Cardiopulmonary resuscitation

Clinical Investigations Extracorporeal membrane oxygenation support can extend the duration of cardiopulmonary resuscitation* Yih-Sharng Chen, MD; Hsi-Yu Yu, MD; Shu-Chien Huang, MD; Jou-Wei Lin, MD; Nai-Hsin Chi, MD; Chih-Hsien Wang, MD; Shoei-Shan Wang, MD; Fang-Yue Lin, MD; Wen-Je Ko, MD Objectives: To evaluate the use of extracorporeal membrane oxygenation in prolonged cardiopulmonary resuscitation and to estimate how long cardiopulmonary resuscitation can be extended with acceptable results. Design: Review of consecutive adult in-hospital cardiopulmonary resuscitation patients without return of spontaneous circulation in 10 mins and with extracorporeal membrane oxygenation rescue, and analysis of the relationship between outcome and cardiopulmonary resuscitation duration and possible etiologies. The data were collected following the Utstein style guidelines on in-hospital cardiopulmonary resuscitation. Two organ dysfunction scores were incorporated into the analysis for outcome prediction. Setting: A university-affiliated tertiary referral medical center and extracorporeal membrane oxygenation center. Patients: An observational cohort study in 135 consecutive adult in-hospital cardiopulmonary resuscitation patients without return of spontaneous circulation who received extracorporeal membrane oxygenation during cardiopulmonary resuscitation. Main Results: The average cardiopulmonary resuscitation duration was 55.7 27.0 mins and 56.3% of patients received subsequent interventions to treat underlying etiologies. The successful weaning rate was 58.5% and the survival-to-discharge rate was 34.1%. The majority of survivors (89%) had an acceptable neurologic status on discharge. Risk factors for hospital mortality included longer cardiopulmonary resuscitation duration, etiology of acute coronary syndrome, and a higher organ dysfunction score in the first 24 hrs. Logistic regression analysis revealed the probability of survival was approximately 0.5, 0.3, or 0.1 when the duration of cardiopulmonary resuscitation was 30, 60, or 90 mins, respectively. Conclusion: Assisted circulation might extend the presently accepted duration of cardiopulmonary resuscitation in adult inhospital cardiopulmonary resuscitation patients. (Crit Care Med 2008; 36:2529 2535) KEY WORDS: extracorporeal membrane oxygenation; cardiopulmonary resuscitation; in-hospital Cardiopulmonary resuscitation (CPR) has much improved over the past decade because of the early applications of basic cardiac life support and the automated external defibrillator. Nevertheless, the outcome of in-hospital CPR is quite varied and major studies have reported a survival-to-discharge of 10% to 20% (1 4). The probability of survival drops rapidly if CPR lasts longer than 10 mins (5 7), and even more if it lasts longer *See also p. 2685. From the Departments of Surgery (Y-SC, H-YY, S-CH, N-HC, C-HW, S-SW, F-YL, W-JK), and Medicine (J-WL), National Taiwan University Hospital, Yunlin Branch, Taipei, Taiwan. Supported, in part, by the following grant: NSC 93-2314-B-002-239, 94-2314-B-002-121, 95-2745- B-002-232, 96-2314-B-002-166, and NTUH 97-097- 000857. The authors have not disclosed any potential conflicts of interest. For information regarding this article, E-mail Dr. Ko at yschen1234@yahoo.com.tw Copyright 2008 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins DOI: 10.1097/CCM.0b013e318183f491 than 30 mins (8, 9). A large series of adult in-hospital CPR patients with only 18% survival had a median CPR duration of 18 mins in all patients combined, which was even shorter in survivors (9). This reflects the fact that most CPR patients can only tolerate a short period of CPR. Because prolonged CPR has such a poor result, several mechanical devices and techniques have been developed to improve outcome (10). Clinical investigation of the use of these adjunctive devices in CPR has demonstrated an improved hemodynamic effect and increased frequency of return of spontaneous circulation, but with no, or only minimal, improvement in survival (10 12). The potential of extracorporeal membrane oxygenation (ECMO) as a resuscitative tool was suggested as early as 1966 (13), but few clinical investigations focused on its use at that time because of its complexity. Emergency extracorporeal circulation data collected from multiple institutions in the early 1990s revealed improved survival in patients who received ECMO for cardiac arrest or shock (14, 15). However, there is no adequate information on how long the victim can tolerate CPR with an acceptable outcome when ECMO is available as an adjunctive device. We performed this observational cohort study of the use of ECMO in prolonged adult in-hospital CPR (ECPR) in a tertiary medical center, and analyzed the results and estimated how long CPR can be extended with acceptable results in the setting of mechanical assisted circulation. PATIENTS AND METHODS This study was approved by our institutional review board. We reviewed our ECMO database which was established in 1994. The database were collected consecutively and reviewed for ECMO quality assurance. Adult patients (aged 18 yrs) experiencing CPR of 10 mins and receiving ECMO were recruited. A CPR event was defined as the documented loss of a pulse and respiration with the patient receiving multiple doses of epinephrine injection and/or undergoing chest compression with or without defibrillation. Monitored status was defined as 2529

being placed on monitoring systems for electrocardiography, blood pressure, or pulse oximetry during CPR (16). CPR duration was defined as the interval between initiation of CPR and initiation of ECMO. The first 57 cases had been presented previously (17). Exclusion Criteria of ECPR. Because ECPR is an aggressive intervention requiring considerable resources and of uncertain effect, we defined some absolute contraindications to avoid initiating inappropriate ECPR, including CPR with traumatic origin unless bleeding was under control, previous severe brain damage, terminal status of malignancy, and an age 75 yrs. From 2001, the age criterion was extended to 80 yrs because of the increased number of older patients and the satisfactory survival rate of the initial ECPR patients (17). Patients experiencing shock necessitating ECMO in an elective condition or emergency situation without cardiac massage or multiple boluses of epinephrine injection were excluded. Patients with a CPR duration 10 mins were excluded because they did not fit the criterion of prolonged CPR. Patients with postcardiotomy shock requiring ECMO because they could not be weaned from cardiopulmonary bypass were also excluded, as were patients who had signed a Do Not Resuscitate consent. Organization of ECMO. The equipment and management had been previously described elsewhere (17). The ECMO equipment consisted of the circuit, a centrifugal pump, and a hollow-fiber oxygenator (Medtronic, Anaheim, CA), all of which had heparin-bound surface. The circuit was primed with normal saline containing 2 units/ml of heparin, which took 5 mins. To simplify the system, we did not use a bridge tube between the arterial and venous circuits. A wheeled ECMO cart equipped with various sizes of cannulas, surgical suits, gloves and drapes, and various surgical instruments for vascular exposure was constructed for the ECPR team and could be rapidly transported to the CPR site to supply all the equipment for ECMO setup. The ECMO cart cost 200 US dollars to build and supplies were replenished from the operation theater. Team Members. Two teams were involved in ECPR, the resuscitation team and the ECMO team. The resuscitation team consisted of duty physicians and nurses responding to in-hospital CPR. The ECMO team consisted of cardiovascular surgeons and ECMO technicians. Duty cardiovascular surgical residents were in the hospital at all times. The technicians were on duty in the hospital day and evening and on call at night. Because the ECMO team did not attend every arrest event, the decision to call for ECPR was dependent on the primary physicians in charge of CPR. In general, the call was initiated after a trial of CPR for 10 to 20 mins without return of spontaneous circulation. Brief discussions between the two teams were held to preclude absolute contraindications. Oral permission was obtained from the families and the documents, including the formal informed consent for ECPR, further interventions, and data collection, were signed later. Cannulation. The femoral vessels were preferred for vascular access and were usually cannulated by the modified Seldinger technique under direct vision by surgical exposure. The percutaneous technique was used only on patients in the catheterization laboratory, whereas transthoracic cannulation was used on patients with open massage or in whom peripheral cannulation had failed. Intravenous heparin, 100 units/kg, was given immediately before cannulation. In general, it took 10 to 30 mins to set up the ECMO equipment once it had been decided to carry out ECPR. Management. For the potential problem of distal limb ischemia, we applied a previously reported salvage method using a distal perfusion catheter (18). Systemic anticoagulation was maintained with continuous heparin infusion to keep the activated clotting time between 180 and 220 sec. The patients were kept at normothermia. We did not apply a hypothermia strategy in the present ECPR study. The possible etiologies leading to CPR were investigated as soon as possible. Diagnosis of, and treatment for, the underlying etiology were performed within 24 hrs. In general, the level of consciousness of the patient was the main decisive factor for further therapies; however, temporary dilated and unresponsive pupils immediately after ECPR was not an indication for withdrawing support or with- Figure 1. Cohort for the ECPR study. ARDS, acute respiratory distress syndrome; CPR, cardiopulmonary resuscitation; ECPR, ECMO for CPR; ECMO, extracorporeal membrane oxygenation. 2530

holding further treatment, as this could be due to a temporary effect of epinephrine. Daily echocardiography was essential to estimate myocardial recoverability. Weaning from ECMO was usually tried 72 hrs after ECMO setup. When the patients had stable hemodynamics on dopamine and dobutamine 10 g/kg/min, a central venous pressure 12 mm Hg, and a left ventricular ejection fraction 40% by echocardiography, ECMO weaning was attempted by decreasing the ECMO blood flow to 0.5 L/min and the patient was observed for at least 10 mins. If the hemodynamics were not maintained during weaning, weaning was discontinued and attempted the following day. A ventricular assist device or heart transplantation was considered in the absence of contraindications if weaning was unsuccessful within 5 to 7 days. The level of consciousness was reexamined every 12 hrs before making further decisions. Data Collection. The data were collected following Utstein style guidelines on inhospital CPR (16, 17). The Sequential Organ Failure Assessment (SOFA) score (19) and the Logistic Organ Dysfunction Score (LODS) (20) were used to measure the severity of organ dysfunction. The efficacy and specificity of the scoring systems for critical patients had been validated (20, 21). Because the ECPR patients were supported by mechanical circulation, return of spontaneous circulation, which is traditionally defined as one of the CPR end points, was not a meaningful end point in this cohort. Instead, we used return of spontaneous heart beating. Weaning was defined as successful separation from ECMO support without mortality for at least 12 hrs. Survival-to-discharge was defined as successful weaning from ECMO followed by discharge from the hospital, and was regarded as the major end point. The functional status of survivors in follow-up was recorded using the Pittsburgh cerebral performance category score (22). Statistical Analysis. The data are shown as the mean SD and the median value is given if necessary. Categorical variables were compared using Fisher s exact test. Continuous variables were compared using the Mann-Whitney U test. Univariate and multivariate analysis were performed to identify risk factors for mortality. A multiple logistic regression model was used to identify predictors of survival in the ECPR cohort. The odds ratio (OR) and 95% confidence interval (CI) are reported for each variable. The probability of survival-to-discharge inferred by logistic regression was plotted against CPR duration. Stata 8.0 (Stata Corp., College Station, TX) was used for the statistical analyses. RESULTS Demographic Data. From December 1994 to July 2005, 529 adult patients received ECMO support at our institute, and 135 who had undergone in-hospital Table 1. Data of ECPR cohort (n 135) Parameters n (%) Male 90 (66.7%) Age (yrs) 54.0 15.7 (56, 16 87) Etiologies of underlying disease ACS 66 (48.9%) Postcardiotomy 23 (17.0%)) CMP 22 (16.3%) Myocarditis 12 (8.9%) PE 5 (3.8%) Mechanical valve obstruction 3 (2.2%) Others 4 (2.9%) Comorbidity DM 39 (28.9%) HT 32 (23.7%) Ischemic heart disease 77 (57.0%) Previous heart surgery 10 (7.4%) Valve disease 6 (4.4%) Congenital heart disease 2 (1.5%) Previous CVA 3 (2.2%) ESRD 2 (1.5%) CPR site ICU 82 (60.7%) Ward 4 (3.0%) OT 2 (1.5%) Cath lab 21 (15.6%) ER 26 (19.3%) Previous CPR 23 (17.0%) Initial rhythm VT/VF 97 (71.9%) PEA/asystole 34 (25.2%) Unknown 5 (3.7%) Defibrillation 114 (84.4%) Witnessed cardiac arrest 133 (98.5%) Monitored during CPR 109 (80.7%) IE at CPR ( g/kg/min) 72.9 77.1 (47, 8 594) CPR duration (mins) 55.7 27.0 (50, 16 150) Subsequent intervention 76 (56.3%) PCI 8 (10.5%) CABG 37 (48.7%) Unloading left heart 8 (10.5%) Valve replacement 2 (2.6%) HTx 8 (10.5%) VAD 4 (5.3%) Ventriculorraphy 5 (4.9%) Pulmonary embolectomy 3 (3.9%) ECMO duration (hrs) 91.1 78.3 (73.5, 1.5 407) SOFA (first 24 hrs) 12.4 3.2 (13, 5 20) LODS (first 24 hrs) 11.2 3.8 (11, 4 20) Outcome Survival 72 hr 102 (75.6%) Wean-off 79 (58.5%) Survival to hospital discharge 46 (34.1%) CPC status at hospital discharge 1/2 41 (89.1%) 3/4 5 (10.1%) Variables are expressed as number (%) or mean standard deviation (medium, range). Percent of each intervention is calculated as percentage of all patients receiving subsequent interventions. ACS, acute coronary syndrome; CABG, coronary artery bypass grafting; Cath lab, catheterization laboratory; CMP, cardiomyopathy; CPC, cerebral performance category; CPR, cardiopulmonary resuscitation; CVA, cerebral vascular accident; DM, diabetes mellitus; ECMO, extracorporeal membrane oxygenation; ECPR, ECMO for CPR; ECLS, extracorporeal life support; ER, emergency room; ESRD, end stage renal disease; HT, hypertension; HTx, heart transplantation; IE, inotropic equivalent; ICU, intensive care unit; LODS, Logistic Organ Dysfunction Score; OT, operative theater; PCI, percutaneous coronary intervention; PE, pulmonary embolism; PEA, pulseless electrical activity; SOFA, Sequential Organ Failure Assessment; VAD, ventricular assist device; VF, ventricular fibrillation; VT, ventricular tachycardia. 2531

Table 2. Weaning rate and survival to hospital discharge rate of extracorporeal membrane oxygenation for cardiopulmonary resuscitation patients with different etiologies Etiology n (%) Weaning Rate, n (%) Survival Rate, n (%) Acute coronary syndrome 66 (48.9) 35 (53.0) 16 (24.2) Postcardiotomy 23 (17.0) 16 (69.6) 11 (47.8) Cardiomyopathy 22 (16.3) 11 (50.0) 7 (31.8) Acute myocarditis 12 (8.9) 11 (91.7) 7 (58.3) Obstructive lesions 8 (5.9) 3 (37.5) 2 (25) Others 4 (3.0) 3 (75) 3 (75) Overall 135 (100) 79 (58.5) 46 (34.1) Weaning rate and survival to hospital discharge rate are expressed as number and percentage of each etiology group of patients. Table 3. Odds ratio (OR) and 95% confidence interval (CI) for weaning and survival to discharge Wean-off CPR and received ECMO during CPR were included in the study (Fig. 1). The demographic, intervention, and outcome data for the study patients are shown in Table 1. Femoral cannulation was performed in the majority of patients (98.5%). One of these had the complication of retroperitoneal hematoma due to vessel perforation and expired soon after initiation of ECMO. Transthoracic cannulation was performed on two patients because of open cardiac massage. Four patients (3.0%) did not show return of spontaneous heart beating despite successful ECMO setup. Seventy-six patients (56.3%) underwent subsequent interventions to treat the underlying etiology; coronary intervention was performed in 59.2% of these patients (Table 1). Outcome. Thirty-three of the 79 successfully weaned patients died 870 3021 (median 201) hours later. Forty-six (34.1%) patients survived to hospital discharge after ECPR, but seven subsequently died (mean follow-up for these seven patients: 57.6 36.9 mos, range 8 143 mos). One patient died of pneumonia 5 months after discharge. Three patients died of cardiac-related events (progressive heart failure or pulmonary hypertension) at 1, 1, and 3 yrs. One patient died of underlying ovarian malignancy 4 yrs after ECPR. Two patients who survived ECPR followed by ventricular assist device and cardiac transplantation died 4 and 8 yrs later due to chronic rejection and posttransplant malignancy, respectively. The 1-yr and 5-yr survival rates of patients who survived to hospital discharge were 93.5% and 85%, respectively. The majority of survivors (89.1%, n 41, Table 1) were in cerebral performance category status 1 or 2 at hospital discharge. This indicates that neurologic function can be preserved in most ECPR survivors. Risk Factor Analysis. The weaning rate and survival-to-discharge rate in patients with different underlying etiologies are shown in Table 2. The univariate analysis for weaning and survival-todischarge, shown in Table 3, revealed that both CPR duration and an etiology of acute coronary syndrome (ACS) were significantly associated with survival (Table 3, p 0.05). However, after multivariate logistic regression analysis, only CPR duration remained a risk factor related to the survival. Underlying Etiologies. As shown in Table 2, patients with ACS had a lower survival-to-discharge rate than those with other etiologies (24.2% vs. 43.5%, OR: 0.42, 95% CI: 0.19 0.92), and myocarditis patients had a higher weaning rate than those with other etiologies (91.7% vs. 55.3%, OR: 8.9, 95% CI: 1.2 389.7). In the ACS group, CPR duration was an important factor for survival (45.1 27.2 mins in survivors vs. 62.2 29.3 mins in nonsurvivors, p 0.04), but not for weaning (p 0.05). Subsequent Interventions. Subsequent interventions after ECPR were performed on 76 patients, of which 30 Survival OR 95% CI OR 95% CI Male 1.38 0.625 3.011 1.22 0.536 2.859 Age 60 yrs 0.655 0.297 1.352 0.536 0.232 1.207 Underlying etiologies ACS 0.641 0.304 1.351 0.42 0.185 0.922 a CMP 0.738 0.266 2.056 0.89 0.281 2.549 Myocarditis 8.897 1.221 389.715 a 3.05 0.771 12.886 Postcardiotomy 1.778 0.628 5.505 2.02 0.724 5.514 Comorbidity DM 0.87 0.358 2.107 0.53 0.189 1.426 HT 0.54 0.223 1.293 0.70 0.257 1.774 Ischemic heart disease 2.03 0.960 4.309 0.74 0.341 1.617 Previous heart surgery 7.07 0.921 315.421 3.19 0.704 16.096 Previous CPR 1.41 0.508 4.150 0.82 0.262 2.330 VT/VF 1.13 0.494 2.556 1.39 0.578 3.479 CPR location ICU/Cath/OT 1.42 0.570 3,519 1.19 0.461 3.282 ER 0.54 0.206 1.393 0.66 0.216 1.832 Monitored 1.27 0.484 3.260 1.93 0.672 6.350 CPR duration 0.99 0.974 1.000 0.97 0.947 0.983 a Defibrillation 0.39 0.104 1.208 0.64 0.225 1.893 Subsequent intervention 1.75 0.828 3.722 1.52 0.694 3.392 CABG 0.27 0.080 0.838 0.59 0.208 1.675 a p 0.05. ACS, acute coronary syndrome; CABG, coronary artery bypass grafting; Cath lab, catheterization laboratory; CMP, cardiomyopathy; CPC, cerebral performance category; CPR, cardiopulmonary resuscitation; CVA, cerebral vascular accident; DM, diabetes mellitus; ECLS, extracorporeal life support; ER, emergency room; ESRD, end stage renal disease; HT, hypertension; HTx, heart transplantation; IE, inotropic equivalent; ICU, intensive care unit; LODS, Logistic Organ Dysfunction Score; OT, operative theater; PCI, percutaneous coronary intervention; PE, pulmonary embolism; PEA, pulseless electrical activity; SOFA, Sequential Organ Failure Assessment; VAD, ventricular assist device; VF, ventricular fibrillation; VT, ventricular tachycardia. (39.5%) survived to discharge. The survival rate of patients with subsequent intervention was not significantly better than that of patients without intervention (39.5% survival with intervention vs. 28.8% survival without intervention, p 0.15). The different intervention procedures did not influence the weaning or survival rate (p 0.05). However, in patients undergoing subsequent interventions, CPR duration was still an important determining factor for survival (41.7 21.6 mins in survivors vs. 60.6 25.1 mins in nonsurvivors, p 0.001). 2532

In the ACS group, patients with subsequent intervention had a better weaning rate (61.2%) than those without intervention (29.4%) (p 0.028, OR: 3.8, 95% CI: 1.0 15.7) and a marginally better survival-to-discharge rate (30.6% survival with intervention vs. 5.9% survival without intervention, p 0.051, OR: 7.1, 95% CI: 0.9 315.4). No difference in the survival or weaning rate was seen in the ACS group when comparing different revascularization methods (surgical coronary artery bypass vs. percutaneous coronary intervention) (p 0.49 for weaning and p 0.27 for survival). Despite the low survival rate of the ACS group compared with other etiology groups, aggressive intervention for revascularization of the ischemic/infarcted myocardium can still be recommended after ECPR because it provides better survival. Duration of ECMO. A statistical difference in ECMO duration was seen between the weaned and nonweaned groups (100.6 76.6 hrs for weaning vs. 77.9 77.8 hrs for nonweaning, p 0.0076), but there was no effect on survival (105.6 88.4 hrs for survivors vs. 83.7 71.5 hrs for nonsurvivors, p 0.15). Organ Dysfunction Parameters. Significant differences in both the SOFA score and LODS in the first 24 hrs were seen between the survival and nonsurvival group and between the weaned and nonweaned group (Table 4). The survival group and weaned group had significantly lower neurologic and renal subscores in the SOFA score and LODS than the nonsurvival and nonweaned group, respectively. These two significant specific factors resulted in the total SOFA score and LODS also being a predictive parameter. Duration of CPR. Patients who survived to discharge had a significantly shorter CPR duration than those who did not (43.2 19.9 mins for survivors vs. Table 4. The Sequential Organ Failure Assessment (SOFA) and Logistic Organ Dysfunction Score (LODS) scores in the first 24 hrs of extracorporeal membrane oxygenation for cardiopulmonary resuscitation group Weaned Nonweaned p Survival Nonsurvival p SOFA total a 11.0 2.8 14.4 2.6 0.001 10.3 2.4 13.6 2.9 0.001 Resp 1.8 1.4 2.4 1.4 0.02 1.9 1.4 2.1 1.5 0.48 Coag 1.3 1.0 1.4 1.0 0.40 1.2 1.1 1.4 1.02 0.23 Cardiac 3.9 0.8 3.9 0.4 0.22 4.0 0.1 3.9 0.5 0.34 Hepatic 1.1 0.9 0.9 0.9 0.27 1.2 0.9 1.0 0.9 0.11 CNS a 1.4 1.5 3.3 1.3 0.001 0.6 0.9 3.0 1.4 0.001 Renal a 1.5 1.0 2.4 1.2 0.001 1.2 0.8 2.2 1.2 0.001 LODS total a 9.6 3.2 13.5 3.4 0.001 8.5 2.5 12.5 3.6 0.001 Resp 1.5 0.9 1.9 1.0 0.02 1.5 0.9 1.8 1.0 0.09 Coag 0.2 0.5 0.2 0.4 0.46 0.2 0.6 0.2 0.4 1.0 Cardiac 3.1 1.0 3.5 1.2 0.02 3.1 0.9 3.4 1.2 0.16 Hepatic 0.7 0.5 0.6 0.5 0.28 0.8 0.4 0.6 0.5 0.07 CNS a 1.4 1.7 3.9 1.7 0.001 0.5 0.8 3.4 1.9 0.001 Renal a 2.7 1.5 3.3 1.4 0.02 2.4 1.4 3.2 1.4 0.002 The total score and the subscore of CNS and renal system consistently revealed difference between weaning and nonweaning or survival and nonsurvival. CNS, central nervous system; coag, coagulation and hematological system; resp, respiratory. a p 0.05. Table 5. Relationship between weaning or survival and cardiopulmonary resuscitation (CPR) duration CPR Duration (min) n (%) Wean-off, n (%) Survival, n (%) OR 95% CI 15 0 (0) 30 19 (14.1) 12 (63.2) 12 a (63.2) 4.134 1.349 13.379 30 116 (85.9) 67 (57.8) 34 a (29.3) 45 58 (43) 37 (63.8) 29 a (50) 3.529 1.574 7.981 45 77 (57) 42 (54.5) 17 a (22.1) 60 88 (65.2) 59 a (67.0) 42 a (47.4) 9.815 3.124 40.234 60 47 (34.8) 20 a (42.6) 4 a (8.5) Total 135 (100) 79 (58.5) 46 (34.1) Odds ratio (OR) and 95% confidence interval (CI) are calculated to compare survival-to-hospitaldischarge rates between each two groups. a p 0.05. 62.1 27.9 mins for nonsurvivors, p 0.001). A detailed analysis of the relationship between CPR duration and survival is shown in Table 5. A logistic regression model was developed to demonstrate the relationship between CPR duration and probability of survival-to-discharge using the equation: Probability of survival exp (1.17 0.036 CPR duration) 1 exp (1.17 0.036 CPR duration) The probability of survival in the ECPR setting was approximately 0.5, 0.3, and 0.1 when CPR was 30, 60, and 90 mins, respectively (Fig. 2). CPR duration is an important factor for survival, but CPR duration did not play such an important role in ECMO weaning (Table 5). Using a cut-off for CPR duration of 60 mins, the weaning rate was significantly higher in the CPR 60 mins group (Table 5, p 0.01). Despite the fact that successfully weaned patients experienced a shorter CPR duration than those who were not weaned (51.8 23.1 mins in weaned patients vs. 61.1 31.0 mins in nonweaned patients), the difference did not reach statistical significance (p 0.059). DISCUSSION With assisted circulation, CPR duration could be extended to 60 mins with acceptable survival and the incidence of major neurologic deficits was relatively low at hospital discharge. The organ dysfunction scores (SOFA score or LODS) reflected the severity of shock damage in the early stage after ECPR and thus helped to predict the outcome. The availability of an ECMO team equipped with a wheeled ECMO cart allowed a rapid response to a call for ECPR and shortened the CPR duration before ECMO setup. Recent improvements in results of the use of ECMO in cardiopulmonary insufficiency have encouraged more physicians to apply ECMO for arrested patients (17, 22 25). An increasing number of sporadic case reports or series with limited cases have demonstrated long CPR tolerance if ECMO is used in CPR (25 28). However, most of these studies were performed on pediatric groups, and none have been carried out on adult in-hospital CPR. No study has tried to delineate CPR duration tolerance in this setting. To our knowledge, the present study is the largest series of ECPR and clearly demon- 2533

Figure 2. Relationship between probability of survival-to-hospital discharge and cardiopulmonary resuscitation (CPR) duration. ECPR, extracorporeal membrance oxygenation for CPR. Table 6. Comparison of conventional CPR groups with ECPR groups In-Hospital Prolonged CPR ( 10 mins) No. Duration, Mins (Mean SD, Median) strates tolerance of a long duration of ECPR in adult in-hospital CPR patients. ECMO team equipped with the wheeled ECMO cart allowed rapid response to the call for ECPR, and shortened the required CPR duration before ECMO setup. However, CPR duration continues to be a key factor in conventional CPR (8) and ECPR. Our study clearly revealed that a longer CPR duration resulted in a lower probability of survival, but ECMO allowed a longer CPR duration than expected in conventional CPR apparently because immediate assisted perfusion prevents progressive acidosis, alleviates post-cpr myocardium stunning, and reverses post-cpr organ dysfunction. ECMO can bring some patients back from irreversibility (29). The use of ECMO has changed previous concepts about the time limitation in Age, Yr (Mean SD, Median) Survival (%) p C1, all causes 243 35.3 23.9, 30 a 56.0 15.8, 59.5 (NS) 9.5 a 0.001 C2, cardiopulmonary 168 36.2 24.2, 30 a 48.8 13.7, 53.5 a 8.9 a 0.001 origin With ECMO (ECPR) 135 55.7 27.0, 50 54.0 15.7, 56.0 34.1 CPR, cardiopulmonary resuscitation; ECMO, extracorporeal membrane oxygenation; ECPR, extracorporeal cardiopulmonary resuscitation; NS, not significant; C1, C2 groups, C1 in-hospital adult CPR ( 10 mins); C2, C1 group with cardiopulmonary origin. a p 0.05 compared with ECPR group. CPR. Based on the present results, ECMO can extend CPR duration to 90 min with a 0.1 probability of survival (Fig. 2). However, the quality of the applied CPR has not been investigated in this study, and might impact the outcome (30). According to our web-based registry for in-hospital CPR between January 2004 and December 2005 (8), 696 adult patient were recorded. After excluding ECPR patients, patients aged 75 yrs, and patients with a CPR duration 10 mins, 243 patients who had received CPR for 10 mins were recruited as a comparison group (C1, Table 6); 168 of these had a cardiopulmonary origin and formed a second comparison group (group C2, Table 6). The comparison demonstrated that the ECPR group, despite being older and having a longer duration of CPR, still had a better survival. Although these groups were not ideal for comparison (age and CPR duration differences), this supplementary data provides evidence supporting the use of ECPR. In the same study period, seven patients with out-of-hospital CPR received ECMO deployment at the emergency service. Only three were able to be weaned off ECMO with a cerebral performance category status of 4, and only one survived to hospital discharge with vegetative status. Based on this experience, we hesitate to recommend ECMO for out-ofhospital CPR because of the uncertain duration of arrest. The organ dysfunction scores (SOFA score or LODS) were able to predict the outcome in the early stage after the event. Both scores reflected the severity of shock damage in the early stage after ECPR and thus helped to predict the outcome. Patients with the etiology of ACS had the worst outcome among the subgroups and this might be related to the severity of the coronary lesion and lower reversibility of the ischemic myocardium. Most of our ACS patients had multiple or proximal lesions and prolonged resuscitation might lead to lower reversibility of the ischemic myocardium. However, half were weaned off ECMO, of whom 50% survived. The extremely ill patients in the SHould we emergently revascularize Occluded Coronaries in cardiogenic shock registry had a 70% mortality (31), comparable with that in our ACS subgroup, despite being less critical than our ACS subgroup. Limitation. This is an observational study, not a randomized trial, and might have selection bias. However, we do not think that it is possible to conduct a prospective randomized study for patients undergoing CPR, as it is not ethical to exclude possibly reversible victims from ECMO, especially after prolonged CPR. The quality of CPR was not well controlled in this study and this may influence the outcome (30). We did not apply a hypothermic strategy in the present ECPR group. Whether a hypothermic strategy in addition to ECMO would improve the outcome may be an attractive topic, and it deserves further investigation. CONCLUSION A shorter CPR duration allows a better survival in ECPR. However, ECMO still offers an acceptable survival rate in pa- 2534

tients with prolonged CPR up to 60 mins with 0.3 probability of survival. REFERENCES 1. Sandroni C, Nolan J, Cavallaro F, et al: Inhospital cardiac arrest: Incidence, prognosis and possible measures to improve survival. Intensive Care Med 2007; 33:237 245 2. Nadkarni VM, Larkin GL, Peberdy MA, et al: First documented rhythm and clinical outcome from in-hospital cardiac arrest among children and adults. JAMA 2006; 295:50 57 3. Cooper S, Janghorbani M, Cooper G: A decade of in-hospital resuscitation: Outcomes and prediction of survival? Resuscitation 2006; 68:231 237 4. Weil MH, Fries M: In-hospital cardiac arrest. Crit Care Med 2005; 33:2825 2830 5. Hajbaghery MA, Mousavi G, Akbari H: Factors influencing survival after in-hospital cardiopulmonary resuscitation. Resuscitation 2005; 66:317 321 6. Pionkowski RS, Thompson BM, Gruchow HW, et al: Resuscitation time in ventricular fibrillation a prognostic indicator. Ann Emerg Med 1983; 12:733 738 7. Schultz SC, Cullinane DC, Pasquale MD, et al: Predicting in-hospital mortality during cardiopulmonary resuscitation. Resuscitation 1996; 33:13 17 8. Shih CL, Lu TC, Jerng JS, et al: A web-based Utstein style registry system of in-hospital cardiopulmonary resuscitation in Taiwan. Resuscitation 2007; 72:394 403 9. Peberdy MA, Kaye W, Ornato JP, 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 10. International Liaison Committee on Resuscitation: 2005 International Consensus on Cardiopulmonary Resuscitation and Emergency Cardiovascular Care Science with Treatment Recommendations. Part 4: Advanced life support. Resuscitation 2005; 67: 213 247 11. Kern KB, Morley PT, Babbs CF, et al: Use of adjunctive devices in cardiopulmonary resuscitation. Ann Emerg Med 2001; 37(4 Suppl): S68 S77 12. Wigginton JG, Miller AH, Benitez FL, et al: Mechanical devices for cardiopulmonary resuscitation. Curr Opin Crit Care 2005; 11: 219 223 13. Kennedy JH: The role of assisted circulation in cardiac resuscitation. JAMA 1966; 197: 615 618 14. Mooney MR, Arom KV, Joyce LD, et al: Emergency cardiopulmonary bypass support in patients with cardiac arrest. J Thorac Cardiovasc Surg 1991; 101:450 454 15. Hill JG, Bruhn PS, Cohen SE, et al: Emergent applications of cardiopulmonary support: A multiinstitutional experience. Ann Thorac Surg 1992; 54:699 704 16. Cummins RO, Chamberlain D, Hazinski MF, et al: Recommended guidelines for reviewing, reporting, and conducting research on in-hospital resuscitation: The inhospital 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 17. Chen YS, Chao A, Yu HY, et al: Analysis and results of prolonged resuscitation in cardiac arrest patients rescued by extracorporeal membrane oxygenation. J Am Coll Cardiol 2003; 41:197 203 18. Huang SC, Yu HY, Ko WJ, et al: Pressure criterion for placement of distal perfusion catheter to prevent limb ischemia during adult extracorporeal life support. J Thorac Cardiovasc Surg 2004; 128:776 777 19. Vincent JL, de Mendonca A, Cantraine F, et al: Use of the SOFA score to assess the incidence of organ dysfunction/failure in intensive care units: Results of a multicenter, prospective study. Working group on sepsis-related problems of the European Society of Intensive Care Medicine. Crit Care Med 1998; 26: 1793 1800 20. Le Gall JR, Klar J, Lemeshow S, et al: The Logistic Organ Dysfunction system. A new way to assess organ dysfunction in the intensive care unit. ICU Scoring Group. JAMA 1996; 276:802 810 21. Ferreira FL, Bota DP, Bross A, et al: Serial evaluation of the SOFA score to predict outcome in critically ill patients. JAMA 2001; 286:1754 1758 22. Tunstall-Pedoe H, Bailey L, Chamberlain DA, et al: Survey of 3765 cardiopulmonary resuscitations in British hospitals (the BRESUS Study): Methods and overall results. BMJ 1992; 304:1347 1351 23. Nagao K, Hayashi N, Kanmatsuse K, et al: Cardiopulmonary cerebral resuscitation using emergency cardiopulmonary bypass, coronary reperfusion therapy and mild hypothermia in patients with cardiac arrest outside the hospital. J Am Coll Cardiol 2000; 36:776 783 24. Chen YS, Ko WJ, Lin FY, et al: Preliminary result of an algorithm to select proper ventricular assist devices for high-risk patients with extracorporeal membrane oxygenation support. J Heart Lung Transplant 2001; 20: 850 857 25. Schwarz B, Mair P, Margreiter J, et al: Experience with percutaneous venoarterial cardiopulmonary bypass for emergency circulatory support. Crit Care Med 2003; 31: 758 764 26. Parra DA, Totapally BR, Zahn E, et al: Outcome of cardiopulmonary resuscitation in a pediatric cardiac intensive care unit. Crit Care Med 2000; 28:3296 3300 27. Aharon AS, Drinkwater DC Jr, Churchwell KB, et al: Extracorporeal membrane oxygenation in children after repair of congenital cardiac lesions. Ann Thorac Surg 2001; 72: 2095 2101; discussion 2101 2092 28. Posner JC, Osterhoudt KC, Mollen CJ, et al: Extracorporeal membrane oxygenation as a resuscitative measure in the pediatric emergency department. Pediatr Emerg Care 2000; 16:413 415 29. Massetti M, Tasle M, Le Page O, et al: Back from irreversibility: Extracorporeal life support for prolonged cardiac arrest. Ann Thorac Surg 2005; 79:178 183; discussion 183 174 30. Abella BS, Alvarado JP, Myklebust H, et al: Quality of cardiopulmonary resuscitation during in-hospital cardiac arrest. JAMA 2005; 293:305 310 31. Hochman JS, Sleeper LA, Webb JG, et al: Early revascularization in acute myocardial infarction complicated by cardiogenic shock. SHOCK Investigators. Should We Emergently Revascularize Occluded Coronaries for Cardiogenic Shock. N Engl J Med 1999; 341:625 634 2535