In this randomized controlled trial that compared a structured lactation consultant support program to the standard of care for hospitalized jaundiced newborns, we did not find any dif- ferences in exclusive or partial breastfeeding proportions up to 6 months. This result is surpris- ing given the published literature. A recent Cochrane review found that both professional and lay support were effective in prolonging breastfeeding, with professional support being the most effective of the two [23]. In a different systematic review, when focusing on random controlled trials combining lactation support with educational programs in developed coun- tries, combined education and support strategies increased short-term breastfeeding rates by 36% [12]. Finally, in the PROBIT trial [13], a randomized controlled trial evaluating the effec- tiveness of a breastfeeding promotion intervention in the Republic of Belarus, an absolute in- crease of almost 37% in the prevalence of exclusive breastfeeding at 3 months was found in the intervention group. It is worth noting the sharp drop in breastfeeding proportions in both par- ticipant groups in the first month as well as between 3 and 4 months (Fig. 2). These time points may well represent crucial times that breastfeeding interventions should focus on.
Despite there being no difference in overall breastfeeding outcomes in our study, the sec- ondary qualitative findings were very positive towards our lactation support intervention. Par- ticipants described an enhanced sense of encouragement and reassurance, increased motivation and empowerment to continue breastfeeding, and an enhanced sense of emotional wellbeing. Mothers in the intervention group felt overall more confident with breastfeeding in comparison to those in the control group. This is crucial as maternal confidence is a strong pre- dictor of breastfeeding outcome [10], with lack of confidence leading to a higher likelihood of weaning in the first 6 weeks post-partum [23]. In the control group, most participants recog- nized that the support provided by nurses and physicians was limited. This is consistent with other studies showing that physicians and residents lack the skills to offer proper guidance to lactating mothers.
Our study has some limitations. Recruitment was much more difficult than anticipated, and we were not able to reach our target sample size of 62 mothers per group. Given the very large confidence interval around our risk ratio (0.56 to 1.24) for exclusive breastfeeding at 3 months, it is possible that the intervention could result in an improvement in breastfeeding in this popula- tion. It would appear however that the effect is likely small, and would require a much larger sample size. Thus, even if we had reached our target sample size, we likely would not have seen an effect of the intervention. The effect of intervention may have been attenuated since it appears that women with a strong desire to breastfeed were more likely to enrol in the study (the breast- feeding proportions in our control group were found to be much higher than those in the general population). We also noted that more mothers in the control group compared to the interven- tion group reported receiving breastfeeding support with previous infants and this may also have increased the proportion of mothers that maintained breastfeeding in the control group.