained bystander who required remote dispatch assistance. This particular finding suggests the potential benefits of widespread community-based CPR training.
Consideration of types of dispatch-assisted CPR systems or interventions to improve the quality of dispatch-assisted CPR was beyond the scope of this review. A limitation of the evidence that forms the basis of these treatment recommendations is that data are only derived from two countries--Japan and Korea. The EMS systems involved may differ in their response to OHCA compared with EMS responses in other regions. For example, the vast majority of systems in Japan and Korea do not use adrenaline to treat pediatric out-of-hospital cardiac arrest, while adrenaline/epinephrine is used to treat pediatric OHCA in many EMS systems in North America and Europe. Thus, caution is required when attempting to extrapolate these results to different EMS systems of care.
Finally, this systematic review did not address the content of CPR instructions offered or provided by the dispatcher, i.e. conventional CPR vs compressions-only CPR. For the included pediatric studies, the content of the CPR instructions varied across studies, depending on: 1- bystander CPR skill level (Akahane 2012 1410, Goto 2014 e000499), 2- etiology of the arrest (Lee), 3- victim age (Ro 2016 1331), or 4- the AHA guidelines that were followed by the EMS agency at the time of the study (Chang 2018 120). Only the Goto study specifically adjusted for content of CPR instructions. We recognize that CPR with ventilations is better for children in cardiac arrest (link to CC vs conventional CPR CoSTR), but we do not know how achievable or effective it is for a dispatcher to train a bystander on how to provide ventilations in pediatric cardiac arrest. This variable continues to be relevant in pediatrics and we advocate for future studies to consider this variable as a key predictor of interest.
Knowledge Gaps
1. Can dispatchers effectively guide untrained bystanders to provide conventional CPR for a child in cardiac arrest?
2. Only one study adjusted for content of CPR/dispatch-assisted CPR instructions provided. All future pediatric OHCA should adjust for this important co-variable, specifically addressing the role of ventilation in infant and child CPR, the language used by dispatchers and its effects on the initiation of bystander CPR, and how CPR instructions are provided (by the phrasing and enunciation of word, (termed wording), video adjuncts via cellphone, etc).
3. All pediatric OHCA studies should include data on certainty of bystander CPR (including when was detection of cardiac arrest made, what was the time to initiation of CPR and whether conventional CPR was successfully given) and subsequent in-hospital (post-arrest) factors.
4. Studies addressing dispatch-assisted CPR when bystander CPR is already initiated—is there specific guidance (e.g., to pace compression rate) that the dispatcher can offer that will improve the outcome of ongoing bystander CPR?
5. As only short term outcomes were evaluated, future studies should document long term outcomes, including Quality of Life outcomes.
6. Future studies of bystander CPR should adjust for bystander CPR characteristics.
7. The effect of EMS response times on outcomes with dispatch-assisted CPR should be considered.
8. The cost-effectiveness studies of dispatch-assisted CPR is unknown.
Attachments
Should patients/cases where DA-CPR is offered vs. patients/ cases where no CPR is offered be used for children with cardiac arrest in out of hospital settings?
References
Akahane M, Ogawa T, Tanabe S, Koike S, Horiguchi H, Yasunaga H, Imamura T, Impact of telephone dispatcher assistance on the outcomes of pediatric out-of-hospital cardiac arrest. Crit Care Med. 2012; 5: 1410-6. doi: 10.1097/CCM.0b013e31823e99ae.
Bohm K, Vaillancourt C, Charette ML, Dunford J, Castren. I, patients with out-of-hospital cardiac arrest, does the provision of dispatch cardiopulmonary resuscitation instructions as opposed to no instructions improve outcome: a systematic review of the literature. Resuscitation. 2011; 82: 1490-5.
Chang I, Ro YS, Shin SD, Song KJ, Park JH, Kong SY, Association of dispatcher-assisted bystander cardiopulmonary resuscitation with survival outcomes after pediatric out-of-hospital cardiac arrest by community property value. Resuscitation. 2018;132:120-126. doi: 10.1016/j.resuscitation.2018.09.008. Epub 2018 Sep 17.
Goto Y, Maeda T, Goto Y, Impact of dispatcher-assisted bystander cardiopulmonary resuscitation on neurological outcomes in children with out-of-hospital cardiac arrests: a prospective, nationwide, population-based cohort study. J Am Heart Assoc. 2014; 3: e000499. doi: 10.1161/JAHA.113.000499.
Lerner EB, Rea TD, Bobrow BJ, Acker JE 3rd, Berg RA, Brooks SC, Cone DC, Gay M, Gent LM, Mears G, Nadkarni VM, O'Connor RE, Potts J, Sayre MR, Swor RA, Travers AH; American Heart Association Emergency Cardiovascular Care Committee; Council on Cardiopulmonary, Critical Care, Perioperative and Resuscitation. Emergency medical service dispatch cardiopulmonary resuscitation prearrival instructions to improve survival from out-of-hospital cardiac arrest: a scientific statement from the American Heart Association. Circulation. 2012; 4: 648-55. doi: 10.1161/CIR.0b013e31823ee5fc. Epub 2012 Jan 9.
Ro YS, Shin SD, Song KJ, Hong SO, Kim YT, Cho SI, Bystander cardiopulmonary resuscitation training experience and self-efficacy of age and gender group: a nationwide community survey. Am J Emerg Med. 2016;8:1331-7. doi: 10.1016/j.ajem.2015.12.001.