The effect of skin-to-skin contact at birth, early versus immediate, on the duration of exclusive human lactancy in full-term newborns treated at the Clínica Universidad de La Sabana: study protocol for a randomized clinical trial

Human lactancy is a simple and cost-effective strategy that influences infant and maternal mortality rates. Skin-to-skin contact (SSC) is an immediate postpartum period strategy that has proven to benefit the initiation and continuation of human lactation and to decrease hospitalization during the first week of life. This study aims to determine the effect of SSC initiation at birth (immediate versus early) in healthy, full-term newborns treated at the Universidad de La Sabana Clinic on the duration of exclusive human lactation. A randomized, blind clinical trial will be performed with full-term healthy newborns born at the Universidad de La Sabana Clinic. The blind trial participants will be those persons measuring the results and analyzing the data. The sample size will be calculated for a type I error of 5 %, a two-tailed type II error of 20 %, and an estimated percentage loss of 30 %; 150 infants will be included in each group. Randomization will be performed using permuted, size-6 blocks. Descriptive analysis will be conducted using central tendency and dispersion measurements. A bivariate analysis will be performed to determine which variables are associated with exclusive lactancy at 6 months. For continuous variables, Student's t test will be used for independent samples, and the Wilcoxon rank sum test will be used if the assumptions of normality for the t tests are not fulfilled. The assumption of normality will be evaluated using the Shapiro-Wilk and Kolmogorov-Smirnov tests. Categorical variables in contingency tables will be constructed to assess the independence between variables using the chi-square test, or Fisher's exact test when the assumption of the number of cases is not met by the values in the contingency tables multiplied by two. This will be calculated as a measurement of the effect of relative risk (RR) with confidence intervals; the adjusted measurements will be calculated using a multivariate regression Poisson model. Variables with significant results will be used in the bivariate analysis, and those with biological plausibility will be used for the adjustment. The analysis will be carried out for a two-tailed type I error level of 5 %. The Stata 11 program will be used for data analysis. An interim analysis will be performed upon the submission of half the expected events (106), setting limits for the early termination of the trial according to the method proposed by Pampallona and Tsiatis (1994). There will be two SSC randomiz