Nynnamm - Effects of Infant Directed Singing on Breastfeeding Difficulties

In Sweden, there is an extensive restructuring of care in relation to the concept "good quality, local health care" (God och nära vård). Most new mothers are cared for at home in direct connection to birth, which places new demands on primary health care providers. A lack of knowledge and training for professionals will affect new mothers' opportunities to breastfeed to the extent that they wish. Furthermore, breastfeeding rates are decreasing in Sweden. Contributors to this trend could be health care-related factors, downsizing, staff shortages, and a lack of breastfeeding knowledge and supportive interventions. There is a risk for even lower breastfeeding rates when professionals lack supportive interventions for advising mothers. This is hazardous since its known that breastfeeding is an effective way to ensure child health and survival. Breastfeeding also provides significant health benefits for mothers and is a cost-effective way to reduce health care costs.

Purpose and aims The aim of this study is to evaluate the effect of a self-care intervention involving maternal infant-directed singing during breastfeeding on maternal stress, experiences, and breastfeeding rates up to six months postpartum among mothers with breastfeeding difficulties.

The specific aim of the study is to:

• investigate whether maternal infant-directed singing during breastfeeding is effective in increasing the breastfeeding rate 4 weeks after the intervention among mothers with breastfeeding difficulties (primary outcome).
• investigate whether maternal infant-directed singing during breastfeeding is effective in decreasing maternal stress and breastfeeding difficulties up to six months postpartum among mothers with breastfeeding difficulties (secondary outcome).
• describe maternal experiences of using infant-directed singing during breastfeeding among mothers who have breastfeeding difficulties (secondary outcome).
• investigate parent-infant moment-to-moment communication to provide a deeper understanding of breastfeeding situations and potential problems in this context (secondary outcome).
• investigate cortisol levels, sodium levels, and the somatic cell count in breast milk and saliva (cortisol only) among the participating mothers (secondary outcome).

Method Design An exploratory multicenter randomized controlled trial with two parallel groups with an allocation ratio of 1:1 is planned. The investigators hypothesize that a combination of standard breastfeeding support and maternal infant-directed singing during breastfeeding will improve the proportion of breastfeeding mothers, including the mother-infant dyad's experience, and decrease stress among mothers who have breastfeeding difficulties.

Participants and setting The study will be performed at different child health care centers and postnatal care facilities in the Dalarna, Uppsala and Värmland regions. Mothers who seek support for breastfeeding difficulties, including experiencing a low milk supply, sore nipples, and breast pain, feeling stressed or unwell while breastfeeding or having an infant who has problems with latching, will be eligible for inclusion. The participants must understand Swedish or English and be over 18 years old. During the health care visit, mothers will be informed about the study, and written informed consent will be obtained from the mothers who want to participate in the study.

Data collection and procedures The mothers will be randomized to one of two arms. The randomization will be performed by the health care staff or research team members using an automated and secure web-based system administered independently of the research team. Block randomization will be used, with blocks of 8 interventions and 8 controls.

The intervention group will receive standard breastfeeding support from health care staff combined with maternal infant-directed singing during breastfeeding. The starting point of the singing will be individually assessed and tailored to the dyads and will start approximately a few minutes before breastfeeding starts. An educational video about how to apply infant-directed singing will be shown to all mothers in the intervention group.

Mothers allocated to the control group will be offered standard breastfeeding support from health care staff, and data on secondary outcome measures will be collected during one breastfeeding session.

The health care staff will provide standard breastfeeding support to all mothers during their visit. Then, the mothers will breastfeed their infants. Before breastfeeding starts, the research team will collect baseline data (questions and biological samples) and connect the mothers to the Galvanic Skin Response (GSR) device, which measures changes in sweat gland activity that are reflective of the intensity of the mothers' emotional state/ arousal. The breastfeeding session will be video-recorded. After breastfeeding, the mothers will answer additional questions, and another saliva cortisol sample will be collected. The order may be different depending on infant hunger signals.

The research team will be available approximately 5-10 days a month during the data collection period, which is expected to last 1 ½ years.

Due to the nature of the intervention, blinding will not be possible for the health care staff and parents during the procedures. The research team will not be completely blinded; however, data collection will be performed by different researchers on different data collection days. Hence, no researcher will have full access to all mothers' randomization groups. The person calling to collect data for the primary outcome will be blinded. Additionally, for the secondary outcomes, the research team will be blinded to a large extent.

Data analysis and statistics A priori power analysis will be used to examine the power necessary to detect an 11% difference between the two groups in the breastfeeding prevalence 4 weeks after the intervention. There is a need to include 121 mothers in the intervention group and 121 mothers in the control group to be able to reject the null hypothesis that the proportion of breastfeeding mothers in the intervention and control groups is equal with a probability (power) of 0.8 using logistic regression. To account for possible dropouts, an additional 18 (nine in each group) mothers will be recruited, leaving us with a total sample of 260 mother-infant dyads. The Type I error probability associated with this test of the null hypothesis is 0.05.

An intention-to-treat approach will be used during the analysis. To analyze the data, several statistical models and qualitative methods will be used. The primary outcome-breastfeeding (yes/no)-will be analyzed with logistic regression analysis presented with odds ratios and 95% confidence intervals. Depending on the distribution of the variables, the secondary outcomes will be analyzed with appropriate comparative statistical models. Quantitative data will also be presented using descriptive statistics. Qualitative data (i.e., open-ended questions, notes, comments, interview data) will be analyzed with a qualitative content analysis. The video recordings and singing will be analyzed with microanalysis. Microanalysis can be defined as a social microscope, an in-depth, hand-coded, second-by-second analysis method of video footage, which may provide a deeper understanding of breastfeeding situations and potential problems in this context. With microanalysis, the investigators will, for example, be able to discern synchronicity and timing or the lack thereof in breastfeeding mothers on a micro level, providing research with knowledge on a macro level on how to better guide mothers in similar situations. Mixed methods will be used when analyzing data. Quantitative and qualitative data will first be analyzed separately and thereafter merged and interpreted as a whole.