1, § Department of Psychiatry, College of Physicians & Surgeons, 1051 Riverside Drive, Unit 40, New York, NY, 10032, USA
In our methods paper , we misstated the primary outcome variable, In addition, the power analyses on which the study Ns are based was not included in the article.
(1) The single primary outcome variable of our study design was Length of Stay (LOS), the number of days in the intensive care unit from day of birth to day of discharge home. All measures described in the paper as primary outcomes are secondary outcomes, including those outcomes listed in Tables 1 and 2.
(2) Previous studies that used LOS for evaluating interventions had total sample sizes of N = 128, 100, and 28, respectively [2-4]. To be conservative, we designated a total N = 150 to increase the prospect of statistical significance within a more acute population from our NICU (lower average birth weight and age) and to allow for attrition (including transfer to local level 2 NICUs). It was calculated that, for a significance level of alpha = .05, one hundred fifty infants would be sufficient to provide 80% power to detect a 5.3 day difference in length of hospital stay.
Welch MG, Hofer MA, Brunelli SA, Stark RI, Andrews HF, Austin J, Myers MM, FNI Trial Group FN: Family nurture intervention (FNI): methods and treatment protocol of a randomized controlled trial in the NICU.
J Paediatr Child H 2007, 43(12):831-836. Publisher Full Text
The pre-publication history for this paper can be accessed here: