[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Figure 1.
Colorimetric Contour Plots Showing the Proportion of Patients With Functionally Favorable Survival and Weighted Survival Among Those in the Overall Cohort at Each Combination of Rate and Depth
Colorimetric Contour Plots Showing the Proportion of Patients With Functionally Favorable Survival and Weighted Survival Among Those in the Overall Cohort at Each Combination of Rate and Depth

A, Left panel, standard cardiopulmonary resuscitation (n = 93); right panel, addition of an active impedance threshold device (n = 93). B, Left panel, standard cardiopulmonary resuscitation; right panel, addition of an active impedance device. Survival scale is not shown because the data are derived from weighted units.

Table 1.  
Characteristics of Patients Receiving Standard CPR (Sham-ITD) Compared With Those Receiving an Active-ITD
Characteristics of Patients Receiving Standard CPR (Sham-ITD) Compared With Those Receiving an Active-ITD
Table 2.  
Persons Falling Within Each of 130 Combinations of Rate and Depth and Those With Functionally Favorable Survival Within Each of 130 Combinations of Rate and Depth
Persons Falling Within Each of 130 Combinations of Rate and Depth and Those With Functionally Favorable Survival Within Each of 130 Combinations of Rate and Depth
Table 3.  
Optimal CCR-CCD Among Cardiac Arrest Survivors Overall and for Predefined Subgroups
Optimal CCR-CCD Among Cardiac Arrest Survivors Overall and for Predefined Subgroups
Table 4.  
Survivors and Survival in the Cohort Observed Within a 20% Range of the Identified Optimal CCR-CCD Combination of 107 cpm and 4.7 cm
Survivors and Survival in the Cohort Observed Within a 20% Range of the Identified Optimal CCR-CCD Combination of 107 cpm and 4.7 cm
1.
Kouwenhoven  WB, Jude  JR, Knickerbocker  GG.  Closed-chest cardiac massage.  JAMA. 1960;173(10):1064-1067. doi:10.1001/jama.1960.03020280004002PubMedGoogle ScholarCrossref
2.
Benjamin  EJ, Blaha  MJ, Chiuve  SE,  et al; American Heart Association Statistics Committee and Stroke Statistics Subcommittee.  Heart disease and stroke statistics-2017 update: a report from the American Heart Association.  Circulation. 2017;135(10):e146-e603. doi:10.1161/CIR.0000000000000485PubMedGoogle ScholarCrossref
3.
Idris  AH, Guffey  D, Aufderheide  TP,  et al; Resuscitation Outcomes Consortium (ROC) Investigators.  Relationship between chest compression rates and outcomes from cardiac arrest.  Circulation. 2012;125(24):3004-3012. doi:10.1161/CIRCULATIONAHA.111.059535PubMedGoogle ScholarCrossref
4.
Stiell  IG, Brown  SP, Christenson  J,  et al; Resuscitation Outcomes Consortium (ROC) Investigators.  What is the role of chest compression depth during out-of-hospital cardiac arrest resuscitation?  Crit Care Med. 2012;40(4):1192-1198. doi:10.1097/CCM.0b013e31823bc8bbPubMedGoogle ScholarCrossref
5.
Stiell  IG, Brown  SP, Nichol  G,  et al; Resuscitation Outcomes Consortium Investigators.  What is the optimal chest compression depth during out-of-hospital cardiac arrest resuscitation of adult patients?  Circulation. 2014;130(22):1962-1970. doi:10.1161/CIRCULATIONAHA.114.008671PubMedGoogle ScholarCrossref
6.
Idris  AH, Guffey  D, Pepe  PE,  et al; Resuscitation Outcomes Consortium Investigators.  Chest compression rates and survival following out-of-hospital cardiac arrest.  Crit Care Med. 2015;43(4):840-848. doi:10.1097/CCM.0000000000000824PubMedGoogle ScholarCrossref
7.
Yannopoulos  D, Aufderheide  TP, Abella  BS,  et al.  Quality of CPR: an important effect modifier in cardiac arrest clinical outcomes and intervention effectiveness trials.  Resuscitation. 2015;94(3):106-113. doi:10.1016/j.resuscitation.2015.06.004PubMedGoogle ScholarCrossref
8.
Nolan  JP, Perkins  GD, Soar  J.  Chest compression rate: where is the sweet spot?  Circulation. 2012;125(24):2968-2970. doi:10.1161/CIRCULATIONAHA.112.112722PubMedGoogle ScholarCrossref
9.
Idris  AH.  The sweet spot: chest compressions between 100-120/minute optimize successful resuscitation from cardiac rest.  JEMS. 2012;37(9):4-9.PubMedGoogle Scholar
10.
Resuscitation Outcomes Consortium. https://devroc.uwctc.org/tiki/roc-background. Accessed July 7, 2019.
11.
Aufderheide  TP, Kudenchuk  PJ, Hedges  JR,  et al; ROC Investigators.  Resuscitation Outcomes Consortium (ROC) PRIMED cardiac arrest trial methods part 1: rationale and methodology for the impedance threshold device (ITD) protocol.  Resuscitation. 2008;78(2):179-185. doi:10.1016/j.resuscitation.2008.01.028PubMedGoogle ScholarCrossref
12.
Aufderheide  TP, Nichol  G, Rea  TD,  et al; Resuscitation Outcomes Consortium (ROC) Investigators.  A trial of an impedance threshold device in out-of-hospital cardiac arrest.  N Engl J Med. 2011;365(9):798-806. doi:10.1056/NEJMoa1010821PubMedGoogle ScholarCrossref
13.
Wik  L, Kramer-Johansen  J, Myklebust  H,  et al.  Quality of cardiopulmonary resuscitation during out-of-hospital cardiac arrest.  JAMA. 2005;293(3):299-304. doi:10.1001/jama.293.3.299PubMedGoogle ScholarCrossref
14.
ECC Committee, Subcommittees and Task Forces of the American Heart Association. 2005 American Heart Association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation. 2005;112(24 suppl):IV 1-203.
15.
Banks  JL, Marotta  CA.  Outcomes validity and reliability of the modified Rankin scale: implications for stroke clinical trials: a literature review and synthesis.  Stroke. 2007;38(3):1091-1096. doi:10.1161/01.STR.0000258355.23810.c6PubMedGoogle ScholarCrossref
16.
Pepe  PE, Scheppke  KA, Antevy  PM,  et al.  Confirming the clinical safety and feasibility of a bundled methodology to improve cardiopulmonary resuscitation involving a head-up/torso-up chest compression technique.  Crit Care Med. 2019;47(3):449-455. doi:10.1097/CCM.0000000000003608PubMedGoogle ScholarCrossref
17.
Adnet  F, Triba  MN, Borron  SW,  et al.  Cardiopulmonary resuscitation duration and survival in out-of-hospital cardiac arrest patients.  Resuscitation. 2017;111:74-81. doi:10.1016/j.resuscitation.2016.11.024PubMedGoogle ScholarCrossref
18.
Aufderheide  TP, Pirrallo  RG, Yannopoulos  D,  et al.  Incomplete chest wall decompression: a clinical evaluation of CPR performance by EMS personnel and assessment of alternative manual chest compression-decompression techniques.  Resuscitation. 2005;64(3):353-362. doi:10.1016/j.resuscitation.2004.10.007PubMedGoogle ScholarCrossref
19.
Wigginton  JG, Pepe  PE, Bedolla  JP, DeTamble  LA, Atkins  JM.  Sex-related differences in the presentation and outcome of out-of-hospital cardiopulmonary arrest: a multiyear, prospective, population-based study.  Crit Care Med. 2002;30(4)(suppl):S131-S136. doi:10.1097/00003246-200204001-00002PubMedGoogle ScholarCrossref
20.
Roppolo  LP, Wigginton  JG, Pepe  PE. Emergency ventilatory management as a detrimental factor in resuscitation practices and clinical research efforts. In: Vincent JL, ed. 2004 Yearbook of Intensive Care and Emergency Medicine. Heidelberg, Germany: Springer-Verlag. 2004;139-151.
Views 3,025
Citations 0
Original Investigation
August 14, 2019

Optimal Combination of Compression Rate and Depth During Cardiopulmonary Resuscitation for Functionally Favorable Survival

Author Affiliations
  • 1Cardiovascular Division, University of Minnesota Medical School, Minneapolis
  • 2Department of Medicine, The University of Texas Southwestern Medical Center, Dallas
  • 3Department of Surgery, The University of Texas Southwestern Medical Center, Dallas
  • 4Department of Pediatrics, The University of Texas Southwestern Medical Center, Dallas
  • 5Department of Emergency Medicine, The University of Texas Southwestern Medical Center, Dallas
  • 6School of Public Health, The University of Texas Southwestern Medical Center, Dallas
  • 7Department of Emergency Medicine, Medical College of Wisconsin, Milwaukee
  • 8Department of Emergency Medicine, University of Oklahoma School of Community Medicine, Tulsa
  • 9Department of Emergency Medicine, University Hospital of Grenoble Alps, Grenoble, France
  • 10Quality of Care Unit, University Hospital of Grenoble Alps, Grenoble, France
  • 11Department of Pharmacology, Faculty of Medicine, Toho University, Tokyo, Japan
JAMA Cardiol. Published online August 14, 2019. doi:10.1001/jamacardio.2019.2717
Key Points

Question  During cardiopulmonary resuscitation, is there an optimal combination of chest compression rate and depth associated with an enhanced likelihood of favorable functional outcome, and does that optimal combination change with respect to age, sex, presenting cardiac rhythm, or use of a cardiopulmonary resuscitation adjunct?

Findings  In this cohort study of data from 3643 individuals in the National Institutes of Health clinical trials network database, the optimal combination of chest compression rate was 107 compressions per minute and chest compression depth of 4.7 cm; this finding remained relatively consistent regardless of age, sex, presenting cardiac rhythm, or cardiopulmonary resuscitation adjunct use. Adjunct use was associated with significant improvements in outcome, but this was dependent on delivering the identified optimal chest compression rate and depth combination.

Meaning  The findings suggest that the combination of 107 compressions per minute and a depth of 4.7 cm may be the optimal target for chest compression rate and depth, and that use of an adjunct may be associated with significantly enhanced outcomes if this target is used.

Abstract

Importance  Previous studies of basic cardiopulmonary resuscitation (CPR) indicate that both chest compression rate (CCR) and chest compression depth (CCD) each are associated with survival probability after out-of-hospital cardiac arrest. However, an optimal CCR-CCD combination has yet to be identified, particularly with respect to age, sex, presenting cardiac rhythm, and CPR adjunct use.

Objectives  To identify an ideal CCR-CCD combination associated with the highest probability of functionally favorable survival and to assess whether this combination varies with respect to age, sex, presenting cardiac rhythm, or CPR adjunct use.

Design, Setting, and Participants  This cohort study used data collected between June 2007 and November 2009 from a National Institutes of Health (NIH) clinical trials network registry of out-of-hospital and in-hospital emergency care provided by 9-1-1 system agencies participating in the network across the United States and Canada (n = 150). The study sample included 3643 patients who had out-of-hospital cardiac arrest and for whom CCR and CCD had been simultaneously recorded during an NIH clinical trial of a CPR adjunct. Subgroup analyses included evaluations according to age, sex, presenting cardiac rhythm, and application of a CPR adjunct. Data analysis was performed from September to November 2018.

Interventions  Standard out-of-hospital cardiac arrest interventions compliant with the concurrent American Heart Association guidelines as well as use of the CPR adjunct device in half of the patients.

Main Outcomes and Measures  The optimal combination of CCR-CCD associated with functionally favorable survival (modified Rankin scale ≤3) overall and by age, sex, presenting cardiac rhythm, and CPR adjunct use.

Results  Of 3643 patients, 2346 (64.4%) were men; the mean (SD) age was 67.5 (15.7) years. The identified optimal CCR-CCD for all patients was 107 compressions per minute and a depth of 4.7 cm. When CPR was performed within 20% of this value, survival probability was significantly higher (6.0% vs 4.3% outside that range; odds ratio, 1.44; 95% CI, 1.07-1.94; P = .02). The optimal CCR-CCD combination remained similar regardless of age, sex, presenting cardiac rhythm, or CPR adjunct use. The identified optimal CCR-CCD was associated with significantly higher probabilities of survival when the CPR device was used compared with standard CPR (odds ratio, 1.90; 95% CI, 1.06-3.38; P = .03), and the device’s effectiveness was dependent on being near the target CCR-CCD combination.

Conclusions and Relevance  The findings suggest that the combination of 107 compressions per minute and a depth of 4.7 cm is associated with significantly improved outcomes for out-of-hospital cardiac arrest. The results merit further investigation and prospective validation.

Introduction

In recent clinical reports regarding cardiac arrest outcomes after closed-chest cardiopulmonary resuscitation (CPR), 1 factor strongly associated with worse outcomes has been inadequate performance of chest compressions.1-7 Recovery with good neurologic function after out-of-hospital cardiac arrest (OHCA) is well-correlated with target ranges of chest compression rate (CCR) and chest compression depth (CCD).3-7 In these studies,3-9 favorable ranges of CCR or CCD were independently identified, with worse outcomes outside each of those respective ranges.

Despite these complementary but independent findings, there are interactions between CCR and CCD, such as a faster CCR being associated with compromised CCD.9 Data are still lacking with respect to specifically identifying the optimal combination of CCR and CCD and whether the same CCR-CCD target combination should be applied to all patients irrespective of sex, age, presenting cardiac rhythm, or CPR adjunct use.3-9 Knowing, monitoring, and confirming target CCR-CCD combinations would not only optimize treatment but also improve the study design and reliability of clinical studies.7

The specific hypothesis was whether a target CCR-CCD combination could be identified that would be associated with improved likelihood of favorable functional outcome after OHCA. It was also hypothesized that a different target combination might be delineated when comparing sex, age, presenting heart rhythms, or application of CPR adjuncts.

Methods

This cohort study used data from the National Institutes of Health (NIH) clinical trials network database. For the past 2 decades, the NIH and partner agencies sponsored multicenter clinical trials managed by the NIH Resuscitation Outcomes Consortium (ROC), which tested pharmacological, procedural, and device interventions for OHCA.10 The ROC PRIMED (ROC Prehospital Resuscitation Impedance Valve and Early Versus Delayed Analysis) trial evaluated a CPR adjunct using a sizeable, diverse cohort of patients with OHCA treated across 150 US and Canadian emergency medical services (EMS) agencies participating in the ROC network between June 2007 and November 2009.10-12 It was the first multicenter trial to use electronically documented measurements of CCR and CCD.10-12 By enrolling 8718 adult-age patients with OHCA, a high percentage of women, and use of a CPR adjunct, the data set from this trial12 was considered an appropriate vehicle for this present investigation.11,12 The present study, undertaken independently of the NIH, involved analyses of data from the ROC PRIMED database that were obtained through the NIH Data Sharing Policy and Freedom of Information Act (https://grants.nih.gov/policy/sharing.htm). Published studies3,5-7,9 examining either optimal ranges of CCR or optimal ranges of CCD have used similar approaches. The Human Subjects Committees at the University of Minnesota, Minneapolis, reviewed and approved the study; the study met exempt qualifications because this was an analytic review of a deidentified public database, and therefore informed consent was not required. Investigators followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline. Data analysis was performed from September to November 2018.

Study Design

For ROC PRIMED, both CCR and CCD data were collected electronically using measurement and recording sensors linked to the EMS agencies’ electrocardiographic monitor defibrillators.11,12 On the basis of previous publications,11,13 the CCR-CCD data used here were the means of measurements taken during the first 5 minutes of recorded CPR, with CCR recorded to the nearest integer and CCD recorded to the nearest 0.5 cm. To avoid detracting outliers with CCR or CCD values indicating negligible odds of survival, data sets were trimmed to only include patients receiving CCR between 60 and 160 compressions per minute (cpm) and CCD between 2.0 and 8.0 cm.

Analyzed data included age, sex, presenting cardiac rhythm, and CPR adjunct use. The adjunct, methods, and primary results of the original trial are described elsewhere.10-12 In brief, patients were assigned randomly in a blinded manner to receive conventional CPR using either an inactive (sham) impedance threshold device (ITD) or active-ITD providing 16 cm H2O resistance (ZOLL Medical).11,12 Each device was labeled with a numerical code known only to the data coordinating center for subsequent identification of sham-ITD or active-ITD assignments.

Patient Care Protocols

The EMS first-responders were instructed to apply ITDs by facemask or advanced airway while providing chest compressions and ventilation according to concurrent American Heart Association recommendations.3,11,12,14 These recommendations stipulated 80 to 100 cpm, a compression depth of 4.0 to 6.0 cm, and using an advanced airway, 10 positive-pressure breaths per minute with approximately 600 mL tidal volume. The breaths were delivered in a 30:2 compression to breath ratio when using basic airways.3,11,12,14 The ROC sites were permitted to only enroll persons after showing for several months that CPR could be delivered with these predefined metrics more than 50% of the time.11,12

Study Participants

Among 8718 ROC PRIMED patients, 6199 had recordings of CCR and 3750 had CCD recordings, but most lacked simultaneous measurements of CCR and CCD during the first 5 minutes of CPR. Eligible study participants were those with intact sets of simultaneous CCR-CCD recordings during the first 5 minutes of EMS-performed CPR. As previously stated, those with CCR-CCD values outside the proscribed ranges (60-160 cpm and 2.0- to 8.0-cm depth) were excluded from analyses.

Statistical Analysis

Statistical techniques were used to calculate the optimal CCR-CCD combination associated with a maximized probability of survival with a modified Rankin scale (mRS) score of 3 or less at the time of hospital discharge, examining the entire cohort of analyzed patients and the subset of survivors.15 For both analyses, a 130-cell grid was constructed with 10 levels of CCR (ranging from 60-69 cpm to 150-160 cpm) and 13 levels of CCD (ranging from 2.0 cm to 8.0 cm using 0.5-cm increments). For the full cohort, the survival probability in each cell was calculated as the numerator of survivors in each cell divided by the denominator of patients within that individual cell. That survival probability was then multiplied by the reciprocal of its variance within each cell to create a set of weighted probabilities for each of those cells. For the survivors-only analyses, each of the similarly constructed 130-grid cells contained the corresponding proportion of survivors (with the sum of all 130 proportions equaling one).

Response Surface Modeling Approach

Both cohort and survivor samples were analyzed using response surface modeling to estimate the combination CCR-CCD values associated with optimized outcome. In these models, CCR was represented by the midpoint of the rate interval (eg, 95 cpm for the interval of 90-99 cpm), whereas CCD was defined by rounding to the nearest 0.5 cm as previously described.

A regression model with a linear and quadratic term for each of the rates and depths (and their interaction) was fitted to the data overall and then separately fitted for sham-ITD (inactive) and active-ITD groups. A stepwise method was used to identify the best-fitting model. From these models, optimal CCR-CCD combination values were calculated using numerical optimization techniques. The proposed optimal combination was evaluated further within a range that was within 20% of the identified CCR-CCD target.

Subgroup Analyses

Analyses were performed to determine whether optimal CCR-CCD targets varied by sex, age (using median age of the overall cohort: <70 years vs ≥70 years), or the presenting cardiac rhythm, specifically comparing ventricular fibrillation or ventricular tachycardia with other presenting rhythms or asystole. In addition, optimal CCR-CCD combinations for sham-ITD (standard CPR) and active-ITD (adjunct CPR) were estimated within each sham-ITD or active-ITD subgroup and across subgroups combined.

Contour Plot Approach

Contour plots were constructed to visually display optimal CCD-CCR combinations colorimetrically with separate displays for sham-ITD and active-ITD groups. These plots were designed to show the relative proportions of survivors across the survivor sample and the weighted survival proportions for the overall cohort within each cell, with the rate and depth categories forming a 2-dimensional plot. Colder zones represent the lowest (negligible) proportion of survivors or survival probability, cool zones represent slightly higher proportions, and warmer and hotter zones represent higher proportions of survivors or survival.

Descriptive statistics for continuous variables are reported as mean (SD) and categorical variables by frequency and percentage. Comparisons are reported as mean difference (95% CI) or odds ratio (OR) (95% CI); P < .05 (2-sided) indicates statistical significance. Analyses were performed in Stata, version 13.1 (StataCorp) and Minitab, version 17.3.1 (Minitab Statistical Software).

Results
Patient Characteristics

Simultaneous measurements of CCR and CCD during the first 5 minutes of CPR efforts were recorded for 3749 patients, with 106 patients (2.8%) having CCR-CCD values outside the trimmed ranges (60-160 cpm; 2.0-8.0 cm), leaving a study cohort of 3643 patients (mean [SD] age, 67.5[15.7] years; 2346 [64.4%] men). Although 35 (0.9%) achieved return of spontaneous circulation within that 5-minute period, their data before return of spontaneous circulation were included. Compared with those achieving return of spontaneous circulation after 5 minutes, these patients remained well-matched in terms of demographics, active ITD use, and survival.

Of the 3643 patients, 1527 (41.9%) had bystanders witness the OHCA with bystander-CPR performed for 1323 (36.3%); 1740 (47.8%) presented with asystole and 893 (24.5%) with ventricular fibrillation or ventricular tachycardia. First-in responders had a mean (SD) response interval of 5.7 (2.0) minutes (dispatch to street location arrival); 3316 patients (91.1%) received at least 1 prehospital dose of epinephrine, and 186 (5.1%; 93 controls and 93 active-ITD patients) had functionally favorable survival to hospital discharge (mRS ≤3).

When comparing 1832 patients (50.3%) assigned to sham-ITD and 1811 (49.7%) receiving the active-ITD, the demographic, clinical presentation, and treatment data confirmed well-matched subgroups and mimicked the overall study group (Table 1). The only statistically significant difference was frequency of epinephrine administration (sham-ITD vs active-ITD: 89.8% vs 92.3%; OR, 1.35; 95% CI, 1.07-1.70; P = .01).

The survivor sample (n = 186) showed similar comparisons except that 1 statistically significant difference was a longer response interval for active-ITD patients (mean [SD], 5.4 [1.5] vs 4.9 [1.7] minutes for sham-ITD; mean difference, 0.49 minutes [95% CI, 0.02-0.95 minutes]; P = .04).

Rate and Depth Data

Across the 130 CCR-CCD combinations, the 100-109 cpm and 4.0 cm combination was the most populated whether for sham-ITD, active-ITD, or the overall cohort (Table 2). In the survivor group (n = 186), the most populated cell was the 90-99 cpm/4.5 cm combination.

Results from Response Surface Models

Table 3 provides the response surface modeling results for the 186 survivors. Terms for rate, depth, and their quadratic forms were kept in the final models for all groups with the interaction term between rate and depth not significant in any model.

The optimal combination of CCR-CCD associated with the greatest probability of favorable functional outcome was identified as 107 cpm and 4.7 cm with little difference across subgroups (age, sex, cardiac rhythm, or adjunct use). With CPR performed within 20% of this identified combination (86-128 cpm; 3.8-5.6 cm), survival probability was significantly higher (6.0% vs 4.3% outside that range; OR, 1.44; 95% CI, 1.07-1.94; P = .02). Corresponding comparisons for sham-ITD and active-ITD survivors (mRS ≤3) showed significantly larger numbers of survivors with the active-ITD (n = 60) vs the sham-ITD (n = 43) (OR, 2.11; 95% CI, 1.17-3.81; P = .01).

Results From Contour Plots

Contour plots were developed for the 93 sham-ITD (standard CPR) survivors (mRS ≤3) (Figure 1A) and 93 active-ITD counterparts (Figure 1A). Optimal CCR-CCD combinations were similar, with the cell with the highest proportion of survivors being 100-109 cpm and 4.5-5.0 cm. However, the peak proportion of survivors for the active-ITD group was significantly higher compared with the corresponding sham-ITD group, indicated by the hotter colorimetric zones and the only red zone findings. A similar pattern was shown when evaluating all 3643 patients combined (Figure 1B). Despite the higher probability of survival with ITD use, the identified optimal CCR-CCD combination remained similar with or without the device.

When evaluating the 4 most populated combinations of CCR-CCD among survivors, survival (mRS ≤3) was 7.4% (Table 4). However, when stratified, survival probability was 9.6% for active-ITD use vs 5.3% for sham-ITD use (OR, 1.90; 95% CI, 1.06-3.38; P = .03).

Subgroup Analyses

Among 186 survivors (mRS ≤3), 133 (71.5%) were men. Although survival differences between men (5.7%) and women (4.1%) were significant (OR, 1.41; 95% CI, 1.02-1.95; P = .04), the identified optimal CCR-CCD combination remained consistent (Table 3). Older individuals (age, ≥70 years) appeared to benefit from a shallower CCD (Table 3), but differences were not statistically significant. Standard CPR (sham-ITD) patients with nonshockable presentations appeared to have a lower optimal CCR compared with counterparts presenting with ventricular fibrillation or ventricular tachycardia (99 vs 109 cpm), but definitive conclusions could not be drawn because of small sample sizes.

In general, there did not appear to be conclusive support for a variable favorable combination for any of the predefined subgroups compared with the overall findings.

Discussion

Despite reported interactive associations between CCR and CCD, data have been lacking with respect to determining a specific optimal CCR-CCD combination. Previous studies generally addressed independent evaluations of optimal ranges for CCR or CCD.3-9 Of importance, whether such an ideal CCR-CCD target would differ significantly depending on sex or age (anatomical and physiologic differences), the presenting cardiac rhythm (possible surrogate for more prolonged hypoxic event), or the use of a CPR adjunct (that might augment flow) has not been specifically addressed.

Although this was a secondary analysis of clinical trial data, the study included prospectively collected, well-defined data points from the OHCA experience of more than 150 EMS agencies in 2 countries including actual simultaneous recordings of CCR and CCD, constituting the best available data from the largest North American database on the subject. Recognizing the limitations of this analysis and that the findings may not be universally applicable, the results suggest a plausible value for an optimal CCR-CCD combination associated with a maximized probability of functionally favorable survival after OHCA. This optimal combination should now be further studied and validated in future prospective investigations. Although the combination may not be the eventual definitive answer regarding optimal rate and depth, it is an important step in the process of finding the best practice and determining whether the combination varies according to various factors.

The data from this analysis showed that, regardless of the presenting cardiac rhythm, age, sex, or use of a particular CPR adjunct, the optimal CCR-CCD combination remained similar. It is still possible that other interventions, such as various mechanical CPR devices or more prolonged arrest intervals, could have altered that finding. Therefore, evaluations of CCR-CCD combinations should be stratified accordingly in future studies of such interventions or conditions.

In this study, ITD use was associated with significantly improved survival likelihood when CPR was performed within or near the identified best combination, and this finding was dependent on that optimal performance of CPR. The other findings, such as the favorable associations for a shallower CCD in older patients or slower CCR for nonshockable rhythms were not conclusive because of the small sample sizes but could be considered hypothesis generating.

The wide variation in both CCR and CCD across the study cohort (Table 2) may indicate the challenges of optimizing manual CPR performance among numerous rescuers whose individual abilities to perform CPR properly may be variable, even in closely monitored EMS systems. One could therefore argue for real-time CPR feedback tools on a day-to-day basis and/or automated CPR devices to better ensure consistent delivery of optimal CCR-CCD combinations. Studies such as this and follow-up investigations may provide presumptive guidance, but evolving factors such as bundled CPR approaches that include mechanical CPR and other adjuncts may also alter that optimal target.16

A unique feature of this analysis was the use of response surface models and contour plots. Response surface models provide a better estimate of the optimal combination of CCR and CCD to achieve the best survival; the contour plots are useful tools that enable direct visualization of the joint associations of CCR and CCD with survival for the whole grid of values.17

If our data are on target and the CCR-CCD combination of 107 cpm and 4.7 cm (within 20%) are proven to be the best approach, the 6% survival among those patients compared with the 4% survival outside the combination zone would translate into several thousands of additional lives saved each year in the United States alone. Furthermore, if the ITD were used within the optimum 4 best cells for survivors, the 9.6% neurologic-intact survival that we detected would conservatively translate into at least 10 000 more lives saved annually.

Limitations

The findings here may not be universally applicable. They need to be further validated and examined for modifications as certain variables change in the future.16,18,19 It also involved EMS systems with presumably seasoned 9-1-1 agencies and well-monitored OHCA cases initially audited by the NIH ROC leadership and therefore not entirely representative of other circumstances. However, even if the results were simply reflective of a subset of EMS personnel more focused and trained well in resuscitative tasks with high-level performance, those factors should not only improve the results, but also serve largely to better reinforce reliability of the study’s findings.

Cardiopulmonary resuscitation was not always performed optimally. Targeting rescuers charged with delivering a rate of 100 cpm (range, 80-100 cpm) and a depth of 4.0 to 6.0 cm might appear to be a form of selection bias. However, previous studies3-9 have shown that even when the CCR was within a preferred range, CCD might not have been, or vice versa. We also sought to find the optimal CCR-CCD combination within that proscribed range and evaluate whether the preferred target changed according to age, sex, electrocardiographic presentation, or use of a flow-enhancing device (eg, ITD).

More than half the patients (53.2%) were found to be in CCR-CCD grids beyond a calculated optimal target combination range of within 20%, and 80% of the study population was outside the 4 most populated grids for survivors; those were grids that closely represented what the rescuers were expected to be providing. Also, most of the patients overall received CCR-CCD combinations that were below what were determined by this analysis to be the optimal grid zones for survivors (Table 2).

This study cohort was comprised of patients who had simultaneous recordings of CCR and CCD performed. This cohort was derived from within a larger cohort of study patients from the selected clinical trial.12 In many settings and certain individual cases, CCR and CCD were not measured simultaneously during the proscribed initial 5-minute period or were not technically retrievable (approximately 57% of the source cohort). Although this might also create the concern for a potential selection bias, the present study cohort was shown to be representative of the entire group when we compared the demographic and clinical presentations of the original clinical trial cohort.12 The analyzed standard CPR and active-ITD groups matched especially in terms of demographics, clinical presentations, and treatment.

Another limitation is that the quality of chest wall recoil was not available and no information regarding the actual performance of assisted ventilation (frequency, tidal volume, timing, and squeeze duration) was provided.13,18-20 All of these variables have been considered to be effect modifiers in terms of outcomes, and the optimal CCR-CCD target described in this study could shift if information related to optimal chest wall recoil, chest compression fraction, ventilatory parameters, or other modifiers were considered simultaneously when determining optimal CCR-CCD targets.16,18-20

With the consideration that the present study groups were so well matched, it is reasonable to assume that recoil and ventilatory aberrations were equally distributed and further optimization of recoil and ventilation would likely serve to improve survival chances even further at the optimal combination of CCR and CCD. Regardless, these measures are recommended factors to capture and evaluate as part of an optimal bundle of care delivery in future investigations.

In addition, although crude surrogates, the sex-based and age-based comparisons were performed to detect any potential anatomic and physiologic differences among those complex subcategories.18,19 In future analyses, investigators might consider collecting more specific data regarding weight and height or body mass indices and document rib fractures occurring during CPR. Also, we used binary evaluations (men vs women; age, <70 vs ≥70 years). Validation studies might be improved with evaluations of more subsegmented or continuum data combinations of age and sex categories.

Conclusions

In this study, the optimal CCR-CCD combination associated with a favorable neurologic outcome after OHCA was 105 to 109 cpm and 4.5 to 5.0 cm, with an estimated peak at or near 107 cpm and depth of 4.7 cm. This same combination generally applied regardless of age, sex, presenting cardiac rhythm, or the use of an ITD. Moreover, improved survival with the use of the ITD appeared to be dependent on providing the optimal combination of CCR and CCD as identified here. Therefore, optimal CCR-CCD combinations merit further validation and should be important considerations in future CPR survival investigations, particularly those involving studies of CPR-dependent interventions.

Back to top
Article Information

Accepted for Publication: May 24, 2019.

Published Online: August 14, 2019. doi:10.1001/jamacardio.2019.2717

Open Access: This article is published under the JN-OA license and is free to read on the day of publication.

Corresponding Author: Paul E. Pepe, MD, MPH, Department of Emergency Medicine, The University of Texas Southwestern Medical Center, Dallas, Texas, 5323 Harry Hines Blvd, MC 8579, Dallas, TX 75390-8579 (paul.pepe@utsw.edu).

Author Contributions: Dr Duval had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Duval, Pepe, Goodloe, Sugiyama, Yannopoulos.

Acquisition, analysis, or interpretation of data: Duval, Pepe, Aufderheide, Goodloe, Debaty, Labarère.

Drafting of the manuscript: Duval, Pepe, Aufderheide, Sugiyama, Yannopoulos.

Critical revision of the manuscript for important intellectual content: Duval, Pepe, Aufderheide, Goodloe, Debaty, Labarère, Sugiyama.

Statistical analysis: Duval.

Supervision: Pepe, Sugiyama, Yannopoulos.

Conflict of Interest Disclosures: Dr Aufderheide reported receiving grants from the National Institutes of Health, Helmsley Trust, and Ortho Clinical Diagnostics during the conduct of the study. Dr Debaty reported receiving personal fees and nonfinancial support from Zoll Medical outside the submitted work. Dr Yannopoulos reported receiving grants from the National Institutes of Health and from Helmsley Trust outside the submitted work. No other disclosures were reported.

Additional Contributions: We thank the National Institutes of Health and the Resuscitation Outcomes Consortium for providing the data used in this research; the thousands of emergency medical services personnel, investigators, trainers, and staff who implemented the studies and made this present investigation possible; and the communities that were served by the advances garnered by the National Institutes of Health Resuscitation Outcomes Consortium team.

References
1.
Kouwenhoven  WB, Jude  JR, Knickerbocker  GG.  Closed-chest cardiac massage.  JAMA. 1960;173(10):1064-1067. doi:10.1001/jama.1960.03020280004002PubMedGoogle ScholarCrossref
2.
Benjamin  EJ, Blaha  MJ, Chiuve  SE,  et al; American Heart Association Statistics Committee and Stroke Statistics Subcommittee.  Heart disease and stroke statistics-2017 update: a report from the American Heart Association.  Circulation. 2017;135(10):e146-e603. doi:10.1161/CIR.0000000000000485PubMedGoogle ScholarCrossref
3.
Idris  AH, Guffey  D, Aufderheide  TP,  et al; Resuscitation Outcomes Consortium (ROC) Investigators.  Relationship between chest compression rates and outcomes from cardiac arrest.  Circulation. 2012;125(24):3004-3012. doi:10.1161/CIRCULATIONAHA.111.059535PubMedGoogle ScholarCrossref
4.
Stiell  IG, Brown  SP, Christenson  J,  et al; Resuscitation Outcomes Consortium (ROC) Investigators.  What is the role of chest compression depth during out-of-hospital cardiac arrest resuscitation?  Crit Care Med. 2012;40(4):1192-1198. doi:10.1097/CCM.0b013e31823bc8bbPubMedGoogle ScholarCrossref
5.
Stiell  IG, Brown  SP, Nichol  G,  et al; Resuscitation Outcomes Consortium Investigators.  What is the optimal chest compression depth during out-of-hospital cardiac arrest resuscitation of adult patients?  Circulation. 2014;130(22):1962-1970. doi:10.1161/CIRCULATIONAHA.114.008671PubMedGoogle ScholarCrossref
6.
Idris  AH, Guffey  D, Pepe  PE,  et al; Resuscitation Outcomes Consortium Investigators.  Chest compression rates and survival following out-of-hospital cardiac arrest.  Crit Care Med. 2015;43(4):840-848. doi:10.1097/CCM.0000000000000824PubMedGoogle ScholarCrossref
7.
Yannopoulos  D, Aufderheide  TP, Abella  BS,  et al.  Quality of CPR: an important effect modifier in cardiac arrest clinical outcomes and intervention effectiveness trials.  Resuscitation. 2015;94(3):106-113. doi:10.1016/j.resuscitation.2015.06.004PubMedGoogle ScholarCrossref
8.
Nolan  JP, Perkins  GD, Soar  J.  Chest compression rate: where is the sweet spot?  Circulation. 2012;125(24):2968-2970. doi:10.1161/CIRCULATIONAHA.112.112722PubMedGoogle ScholarCrossref
9.
Idris  AH.  The sweet spot: chest compressions between 100-120/minute optimize successful resuscitation from cardiac rest.  JEMS. 2012;37(9):4-9.PubMedGoogle Scholar
10.
Resuscitation Outcomes Consortium. https://devroc.uwctc.org/tiki/roc-background. Accessed July 7, 2019.
11.
Aufderheide  TP, Kudenchuk  PJ, Hedges  JR,  et al; ROC Investigators.  Resuscitation Outcomes Consortium (ROC) PRIMED cardiac arrest trial methods part 1: rationale and methodology for the impedance threshold device (ITD) protocol.  Resuscitation. 2008;78(2):179-185. doi:10.1016/j.resuscitation.2008.01.028PubMedGoogle ScholarCrossref
12.
Aufderheide  TP, Nichol  G, Rea  TD,  et al; Resuscitation Outcomes Consortium (ROC) Investigators.  A trial of an impedance threshold device in out-of-hospital cardiac arrest.  N Engl J Med. 2011;365(9):798-806. doi:10.1056/NEJMoa1010821PubMedGoogle ScholarCrossref
13.
Wik  L, Kramer-Johansen  J, Myklebust  H,  et al.  Quality of cardiopulmonary resuscitation during out-of-hospital cardiac arrest.  JAMA. 2005;293(3):299-304. doi:10.1001/jama.293.3.299PubMedGoogle ScholarCrossref
14.
ECC Committee, Subcommittees and Task Forces of the American Heart Association. 2005 American Heart Association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation. 2005;112(24 suppl):IV 1-203.
15.
Banks  JL, Marotta  CA.  Outcomes validity and reliability of the modified Rankin scale: implications for stroke clinical trials: a literature review and synthesis.  Stroke. 2007;38(3):1091-1096. doi:10.1161/01.STR.0000258355.23810.c6PubMedGoogle ScholarCrossref
16.
Pepe  PE, Scheppke  KA, Antevy  PM,  et al.  Confirming the clinical safety and feasibility of a bundled methodology to improve cardiopulmonary resuscitation involving a head-up/torso-up chest compression technique.  Crit Care Med. 2019;47(3):449-455. doi:10.1097/CCM.0000000000003608PubMedGoogle ScholarCrossref
17.
Adnet  F, Triba  MN, Borron  SW,  et al.  Cardiopulmonary resuscitation duration and survival in out-of-hospital cardiac arrest patients.  Resuscitation. 2017;111:74-81. doi:10.1016/j.resuscitation.2016.11.024PubMedGoogle ScholarCrossref
18.
Aufderheide  TP, Pirrallo  RG, Yannopoulos  D,  et al.  Incomplete chest wall decompression: a clinical evaluation of CPR performance by EMS personnel and assessment of alternative manual chest compression-decompression techniques.  Resuscitation. 2005;64(3):353-362. doi:10.1016/j.resuscitation.2004.10.007PubMedGoogle ScholarCrossref
19.
Wigginton  JG, Pepe  PE, Bedolla  JP, DeTamble  LA, Atkins  JM.  Sex-related differences in the presentation and outcome of out-of-hospital cardiopulmonary arrest: a multiyear, prospective, population-based study.  Crit Care Med. 2002;30(4)(suppl):S131-S136. doi:10.1097/00003246-200204001-00002PubMedGoogle ScholarCrossref
20.
Roppolo  LP, Wigginton  JG, Pepe  PE. Emergency ventilatory management as a detrimental factor in resuscitation practices and clinical research efforts. In: Vincent JL, ed. 2004 Yearbook of Intensive Care and Emergency Medicine. Heidelberg, Germany: Springer-Verlag. 2004;139-151.
×