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

Neurodevelopmental outcome of infants supported with extracorporeal membrane oxygenation after cardiac surgery.

Hamrick SE, Gremmels DB, Keet CA, et al.

Pediatrics. 2003 Jun;111(6 Pt 1):e671-5 [full-text]

Reviewed by Sian Cooper MBChB MRCPCH, Royal Children's Hospital, Melbourne, Victoria, Australia

Review posted February 20, 2005

I. What is being studied?:

The study objective:

To evaluate the long term neurodevelopmental outcome of infants who underwent cardiac surgery and required postoperative support with extracorporeal membrane oxygenation (ECMO). The authors also evaluated possible predictors for death or disability.

The study design:

A retrospective case note review of 53 infants with cardiac disease who were supported postoperatively with ECMO between 1990 and 2001. Neurodevelopmental data was collected prospectively in a follow up clinic at 1, 1.5, 2.5 and 4.5 years of age. Chart review was approved by the Committee on Human Research at the University of California San Francisco, and consent was also obtained at the initial follow up visit.

The outcomes assessed:

The primary outcome measures were neuromotor outcome and cognitive outcome. Complete physical, neurological and age-appropriate developmental examinations were performed at follow up visits. The authors also assessed predictors for survival or disability.

II. Are the results in the study valid?

Primary questions:

1. Was there a representative and well-defined sample of patients at a similar point in the course of the disease?

Yes and no. The patients included were infants less than 1 year of age who received cardiac surgery for a congenital heart lesion before the initiation of ECMO. From a neurodevelopmental standpoint, this is a wide age range. There may be significant differences in the neurologic outcome following ECMO for neonates compared to older infants. In addition, the variety of cardiac lesions and palliations included here, as well as the varied indications for postoperative ECMO, make it reasonable to expect different neurologic outcomes among the study infants. Infants were excluded if they were supported for ECMO before surgical repair, or if they had congenital heart disease but were placed on ECMO for respiratory failure secondary to congenital diaphragmatic hernia. There is no mention of the infants' neurologic status before surgery, except noting the presence of DiGeorge in two patients and perinatal asphyxia in another.

2. Was follow-up sufficiently long and complete?

Patients were studied until 4.5 years of age. This follow up period is adequate to detect significant neuromotor and cognitive impairment. A longer follow up period may be useful to assess the function of those patients with "suspect" outcomes. Only one patient of 17 survivors was lost to follow up. Also to note is that 3 survivors were followed up in a different state, and so neurodevelopmental assessment was based on physical examinations performed by their local pediatrician; cognitive outcome was assessed using the Child Development Chart and Review, a different measure than was used for the other study patients.

Secondary questions:

3. Were objective and unbiased outcome criteria used?

A clear definition of adverse outcomes was provided at the beginning of the study. Neuromotor outcome was classified as normal, suspect (clumsiness, tremor, or mild tone and reflex changes without functional impairment) or abnormal (cerebral palsy or abnormalities associated with functional impairment such as diplegia or hemiplegia). Cognitive outcome was determined by scores obtained using the Bayley Scales of Infant Development II, the Stanford Binet Intelligence Scale and the McCarthy Scales of Children's Abilities or Wechsler Scale of Intelligence, according to the age of the child. Cognitive outcome was classified as normal, suspect (score on age-appropriate scale between 1 and 2 standard deviations [SD] below the mean) or abnormal (score 2 or more SD below the mean).

The authors do not say whether those assessing neuromotor and cognitive outcome were blinded to possible prognostic factors. For example, they may have been aware that the child had a cardiac arrest, or required haemofiltration.

4. Was there adjustment for important prognostic factors?

The authors have compared survivors with nonsurvivors, and identified predictors for in hospital mortality. They identified three predictors for in hospital mortality; there was no attempt to adjust for each of these and so we do not know whether they are independent predictors for mortality. And as noted above, there was no adjustment for potential pre-existing factors that could affect neurodevelopmental outcome.

III. What are the results?

1. How large is the likelihood of the outcome event(s) in a specified period of time?

Of the 53 infants in the study, 17 (32%) survived to hospital discharge. Thirty-six (68%) died in hospital.

Of 17 survivors to hospital discharge, there were 2 late deaths and 1 lost to follow up. Of the remaining 14 survivors, at 4.5 years of age, 10 (71%) had a normal neuromotor outcome, 1 was suspect and 3 were abnormal. Seven (50%) had a normal cognitive outcome, 3 were suspect and 4 were abnormal.

Predictors for in hospital mortality were cardiac arrest as an indication for ECMO (6% of survivors vs 31% of nonsurvivors, p=0.05), a longer aortic cross-clamp time (mean 32 minutes in survivors vs 73 minutes in non survivors, p=0.01) and requirement for continuous arteriovenous haemofiltration (CAVH; 35% of survivors vs 78% of non survivors, p=0.01). There was a trend toward statistical significance for length of aortic cross-clamp time as a predictor for cognitive outcome (normal group mean 23 minutes vs abnormal mean 47 minutes, p=0.06). No survivor with cross-clamp > 40 minutes had a normal cognitive outcome.

2. How precise are the estimates of likelihood?

The authors have not given any 95% confidence intervals in their paper. The relative risk of mortality associated with cardiac arrest as an indication for ECMO was 1.5 (95% CI 1.1-2.0), and the relative risk for CAVH as a predictor for mortality was 2.0 (95% CI 1.1-3.4). Confidence intervals cannot be calculated for aortic cross-clamp time without the raw data, which has not been provided.

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

1. Were the study patients and their management similar to my own?

Yes. The clinical characteristics of the patients are similar to those in other paediatric cardiac intensive care units. Detailed information regarding delivery of ECMO and intensive care management is not given; however it is likely we can apply this study to our own practice.

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

This study confirms that there is a high mortality rate (68%) in infants supported postoperatively with ECMO after surgery for congenital heart disease, consistent with reported rates of 50-61% in other studies (1-3). Provision of ECMO is expensive, requires highly specialized equipment and staff, and carries the risk of complications (bleeding, thrombosis, technical problems) which may be terminal events. However in this study population it is encouraging that of the survivors, two thirds had a normal neuromotor outcome, and half had a normal cognitive outcome at the age of 4.5 years. So although the risk of death is high, the results suggest that this treatment should be offered as there is more than a 70% chance of normal or mildly impaired neuromotor and cognitive function in survivors.

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

Yes. It is important for the family to be aware of the high mortality in this group of patients, and helpful to know about prognostic factors for mortality when counseling them. When infants have survived to hospital discharge, we can use this data when talking to their parents about longer term neurodevelopmental outcome. We can tell them that there is no single prognostic factor shown to influence outcome, but that of infants who have required postoperative support with ECMO and survive, around two thirds will have a normal or only mildly abnormal neurodevelopmental outcome at 4.5 years of age.

References:

1. Chatzis AC, Giannopoulos NM, Tsoutsinos AJ, Zavaropoulos PN, Kirvassilis GV, Sarris GE. Extracorporeal membrane oxygenation circulatory support after cardiac surgery. Transplant Proc 2004;36(6):1763-5. [abstract]
2. Morris MC, Ittenbach RF, Godinez RI, Portnoy JD, Tabbutt S, Hanna BD, Hoffman TM, Gaynor JW, Connelly JT, Helfaer MA, Spray TL, Wernovsky G. Risk factors for mortality in 137 pediatric cardiac intensive care unit patients managed with extracorporeal membrane oxygenation. Crit Care Med 2004;32(4):1061-9. [abstract]
3. Kolovos NS, Bratton SL, Moler FW, Bove EL, Ohye RG, Bartlett RH, Kulik TJ. Outcome of pediatric patients treated with extracorporeal life support after cardiac surgery. Ann Thorac Surg 2003;76(5):1435-41. [abstract]

 

Comments

Back to the EB Journal Club Index