Advertisement

Physiologic monitoring of CPR quality during adult cardiac arrest: A propensity-matched cohort study

      Abstract

      Aim

      The American Heart Association (AHA) recommends monitoring cardiopulmonary resuscitation (CPR) quality using end tidal carbon dioxide (ETCO2) or invasive hemodynamic data. The objective of this study was to evaluate the association between clinician-reported physiologic monitoring of CPR quality and patient outcomes.

      Methods

      Prospective observational study of index adult in-hospital CPR events using the AHA's Get With The Guidelines – Resuscitation Registry. Physiologic monitoring was defined using specific database questions regarding use of either ETCO2 or arterial diastolic blood pressure (DBP) to monitor CPR quality. Logistic regression was used to evaluate the association between physiologic monitoring and outcomes in a propensity score matched cohort.

      Results

      In the matched cohort, (monitored n = 3032; not monitored n = 6064), physiologic monitoring of CPR quality was associated with a higher rate of return of spontaneous circulation (ROSC; OR 1.22, CI95 1.04–1.43, p = 0.017) compared to no monitoring. Survival to hospital discharge (OR 1.04, CI95 0.91–1.18, p = 0.57) and survival with favorable neurological outcome (OR 0.97, CI95 0.75–1.26, p = 0.83) were not different between groups. Of index events with only ETCO2 monitoring indicated (n = 803), an ETCO2 >10 mmHg during CPR was reported in 520 (65%), and associated with improved survival to hospital discharge (OR 2.41, CI95 1.35–4.30, p = 0.003), and survival with favorable neurological outcome (OR 2.31, CI95 1.31–4.09, p = 0.004) compared to ETCO2 ≤10 mmHg.

      Conclusion

      Clinician-reported use of either ETCO2 or DBP to monitor CPR quality was associated with improved ROSC. An ETCO2 >10 mmHg during CPR was associated with a higher rate of survival compared to events with ETCO2 ≤10 mmHg.

      Abbreviations:

      AHA (American Heart Association), CPR (cardiopulmonary resuscitation)

      Keywords

      Introduction

      In-hospital cardiac arrest affects approximately 200,000 patients each year in the United States.
      • Merchant R.M.
      • Yang L.
      • Becker L.B.
      • et al.
      Incidence of treated cardiac arrest in hospitalized patients in the United States.
      Although outcomes have been improving over the last decade, more than half of adults who have a cardiac arrest during their hospitalization will not survive.
      • Meaney P.A.
      • Nadkarni V.M.
      • Kern K.B.
      • Indik J.H.
      • Halperin H.R.
      • Berg R.A.
      Rhythms and outcomes of adult in-hospital cardiac arrest.
      • Girotra S.
      • Nallamothu B.K.
      • Spertus J.A.
      • et al.
      Trends in survival after in-hospital cardiac arrest.
      • Goldberger Z.D.
      • Chan P.S.
      • Berg R.A.
      • et al.
      Duration of resuscitation efforts and survival after in-hospital cardiac arrest: an observational study.
      Many of these arrests now occur in highly monitored intensive care units (ICUs), where patients often have invasive monitoring in place at the time of arrest.
      • Meaney P.A.
      • Nadkarni V.M.
      • Kern K.B.
      • Indik J.H.
      • Halperin H.R.
      • Berg R.A.
      Rhythms and outcomes of adult in-hospital cardiac arrest.
      A resuscitation strategy that takes advantage of this monitoring and uses it to incorporate a patient's physiologic response into the ongoing resuscitation can be expected to save more lives.
      • Sutton R.M.
      • Friess S.H.
      • Naim M.Y.
      • et al.
      Patient-centric blood pressure-targeted cardiopulmonary resuscitation improves survival from cardiac arrest.
      • Friess S.H.
      • Sutton R.M.
      • French B.
      • et al.
      Hemodynamic directed CPR improves cerebral perfusion pressure and brain tissue oxygenation.
      Experimental literature has established that survival following cardiopulmonary resuscitation (CPR) for cardiac arrest depends on provision of adequate myocardial blood flow.
      • Kern K.B.
      • Ewy G.A.
      • Voorhees W.D.
      • Babbs C.F.
      • Tacker W.A.
      Myocardial perfusion pressure: a predictor of 24-hour survival during prolonged cardiac arrest in dogs.
      • Ralston S.H.
      • Voorhees W.D.
      • Babbs C.F.
      Intrapulmonary epinephrine during prolonged cardiopulmonary resuscitation: improved regional blood flow and resuscitation in dogs.
      • Halperin H.R.
      • Lee K.
      • Zviman M.
      • et al.
      Outcomes from low versus high-flow cardiopulmonary resuscitation in a swine model of cardiac arrest.
      However, measurements of myocardial blood flow during CPR are not available to the rescuer. Therefore, the American Heart Association (AHA) now recommends using clinical surrogates closely related to myocardial blood flow (end tidal carbon dioxide (ETCO2) or diastolic blood pressure (DBP)) to monitor resuscitation quality.
      • Meaney P.A.
      • Bobrow B.J.
      • Mancini M.E.
      • et al.
      Cardiopulmonary resuscitation quality: Improving cardiac resuscitation outcomes both inside and outside the hospital: a consensus statement from the American Heart Association.
      Although the conceptual relevance of hemodynamic and ETCO2 monitoring during CPR is well established, clinical studies supporting the titration of these parameters during human CPR are lacking.
      The primary objective of this study was to evaluate the association between clinician-reported physiologic monitoring of CPR quality using either ETCO2 or DBP and survival outcomes. We hypothesized that use of physiology to monitor resuscitation quality would be associated with higher rates of return of spontaneous circulation (ROSC). We addressed this hypothesis in a propensity-score-matched cohort study of adult in-hospital CPR events reported to the AHA's large multicenter Get With The Guidelines®-Resuscitation (GWTG-R) registry database.

      Methods

      The GWTG-R registry is an AHA sponsored prospective multisite database of patients undergoing in-hospital resuscitation that utilizes Utstein-style data reporting.
      • Jacobs I.
      • Nadkarni V.
      • Bahr J.
      • et al.
      Cardiac arrest and cardiopulmonary resuscitation outcome reports: update and simplification of the Utstein templates for resuscitation registries. A statement for healthcare professionals from a task force of the International Liaison Committee on Resuscitation (American Heart Association, European Resuscitation Council, Australian Resuscitation Council, New Zealand Resuscitation Council, Heart and Stroke Foundation of Canada, InterAmerican Heart Foundation, Resuscitation Council of Southern Africa).
      • Cummins R.O.
      • Chamberlain D.
      • Hazinski M.F.
      • et al.
      Recommended guidelines for reviewing, reporting, and conducting research on in-hospital resuscitation: The in-hospital ‘Utstein style’. A statement for healthcare professionals from the American Heart Association, the European Resuscitation Council, the Heart and Stroke Foundation of Canada, the Australian Resuscitation Council, and the Resuscitation Councils of Southern Africa.
      • Peberdy M.A.
      • Kaye W.
      • Ornato J.P.
      • et al.
      Cardiopulmonary resuscitation of adults in the hospital: A report of 14720 cardiac arrests from the National Registry of Cardiopulmonary Resuscitation.
      Hospitals voluntarily participate in the registry for the primary purpose of quality improvement. The design and reporting of GWTG-R has been described in detail previously (www.heart.org/resuscitation).
      • Girotra S.
      • Nallamothu B.K.
      • Spertus J.A.
      • et al.
      Trends in survival after in-hospital cardiac arrest.
      • Chan P.S.
      • Krumholz H.M.
      • Nichol G.
      • Nallamothu B.K.
      • American Heart Association National Registry of Cardiopulmonary Resuscitation Investigators
      Delayed time to defibrillation after in-hospital cardiac arrest.
      • Nadkarni V.M.
      • Larkin G.L.
      • Peberdy M.A.
      • et al.
      First documented rhythm and clinical outcome from in-hospital cardiac arrest among children and adults.
      Quintiles (Cambridge Massachusetts), through their online, interactive case report form and Patient Management Tool (PMT), serves as the data collection coordination center for the AHA/American Stroke Association Get With The Guidelines programs. The University of Pennsylvania acts as the data analytic center and prepares the data for research purposes. Institutional review board (IRB) approval is not required for hospital participation in this quality improvement database; however, the present research investigation was reviewed by the IRB at The Children's Hospital Of Philadelphia and determined to be IRB exempt.

      Cohort development – inclusion/exclusion criteria

      All adult patients (≥18 years of age) with a CPR event requiring chest compressions with an invasive airway or arterial catheter in place at the time of arrest were eligible for inclusion. We then subsequently excluded events lasting <1 min, any delivery room events, and any events missing either outcome or important confounding variables necessary for propensity matching (Fig. 1).
      Figure thumbnail gr1
      Fig. 1Consort-like diagram. *The primary predictor variables were taken from questions added to the data collection form in 2007. All data for pre-existing medical conditions and cause of arrest were missing which prohibited adjustment for missing in the propensity model as we did with variables with sporadic missingness.

      Study variables

      The primary independent variable was taken from specific questions regarding use of physiologic monitors in the GWTG-R database optional section on resuscitation quality (Fig. 2). Physiology was deemed to have been used to monitor resuscitation quality if either of the following data fields were answered in the affirmative: (1) “Was continuous end tidal CO2 monitoring used to monitor quality of CPR?” or (2) “Was arterial line diastolic pressure used to monitor compression quality?” In the subset of events in which CPR quality monitoring with ETCO2 was indicated, we evaluated the association between the level of ETCO2 achieved during CPR and patient survival. In this analysis, the primary independent variable was ETCO2 >10 mmHg vs. ETCO2 ≤10 mmHg
      • Sanders A.B.
      • Kern K.B.
      • Otto C.W.
      • Milander M.M.
      • Ewy G.A.
      End-tidal carbon dioxide monitoring during cardiopulmonary resuscitation. A prognostic indicator for survival.
      • Levine R.L.
      • Wayne M.A.
      • Miller C.C.
      End-tidal carbon dioxide and outcome of out-of-hospital cardiac arrest.
      , the only question specifically asked in the quality section of the GWTG-R registry regarding level of ETCO2 achieved (“was an end tidal CO2 value of >10 mmHg achieved?”). The primary outcome for all analyses was ROSC. This acute resuscitation outcome was chosen a priori because: (1) it is the most proximate outcome to the monitored CPR (the primary predictor variable); and (2) we prospectively determined that we would be underpowered to detect survival differences between monitoring groups. Therefore, secondary outcomes included survival to hospital discharge and survival to discharge with a favorable neurological outcome. Survival with a favorable neurological outcome was defined as a cerebral performance category (CPC) score of 1, 2 or no change from baseline.
      • Jacobs I.
      • Nadkarni V.
      • Bahr J.
      • et al.
      Cardiac arrest and cardiopulmonary resuscitation outcome reports: update and simplification of the Utstein templates for resuscitation registries. A statement for healthcare professionals from a task force of the International Liaison Committee on Resuscitation (American Heart Association, European Resuscitation Council, Australian Resuscitation Council, New Zealand Resuscitation Council, Heart and Stroke Foundation of Canada, InterAmerican Heart Foundation, Resuscitation Council of Southern Africa).
      • Nadkarni V.M.
      • Larkin G.L.
      • Peberdy M.A.
      • et al.
      First documented rhythm and clinical outcome from in-hospital cardiac arrest among children and adults.
      In accordance with standard Utstein-style definitions
      • Jacobs I.
      • Nadkarni V.
      • Bahr J.
      • et al.
      Cardiac arrest and cardiopulmonary resuscitation outcome reports: update and simplification of the Utstein templates for resuscitation registries. A statement for healthcare professionals from a task force of the International Liaison Committee on Resuscitation (American Heart Association, European Resuscitation Council, Australian Resuscitation Council, New Zealand Resuscitation Council, Heart and Stroke Foundation of Canada, InterAmerican Heart Foundation, Resuscitation Council of Southern Africa).
      , potential confounders were: patient factors, such as demographic characteristics (e.g., race
      • Razi R.R.
      • Churpek M.M.
      • Yuen T.C.
      • et al.
      Racial disparities in outcomes following PEA and asystole in-hospital cardiac arrests.
      , gender
      • Topjian A.A.
      • Localio A.R.
      • Berg R.A.
      • et al.
      Women of child-bearing age have better inhospital cardiac arrest survival outcomes than do equal-aged men.
      ), preexisting conditions, and illness categories; and arrest characteristics, such as interventions in place at time of arrest, first documented rhythm, immediate cause of arrest, time of arrest, duration of CPR, and pharmacologic interventions (Supplementary Table 1). “First documented rhythm” was defined as the first electrocardiographic rhythm documented during a CPR event. All patients with CPR events regardless of the presence of pulse and rhythm at the onset of CPR were included.
      Figure thumbnail gr2
      Fig. 2Get with the Guidelines® – Resuscitation Patient Management Tool (PMT), Section 7.1 CPR Quality Tab. Arterial line response is optional for the adult database.

      Statistical methods

      The following analysis plan was prospectively developed, and was approved by both the Adult Research Task Force of the AHA's GWTG-R registry and the Executive Database Steering Committee of the AHA prior to the authors receiving the data for analysis. In this prospective analysis, propensity score matching
      • Brookhart M.A.
      • Schneeweiss S.
      • Rothman K.J.
      • Glynn R.J.
      • Avorn J.
      • Sturmer T.
      Variable selection for propensity score models.
      was used to balance the distribution of potential confounders between patients with vs. patients without physiologic monitoring (either ETCO2 or DBP) at the index event. Propensity scores were estimated as a function of the observed covariates (listed in Supplementary Table 1). For any covariate with missing values, ‘missing’ was coded as a separate category so that the propensity score balanced the distribution of missingness between groups.
      • D’Agostino R.B.
      • Rubin D.B.
      Estimating and using propensity scores with partially missing data.
      Propensity score models only included the main effects for covariates because there were no suspected effect modifiers. Participants were matched 2:1 using nearest neighbor matching, with distance based on the log odds. Matching was performed without consideration of outcome values.
      Within the matched cohort, standardized mean differences were used to evaluate balance between groups, with balance achieved if the standardized mean difference was <0.10.
      • Harder V.S.
      • Stuart E.A.
      • Anthony J.C.
      Propensity score techniques and the assessment of measured covariate balance to test causal associations in psychological research.
      Logistic regression models were used to compare resuscitation outcomes between matched groups. A robust variance estimator was used to account for the correlation due to clustering of patients within hospitals.
      • Liang K.Y.
      • Zeger S.L.
      Longitudinal data analysis using generalized linear models.
      In sensitivity analyses, we adjusted for hospital characteristics: trauma level designation, teaching vs. non-teaching hospital, urban vs. rural location, and presence vs. absence of an approved residency program. These sensitivity analyses were limited to hospitals with available data.

      Secondary analyses

      In the subset of patients with only ETCO2 monitoring reported at the index event, we compared patient survival outcomes between events where the reported ETCO2 was >10 mmHg during CPR vs. events where the reported ETCO2 was lower (≤10 mmHg) using a logistic regression model with a robust variance estimator to account for within-hospital correlation and adjusted for age, sex, race, patient illness category, night or weekend event
      • Peberdy M.A.
      • Ornato J.P.
      • Larkin G.L.
      • et al.
      Survival from in-hospital cardiac arrest during nights and weekends.
      , and first documented rhythm.
      • Meaney P.A.
      • Nadkarni V.M.
      • Kern K.B.
      • Indik J.H.
      • Halperin H.R.
      • Berg R.A.
      Rhythms and outcomes of adult in-hospital cardiac arrest.
      • Nadkarni V.M.
      • Larkin G.L.
      • Peberdy M.A.
      • et al.
      First documented rhythm and clinical outcome from in-hospital cardiac arrest among children and adults.
      We also evaluated the effect of the release of the 2010 AHA guidelines – which recommended continuous ETCO2 monitoring during CPR
      • Neumar R.W.
      • Otto C.W.
      • Link M.S.
      • et al.
      Part 8: Adult advanced cardiovascular life support: 2010 American Heart Association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care.
      – on the proportion of patients with clinician-reported ETCO2 monitoring at the index event. The independent variable in these models was 2006–2010 period vs. 2011–2012. A similar analysis was not completed for DBP monitoring because it was not part of the AHA recommendations until the 2013 CPR Quality Consensus Statement;
      • Meaney P.A.
      • Bobrow B.J.
      • Mancini M.E.
      • et al.
      Cardiopulmonary resuscitation quality: Improving cardiac resuscitation outcomes both inside and outside the hospital: a consensus statement from the American Heart Association.
      however, we performed an exploratory analysis to see if the reported use of DBP monitoring changed over time in those patients with an arterial catheter in place at the time of the arrest.
      All hypothesis tests were two-sided. All secondary analyses were performed among the unmatched cohort using logistic regression models with a robust variance estimator, as above. All analyses were completed using R 3.1.1 (R Development Core Team, Vienna, Austria), including the MatchIt and geepack extension packages.

      Data integrity

      To ensure integrity of the data, the GWTG-R registry uses a detailed periodic reabstraction process, whereby participating hospitals submit randomly selected records that are reabstracted and reviewed for errors by the registry's oversight committee. Mean (standard deviation) error rates for all data were 2.5% (2.7%). Remediation through a web-based process to support data integrity was continuously available for all sites. Any new hospitals were certified by testing of the accuracy of their data collection prior to data submission centrally.

      Results

      Between January 1, 2006 and September 7, 2012, there were 64,556 index adult CPR events with either an invasive airway or arterial catheter in place at the time of the arrest, of which 23,429 (36%) met inclusion criteria (Fig. 1). Of 21,375 index events with an invasive airway in place at the time of the arrest, clinicians reported using ETCO2 to monitor quality in 803 (4%). Compared to earlier events, clinician-reported use of ETCO2 to monitor resuscitation quality was more common in index events occurring in 2011 and later (5% vs. 3%; OR 1.59, CI95 1.30–1.94, p < 0.001). Of 7,260 events with an arterial catheter in place at the time of the arrest, clinicians reported using DBP to monitor quality in 2145 (30%). There was no trend in clinician-reported DBP use over time, which ranged between 25% and 35% per year (p = 0.77).

      Physiologic monitoring of CPR quality

      The propensity score balanced the distribution of potential confounders in the matched cohort of index events (Table 1; monitored n = 3032, not monitored n = 6064), with all standardized mean differences <0.1 (Supplementary Table 1). Clinician-reported physiologic monitoring of CPR quality was associated with higher odds of ROSC (OR 1.22, CI95 1.04–1.43, p = 0.017) compared to no monitoring with either ETCO2 or DBP. Survival to hospital discharge (OR 1.04, CI95 0.92–1.18, p = 0.57) and survival with favorable neurological outcome (OR 0.97, CI95 0.75–1.26, p = 0.83) were not different between groups (Table 2).
      Table 1Characteristics of the study cohort. Summaries presented as median (25th, 75th percentile) or n (%).
      Physiologic monitoring of CPR quality
      Yes
      Each participant with either physiologic monitoring of CPR quality was matched to two participants without monitoring based on a propensity score that included characteristics listed in Table 1 and calendar year. All covariates were balanced within the matched cohorts (all standardized mean differences >0.1; see Supplementary Table 1).
      (N = 3032)
      No, matched
      Each participant with either physiologic monitoring of CPR quality was matched to two participants without monitoring based on a propensity score that included characteristics listed in Table 1 and calendar year. All covariates were balanced within the matched cohorts (all standardized mean differences >0.1; see Supplementary Table 1).
      (N = 6064)
      No, all (N = 20,397)
      Demographic characteristics
      Age, y62 (50, 73)62 (51, 73)63 (52, 75)
      Male sex1877 (62)3701 (61)11,824 (58)
      Race
       White1919 (63)3846 (63)12,643 (62)
       Black687 (23)1326 (22)4485 (22)
       Hispanic199 (7)410 (7)1427 (7)
       Other or unknown227 (7)482 (8)1842 (9)
      Pre-existing medical conditions
      Congestive heart failure480 (16)957 (16)2902 (14)
      Acute myocardial infarction553 (18)1167 (19)3332 (16)
      Arrhythmia929 (31)1910 (31)6167 (30)
      Hypotension or hypoperfusion1559 (51)2958 (49)7711 (38)
      Respiratory insufficiency1732 (57)3524 (58)11,770 (58)
      Sepsis or pneumonia945 (31)1839 (30)6689 (33)
      Metabolic or electrolyte abnormality599 (20)1310 (22)4112 (20)
      Renal insufficiency1090 (36)2182 (36)6947 (34)
      Hepatic insufficiency295 (10)682 (11)1966 (10)
      Central nervous system depression354 (12)711 (12)2741 (13)
      Major trauma314 (10)557 (9)1625 (8)
      Cancer346 (11)670 (11)2328 (11)
      Characteristics before cardiac arrest event
      Patient illness category
       Medical cardiac843 (28)1785 (29)5840 (29)
       Medical non-cardiac944 (31)1894 (31)8836 (43)
       Surgical cardiac440 (15)898 (15)1790 (9)
       Surgical non-cardiac525 (17)1006 (17)2576 (13)
       Trauma270 (9)463 (8)1325 (6)
       Other or unknown10 (<1)18 (<1)30 (<1)
      Pulse oximeter placement2948 (97)5892 (97)19014 (93)
      Electrocardiography monitoring3006 (99)6012 (99)20,002 (98)
      Arterial catheter placement2224 (73)4465 (74)5481 (27)
      Assisted or mechanical ventilation2410 (79)4704 (78)15,070 (74)
      Vascular access2992 (99)5973 (99)19,827 (97)
      Vasoactive infusion1998 (66)3948 (65)11,023 (54)
      Characteristics of cardiac arrest event
      Location of event
       Intensive care unit2279 (75)4516 (75)13,540 (66)
       Inpatient monitored612 (20)1240 (20)3868 (19)
       Inpatient ward14 (<1)33 (1)422 (2)
       Emergency department85 (3)183 (3)2151 (11)
       Other or unknown42 (1)92 (2)416 (2)
      Night or weekend event
      Night: 11:00pm to 6:59am; Weekend: Friday 11:00pm to Monday 6:59am.
      1434 (47)2882 (48)10,117 (50)
      Immediate cause of arrest
       Acute myocardial infarction248 (8)526 (9)1684 (8)
       Hypotension or hypoperfusion1931 (64)3795 (63)10,843 (53)
       Acute respiratory insufficiency842 (28)1755 (29)6572 (32)
       Inadequate invasive airway56 (2)120 (2)686 (3)
       Acute pulmonary edema54 (2)105 (2)345 (2)
       Acute pulmonary embolism61 (2)100 (2)364 (2)
       Metabolic or electrolyte abnormality598 (20)1322 (22)3980 (20)
      First documented rhythm
       Asystole or pulseless electrical activity2362 (78)4615 (76)15,936 (78)
       Ventricular fibrillation or tachycardia576 (19)1239 (20)3485 (17)
       Number of shocks, #2 (1, 4)2 (1, 4)12 (5, 24)
       Bradycardia21 (1)43 (1)104 (1)
       Other or unknown73 (2)167 (3)872 (4)
       Duration of resuscitation, min15 (7, 27)14 (6, 27)13 (6, 25)
      Pharmacologic interventions during arrest
      Epinephrine bolus2719 (90)5393 (89)18,408 (90)
      Sodium bicarbonate1797 (59)3566 (59)11,683 (57)
      Calcium chloride or gluconate1146 (38)2227 (37)6542 (32)
      Atropine1889 (62)3690 (61)13,282 (65)
      Fluid bolus1104 (36)2115 (35)6380 (31)
      Lidocaine244 (8)502 (8)1383 (7)
      Amiodarone675 (22)1406 (23)3949 (19)
      Magnesium sulfate366 (12)702 (12)1927 (9)
      Dextrose bolus147 (5)334 (6)1032 (5)
      a Each participant with either physiologic monitoring of CPR quality was matched to two participants without monitoring based on a propensity score that included characteristics listed in Table 1 and calendar year. All covariates were balanced within the matched cohorts (all standardized mean differences >0.1; see Supplementary Table 1).
      b Night: 11:00 pm to 6:59 am; Weekend: Friday 11:00 pm to Monday 6:59 am.
      Table 2Comparison of resuscitation outcomes among adults with and without monitoring.
      Each participant with monitoring was matched to two participants without monitoring based on a propensity score that included all variables listed in Supplementary Table 1.
      OutcomeWith n (%)Without n (%)OR (CI95)
      Odds ratios compare the odds of an outcome between index events with vs. without monitoring, estimated from a logistic regression model with a robust variance estimator to account for within-hospital correlation.
      p Value
      Return of spontaneous circulation2117/3032 (70)3973/6064 (66)1.22 (1.04, 1.43)0.017
      Survival to hospital discharge527/3020 (17)1019/6022 (17)1.04 (0.91, 1.18)0.57
      Survival with favorable neurological outcome368/2911 (13)757/5840 (13)0.97 (0.75, 1.26)0.83
      OR, odds ratio; CI, confidence interval.
      a Each participant with monitoring was matched to two participants without monitoring based on a propensity score that included all variables listed in Supplementary Table 1.
      b Odds ratios compare the odds of an outcome between index events with vs. without monitoring, estimated from a logistic regression model with a robust variance estimator to account for within-hospital correlation.
      Of the 334 hospitals that contributed data to this analysis, 262 (78%) had data available on hospital characteristics. In this subset of hospitals, monitoring was associated with increased odds of ROSC (OR 1.23, CI95 1.04–1.45, p = 0.015). In sensitivity analyses, adjustment for hospital characteristics did not impact the association between monitoring and ROSC (OR 1.22, CI95 1.04–1.43, p = 0.013).
      In the subset of index events with ETCO2 CPR quality monitoring indicated (n = 803), an ETCO2 >10 mmHg during CPR was reported in 520 (65%) index events, and was associated with improved survival to hospital discharge (24% vs. 11%; OR 2.41, CI95 1.35–4.30, p = 0.003), and improved survival with favorable neurological outcome (18% vs. 8%; OR 2.31, CI95 1.31–4.09, p = 0.004) compared to index events with ETCO2 ≤10 mmHg after accounting for within-hospital correlation and adjusting for age, sex, race, patient illness category, night or weekend event, and first pulseless rhythm (Table 3).
      Table 3Comparison of resuscitation outcomes among index events with ETCO2 monitoring (n = 803 index events).
      OutcomeETCO2 >10 mmHg n (%)ETCO2 ≤10 mmHg n (%)OR (CI95)
      In the subset of index events with ETCO2 monitoring indicated, odds ratios compare the odds of an outcome between index events with a reported ETCO2 >10mmHg vs. events with a reported ETCO2 ≤10mmHg, estimated from a logistic regression model with a robust variance estimator to account for within-hospital correlation and adjusted for age, sex, race, patient illness category, night or weekend event, and first pulseless rhythm.
      p-Value
      Survival to hospital discharge125/520 (24)31/283 (11)2.41 (1.35, 4.30)0.003
      Survival with favorable neurological outcome94/520 (18)23/283 (8)2.31 (1.31, 4.09)0.004
      OR, odds ratio; CI, confidence interval.
      a In the subset of index events with ETCO2 monitoring indicated, odds ratios compare the odds of an outcome between index events with a reported ETCO2 >10 mmHg vs. events with a reported ETCO2 ≤10 mmHg, estimated from a logistic regression model with a robust variance estimator to account for within-hospital correlation and adjusted for age, sex, race, patient illness category, night or weekend event, and first pulseless rhythm.

      Discussion

      In this propensity-matched-cohort study of adults using the AHA's multicenter GWTG-R registry, clinician-reported use of physiology to monitor CPR quality with either ETCO2 or DBP was associated with improved rates of ROSC. Survival to hospital discharge and survival with favorable neurological outcome were not different between groups. In an adjusted model using the subset of 803 index events during which ETCO2 monitoring was reported as used, an ETCO2 >10 mmHg during CPR was associated with a higher rate of patient survival to hospital discharge and survival with favorable neurological outcome compared to events when ETCO2 >10 mmHg during CPR was not reported. Of note, in the subset of events with an invasive airway in place at the time of arrest, it was not common (4%) for clinicians to report using ETCO2 to monitor CPR quality.
      In a recent consensus statement published in Circulation, the AHA now recommends monitoring the patient's physiologic response to the resuscitation effort using ETCO2 and invasive hemodynamic data.
      • Meaney P.A.
      • Bobrow B.J.
      • Mancini M.E.
      • et al.
      Cardiopulmonary resuscitation quality: Improving cardiac resuscitation outcomes both inside and outside the hospital: a consensus statement from the American Heart Association.
      In a hierarchal fashion depending on what is available at the time of the arrest, hemodynamic goals were prioritized over capnography goals. Despite the substantial amount of data that was evaluated in this consensus statement, the authors conceded that clinical studies “supporting the optimal titration of these parameters during human CPR are lacking.” Therefore, this study takes an important first step by providing prospectively collected human data and associating clinician-reported intra-arrest physiologic monitoring of CPR quality with superior clinical outcomes.
      Numerous experimental models have demonstrated that ETCO2 correlates well with pulmonary blood flow, cardiac output, and resuscitation success.
      • Levine R.L.
      • Wayne M.A.
      • Miller C.C.
      End-tidal carbon dioxide and outcome of out-of-hospital cardiac arrest.
      • Sanders A.B.
      • Atlas M.
      • Ewy G.A.
      • Kern K.B.
      • Bragg S.
      Expired PCO2 as an index of coronary perfusion pressure.
      • Sanders A.B.
      • Ewy G.A.
      • Bragg S.
      • Atlas M.
      • Kern K.B.
      Expired PCO2 as a prognostic indicator of successful resuscitation from cardiac arrest.
      • Weil M.H.
      • Bisera J.
      • Trevino R.P.
      • Rackow E.C.
      Cardiac output and end-tidal carbon dioxide.
      Clinical data supporting ETCO2 monitoring during CPR dates back to the 1970s when Kalenda described using ETCO2 levels to monitor for rescuer fatigue (declining ETCO2) and for ROSC (sudden rise in ETCO2).
      • Kalenda Z.
      The capnogram as a guide to the efficacy of cardiac massage.
      Similarly, but more recently, Sheak et al. demonstrated that CPR quality is associated with ETCO2 levels during both in- and out-of-hospital resuscitation attempts.
      • Sheak K.R.
      • Wiebe D.J.
      • Leary M.
      • et al.
      Quantitative relationship between end-tidal carbon dioxide and CPR quality during both in-hospital and out-of-hospital cardiac arrest.
      Other studies have suggested that ETCO2 can be used as a prognostic tool during cardiac arrest, as failure to achieve an ETCO2 of at least 10 mmHg is rarely associated with good outcome.
      • Sanders A.B.
      • Kern K.B.
      • Otto C.W.
      • Milander M.M.
      • Ewy G.A.
      End-tidal carbon dioxide monitoring during cardiopulmonary resuscitation. A prognostic indicator for survival.
      • Levine R.L.
      • Wayne M.A.
      • Miller C.C.
      End-tidal carbon dioxide and outcome of out-of-hospital cardiac arrest.
      Similar to these previous reports, we now provide data suggesting that when clinicians monitor resuscitation quality with ETCO2, rates of ROSC are higher, and further, than when an ETCO2 >10 mmHg was reported during CPR, rates of survival to discharge and survival with good neurological are also higher.
      In the AHA consensus statement, hemodynamic goals were prioritized over ETCO2 based upon a large body of experimental literature that has established myocardial blood flow as the primary determinant of successful cardiac arrest resuscitation.
      • Kern K.B.
      • Ewy G.A.
      • Voorhees W.D.
      • Babbs C.F.
      • Tacker W.A.
      Myocardial perfusion pressure: a predictor of 24-hour survival during prolonged cardiac arrest in dogs.
      • Ralston S.H.
      • Voorhees W.D.
      • Babbs C.F.
      Intrapulmonary epinephrine during prolonged cardiopulmonary resuscitation: improved regional blood flow and resuscitation in dogs.
      • Halperin H.R.
      • Lee K.
      • Zviman M.
      • et al.
      Outcomes from low versus high-flow cardiopulmonary resuscitation in a swine model of cardiac arrest.
      Measures of myocardial blood flow are not easily made available during the resuscitation attempt; therefore, coronary perfusion pressure (CPP; the primary determinant of myocardial blood flow) and arterial DBP (the primary driving force of CPP) were recommended first line as physiologic resuscitation quality monitors because they are most closely related to myocardial blood flow.
      • Sanders A.B.
      • Ewy G.A.
      • Taft T.V.
      Prognostic and therapeutic importance of the aortic diastolic pressure in resuscitation from cardiac arrest.
      Supporting this recommendation, there is also a growing body of pre-clinical evidence demonstrating that titration of CPR efforts to arterial blood pressure can improve survival.
      • Sutton R.M.
      • Friess S.H.
      • Naim M.Y.
      • et al.
      Patient-centric blood pressure-targeted cardiopulmonary resuscitation improves survival from cardiac arrest.
      • Friess S.H.
      • Sutton R.M.
      • French B.
      • et al.
      Hemodynamic directed CPR improves cerebral perfusion pressure and brain tissue oxygenation.
      • Sutton R.M.
      • Friess S.H.
      • Bhalala U.
      • et al.
      Hemodynamic directed CPR improves short-term survival from asphyxia-associated cardiac arrest.
      Highlighting a potential limitation of these data, we would have been underpowered to detect an association between individual monitoring modalities and outcome in our study cohorts; therefore, we a priori chose to define physiologic monitoring as use of either ETCO2 or DBP. Consequently, further quantitative human investigations are warranted to not only determine the effectiveness of physiologic monitoring to improve survival outcomes, but also which monitor, and further, which physiologic target is most associated with outcome.
      The relatively low reported use of physiologic monitoring to guide resuscitation quality deserves comment. Of index events with an invasive airway in place at the time of the arrest, clinicians reported using ETCO2 to monitor quality in only 803 (4%). Because our dataset was comprised of many events that occurred prior to release of the 2010 AHA guidelines
      • Neumar R.W.
      • Otto C.W.
      • Link M.S.
      • et al.
      Part 8: Adult advanced cardiovascular life support: 2010 American Heart Association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care.
      , which recommended the use of ETCO2 to monitor quality, low rates of clinician reported ETCO2 monitoring are not surprising. Similarly, among events with an arterial catheter in place, only about 1/3 (30%) of clinicians reported using DBP to monitor CPR quality. DBP monitoring has also not changed significantly over time, likely because all of the index events in the dataset occurred before the release of the 2013 AHA consensus statement which prioritized hemodynamic monitoring.
      • Meaney P.A.
      • Bobrow B.J.
      • Mancini M.E.
      • et al.
      Cardiopulmonary resuscitation quality: Improving cardiac resuscitation outcomes both inside and outside the hospital: a consensus statement from the American Heart Association.
      Interestingly, there was an approximately 50% relative increase in the reported use of ETCO2 to monitor quality at index events that occurred after release of the 2010 guidelines. However, the overall percentage for both still remains low, and suggests that educational efforts to increase physiologic monitoring during resuscitation are warranted.
      This study has limitations. First, as with all studies of large multicenter databases, analysis of the data may be limited by data integrity and validation issues due to multiple sites submitting data to the registry. However, the aforementioned rigorous training and certification process, reabstraction data validation processes, and large sample size, which are unique attributes of the AHA GWTG-R registry, minimize these sources of bias. Second, because important confounding post-resuscitation variables
      • Kilgannon J.H.
      • Roberts B.W.
      • Reihl L.R.
      • et al.
      Early arterial hypotension is common in the post-cardiac arrest syndrome and associated with increased in-hospital mortality.
      • Kilgannon J.H.
      • Jones A.E.
      • Parrillo J.E.
      • et al.
      Relationship between supranormal oxygen tension and outcome after resuscitation from cardiac arrest.
      • Kilgannon J.H.
      • Jones A.E.
      • Shapiro N.I.
      • et al.
      Association between arterial hyperoxia following resuscitation from cardiac arrest and in-hospital mortality.
      • Roberts B.W.
      • Kilgannon J.H.
      • Chansky M.E.
      • Mittal N.
      • Wooden J.
      • Trzeciak S.
      Association between postresuscitation partial pressure of arterial carbon dioxide and neurological outcome in patients with post-cardiac arrest syndrome.
      • Arrich J.
      • Holzer M.
      • Havel C.
      • Mullner M.
      • Herkner H.
      Hypothermia for neuroprotection in adults after cardiopulmonary resuscitation.
      were not available for the analysis, the effect of physiologic monitoring during CPR on long term survival was not evaluated. Third, actual values of ETCO2 or DBP, which could have been used to verify the presence of an end tidal or arterial line monitor, and also the level of ETCO2 achieved during CPR, are not routinely recorded in the GWTG-R database. Next, as the section on CPR quality in the GWTG-R database is optional, generalizability outside of hospitals that report this data may be limited. Fifth, it is worth noting that ventilation rate is an important unmeasured confounder. It is both associated with ETCO2 levels during human CPR
      • Sheak K.R.
      • Wiebe D.J.
      • Leary M.
      • et al.
      Quantitative relationship between end-tidal carbon dioxide and CPR quality during both in-hospital and out-of-hospital cardiac arrest.
      and cardiac arrest survival in animal models
      • Aufderheide T.P.
      • Sigurdsson G.
      • Pirrallo R.G.
      • et al.
      Hyperventilation-induced hypotension during cardiopulmonary resuscitation.
      , with overventilation associated with mortality. Therefore, the association between higher reported ETCO2 levels and improved survival may be driven partly by the avoidance of excessive ventilation during CPR. Finally, because only about 15% of hospitals in the US are represented, the generalizability of our findings outside of GWTG-R hospitals remains an unanswered question.

      Conclusions

      In this propensity-matched-cohort study of adult in-hospital cardiac arrest using the AHA's GWTG-R registry, clinician-reported use of physiologic monitoring of CPR quality with either continuous ETCO2 or diastolic blood pressure was associated with an improved rate of ROSC compared to no reported physiologic monitoring. However, survival to hospital discharge and survival with favorable neurological outcome were not different between groups. In the subset of index events where CPR quality was monitored with only ETCO2, there were improved rates of patient survival to hospital discharge and survival with favorable neurological outcome when ETCO2 was reported to be >10 mmHg during CPR monitoring. It was not common for clinicians to report the use of physiology to guide resuscitation quality.

      Conflicts of interest statement

      Robert Sutton has received a speaker's honoraria from the Zoll Medical Corporation. Raina Merchant has received grant/research support from the NIH (K23 10714038; R01 HL122457), and pilot funding from Physio-Control (Seattle, WA), Zoll Medical (Boston MA), Cardiac Science (Bothell, WA), and Philips Medical (Seattle, WA). Benjamin Abella has received research funding from the NIH, AHA, PCORI, and Medtronic Foundation, speaking honoraria from Philips Healthcare (Seattle, WA) and CR Bard (Murray Hill, NJ), and has an advisory board relationship with CardioReady (Conshohocken, PA). Dana Edelson has received research funding from Philips Healthcare (Seattle, WA) and speaking fees from Physio-Control (Seattle, WA). Christopher Parshuram is a named inventor of the Bedside Paediatric Early Warning System, and holds shares in Bedside Clinical Systems. Benjamin French, Peter Meaney, Stephen Schexnayder, Robert Berg, Melania Bembea, and Vinay Nadkarni have no disclosures or funding related to the project.

      Acknowledgements

      Robert Sutton, Alexis Topjian, and Melania Bembea are supported by career development awards (RMS: National Institute of Child Heath and Human Development; AAT and MB National Institute of Neurological Disorders and Stroke).

      Appendix A. Supplementary data

      The following are the supplementary data to this article:

      References

        • Merchant R.M.
        • Yang L.
        • Becker L.B.
        • et al.
        Incidence of treated cardiac arrest in hospitalized patients in the United States.
        Crit Care Med. 2011; 39: 2401-2406
        • Meaney P.A.
        • Nadkarni V.M.
        • Kern K.B.
        • Indik J.H.
        • Halperin H.R.
        • Berg R.A.
        Rhythms and outcomes of adult in-hospital cardiac arrest.
        Crit Care Med. 2010; 38: 101-108
        • Girotra S.
        • Nallamothu B.K.
        • Spertus J.A.
        • et al.
        Trends in survival after in-hospital cardiac arrest.
        N Engl J Med. 2012; 367: 1912-1920
        • Goldberger Z.D.
        • Chan P.S.
        • Berg R.A.
        • et al.
        Duration of resuscitation efforts and survival after in-hospital cardiac arrest: an observational study.
        Lancet. 2012; 380: 1473-1481
        • Sutton R.M.
        • Friess S.H.
        • Naim M.Y.
        • et al.
        Patient-centric blood pressure-targeted cardiopulmonary resuscitation improves survival from cardiac arrest.
        Am J Respir Crit Care Med. 2014; 190: 1255-1262
        • Friess S.H.
        • Sutton R.M.
        • French B.
        • et al.
        Hemodynamic directed CPR improves cerebral perfusion pressure and brain tissue oxygenation.
        Resuscitation. 2014; 85: 1298-1303
        • Kern K.B.
        • Ewy G.A.
        • Voorhees W.D.
        • Babbs C.F.
        • Tacker W.A.
        Myocardial perfusion pressure: a predictor of 24-hour survival during prolonged cardiac arrest in dogs.
        Resuscitation. 1988; 16: 241-250
        • Ralston S.H.
        • Voorhees W.D.
        • Babbs C.F.
        Intrapulmonary epinephrine during prolonged cardiopulmonary resuscitation: improved regional blood flow and resuscitation in dogs.
        Ann Emerg Med. 1984; 13: 79-86
        • Halperin H.R.
        • Lee K.
        • Zviman M.
        • et al.
        Outcomes from low versus high-flow cardiopulmonary resuscitation in a swine model of cardiac arrest.
        Am J Emerg Med. 2010; 28: 195-202
        • Meaney P.A.
        • Bobrow B.J.
        • Mancini M.E.
        • et al.
        Cardiopulmonary resuscitation quality: Improving cardiac resuscitation outcomes both inside and outside the hospital: a consensus statement from the American Heart Association.
        Circulation. 2013; 128: 417-435
        • Jacobs I.
        • Nadkarni V.
        • Bahr J.
        • et al.
        Cardiac arrest and cardiopulmonary resuscitation outcome reports: update and simplification of the Utstein templates for resuscitation registries. A statement for healthcare professionals from a task force of the International Liaison Committee on Resuscitation (American Heart Association, European Resuscitation Council, Australian Resuscitation Council, New Zealand Resuscitation Council, Heart and Stroke Foundation of Canada, InterAmerican Heart Foundation, Resuscitation Council of Southern Africa).
        Resuscitation. 2004; 63: 233-249
        • Cummins R.O.
        • Chamberlain D.
        • Hazinski M.F.
        • et al.
        Recommended guidelines for reviewing, reporting, and conducting research on in-hospital resuscitation: The in-hospital ‘Utstein style’. A statement for healthcare professionals from the American Heart Association, the European Resuscitation Council, the Heart and Stroke Foundation of Canada, the Australian Resuscitation Council, and the Resuscitation Councils of Southern Africa.
        Resuscitation. 1997; 34: 151-183
        • Peberdy M.A.
        • Kaye W.
        • Ornato J.P.
        • et al.
        Cardiopulmonary resuscitation of adults in the hospital: A report of 14720 cardiac arrests from the National Registry of Cardiopulmonary Resuscitation.
        Resuscitation. 2003; 58: 297-308
        • Chan P.S.
        • Krumholz H.M.
        • Nichol G.
        • Nallamothu B.K.
        • American Heart Association National Registry of Cardiopulmonary Resuscitation Investigators
        Delayed time to defibrillation after in-hospital cardiac arrest.
        N Engl J Med. 2008; 358: 9-17
        • Nadkarni V.M.
        • Larkin G.L.
        • Peberdy M.A.
        • et al.
        First documented rhythm and clinical outcome from in-hospital cardiac arrest among children and adults.
        JAMA. 2006; 295: 50-57
        • Sanders A.B.
        • Kern K.B.
        • Otto C.W.
        • Milander M.M.
        • Ewy G.A.
        End-tidal carbon dioxide monitoring during cardiopulmonary resuscitation. A prognostic indicator for survival.
        JAMA. 1989; 262: 1347-1351
        • Levine R.L.
        • Wayne M.A.
        • Miller C.C.
        End-tidal carbon dioxide and outcome of out-of-hospital cardiac arrest.
        N Engl J Med. 1997; 337: 301-306
        • Razi R.R.
        • Churpek M.M.
        • Yuen T.C.
        • et al.
        Racial disparities in outcomes following PEA and asystole in-hospital cardiac arrests.
        Resuscitation. 2015; 87: 69-74
        • Topjian A.A.
        • Localio A.R.
        • Berg R.A.
        • et al.
        Women of child-bearing age have better inhospital cardiac arrest survival outcomes than do equal-aged men.
        Crit Care Med. 2010; 38: 1254-1260
        • Brookhart M.A.
        • Schneeweiss S.
        • Rothman K.J.
        • Glynn R.J.
        • Avorn J.
        • Sturmer T.
        Variable selection for propensity score models.
        Am J Epidemiol. 2006; 163: 1149-1156
        • D’Agostino R.B.
        • Rubin D.B.
        Estimating and using propensity scores with partially missing data.
        J Am Stat Assoc. 2000; 95: 749-759
        • Harder V.S.
        • Stuart E.A.
        • Anthony J.C.
        Propensity score techniques and the assessment of measured covariate balance to test causal associations in psychological research.
        Psychol Methods. 2010; 15: 234-249
        • Liang K.Y.
        • Zeger S.L.
        Longitudinal data analysis using generalized linear models.
        Biometrika. 1986; 73: 13-22
        • Peberdy M.A.
        • Ornato J.P.
        • Larkin G.L.
        • et al.
        Survival from in-hospital cardiac arrest during nights and weekends.
        JAMA. 2008; 299: 785-792
        • Neumar R.W.
        • Otto C.W.
        • Link M.S.
        • et al.
        Part 8: Adult advanced cardiovascular life support: 2010 American Heart Association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care.
        Circulation. 2010; 122: S729-S767
        • Sanders A.B.
        • Atlas M.
        • Ewy G.A.
        • Kern K.B.
        • Bragg S.
        Expired PCO2 as an index of coronary perfusion pressure.
        Am J Emerg Med. 1985; 3: 147-149
        • Sanders A.B.
        • Ewy G.A.
        • Bragg S.
        • Atlas M.
        • Kern K.B.
        Expired PCO2 as a prognostic indicator of successful resuscitation from cardiac arrest.
        Ann Emerg Med. 1985; 14: 948-952
        • Weil M.H.
        • Bisera J.
        • Trevino R.P.
        • Rackow E.C.
        Cardiac output and end-tidal carbon dioxide.
        Crit Care Med. 1985; 13: 907-909
        • Kalenda Z.
        The capnogram as a guide to the efficacy of cardiac massage.
        Resuscitation. 1978; 6: 259-263
        • Sheak K.R.
        • Wiebe D.J.
        • Leary M.
        • et al.
        Quantitative relationship between end-tidal carbon dioxide and CPR quality during both in-hospital and out-of-hospital cardiac arrest.
        Resuscitation. 2015; 89: 149-154
        • Sanders A.B.
        • Ewy G.A.
        • Taft T.V.
        Prognostic and therapeutic importance of the aortic diastolic pressure in resuscitation from cardiac arrest.
        Crit Care Med. 1984; 12: 871-873
        • Sutton R.M.
        • Friess S.H.
        • Bhalala U.
        • et al.
        Hemodynamic directed CPR improves short-term survival from asphyxia-associated cardiac arrest.
        Resuscitation. 2013; 84: 696-701
        • Kilgannon J.H.
        • Roberts B.W.
        • Reihl L.R.
        • et al.
        Early arterial hypotension is common in the post-cardiac arrest syndrome and associated with increased in-hospital mortality.
        Resuscitation. 2008; 79: 410-416
        • Kilgannon J.H.
        • Jones A.E.
        • Parrillo J.E.
        • et al.
        Relationship between supranormal oxygen tension and outcome after resuscitation from cardiac arrest.
        Circulation. 2011; 123: 2717-2722
        • Kilgannon J.H.
        • Jones A.E.
        • Shapiro N.I.
        • et al.
        Association between arterial hyperoxia following resuscitation from cardiac arrest and in-hospital mortality.
        JAMA. 2010; 303: 2165-2171
        • Roberts B.W.
        • Kilgannon J.H.
        • Chansky M.E.
        • Mittal N.
        • Wooden J.
        • Trzeciak S.
        Association between postresuscitation partial pressure of arterial carbon dioxide and neurological outcome in patients with post-cardiac arrest syndrome.
        Circulation. 2013; 127: 2107-2113
        • Arrich J.
        • Holzer M.
        • Havel C.
        • Mullner M.
        • Herkner H.
        Hypothermia for neuroprotection in adults after cardiopulmonary resuscitation.
        Cochrane Database Syst Rev. 2012; 9: CD004128
        • Aufderheide T.P.
        • Sigurdsson G.
        • Pirrallo R.G.
        • et al.
        Hyperventilation-induced hypotension during cardiopulmonary resuscitation.
        Circulation. 2004; 109: 1960-1965