Delays in Cardiopulmonary Resuscitation, Defibrillation, and Epinephrine Administration All Decrease Survival in In-hospital Cardiac Arrest

Although in-hospital cardiac arrest is a common event in U.S. hospitals, survival remains as low as about 20%.^1,2  Prior studies for out-of-hospital cardiac arrests have emphasized the critical importance of prompt initiation of cardiopulmonary resuscitation (CPR). Others have documented the importance of prompt treatment with defibrillation for patients with shockable in-hospital cardiac arrest³  and with epinephrine for those with nonshockable in-hospital cardiac arrest.⁴  However, the relationship between time to initiation of CPR and survival for in-hospital cardiac arrest is not well understood. Moreover, the total time between pulselessness and defibrillation or epinephrine treatment comprises both time to initiation of CPR and time from CPR to either treatment. The effect on survival of each of these intervals has not been previously characterized.

Accordingly, we examined the association between time to initiation of CPR and time from CPR to either defibrillation or epinephrine treatment on in-hospital cardiac arrest outcomes using data from Get With The Guidelines–Resuscitation, a large prospective, hospital-based, multicenter clinical registry that uses standardized definitions to assess both care processes and outcomes.⁵  We hypothesized that delays in the initiation of CPR and from time of CPR to defibrillation or epinephrine treatment are each associated with lower in-hospital cardiac arrest survival.

The Get With The Guidelines–Resuscitation (formerly known as the National Registry of Cardiopulmonary Resuscitation) is an American Heart Association–sponsored prospective multicenter observational registry of in-hospital cardiac arrest. The design of the Get With The Guidelines–Resuscitation database has been previously described.²  Briefly, all patients with cardiac arrest (defined as the absence of a palpable central pulse, apnea, and unconsciousness) and without do-not-resuscitate orders are enrolled by hospital quality improvement personnel who have received specialized training. Patients eligible for enrollment are identified from multiple sources, including but not limited to cardiac arrest flow sheets, hospital paging system logs, and routine checks of code carts. Standardized reporting using Utstein-style definitions⁵  are used for patient variables and outcomes. This study was approved by the Institutional Review Board of the University of Pittsburgh, Pittsburgh, Pennsylvania. Our statistical analysis plan was approved the National Registry of Cardiopulmonary Resuscitation Adult Research Task Force on April 9, 2009, before accessing the data. This analysis and manuscript were approved by the Executive Database Steering Committee in accordance with the Get With The Guidelines Publication Policy.⁶ 

Between 2000 and 2008, we identified 132,950 patients with an in-hospital cardiac arrest within Get With The Guidelines–Resuscitation with complete comorbidity data for our model. We excluded 21,212 episodes of recurrent arrest to focus on index in-hospital cardiac arrest events (fig. 1). As we evaluated the effect of time to initiation of CPR on outcomes, we excluded 21,334 patients with an unwitnessed in-hospital cardiac arrest and 6,998 patients without information on time to initiation of CPR. We also excluded 5,549 patients with implausible time to initiation of CPR (i.e., negative times [n = 5,036] and time to initiation of CPR of 7 min or more [n = 513]). Our study population comprised 77,857 patients with a witnessed in-hospital cardiac arrest with time to initiation of CPR of 0 to 6 min. We excluded 8,954 patients for missing values for times from CPR to defibrillation or epinephrine treatment. We also excluded patients missing survival data and for negative or outlier times from CPR to defibrillation or epinephrine treatment (11,528 excluded patients, outliers as defined by Tukey box plot of more than 11 min in the defibrillation group or more than 9 min in the epinephrine group). The final sample sizes were 11,002 in the defibrillation group and 46,310 in the epinephrine group.

Our two main independent variables were established a priori and were (1) time to initiation of CPR and (2) time from the initiation of CPR to treatment, defined as either defibrillation or epinephrine. Time to initiation of CPR was defined as the difference between the recorded clock time for the determination of pulselessness and the recorded clock time for the beginning of chest compressions. Similarly, because we aimed to study the influence of delay (as opposed to the influence of shockable vs. nonshockable rhythms), time to treatment was defined as the difference between the recorded clock time for either the first defibrillation attempt or the administration of epinephrine and time for initiation of CPR. For patients who received both defibrillation and epinephrine treatments, this interval was defined by whichever intervention was recorded as being given first. Survival to discharge was established a priori as our primary outcome.

Baseline differences between the defibrillation- and epinephrine-treated groups were examined. Continuous variables were compared with the median and Kruskal–Wallis tests, and categorical variables were compared using the chi-square test.

We then constructed multivariable logistic regression models to examine the associations between survival as an outcome and time to initiation of CPR and time from CPR to defibrillation or epinephrine treatment as ordinal categorical predictors. To maximize statistical power and to allow comparison of the two groups, we developed models in which the groups were combined, as well as separate models for each group. To further enhance statistical power, in post hoc exploratory analysis, we evaluated various binning strategies (supplemental table 1, https://links.lww.com/ALN/B828, describing the partitioning of categories for each binning strategy) in addition to univariate and pointwise analysis for time to initiation of CPR and time to treatment. We dichotomized time to initiation of CPR into ranges of 0 to 2 and 3 to 6 min. We also categorized time to defibrillation and time to epinephrine treatment into ranges 0 to 2, 3 to 5, 6 to 8, and 9 to 11 min.

After screening study variables for collinearity, we included the following covariates in our model: age, sex, race, whether the patient was monitored, location of cardiac arrest, initial rhythm, illness category (medical cardiac, medical noncardiac, surgical cardiac, surgical noncardiac), and comorbidities present within 24 h of cardiac arrest (table 1; supplemental table 2, https://links.lww.com/ALN/B829, depicting a complete list of group baseline characteristics). Moreover, we included in the model interventions in place at the time of cardiac arrest, including mechanical ventilation, various vasopressors, and other invasive procedures (supplemental table 2, https://links.lww.com/ALN/B829, depicting a complete list of group baseline characteristics). In post hoc testing, we assessed possible clustering effects at the hospital level (i.e., between hospitals) in three separate analyses: (1) adding facility as a covariate to our model, (2) our model using the generalized estimating equation (details in supplemental table 3, https://links.lww.com/ALN/B830, detailing model evaluation), and (3) a two-stage hierarchical analysis using facility and location (within the hospital), as well as the other covariates in our model. We also did post hoc sensitivity analysis by excluding patients who received defibrillation for a nonshockable rhythm or epinephrine for a shockable rhythm. In addition, we performed fractional polynomial analysis and logistic regression diagnostics using Stata/SE 15.1 (StataCorp LLC, USA). All other analyses were conducted using SPSS 22 to 25 (IBM SPSS, USA) and were assessed at a two-sided significance level of 0.05.

Selected baseline characteristics of the patient groups treated with defibrillation and epinephrine are provided in table 1 (for a complete list see supplemental tables 2 and 4, https://links.lww.com/ALN/B829 and https://links.lww.com/ALN/B831, depicting a complete list of group baseline characteristics and their influence on survival, respectively). The median age in the overall cohort was 67 yr (interquartile range, 54, 78), 70.0% (40,130 of 57,312) were of white race; for 61.0% (34,935 of 57,312), in-hospital cardiac arrest occured in an intensive care unit. Of 57,312 patients, 44,241 were medical (77.2%) and 10,720 (18.7%) were surgical. Of patients treated initially with epinephrine, 94.3% (43,622 of 46,310) had a nonshockable cardiac arrest rhythm, whereas 83.0% (9,134 of 11,002) of defibrillated patients had a shockable cardiac arrest rhythm. A greater proportion of the epinephrine-treated group (table 1) were hypotensive at the time of cardiac arrest (defibrillation group 27.2% [2,990 of 11,002] vs. epinephrine group 36.7% [16,980 of 46,310]), had respiratory insufficiency (defibrillation group 36.6% [4,025 of 11,002] vs. epinephrine group 50.5% [23,360 of 46,310]), or required mechanical ventilation (defibrillation group 34.5% [3,799 of 11,002] vs. epinephrine group 41.6% [19,270 of 46,310]). The results of our exploratory analysis are presented in supplemental table 3 (https://links.lww.com/ALN/B830, detailing model evaluation), supplemental table 5 (https://links.lww.com/ALN/B832, detailing univariate analysis), supplemental tables 6–13 (https://links.lww.com/ALN/B833, https://links.lww.com/ALN/B834, https://links.lww.com/ALN/B835, https://links.lww.com/ALN/B836, https://links.lww.com/ALN/B837, https://links.lww.com/ALN/B838, https://links.lww.com/ALN/B839, https://links.lww.com/ALN/B840), and supplemental figs. 1 and 2 (https://links.lww.com/ALN/B841 and https://links.lww.com/ALN/B842, depicting multivariable pointwise analysis).

In our combined model, after multivariable adjustment, increasing time to initiation of CPR and time from CPR to treatment were associated with decreased survival (table 2). In the overall cohort of 57,312 patients, there were 9,802 survivors (17.1%; table 3). Times to initiation of CPR greater than 2 min were associated with a survival of 14.7% (91 of 618) as compared with 17.1% (9,711 of 56,694) if CPR was begun in 2 min or less (adjusted odds ratio [CI], 0.68 [0.54 to 0.87]; P < 0.002; table 2; fig. 2). Times from CPR to either defibrillation or epinephrine treatment of 2 min or less were associated with a survival of 18.0% (7,654 of 42,475), as compared with 15.0% (1,680 of 11,227) for 3 to 5 min, 12.8% (382 of 2,983) for 6 to 8 min, and 13.7% (86 of 627) for 9 to 11 min (reference, 0 to 2 min; for 3 to 5 min adjusted odds ratio, 0.83; CI, 0.78 to 0.88; P < 0.001, for 6 to 8 min adjusted odds ratio, 0.67; CI, 0.60 to 0.76; P < 0.001, and for 9 to 11 min adjusted odds ratio, 0.54; CI, 0.42 to 0.69; P < 0.001; table 2; fig. 3). There was a substantial difference between groups not only with respect to survival (38% [4,178 of 11,002] for patients treated with defibrillation vs. 12.1% [5,624 of 46,310] for patients treated with epinephrine, adjusted odds ratio, 0.41; CI, 0.37 to 0.44; P < 0.001; tables 2 and 3), but also in the rate at which survival is diminished with respect to time from CPR to either defibrillation or epinephrine therapy (overall effect P < 0.001, reference 0 to 2 min, for 3 to 5 min adjusted odds ratio, 0.66; CI, 0.59 to 0.75; P < 0.001, for 6 to 8 min adjusted odds ratio, 0.44; CI, 0.34 to 0.55; P < 0.001, and for 9 to 11 min, adjusted odds ratio, 0.31; CI, 0.25 to 0.44; P < 0.001; tables 2 and 3; supplemental figs. 4 and 6, https://links.lww.com/ALN/B844 and https://links.lww.com/ALN/B846, depicting the stepwise reduction in survival with increasing time to defibrillation and epinephrine treatment, respectively). This same model was tested using the generalized estimating equation and yielded very similar results (details in supplemental table 3, https://links.lww.com/ALN/B830, detailing model evaluation).

If CPR was begun in 2 min or less, survival was 38.1% (4,143 of 10,880) as compared with 28.7% (35 of 122) if CPR was begun in 3 to 6 min (adjusted odds ratio, 0.60; CI, 0.39 to 0.93, P = 0.023; table 2; supplemental fig. 3, https://links.lww.com/ALN/B843, depicting the reduction in survival with delayed CPR). Similarly, if defibrillation was attempted in 2 min or less, survival was 40.5% (3,530 of 8,713), as compared with 31.6% (508 of 1,608) at 3 to 5 min, 22.4% (100 of 447) at 6 to 8 min, and 17.1% (40 of 234) at 9 to 11 min (adjusted odds ratio, 0.79; CI, 0.69 to 0.90 for 3 to 5 min, adjusted odds ratio, 0.67; CI, 0.52 to 0.87 for 6 to 8 min, adjusted odds ratio, 0.51; CI, 0.35 to 0.75, overall effect; P < 0.001; table 2; supplemental fig. 4, https://links.lww.com/ALN/B844, depicting the decrease in survival with delayed defibrillation). If CPR and defibrillation were both delivered promptly (i.e., within 2 min), survival was 40.6% (3,503 of 8,628; table 3). If CPR was begun promptly, but defibrillation was delayed (followed CPR by more than 3 min), survival was 31.9% (504 of 1,582) for 3- to 5-min delay, 21.9% (97 of 442) for 6- to 8- min delay, and 17.1% (39 of 228) for 9- to 11-min delay (table 3). If CPR was delayed (i.e., begun after 3 to 6 min), survival was reduced to 31.8% (27 of 85) if defibrillation followed CPR by 0 to 2 min and 15.4% (4 of 26) if defibrillation followed CPR by 3 to 5 min (table 3).

There was no difference in survival between patients who received CPR in 3 to 6 min (11.3% [56 of 496]) as compared with within 2 min (12.2% [5,568 of 45,814]; adjusted odds ratio, 0.75; CI, 0.56 to 1.00; P = 0.051; table 2; supplemental fig. 5, https://links.lww.com/ALN/B845, depicting survival with prompt vs. delayed CPR). There was a stepwise reduction in survival with each additional interval of delay from the initiation of CPR to epinephrine treatment: if epinephrine was administered within 2 min of initiation of CPR, survival was 12.2% (4,124 of 33,762) as compared with 12.2% (1,172 of 9,619; adjusted odds ratio, 0.88; CI, 0.82 to 0.95; P = 0.001) for 3 to 5 min and 11.2% (328 of 2,929; adjusted odds ratio, 0.75; CI, 0.66 to 0.85; P < 0.001; table 2; supplemental fig. 6, https://links.lww.com/ALN/B846, depicting the reduction in survival with increasing delay in epinephrine treatment). If both CPR and epinephrine were delivered promptly, survival was 12.2% (4,078 of 33,402), and if epinephrine was delayed by 3 to 5 or 6 to 9 min, survival was 12.2% (1,163 of 9,516) or 11.3% (327 of 2,896), respectively (table 3). If CPR was delayed (more than 2 min), survival was 12.8 (46 of 360), 8.7 (9 of 103), and 3.0% (1 of 33) for times to epinephrine of 0 to 2, 3 to 5, and 6 to 9 min, respectively (table 3).

Our study contains several novel results. First, the frequency of delay between the confirmation of pulselessness and the initiation of CPR was greater than our a priori expectation that CPR would begin immediately. We found that 5.7% (3,283 of 57,312) of patients did not have instantaneous initiation of CPR upon determination of a pulseless cardiac arrest (i.e., time to initiation of CPR of more than 0 min). Our second and most important finding was that delay in initiation of CPR reduces survival independent of subsequent delays in defibrillation or epinephrine administration. Although several other studies in this database have examined a potential effect of delay in the initiation of CPR,^4,7–10  none have found that effect.

The third novel finding is that both time to initiation of CPR and time from CPR to defibrillation are determinants of survival in patients with shockable in-hospital cardiac arrests. Thus, the expected survival advantage from early CPR can be severely reduced by subsequent delay in defibrillation, i.e., an inefficient resuscitation. Fourth, in the epinephrine group, although the influence of arrest interval¹¹  and delay in defibrillation^3,12  are well recognized, our analysis is the first to suggest that increasing the time from the initiation of CPR to the administration of epinephrine was associated with lower survival. It is well known that patients in the epinephrine group start with a very poor prognosis as compared with patients in the defibrillation group.^1–4  Attention to both prompt CPR and prompt epinephrine administration are particularly important for the management of in-hospital cardiac arrest, because pulseless electrical activity and asystole comprise up to 82% of all such arrests.^2,13  Fifth, the rate of decline in survival with time from CPR to therapy is quite large in the defibrillation group as compared with the epinephrine group (table 2; supplemental figs. 4 and 6, https://links.lww.com/ALN/B844 and https://links.lww.com/ALN/B846, depicting the stepwise reduction in survival with increasing time from CPR to defibrillation and epinephrine treatment respectively).

The total time from determination of pulselessness to defibrillation in the setting of ventricular fibrillation has long been known to be a determinant of outcome both for in-hospital³  and out-of-hospital cardiac arrest.^5,11,14–17  In the pre-hospital setting, as the total time to defibrillation increases, even though survival decreases, relative improvement associated with bystander CPR increases.¹⁸  We observed a similar context sensitivity for in-hospital cardiac arrest, i.e., delay in CPR reduces the survival benefit of defibrillation even if the total time to defibrillation remains the same. Similarly, total time to epinephrine treatment is also known to be a determinant of outcome for out-of-hospital^19,20  as well as in-hospital cardiac arrest both in adults⁴  and in children.¹⁰  Our results are consistent with prior studies. The novel feature of our work is the explicit demonstration that after adjustment for time to initiation of CPR, the time from the initiation of CPR to epinephrine treatment is a determinant of survival.

Although there is considerable observational evidence from the prehospital setting that increasing duration of cardiac arrest before CPR lowers survival,^15–17,21  relatively few studies have examined delays in CPR in the in-hospital setting. Herlitz et al.²²  reported that if CPR was started within 1 min, survival was 33% as compared with 14% if CPR was started later. Hajbaghery et al.²³  reported that in all patients that survived to hospital discharge and all patients on the morning shift, CPR was started in 1 to 6 min. For the evening and night shifts, CPR was started in 1 to 6 min in 92 and 89% of patients, respectively. Survival to hospital discharge was 8.3, 4.8, and 3.6%, respectively, for day, evening, and night shifts. Forcina et al.²⁴  reported that in nursing units using standard defibrillators, median time to initiation of CPR was 0 (interquartile range, 0, 1), but in those units using automatic external defibrillators, median time to initiation of CPR was 0 (interquartile range, 0, 2; P = 0.08). Although they found that this trend toward increased time to initiation of CPR did not correlate directly with survival, there was a trend toward decreased survival in the automatic external defibrillator units (18%) as compared with the standard defibrillator units (23%, P = 0.09). Although the reported delays in our study are comparable to those in the literature, the use of a large database and risk-adjusted model provides stronger evidence that delay in CPR in in-hospital cardiac arrest decreases survival.

The total time from determination of pulselessness to either the first defibrillation attempt or to epinephrine treatment is a measure of two separate processes. The total time for each includes the time from pulselessness to the initiation of CPR and then the time from CPR initiation to either defibrillation or epinephrine administration. In our study, there was a graded reduction in survival for delays in defibrillation and epinephrine treatment, and the reduction in survival was made worse if CPR was also delayed. Delivery of CPR, defibrillation, and epinephrine treatment are team and system processes, as well as context-sensitive, i.e., the potential benefit of each therapy is partially dependent on the other therapies rendered concurrently or subsequently. Previous analyses have tended to focus on individual therapies rather than consider the relationship between therapies. A well-functioning team, however, will have been trained to provide CPR, defibrillation, and epinephrine administration in a rapid fashion. Reduction of delays requires prompt action, particularly by ward staff while awaiting the arrival of the code response team. We chose a simple performance-based (i.e., time to therapy) model to examine the impact of delays on survival. Although this model has the minor disadvantage of not grouping patients by initial rhythm, the necessary risk adjustment was accomplished by including initial rhythm as a covariate in our model. This model has the advantage that it identifies therapy that does not match the initial rhythm, i.e., defibrillation for nonshockable rhythms and epinephrine for shockable rhythms. Both delays and mismatch of therapy represent opportunities for both research to understand these problems as well as education to ameliorate them.

Limitations of this study include the absence of independent verification of the times recorded, as well as exclusions either because the computed values for times were beyond the range or because of missing values for survival. The lack of synchronization of clocks in hospitals may also lead to errors in times. Our analysis also was not designed to establish causal factors for delays and did not include other unknown factors that may influence timeliness of CPR or defibrillation and epinephrine treatment. These remain areas of active investigation within Get With The Guidelines–Resuscitation. In addition, because of the curvilinear nature of the relationships between delays and survival, the Hosmer–Lemeshow test (supplemental table 3, https://links.lww.com/ALN/B830, detailing model evaluation) suggests that alternative statistical methods might yield a better model fit. Other limitations include lack of extensibility of our results to all hospitals based on the subgroup of hospitals represented in quality improvement registries such as Get With The Guidelines–Resuscitation.

In conclusion, we found that both delays in time to initiation of CPR and time from CPR to treatment with either defibrillation or epinephrine are associated with lower survival for patients with in-hospital cardiac arrest. Further research is needed to determine the impact of both benchmarking and training efforts for in-hospital cardiac arrest focused on accurately measuring and reducing delays in CPR and from CPR to defibrillation or epinephrine administration.

The authors thank Sandra C. Hirsch, M.B.A., (Department of Anesthesiology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania) for her kind assistance with manuscript preparation and correspondence as well as coordination with the Get With The Guidelines–Resuscitation administrative staff. The authors also thank Danette Jordan, M.D., M.P.H., and Nicole E. Scouras, M.D., M.P.H., for their efforts on this project, including assistance with the original proposal and development of the initial model.

Supported by partial salary support from the Department of Anesthesiology and Perioperative Medicine of the University of Pittsburgh School of Medicine (Pittsburgh, Pennsylvania; to Dr. Bircher), and National Institutes of Health (Bethesda, Maryland) grant Nos. R01NS36124, R01GM114851, and T32GM075770 (to Dr. Xu) and 1R01HL123980 and K23HL102224 (to Dr. Chan).

The authors declare no competing interests.

In addition to the author ChanPaul S. M.D., M.Sc., members of the Get With The Guidelines–Resuscitation/National Registry of Cardiopulmonary Resuscitation Science Advisory Board/Clinical Working Group and Adult Task Force include FaillaceRobert T. M.D., Albert Einstein College of Medicine ManciniMary E. R.N., Ph.D., University of Texas at Arlington, Arlington, Texas BergRobert A. M.D., Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania; Emilie Allen, M.S.N., R.N., Parkland Health and Hospital System, Dallas, Texas HuntElizabeth A. M.D., M.P.H., Ph.D., Johns Hopkins Medicine Simulation Center, Baltimore, Maryland NadkarniVinay M. M.D., University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania Ann PeberdyMary M.D., and OrnatoJoseph P. M.D., Virginia Commonwealth University Health System, Richmond, Virginia BraithwaiteScott M.D., New York University School of Medicine, New York, New York NicholGraham M.D., M.P.H., and WarrenSamuel M.D., University of Washington, Seattle, WashingtonDuncan, R.N., Institute for Healthcare Improvement, Boston, Massachusetts KennethLaBreshKathy M.D., Research Triangle Institute International, Research Triangle Park, North Carolina SassonComilla M.D., M.S., University of Colorado, Aurora, Colorado KnightLynda R.N., Lucile Packard Children’s Hospital at Stanford, Palo Alto, California DonninoMichael W. M.D., Beth Israel Deaconess Medical Center, Boston, Massachusetts SmythMindy M.S.N., R.N. EigelBrian Ph.D., and GentLana Ph.D., American Heart Association, Dallas, Texas MaderTimothy J. M.D., Baystate Medical Center/Tufts University School of Medicine, Springfield, Massachusetts KernKarl B. M.D., University of Arizona Medical Center, Tucson, Arizona; and GeocadinRomergryko G. M.D., Johns Hopkins School of Medicine, Baltimore, Maryland