Corrigendum to effect of human milk-based fortification in extremely preterm infants fed exclusively with breast milk: a randomised controlled trial

We acknowledge that the description of the eligibility and inclusion criteria in our article “Effect of human milk-based fortification in extremely preterm infants fed exclusively with breast milk: a randomised controlled trial” published in this journal¹ may have been open to misinterpretation. There has been a question regarding the inclusion of infants before 3 days of age. We therefore appreciate the opportunity to clarify this.

In the method section of the published paper we wrote “Infants were eligible if they were born between 22 + 0 and 27 + 6 weeks of gestation, had survived the first three days of life, and the home clinic of the infant had the logistics of maintaining the study intervention until postmenstrual week (PMW) 34 + 0”. This sentence should be corrected to “Infants were eligible if they were born between 22 + 0 and 27 + 6 weeks of gestation and the home clinic of the infant had the logistics of maintaining the study intervention until postmenstrual week (PMW) 34 + 0”.

Thus according to the study protocol,² postnatal age was not an inclusion or exclusion criterion, so even though the vast majority of infants for practical reasons were included in this fortification trial after three days of age, being alive at three days of age was not an inclusion criterion and death before three days of age was not an exclusion criterion.

The original description was the criteria for inclusion in the screening log. Since mortality is high during the first days of life, families were normally approached for informed consent only when the child was in a stable condition, and since fortification most often is started well after three days of age, the study protocol stated that, all infants born before gestational week 28 + 0 at an inclusion site who survived the first three days should be entered anonymously into the screening log. These infants (n = 405) are the ones presented as eligible in the CONSORT flow chart in the original article (Figure 1 in the original article).

The study protocol has previously been published² and is also available on the webpage of the original article in eClinicalMedicine. Primary and secondary outcomes were analysed for an intention-to-treat population. There was one infant that died at two days of age after being randomised to the human milk-based fortifier group. In accordance with the stipulated study protocol, this infant was included into the intention-to-treat analyses. A per-protocol analysis was also performed which only included events with an onset from the first day of fortification, excluding infants with protocol violation (withdrawal of consent before PMW 34 + 0, formula before PMW 34 + 0, lost to follow-up), or infants that did not receive the study product. The infant mentioned above was not included in the per protocol analyses. As described in the original article, the results of the per-protocol and intention-to-treat analyses were similar, with no significant differences observed in any clinical outcome.

In summary, the correction of our eligibility and inclusion criteria in the paper does not affect the results, interpretation, or conclusions of the study, since all procedures and analyses strictly adhered to the pre-specified inclusion and exclusion criteria outlined in the study protocol.