Survival and outcomes of extremely immature babies, glass half empty or…

Anderson JG, et al. Survival and Major Morbidity of Extremely Preterm Infants: A Population-Based Study. Pediatrics. 2016. This publication is derived from data in a public health database, the data for which comes from ICD-9 codes of the hospital discharge records. There were over 16,000 babies born in California between 22 and 28 weeks gestation in the years 2007-2011, but only 7000 are included as they had linked birth records and hospital discharge records. Of these 7000, there were a surprisingly large number, about 1,000, with major birth defects or chromosomal anomalies, which makes me wonder whether these babies are representative of the whole 16,000. The babies with major birth defects or chromosomal anomalies were not included in the remainder of the publication.

The first survival figures presented are of very little interest “Among the infants born at 22, 23, and 24 weeks, survival to 1 year of age was 6%, 27%, and 60%, respectively”, I say “of very little interest” because these figures include babies who were left to die in the delivery room. As the mortality of the babies of this gestation who don’t get intensive care is 100% there is little point including them in any survival statistics, the proportion who receive active care, and the proportion of those who survive are what is important. Which the authors do subsequently: “When examining survival after attempted resuscitation, the rate of survival to 1 year at 22, 23, and 24 weeks was 31%, 42%, and 64%, respectively”.

Morbidities were also more frequent with decreasing gestation, but I am not entirely sure how accurate all the diagnoses are, discharge record codes are not necessarily always reliable. The 22 and 23 week babies had severe IVH, PVL, NEC and BPD more frequently than at 24 weeks or later, RoP requiring surgery was more frequent at 22, 23 and 24 weeks.

Sepsis occurred in around 2/3 of the babies under 25 weeks, and in an extremely high proportion of all their babies (48% of all the infants from 22 to 28 weeks). They used the ICD-9 code for neonatal sepsis, which seems to include early and late onset sepsis, and doesn’t seem to require a positive culture. In one source I saw it seems to also include NEC, so NEC might have been counted twice in some of these records. In any case, this is an extremely high proportion of babies with a discharge diagnosis of sepsis, In comparison the CNN data in 2014 show about 20% of the same gestational age group with late onset sepsis (and about 2.5% with early onset). I don’t know why these sepsis figures are so high, I haven’t seen rates this high from anywhere in the world that I can think of, which makes me worry about the reliability of these figures. As funding may be related to the ICD-9 codes then there is a pressure on discharge coders to include anything that can be put under the code, and as there are no clear definitions for the neonatal sepsis code, could performing a septic work-up for a suspicion of sepsis end up being tagged?

Similarly the code for NEC includes the notoriously difficult to define stage 1 NEC, and may well be an over-estimate of what would be considered definite NEC in other reports, stage 2 and higher (although, even then, diagnostic accuracy of pneumatosis is poor).

I am emphasizing these points as the authors present the proportion of babies who survived without these major morbidities, but if the diagnoses are inflated then these proportions are much lower than the reality. Some of the diagnoses are probably accurate (coding for surgery for retinopathy for example) for others perhaps less so.

The authors also do not give an estimate of the precision of these numbers, as you will all see in an upcoming publication in Pediatrics, we (that is, a group of investigators led by Matt Rysavy) propose, among other things, that such reports should always include the confidence intervals of the outcomes.

Consistent with data from EpiCure2, and consistent with common sense, babies born at 22 to 28 weeks in a hospital with a regional NICU were more likely to survive than if they were born in hospital with an intermediate level NICU. The differences were even greater for infants from 22 to 24 weeks gestation.

Holsti A, et al. More than two-thirds of adolescents who received active perinatal care after extremely preterm birth had mild or no disabilities. Acta Paediatrica. 2016

In contrast to those data are these very optimistic results from Sweden, they have already published details of their approach which they refer to as “active perinatal care” a co-ordinated approach to supporting the extremely immature baby, just before and after birth. They describe it like this “a universal and consistent policy of APC after all deliveries at the threshold of viability at 23-25 weeks of gestation. This management includes the centralisation of all deliveries at less than 28 weeks whenever possible, the administration of antenatal corticosteroids and tocolytics, the presence of a certified neonatologist at all deliveries at less than 28 weeks of gestation and resuscitation of all infants at 23-25 weeks of gestation with any signs of life”. They have already shown that this policy increases survival, without increasing short-term complications and they now present the outcomes in early adolescence of 132 children born from 23 to 25 weeks in the mid 1990’s.

3.8% of the extremely preterm adolescents had disabling CP (unable to walk), and no controls, 1.5% (2 subjects) had severe visual impairment, and 1.5% were deaf. As usual in these cohorts the most frequent impairments were cognitive; they performed Wechsler IQ tests on the preterm babies and the controls, 31% of the preterms and 5% of the controls were more than 2SD below the mean on full-scale IQ scores; 18% of them were more than 3SD below the mean (and no controls).

So you could say that these results are much worse than the general population, the glass is nearly 1/3 empty, or you could also say that they are remarkably good, and the large majority of the patients are doing extremely well, the glass is over 2/3 full.

About Keith Barrington

I am a neonatologist and clinical researcher at Sainte Justine University Health Center in Montréal
This entry was posted in Neonatal Research. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.