Correspondence

Neonatal MRI and Neurodevelopmental Outcomes

N Engl J Med 2006; 355:2373-2375November 30, 2006

Article

To the Editor:

Magnetic resonance imaging (MRI) is expensive and time-consuming, and it involves risks. From a Bayesian perspective,1 the use of MRI at term equivalent in preterm infants, as reported by Woodward et al. (Aug. 17 issue),2 is useful only if the results lead to revision of the pretest probability of neurodevelopmental impairment to a level above or below the threshold for follow-up or early intervention in infants at high risk.

The inclusion criteria and the 28% prevalence of neurodevelopmental impairment (the pretest probability) indicate that the study patients were already above the accepted risk threshold for follow-up3 and possibly for early intervention. Table 1Table 1Probabilities of Neurodevelopmental Impairment. shows post-test probabilities of neurodevelopmental impairment, calculated using likelihood ratios4 from the data of Woodward et al. The absence of findings on MRI does not appear to decrease the probability of neurodevelopmental impairment enough to dispense with follow-up or early intervention in these infants. Abnormal findings on MRI have wide ranges of probabilities of neurodevelopmental impairment (calculated from 95% confidence intervals for sensitivity and specificity in the study data), which imply uncertainty about prediction and limit the value of MRI in counseling. The degree of usefulness of routine MRI in preterm infants appears low.

Gautham K. Suresh, M.D.
Medical University of South Carolina, Charleston, SC 29425
suresh@musc.edu

4 References

1. 1

   Bianchi MT, Alexander BM. Evidence based diagnosis: does the language reflect the theory? BMJ 2006;333:442-445
   CrossRef | Web of Science | Medline

2. 2

   Woodward LJ, Anderson PJ, Austin NC, Howard K, Inder TE. Neonatal MRI to predict neurodevelopmental outcomes in preterm infants. N Engl J Med 2006;355:685-694
   Full Text | Web of Science | Medline

3. 3

   Follow-up care of high-risk infants. Pediatrics 2004;114:Suppl 1:1377-1397
   Web of Science

4. 4

   Straus SE, Richardson WS, Glasziou P, Haynes RB. Evidence-based medicine: how to practice and teach EBM. 3rd ed. Edinburgh: Churchill Livingstone, 2005:74-5.
   

To the Editor:

Woodward et al. described the use of MRI to screen for neurodevelopmental outcomes in premature infants. Although the study has merit, the suggestion of universal MRI screening of premature infants raises questions of feasibility, utility, safety, and cost. In the United States, 2% of all infants are born before 32 weeks of gestation.1 Universal screening of these infants could result in more than 80,000 MRI scans per year at a cost of more than $80 million. The authors avoided using sedation, but their approach is less feasible on a larger scale. Most centers sedate infants for MRI despite the well-documented risk.2 Therefore, universal MRI screening should be considered only if early detection of risk has the potential to affect the neurodevelopmental outcome significantly. Given the high risk of developmental delay, all infants born before 32 weeks of gestation deserve early intervention with physical, occupational, and speech therapies and care by a developmental pediatrician. Since careful clinical monitoring and appropriate intervention are already the standard of care, early MRI screening is unlikely to affect clinical outcome.

David T. Teachey, M.D.
Children's Hospital of Philadelphia, Philadelphia, PA 19104
teacheyd@email.chop.edu

2 References

1. 1

   Horbar JD, Badger GJ, Carpenter JH, et al. Trends in mortality and morbidity for very low birth weight infants, 1991-1999. Pediatrics 2002;110:143-151
   CrossRef | Web of Science | Medline

2. 2

   Stokowski LA. Ensuring safety for infants undergoing magnetic resonance imaging. Adv Neonatal Care 2005;5:14-27
   CrossRef | Medline

Author/Editor Response

The cost of preterm birth is substantial and growing, particularly in relation to the very high rates of disabilities among preterm infants, which are estimated to cost more than $10 billion per annum.1 The early identification of infants who are at greatest risk and the provision of effective interventions are important. We have demonstrated that cerebral abnormalities on neonatal MRI were stronger predictors of subsequent severe neurodevelopmental disability than traditional clinical risk factors such as gestational age at birth, birth weight, and findings on ultrasonography. We agree with Teachey that all children born before 32 weeks of gestation ought to receive high-quality, multidisciplinary monitoring, early intervention, or both. However, in many states and countries, the use of such services is determined by “risk” criteria or resource limitations; thus, it is rational to target, to the best of our ability, those infants who are at greatest “risk.” MRI during the neonatal period will significantly improve risk prediction and can easily and safely be undertaken without sedation at a cost less than that of 1 day of care in the neonatal intensive care unit.

Finally, as we suggest in a footnote to Table 2 and in the Discussion section of our original article, and as the calculations of Suresh indicate, findings on MRI are not sufficient for accurate risk prediction. This is not surprising, given strong evidence showing that the development of very preterm children, like that of all children, is influenced by multiple factors including genetic, neurologic, medical, social, and familial factors. Thus, we recommend that the MRI of an infant be interpreted alongside a structured neurologic examination, a clinical history, and assessments of family factors, such as maternal well-being and parenting skills, that are also known to influence child development.2

Terrie E. Inder, M.D.
St. Louis Children's Hospital, St. Louis, MO 63110

Lianne J. Woodward, Ph.D.
University of Canterbury, Christchurch 8140, New Zealand
lianne.woodward@canterbury.ac.nz

Peter J. Anderson, Ph.D.
Murdoch Children's Research Institute, Melbourne, VIC 3052, Australia

2 References

1. 1

   Economic costs associated with mental retardation, cerebral palsy, hearing loss, and vision impairment -- United States, 2003. MMWR Morb Mortal Wkly Rep 2004;53:57-59
   Medline

2. 2

   Vohr BR, Msall ME. Follow-up of high risk infants. In: Vergara ER, Bigsby R, eds. Developmental and therapeutic interventions in the NICU. Baltimore: Paul H. Brookes, 2004:267-92.