Published online September 30, 2005
PEDIATRICS
Vol. 116
No. 4
October 2005, pp.
1054
(doi:10.1542/peds.2005-1601)
Ensuring Accurate Knowledge of Prematurity Outcomes for Prenatal Counseling: In Reply
Roger Soll, MD
Fermin Blanco, MD
Department of Pediatrics,
University of Vermont,
Burlington, VT 05405
Gautham Suresh, MD
Department of Pediatrics,
Medical University of South Carolina Children's Hospital,
Charleston, SC 29425
In Reply.—
MacKendrick and Caplan address 2 issues regarding our recent publication, "Ensuring Accurate Knowledge of Prematurity Outcomes for Prenatal Counseling."1 They raise concerns regarding the use of multicenter, aggregated outcomes data for counseling at a particular institution. They note that, for a variety of reasons, outcomes data at a particular institution may vary widely from those reported in multicenter databases. They suggest that individual-center data be used "if sufficiently robust and reliable" in counseling patients. This is precisely the problem. On a yearly basis, most centers have a small sample of patients from whom to draw their inferences. Statistically, there is a broad range of possible results that this could represent. The smaller the sample of local patients, the wider the confidence intervals and the range of possible outcomes. Almost no centers would have a sufficient number of patients to evaluate in each gestational-age category. For most, if not all, centers, the small sample size and the year-to-year variation in outcome makes reliable prediction extremely difficult. In fact, this may be part of the problem, because centers do not know which of the many statistics to quote. Do we quote from our own center? Do we combine gestational-age categories or data across several years to add to precision? For the vast majority of centers, aggregate data from multicenter networks will be much more reliable and realistic. In addition, MacKendrick and Caplan note that synthesized long-term disability rates from combined data from various outcome studies may lead to false conclusions. Again, the argument can be made that these data represent the most likely outcome of these patients. Few centers have reliable data from their own follow-up to address this point adequately. The false security in believing that your data are better than your neighbor's data and the fallacy of overloading patients with too much information that cannot be processed (developmental outcome at a variety of ages and the specific domains that may be affected) is as much a disservice to patients as giving unreliable data.
In our article we stated that "[t]here seems to be sufficient evidence now to establish the limit of viability in the United States at 23 weeks of gestation."1 The authors note that, in fact, decision-making for resuscitation of infants born at 23 and 24 weeks' gestation is a joint process to be worked out with the family. Decisions reached will vary and will be heavily influenced by not only the data but by the values and perceptions of the family. This is clearly true and should apply to infants of even slightly older gestational age. Parental decision-making throughout the period of limited viability (<25 weeks' gestation) is far from black and white and requires detailed discussion with parents and caregivers. Reliable data are just the starting point.
REFERENCE
- Blanco F, Suresh G, Howard D, Soll RF. Ensuring accurate knowledge of prematurity outcomes for prenatal counseling.
Pediatrics. 2005;115
(4). Available at: www.pediatrics.org/cgi/content/full/115/4/e478
PEDIATRICS (ISSN 1098-4275). ©2005 by the American Academy of Pediatrics
