Criteria abstracted from The Users' Guide to Medical Literature, from the Health Information Research Unit and Clinical Epidemiology and Biostatistics, McMaster University

Highlighted lines and questions below provide links to the pertinent description of criteria in The EBM User's Guide, now available at the Canadian Centres for Health Evidence

Article Reviewed:

Please visit the new Evidence Based Journal Club Reviews

Survival outcomes after extracorporeal cardiopulmonary resuscitation instituted during active chest compressions following refractory in-hospital pediatric cardiac arrest.

Morris MC, Wernovsky G, Nadkarni VM.

Pediatr Crit Care Med. 2004 Sep;5(5):440-446. [abstract]

Reviewed by Marie Frazier MD, Ira M. Cheifetz MD, Duke University Medical Center, Durham, NC, and Angela T. Wratney MD MHSc, Children's National Medical Center, Washington, DC

Review posted September 16, 2005

I. What is being studied?

Study objective:

To report survival outcomes and to identify factors associated with survival following extracorporeal cardiopulmonary resuscitation (ECPR) for in-hospital pediatric cardiac arrest.

Study design

Retrospective review of medical charts, the institutional ECMO database, and the records of the institutional resuscitation committee records at a single, large tertiary care institution.

Patients:

Study Cohort: Children in the PICU or CICU (outside of the neonatal intensive care unit) admitted to the Children's Hospital of Philadelphia who were resuscitated from cardiac arrest (defined as external chest compressions or internal cardiac massage performed for > 60 sec. because of a profound low cardiac output state per the American Heart Association National Registry of CPR (1).) during active chest compressions by means of venoarterial (VA) ECMO between January, 1995 and March, 2002. Only those children who were undergoing mechanical ventilation and active chest compressions or internal cardiac massage at the time of ECMO cannulation were included. 217 ECMO episodes in 209 patients occurred outside of the neonatal ICU. Sixty-six sequential cardiac arrest events meeting the criteria were identified involving 64 patients.

Control group: Inpatient cardiac arrest events managed with conventional CPR (outside of the neonatal ICU) from March, 2000 through April, 2002. They reviewed the medical chart records of 79 events in 73 patients to report the duration of chest compressions, whether there was return of spontaneous circulation, and whether the patient survived to hospital discharge.

Outcome:

The primary outcome measure was survival to hospital discharge. A secondary outcome was the development of neurologic impairment prior to hospital discharge. This was defined as new seizure disorder, loss of developmental milestones, loss of muscle tone, or focal weakness. Retrospective chart review was used to assign admission and discharge Pediatric Cerebral Performance Category (PCPC) and Pediatric Overall Performance Category (POPC) in surviving children > two months of age.

II. Are the results of the study valid?

Note: These questions follow from Randolph AG et al. Understanding articles describing clinical prediction tools. Crit Care Med 1998;26:1603-1612. [abstract]

1. Was a representative and well-defined sample of patients at a similar point in the course of the disease? Was follow-up sufficiently long and complete?

Complete follow-up? Partially. The primary outcome was hospital discharge, and all 64 patients were followed to hospital discharge. The secondary outcome was neurologic impairment. Admission and discharge Pediatric Cerebral Performance Category and Pediatric Overall Performance Category were assigned in surviving children greater than two months old through retrospective chart review. Given the limitations of retrospective chart review, obtaining complete and accurate neurologic assessments is not always possible. Of the 21 survivors, Pediatric Cerebral Performance Category and Pediatric Overall Performance Category were obtained in 10 of 13 survivors who were greater than two months of age at discharge. Many of the patients did not have formal pediatric neurology consult documentation available for review. Many of the patients who underwent ECPR were infants who cannot be properly assessed by these scoring measures and would require a detailed neurologic evaluation to detect significant neurologic impairment.

Follow-up sufficiently long? Not ideally. The authors chose to assess the primary outcome as survival to hospital discharge which is a typical measure of survival post-resuscitation with ECMO. However, when assessing survival outcomes one would also want to assess other outcomes such as long-term survival, functional status, and long-term neurologic outcome. Notably, despite successful resuscitation following in-hospital cardiac arrests, as few as 15-20% will survive long-term (> 30 days) (2). Furthermore, developmental delays and functional impairment may not be fully appreciated in short-term follow up. Other researchers believe that longer follow-up is necessary as children may have cognitive improvement even up to a year after serious traumatic brain injury (3). This is supported by the recommendations of the Utstein guidelines for uniform reporting which state that the Pediatric Outcome Categories should be used to record pre-arrest status, status at time of discharge, and status after one year (4,5).

2. Were all potential predictors included?

No. Predictors evaluated were: the presence of isolated heart disease; age; duration of mechanical ventilation before ECMO cannulation; duration of conventional CPR before ECMO; and the duration of ECMO management. Predictors not included in this study but found to be predictors of mortality specifically in the pediatric cardiac ECMO population include: postoperative residual cardiac defects (6); elevated right atrial pressure during termination from ECMO (7); epinephrine requirement at the conclusion of ECMO (8); hypoxia and elevated mean airway pressures at the termination of ECMO (8); infection/sepsis on ECMO (8-10); mediastinal bleeding (11); renal failure on ECMO (9,11-12); multiple system organ failure (10,12,13); and ECMO circuit complications (11). Patients with a single ventricle and cavopulmonary anastomosis have the highest mortality (8,11,13).

3. Did the investigators test the independent contribution of each predictor variable?

No. A bivariant analysis was conducted to compare the variables between study groups. Predictor variables were tested in the comparison groups: survivors vs. nonsurvivors, patients with isolated heart disease vs. those with other medical conditions, and patients managed with ECPR vs. those managed with conventional CPR.

4. Were outcome variables clearly and objectively defined?

Yes. The primary outcome measure was survival to hospital discharge. This was reported as a function of the number of patients who survived to hospital discharge. The secondary outcome was meaningful neurologic impairment that was prospectively defined to include a new seizure disorder, loss of developmental milestones, loss of muscle tone, or focal weakness. Although operationally defined, the attribution of these impairments was not clearly described as many of the patients did not have neurologic consultations performed and the records were retrospectively reviewed.

III. What are the results?

1. What is(are) the prediction tool(s)?

No prediction tool was generated as many of the predictors tested did not predict survival to hospital discharge. The authors report the results of their primary outcome and a number of descriptive analyses.

Survival to hospital discharge: Thirty-three of 66 ECPR events (50%; 95%CI 37% to 63%) resulted in decannulation and survival at least 24 hours, and 21 of 64 children (33%; 95%CI 22% to 46%) survived to hospital discharge. This compares with 56 of 79 (71%) cardiac arrest events not treated with ECPR with a return of spontaneous circulation, and 48 of 79 (61%) survived for at least 24 hours. Twenty-six of 73 (36%; 95% CI, 25% to 48%) non-ECPR cardiac arrest patients survived to hospital discharge, which is not statistically different when compared with the ECPR population.

The only statistically significant difference between groups appeared to be the presence of isolated heart disease vs. other medical conditions. Survival to hospital discharge was much higher (44%) in the isolated heart disease group than among those with other medical conditions (9.5%). Weight, age, duration of CPR, duration of ECMO management, and days of mechanical ventilation prior to ECMO cannulation were all not statistically different between groups.

Descriptive Sub-analysis:

Hypothesis 1: Children with isolated heart disease would be more likely to survive following ECPR than children with other medical conditions.

True. Nineteen of 43 ECPR patients with isolated heart disease (44%; 95% CI, 29% to 60%) survived to hospital discharge compared with 2 of 21 patients with other medical conditions (9.5%; 95% CI, 1% to 30%; p < 0.01).

Hypothesis 2: Children who had a cardiac arrest in the early postoperative period following cardiothoracic surgery would have a higher survival rate than those who had not had cardiac surgery in the preceding 24 hrs.

Possibly. The authors found the survival rate in children < one day or < seven days following cardiac surgery did not differ from those children not in the immediate postoperative period. In 27 patients, ECPR was initiated within seven days following cardiac surgery performed using cardiopulmonary bypass (CPB). Eleven (41%) of these children survived to hospital discharge. In 22 patients, ECPR was initiated within the 24 hours following cardiac surgery performed using CPB. Nine (41%) of these children survived to hospital discharge. Previous studies have noted a possible trend toward improved outcomes when patients were placed on ECMO earlier rather than later in the postoperative period [7-9,13]. Due to the small number of patients enrolled, this study may have limited power to detect a significant difference (Type II error) in the survival rate of children who had a cardiac arrest in the early postoperative period and those who were not in the immediate postoperative period.

Hypothesis 3: A longer duration of conventional CPR would correlate with a decreased likelihood of survival but grossly intact neurologic survival remains possible after > 30 min. of CPR followed by ECPR.

True. Zero of 10 non-ECPR patients and 14 of 43 ECPR patients survived after > 30 min. of chest compressions (p = 0.05). The median duration of chest compressions in the ECPR group was longer than the conventional CPR group, 50 min. (range 5-105 min.) vs. 8 min. (range 1-62 min.; p = 0.07). Comparing survivors and nonsurvivors, patients with a prolonged duration of CPR when followed by ECPR had an increased likelihood of survival than if they were managed with conventional CPR alone. The median duration of chest compressions in the 23 cardiac arrest ECPR events in which the child survived was 50 min. (range 5 to 105 min.). Among the 43 ECPR events in nonsurvivors, the median duration of chest compressions was 46 min. (range 15 to 90 min.; p = 0.88). In contrast, the median duration of chest compressions in the conventional CPR events in which the child survived was 5 min. (range 1 to 30 min.) and 10 min. (range 1 to 62 min.) in those who did not survive to hospital discharge. The authors were not able to demonstrate that a shorter duration of chest compressions was associated with an increased likelihood of survival, but this may have been a type II error due to the limited power of the study.

The authors also found that a grossly intact neurological outcome was possible after prolonged CPR followed by ECPR. Of 21 survivors, 5 patients (24%) had significant neurologic impairment, including three with clinical and radiographic evidence of stroke and three with seizure disorder. For 10 of the 13 survivors greater than two months of age at discharge, admission and discharge PCPC and POPC scores were available. Of the six patients who survived after receiving > 60 min. of CPR, three were greater than two months old. Of these three survivors, two patients had no change in the PCPC or POPC scores and one1 patient had a change of one point in each category. Of the three patients less than two months of age, one patient had no sequelae, two patients had new seizures, and one patient had poor tone and leukomalacia on MRI. Thus, three of the six children who had received greater than 60 min. of CPR prior to ECPR survived without gross neurologic injury.

2. How well does the model categorize patients into different levels of risk?

There is no model defined which provides different levels of risk assessment. The data suggest that those patients with isolated heart disease are more likely to survive to hospital discharge following a cardiac arrest event compared with cardiac arrests events in patients with other medical conditions. This may not be applicable to all patients with heart disease, as the authors were clear to discriminate among those patients who had isolated heart disease. In the group of children who had other medical conditions, many of the nonsurvivors had heart disease: cardiomyopathy, Pentalogy of Cantrell, heart and lung transplant, and hypoplastic left heart syndrome.

3. How confident are you in the estimates of risk?

Forty-four percent (44%) of patients with isolated heart disease vs. 9.5% of patients with other medical conditions survived to hospital discharge following a cardiac arrest event and ECPR. The confidence intervals for these two groups overlapped but were significantly different. The estimated risk of survival to hospital discharge in those patients with isolated heart disease fell between 29% and 60%, and between 1% and 30% (p<0.01) in those patients with other medical conditions.

IV. Will the results help me in caring for my patients?

1. Does the tool maintain its prediction power in a new sample of patients?

Not assessed. This was a single-center retrospective study.

2. Are your patients similar to those patients used in deriving and validating the tool(s)?

Yes. This study evaluated pediatric patients outside the neonatal ICU who suffered an in-hospital cardiac arrest defined as > 60 sec. of external chest compressions or internal cardiac massage managed with conventional CPR or ECPR. Importantly, this study was conducted in a large academic hospital with the resources to support rapid ECMO deployment and cannulation of pediatric patients during active chest compressions. As noted, all of the children with isolated heart disease in this study were managed in a dedicated pediatric cardiac ICU, whereas those with other medical conditions were divided between the pediatric and pediatric cardiac ICUs. The management styles of the two units may differ, and it is unclear how much this may have contributed to the resulting survival outcome.

Furthermore, there is a selection bias imposed by the physicians in determining which patients to manage with ECPR. Patients are selected for ECPR based on suffering a low-cardiac output state which deteriorates to cardiac arrest and based on the presumed reversibility of the underlying cause for cardiac arrest. The authors note that those patients managed with ECPR and those with conventional CPR differed in important ways. The ECPR group included significantly more patients with primary heart disease than the non-ECPR group (54 of 64 (84%) vs. 22 of 73 (30%); p < 0.01).

3. Will the results lead directly to selecting or avoiding therapy?

Yes. The results indicate ECPR is particularly helpful when: the child suffers an in-hospital cardiac arrest, when the underlying process is presumably reversible, when conventional CPR has failed (even after > 60 min. of CPR), and when ECMO cannulation can be implemented rapidly. Although patients managed without ECPR had a similar survival outcome, some of those patients managed with ECPR had a good neurologic outcome. It is unclear if ECPR alone preserved good neurologic function in these children; however, even after prolonged CPR a child may have the potential for survival with good neurologic outcome.. The duration of CPR administered prior to ECMO cannulation in the survivors was 5 to 105 min.

During the seven-year period, the overall survival to hospital discharge for these patients was 33%. Based on data from the National Registry of CPR (NRCPR) inpatient cardiac arrest survival is approximately 26% (1). The authors result is comparable to published survival outcomes for in-hospital cardiac arrest events in children with primary cardiac disease managed in a pediatric cardiac ICU (14,15). It is notable that previously published reports do not elaborate on whether ECPR was employed or if ECMO was initiated after the return of spontaneous circulation. No direct evidence supports avoiding ECPR in patients with other medical conditions.

4. Are the results useful for reassuring or counseling patients?

Yes. Parents of children who have suffered an in-hospital cardiac arrest - even a prolonged one - treated with ECPR, may be reassured that survival to hospital discharge with good neurologic outcome may be possible. Although this was a large cohort of patients managed with ECPR, the numbers were small enough to limit the power of the study to define risk factors for mortality or predictors of survival in this population. This study noted what others have reported in the literature, that children with multiple medical conditions and multiple organ dysfunction have a higher risk for increased morbidity and mortality (10,12,13). ECMO is an expensive and invasive procedure. Therefore, physicians must continue to use their best judgment to determine the potential to maximize benefit and to minimize foreseeable harm in employing ECPR following cardiac arrest events.

Although several patients had a good neurologic outcome at hospital discharge, data regarding long-term neurologic outcome was not available. In counseling parents of patients who have suffered a cardiac arrest event, it is important to note that this study does not provide us with long-term neurologic outcome for the survivors of pediatric cardiac arrest managed with ECPR.

References

1. NRCPR: National Registry of CPR SAB Participant Report. American Heart Association, 2002. 1/1/2000-12/21/2001.
2. Reis AG, Nadkarni V, Perondi MB, et al. A prospective investigation into the epidemiology of in-hospital pediatric cardiopulmonary resuscitation using the international Utstein reporting style. Pediatrics 2002; 109:200-209.[abstract]
3. Carter BG, Taylor A, Butt W. Severe brain injury in children: long-term outcome and its prediction using somatosensory evoked potentials (SEPs). Intensive Care Med. 1999 Jul;25(7):722-8. [abstract]
4. Fiser DH. Assessing the outcome of pediatric intensive care. Journal of Pediatrics 1992; 121:68-74. [abstract]
5. Zaritsky A, Nadkarni V, Hazinski MF, Foltin G, et al. Recommended guidelines for uniform reporting of pediatric advanced life support: the pediatric Utstein style. Task force writing group. Circulation 1995; 92:2006-2020.[abstract]
6. Black MD, Coles JG, Williams WG, et al. Determinants of success in pediatric cardiac patients undergoing extracorporeal membrane oxygenation. Annals of Thoracic Surgery 1995; 60:133-138.[abstract]
7. Walters HL, Hakimi M, Rice MD, et al. Pediatric cardiac surgical ECMO: multivariate analysis of risk factors for hospital death. Annals of Thoracic Surgery 1995; 60:329-336. [abstract]
8. Ziomek S, Harrell JE Jr, Fasules JW, et al. Extracorporeal membrane oxygenation for cardiac failure after congenital heart operation. Annals of Thoracic Surgery 1992; 54:861-867.[abstract]
9. Raithel SC, Pennington DG, Boegner E, et al. Extracorporeal membrane oxygenation in children after cardiac surgery. Circulation 1992; 86(5 Suppl):II305-II310.[abstract]
10. Montgomery VL, Strotman JM, Ross MP. Impact of multiple organ system dysfunction and nosocomial infections on survival of children treated with extracorporeal membrane oxygenation after heart surgery. Critical Care Medicine 2000; 28:526-531.[abstract]
11. Meliones JN, Custer JR, Snedecor S, et al. Extracorporeal life support for cardiac assist in pediatric patients: Review of ELSO registry data. Circulation 1991; 84(Suppl III): III-168-III-172.[abstract]
12. Morris MC, Ittenbach RF, Godinez RI, Portnoy JD, et al. Risk factors for mortality in 137 pediatric cardiac intensive care unit patients managed with extracorporeal membrane oxygenation. Critical Care Medicine 2004; 32:1061-1069.[abstract]
13. Kulik TJ, Moler FW, Palmisano JM, et al Outcome-associated factors in pediatric patients treated with extracorporeal membrane oxygenator after cardiac surgery. Circulation 1996; 94(9 Suppl):II63-II68.[abstract]
14. Parra DA, Totapally BR, Zahn E, et al. Outcome of cardiopulmonary resuscitation in a pediatric cardiac intensive care unit. Critical Care Medicine 2000; 28:3296-3300.[abstract]
15. Rhodes JF, Blaufox D, Seiden H, Asnes JD, et al. Cardiac arrest in children after congenital heart surgery. Circulation 1999; 100 (19) Suppl: II 194-II 199.[abstract]

 

Comments

Back to the EB Journal Club Index