Secondary Logo

Journal Logo

Critical Care and Trauma: Research Report

Outcomes of Cardiopulmonary Resuscitation and Predictors of Survival in Patients Undergoing Coronary Angiography Including Percutaneous Coronary Interventions

Sprung, Juraj MD, PhD*; Ritter, Matthew J. MD*; Rihal, Charanjit S. MD; Warner, Mary E. MD*; Wilson, Gregory A. CCRP*; Williams, Brent A. MS; Stevens, Susanna R. BS; Schroeder, Darrell R. MS; Bourke, Denis L. MD§; Warner, David O. MD*

Editor(s): Takala, Jukka

Author Information
doi: 10.1213/01.ane.0000189082.54614.26
  • Free

Analysis of cardiac arrests that occur during invasive procedures has proved useful in evaluating and improving our medical practice. For example, studies of cardiac arrests in anesthetized surgical patients have provided valuable insights into anesthesia practice (1–5), including risk factors associated with arrests (5). This information has led to changes in practice that have resulted in a reduced frequency of cardiac arrests (4,6). A similar analysis has heretofore not been undertaken for cardiac arrests occurring during coronary angiography (CA) or percutaneous coronary intervention (PCI), and long-term outcomes of these patients have not been reported. This is of particular interest to anesthesiologists because the cardiac arrest rate is considerably more frequent during CA (22 per 10,000) than in the operating room (4.2 per 10,000) (5). The overall goal of this study was to analyze outcomes of cardiac arrests that occurred during or immediately after CA/PCI over an 11-yr period in a tertiary cardiovascular referral center. During this study period, 85% of all CA/PCI procedures were attended by a team from the Department of Anesthesiology. We hypothesized that the immediate presence of both an anesthesiologist (for airway management) and an interventional cardiologist (e.g., for defibrillation, pacing, stenting) would result in a relatively infrequent cardiac arrest and a generally favorable outcome, despite the presence of significant coronary artery disease in many of these patients and the invasive nature of these procedures.

Our specific aims were 1) to determine if the frequency of arrests requiring cardiopulmonary resuscitation (CPR) has changed over time; 2) to identify the characteristics associated with survival after arrest; and 3) to determine the long-term prognostic significance of surviving a periprocedural cardiac arrest.


In our institution, each cardiac arrest that occurs while a patient is receiving anesthesia services is reviewed as a part of quality improvement activities. Shortly after the event, participants are interviewed to ensure the accuracy of the information. With IRB approval, the records of all patients experiencing a cardiac arrest requiring full CPR during CA/PCI between January 1, 1990, and December 31, 2000, were examined. For the first 8 yr of the study all CA/PCI catheterizations were attended by staff from the Department of Anesthesiology (resident or certified nurse anesthetist [CRNA] supervised by anesthesiologist). CRNA providers have 2-yr formal anesthesia training. In 1998, “sedation nurses” (nonanesthesia trained personnel working under the direction of the interventional cardiologist) started to attend cardiac catheterizations. In 1998, 1.4% of all CA/PCIs (95 of 6661 procedures) were performed by sedation nurses. That percentage increased to 40.9% (2727 of 6674 procedures) in 1999 and 75.3% (4983 of 6617 procedures) in 2000. None of the catheterizations attended by nonanesthesia trained sedation nurses were included in the analyses. To ensure that all eligible cardiac arrests were included, the Institutional Clinic Diagnostic Index was searched for patients who were coded as “death,” “unexplained intensive care unit admission,” or “unexpected intubation.” This review did not discover any additional eligible cardiac arrests.

“Cardiac arrest” was defined as an event that required closed-chest compressions (full CPR). Treatment of cardiac arrests followed the American Heart Association Advanced Cardiac Life Support guidelines that were current at the time. We did not include patients who experienced arrhythmias not requiring chest compressions, those who were brought to the catheterization lab in extremis, or those with whom the anesthesia team became involved only after an arrest had occurred. To assess inter-rater reliability, 2 of the authors collected data independently from 10 charts and confirmed the uniformity of data extraction. The same authors reviewed the rest of the medical records and resolved any conflicts by consensus. The two main outcome variables studied were survival to discharge from the postanesthesia recovery room, which we termed procedural survival (PS), and survival to hospital discharge (HS). Using this information, two separate analyses were performed. First, trends in the incidence of cardiac arrest and the characteristics associated with PS and HS were examined. Then, a matched cohort analysis was undertaken to determine the long-term prognostic significance of a procedural cardiac arrest.

We analyzed patient and procedural characteristics associated with PS and HS after cardiac arrest during or immediately after CA/PCI. Patient-related factors included age, gender, ASA physical status score (ASA-PS), and primary comorbid conditions as defined by Hosking et al. (7). ASA-PS scores were categorized as 1, 2, or 3 versus 4 or 5. Diagnostic categories of comorbid conditions included the type of cardiovascular disease (congestive heart failure [CHF], hypertension, significant valvular disease, dysrhythmias, current angina, prior myocardial infarction, prior coronary artery bypass grafting [CABG], or prior percutaneous coronary angioplasty [PTCA]) and diabetes mellitus. The extent of coronary artery disease was quantified as significant left main artery stenosis or significant disease present in one, two, or three main vessels. Only occlusions ≥70% were considered, and multiple lesions in a single artery were counted only once. Procedure-related factors included the type of anesthesia (general anesthesia or monitored anesthesia care with sedation) and the urgency of the procedure (elective or emergent). Procedural categories included intracoronary stenting, PTCA (including atherectomy), or diagnostic angiography only. Combined PTCA and stenting was categorized as a stenting procedure for the purpose of the analysis. Other variables considered included the length of the procedure before arrest and hemodynamic instability on arrival to the catheterization lab, defined as a hemodynamic condition requiring the use of aggressive pharmacologic support and/or intraaortic balloon pump (IABP) before CA/PCI. In addition we studied two subcategories of hemodynamic instability that occurred during the procedure: a) hypotension, defined as a sudden decrease in systolic blood pressure to ≤80 mm Hg that lasted for more than 10 min before the arrest; and b) insertion of IABP before the end of CA/PCI. The timing of the cardiac arrest, either during the procedure or later in the recovery room, was also noted. Finally, the initial electrocardiogram tracing in the absence of a pulse (ventricular fibrillation, asystole, or pulseless electrical activity [PEA]) was noted.

The frequency of cardiac arrests and HS were calculated by year over the study period for all patients. Logistic regression models were developed to determine characteristics associated with PS and HS. Patient age and procedure duration were treated as continuous variables, with all other characteristics treated as categorical variables. For each characteristic, a logistic regression model assessed the association between the characteristic and the outcome. A multivariate logistic regression model was developed using a stepwise variable selection procedure to identify those variables demonstrating the strongest independent associations with the outcomes. All reported characteristics (except hemodynamic instability) were considered in the multivariate models. Hemodynamic instability could not be incorporated in a multivariate model for PS because all the patients who did not experience hemodynamic instability survived the procedure and only 4 of the 33 hemodynamically stable patients did not survive to hospital discharge. Therefore, odds ratio estimates would be very difficult to interpret, and as a consequence hemodynamic instability was not considered in the multivariate models. However, this variable clearly has strong predictive power with respect to both PS and HS. In selecting variables for the multivariate analysis, a significance level of 0.10 in the univariate analysis was required for inclusion in the final model. In other analyses, P values were considered significant if <0.05.

To determine the prognostic significance of procedural cardiac arrest, we compared the long-term survival of HS patients to a matched group of patients who underwent catheterization but did not experience a cardiac arrest. Patients were matched 2:1 for age (±5 yr), gender, date of specific procedure (±1 yr), and emergency status. The medical records of this control patient cohort were reviewed for variables listed in Table 1. Survival was quantified as the time in days from procedure to death or the last date the patient was known to be alive.

Table 1:
Patient Characteristics, Comorbidities, Procedures, and Long-Term Survival in Patients Who Experienced Cardiac Arrest and Survived to Hospital Discharge Compared with Matched Controls

Survival in the two groups was compared using a two-sample log-rank test and then compared with the expected survival in a similarly matched control group from the general Minnesota population using one-sample log-rank tests. Percent survival was estimated at multiple time points using the Kaplan-Meier method, and survival curves were plotted separately for those with and without cardiac arrest. A Cox proportional hazards model was developed to estimate the hazard ratio of mortality between the two groups. An adjusted hazard ratio was estimated after including potential confounders in the model, namely age, comorbidities (hypertension, past or present CHF, past myocardial infarction, past CABG, past PTCA, diabetes), the type of procedure (stent, PTCA, or neither), and the need for emergency CABG. All analyses were performed with use of SAS Software Release 8.2 (SAS Institute Inc., Cary, NC).


Between January 1, 1990, and December 31, 2000, 51,985 patients underwent CA or PCI with the attendance of an anesthesia team. This represents 87% of all CA/PCI procedures performed in our cardiac catheterization lab during that period. During that period the cardiac arrest rate requiring full CPR was 21.9 per 10,000 (n = 114). The frequency of cardiac arrest decreased over the 11-yr study period from 33.9 per 10,000 between January 1, 1990, and December 31, 1994 to 13.1 per 10,000 between 1995 and 2000 inclusive. Of the 114 patients who had CPR, 77.2% were PS and 56.1% were HS. Survival rates did not show consistent temporal trends over the study period.

The results of univariate analysis of characteristics associated with survival in our patients are shown in Table 2. Patient characteristics predicting both PS and HS included lower ASA-PS classification and the absence of a previous CABG. The absence of CHF predicted HS. Several factors were not associated with survival after cardiac arrest. These included 1) the number of major coronary artery vessels affected by significant atherosclerotic disease; 2) prior myocardial infarction; 3) angina; 4) diabetes; 5) age; or 6) gender (Table 2). Procedural characteristics that were associated with decreased HS and PS included longer duration of catheterization, hemodynamic instability, and an initial rhythm other than ventricular fibrillation (Table 3). A significant episode of hypotension (duration ≥10 min before the arrest) and an emergent procedure were related to decreased HS only. Finally, diagnostic-only procedures had a significantly better HS.

Table 2:
Pre-Procedure Patient Characteristics Assessed as Potential Predictors of Survival After Cardiac Arrest (N = 114)
Table 3:
Procedural Characteristics, Timing and Etiology Assessed as Potential Predictors of Survival After Cardiac Arrest (N = 114)

Three characteristics were independent risk factors for decreased PS and HS in the multivariate analysis: prior CABG, emergency catheterization, and a prolonged procedure (Table 4). In addition, an initial rhythm of PEA predicted much less PS but not HS. In contrast, age, hypotension before arrest, and timing of arrest (during the procedure versus in the recovery area) predicted only less HS.

Table 4:
Multivariate Predictors of Survival

Fifty-eight of the 64 HS patients returned to their homes and 6 were admitted to nursing homes. None of these patients exhibited neurologic damage. Long-term survival was compared between the patients who suffered cardiac arrest during catheterization and the matched control group (patients who underwent uneventful catheterization). The two groups were similar in all respects except that the cardiac arrest group was more likely to have had a prior PTCA and was also more likely to have had a PTCA performed during the index catheterization (Table 1). Although survival may appear to be less frequent for the patients who had a cardiac arrest (Fig. 1), that was not the case. Even adjusting for potential confounders, the difference in long-term survival remained statistically insignificant (hazard ratio 1.47, 95% confidence interval 0.88, 2.46, P = 0.14). For comparison, survival for both of these groups was worse (P < 0.01) than predicted for the general Minnesota population of similar age, gender, and calendar year (Fig. 1).

Figure 1.:
Long-term survival in patients after hospital discharge after coronary angiography (CA) and/or percutaneous coronary intervention (PCI). The solid line denotes survival of 64 patients who had periprocedural cardiac arrest (study cohort). The dashed line is the case-control group (patients who had CA/PCI but did not experience cardiac arrest) matched to study cohort according to age, gender, and the level of procedure urgency. The dotted line represents an age- and sex-matched general Minnesota population. Study cohort patients appear to have shorter long-term survival compared with the matched control group, but this was not statistically significant (P = 0.1). Both CA/PCI groups had worse survival rates than the matched general Minnesota population (log-rank test, P < 0.01).


The major findings of this study are 1) the incidence of periprocedural cardiac arrest decreased over the course of the study; 2) more than half the patients who experienced a cardiac arrest survived to be discharged from the hospital, and all were without neurologic damage; 3) survival after cardiac arrest did not depend on the extent of coronary artery disease; 4) for HS patients, the fact that they experienced a cardiac arrest did not adversely affect long-term survival; and 5) the most significant predictors of mortality were prior CABG, hemodynamic instability or hypotension during catheterization, an emergency catheterization, or a procedure of prolonged duration.

With regard to the remarkably frequent rate of short-term survival for the patients in this study as compared with survival after other in-hospital arrests (5,8), we believe two factors are important. First, the patients receive immediate treatment by two specialists experienced at performing successful resuscitations, both with all necessary equipment and medications immediately available to them. Second, for most of the patients in this study, the primary disease process was cardiovascular. In contrast, cardiac arrests in other patient populations may be associated with another terminal condition (malignancy) or be attributable to a comorbid condition (e.g., trauma, sepsis) that may not be effectively reversed with CPR.

Between 1990 and 2001 the incidence of cardiac arrest in our cohort of patients was 21.9 per 10,000. Although direct comparisons are difficult to make, in a similar study of anesthetized noncardiac surgical patients over the same period, the frequency of cardiac arrests was 4.2 per 10,000, with only 35% surviving to hospital discharge (5). When examining the frequency of arrests by year, a dramatic decrease was recorded for the years 1995 to 2000; however, we can only speculate as to possible factors to explain these findings. During this period there was a change from PTCA to intracoronary stenting as the predominant procedure used to treat coronary artery stenosis (9,10). Several authors have suggested that newer coronary procedures with better equipment and techniques have improved various catheterization outcomes (9,11–15).

Of the patients who experienced a cardiac arrest, 56.1% were HS. Although there was year-to-year variation in survival rates, hospital survival between 1990 and 2000 did not show any consistent temporal trend. In the present study, overall in-hospital mortality was 0.96 per 10,000 CA/PCIs. Reported mortality rates from various cardiac catheterization procedures range from 0.07% to 1.9% (16–21), probably reflecting wide variations in patient and procedural mix.

Several of the patient and procedural characteristics associated with HS were consistent with prior studies (21–24). Three of the factors independently associated with decreases in both PS and HS in the multivariate analysis (prior CABG emergency procedure and a prolonged procedure) likely reflect the severity of underlying disease and the acuity of the process necessitating intervention. Hemodynamic instability requiring IABP or vasopressor support before arrest was not considered in the final multivariate analysis because of statistical limitations. However, hemodynamic instability clearly had strong predictive value unlikely to be substantially attenuated by potential confounding factors. Interestingly, the ASA-PS, which predicts the risk of arrest and survival in surgical patients (2,5,25,26) was not an independent risk factor. Even more surprising is our finding that the severity of coronary artery disease, as defined by CA, was not a significant predictor of survival. In other settings, the severity of coronary artery disease has been thought to play a role in survival after cardiac arrest (27). Lozner et al. (21) reported that factors such as New York Heart Association Class IV function left main coronary artery disease and ejection fraction <30% significantly increased the risk of catheterization. Others report that significant left main coronary artery disease may be the most important risk factor for periprocedural mortality, increasing mortality by as much as 20 times (20–22,27,28). Based on our data we cannot comment on how the severity of disease may contribute to the frequency of arrest, but it did not affect short- or long-term survival when an arrest occurred. Although a significant predictor in the univariate analysis, the type of coronary artery procedure (PTCA or stents versus diagnostic procedure only) did not predict either PS or HS in the multivariate analysis. Furthermore, although predictive of PS, PEA (versus asystole or ventricular fibrillation) was not an independent predictor of HS in multivariate analysis. This may be the result of low statistical power, as only 11 patients experienced PEA and just one of these patients survived to discharge. In the univariate analysis, the most commonly occurring arrhythmia, ventricular fibrillation, was associated with the most frequent survival. These observations are comparable to other in-hospital cardiac arrests (8).

We were able to obtain information regarding long-term follow-up for all 64 patients who were discharged after experiencing cardiac arrest during catheterization. Using a matched-cohort design, we demonstrated that if patients survive to hospital discharge, characteristics associated with the occurrence of cardiac arrest do not significantly affect the patients' long-term survival. Indeed, considering the severity of cardiac disease that prompts initial coronary catheterization, the predicted survival rate in cardiac arrest survivors is reassuring (e.g., 71% at 4 years after discharge compared with 79% for patients undergoing catheterization but not experiencing an arrest) (Table 1). When comparing our patient population to those who experienced out-of-hospital cardiac arrests, several interesting points can be noted. First, the reported hospital survival of out-of-hospital cardiac arrest patients is 42% (29), compared with the 56% survival rate seen in our patient population. This increase in survival may reflect the shorter downtime that results from the patients arresting in the catheterization lab. However, of the 42% of survivors who experienced out-of-hospital arrests, 79% were alive at 5 years postdischarge (29). This survival is 8% more than the 4-year survival in our patient population (71%).

Although risk factors for the current study were collected using formal definitions, the completeness and accuracy of these data are limited by the amount of documentation available in the medical records. Also, the statistical power for assessing the association of a given risk factor with survival is dependent on the prevalence of the risk factor among patients who experienced arrest. Therefore, nonsignificant findings should be interpreted with caution when the number of patients with or without the given risk factor is relatively small.

Because of the change in practice in our cardiac catheterization lab in 1998, we cannot exclude the possibility that the underlying characteristics of the population examined may have changed after 1998, thus affecting the incidence and survival data. Also, if the population in the later study period (attended by an anesthesiologist) was biased towards those with more severe disease, this should tend to make cardiac arrest more likely. However, our post hoc analysis showed that this was not the case. Thus, there is little evidence that the decrease in the frequency of cardiac arrest over the observed time period could be attributed to changes in the sampled population.


The overall frequency of cardiac arrest was 21.9 per 10,000 procedures between 1990 and 2000 (inclusive). The frequency of cardiac arrest decreased over this period, but the hospital survival rate after arrest did not change significantly. The poorest hospital survival rates were for patients with a prior CABG operation and emergent or prolonged catheterizations. All three factors indicate more severe coronary artery disease and imply increased complexity of the percutaneous procedure. Moreover, hospital mortality after a cardiac arrest increases 38% for every 10 years of a patient's age. However, for those who survived cardiac arrest and were subsequently discharged from the hospital, the fact that they experienced cardiac arrest was not associated with significantly decreased long-term survival compared with patients who did not experience periprocedural cardiac arrest.

We thank members of the Department of Anesthesiology who participated in collecting and maintaining information in the Performance Improvement Database. We also thank Christopher Beighley, MS (statistics), and Jack Cusma, Ph.D. (cardiology PTCA registry).


1. Beecher HS, Todd DP. A study of the deaths associated with anesthesia and surgery. Ann Surg 1954;140:2–34.
2. Aubas S, Biboulet P, Daures JP, du Cailar J. Incidence and etiology of cardiac arrest occurring during the perioperative period and in the recovery room. Apropos of 102,468 anesthesia cases [in French]. Ann Fr Anesth Reanim 1991;10:436–42.
3. Auroy Y, Narchi P, Messiah A, et al. Serious complications related to regional anesthesia: results of a prospective survey in France. Anesthesiology 1997;87:479–86.
4. Cohen MM, Duncan PG, Pope WD, Wolkenstein C. A survey of 112,000 anaesthetics at one teaching hospital (1975–83). Can Anaesth Soc J 1986;33:22–31.
5. Sprung J, Warner ME, Contreras MG, et al. Predictors of survival following cardiac arrest in patients undergoing noncardiac surgery: a study of 518,294 patients at a tertiary referral center. Anesthesiology 2003;99:259–69.
6. Keenan RL, Boyan CP. Decreasing frequency of anesthetic cardiac arrests. J Clin Anesth 1991;3:354–7.
7. Hosking MP, Warner MA, Lobdell CM, et al. Outcomes of surgery in patients 90 years of age and older. JAMA 1989;261:1909–15.
8. Dumot JA, Burval DJ, Sprung J, et al. Outcome of adult cardiopulmonary resuscitations at a tertiary referral center including results of “limited” resuscitations. Arch Intern Med 2001;161:1751–8.
9. Hasdai D, Berger PB, Bell MR, et al. The changing face of coronary interventional practice: the Mayo Clinic experience. Arch Intern Med 1997;157:677–82.
10. Peterson ED, Lansky AJ, Anstrom KJ, et al. Evolving trends in interventional device use and outcomes: results from the National Cardiovascular Network Database. Am Heart J 2000;139:198–207.
11. Bittl JA. Advances in coronary angioplasty. N Engl J Med 1996;335:1290–302.
12. Holmes DR, Vlietstra RE, Reiter SJ, Bresnahan DR. Advances in interventional cardiology. Mayo Clin Proc 1990;65:565–83.
13. Holmes DR, Vlitstra RE. Percutaneous transluminal coronary angioplasty: current status and future trends. Mayo Clin Proc 1986;61:865–76.
14. Detre K, Holubkov R, Kelsey SF, et al. Percutaneous transluminal coronary angioplasty in 1985–1986 and 1977–1981. The National Heart, Lung and Blood Institute Registry. N Engl J Med 1988;318:265–70.
15. Mercado N, Maier W, Boersma E, et al. Clinical and angiographic outcome of patients with mild coronary lesions treated with balloon angioplasty or coronary stenting. Implications for mechanical plaque sealing. Eur Heart J 2003;24:541–51.
16. Laskey WK, Kimmel S, Krone RJ. Contemporary trends in coronary intervention: a report from the Registry of the Society for Cardiac Angiography and Interventions. Catheter Cardiovasc Interv 2000;49:19–22.
17. de Bono D. Complications of diagnostic cardiac catheterisation: results from 34,041 patients in the United Kingdom confidential enquiry into cardiac catheter complications. The Joint Audit Committee of the Br Cardiac Society and Royal College of Physicians of London. Br Heart J 1993;70:297–300.
18. Williams DO, Holubkov R, Yeh W, et al. Percutaneous coronary intervention in the current era compared with 1985–1986: the National Heart, Lung, and Blood Institute Registries. Circulation 2000;102:2945–51.
19. Morton BC, Higginson LA, Beanlands DS. Death in a catheterization laboratory. CMAJ 1993;149:165–9.
20. Devlin G, Lazzam L, Schwartz L. Mortality related to diagnostic cardiac catheterization: the importance of left main coronary disease and catheter induced trauma. Int J Card Imaging 1997; 13:379–84.
21. Lozner EC, Johnson LW, Johnson S, et al. Coronary arteriography 1984–1987: a report of the Registry of the Society for Cardiac Angiography and Interventions. II. An analysis of 218 deaths related to coronary arteriography. Cathet Cardiovasc Diagn 1989;17:11–4.
22. Kennedy JW. Complications associated with cardiac catheterization and angiography. Cathet Cardiovasc Diagn 1982;8:5–11.
23. Jackson JL, Meyer GS, Pettit T. Complications from cardiac catheterization: analysis of a military database. Mil Med 2000;165:298–301.
24. Clark VL, Khaja F. Risk of cardiac catheterization in patients aged > or = 80 years without previous cardiac surgery. Am J Cardiol 1994;74:1076–7.
25. Newland MC, Ellis SJ, Lydiatt CA, et al. Anesthetic-related cardiac arrest and its mortality: a report covering 72,959 anesthetics over 10 years from a US teaching hospital. Anesthesiology 2002;97:108–15.
26. Biboulet P, Aubas P, Dubourdieu J, et al. Fatal and non-fatal cardiac arrests related to anesthesia. Can J Anaesth 2001;48:326–32.
27. Boehrer JD, Lange RA, Willard JE, Hillis LD. Markedly increased periprocedure mortality of cardiac catheterization in patients with severe narrowing of the left main coronary artery. Am J Cardiol 1992;70:1388–90.
28. Jansson K, Fransson SG. Mortality related to coronary angiography. Clin Radiol 1996;51:858–60.
29. Bunch TJ, White RD, Gersh BJ, et al. Long-term outcomes of out-of-hospital cardiac arrest after successful early defibrillation. N Engl J Med 2003;348:2626–33.
© 2006 International Anesthesia Research Society