Design/ Participants/ Setting
This was a randomized [1:1] controlled pilot study conducted in a primary health care center in Tijuana, México, from October 2018 to March 2019. The study was approved by the Ethics Committee from the Facultad de Medicina y Psicología (FMyP) at Universidad Autonoma de Baja California (UABC) and by the Sanitary Jurisdiction No. 2.
The health center is part of Baja California´s Health System Services, which are generally geared to assist people without health services provided by the social security system and people unable to afford private medical centers. To carry out this pilot, a center with a good inflow of people was selected, located in a strategic area where people from various socioeconomic backgrounds attend. The intervention was conducted in the same place where the immunization services were provided.
Mother-infant dyads that met the following inclusion criteria were selected: a) Request for immunization service at the health center, b) Mothers ≥ 18 years of age, c) Infants ≤ 4 months. This last criterion was established, considering the outcome would be measured 2 months after the intervention and the main interest was EBF, in that way if a baby had a maximum of 4 months at enrollment, after the 2 month-period EBF could still be assessed. Exclusion criteria were: a) Infants with low birth weight (<2500g) b) Infants with <38 gestational weeks, c) twins, and d) Maternal or infant contraindications to breastfeeding.
Intervention
The intervention consisted in providing interpersonal counseling, which focused on solving the main breastfeeding obstacle that a mother might had. The information provided included brochures and / or posters (depending on the topic) that a nurse explained to the mother before the infant was immunized. The control group received immunization counseling as well as standard breastfeeding support.
Five nurses were trained for the intervention. First, a 4-hour session was held, where 2 medical interns provided general information along with the 10 main obstacles to breastfeed as well as recommendations to overcome them. A second 4-hour session was held, with a community promoter, focused on communication skills based on the breastfeeding counseling WHO guidelines. Supervised practice sessions were carried out after the 2 training sessions with women from the target population. After it was determined that there was consistency between all the trainees, they were set ready to start the intervention. Due to the workload of the health center, only 3 nurses completed the training and only 1 was able to participate in the intervention.
Study process: Recruitment, randomization, follow-up
Mothers that brought their babies for vaccination services were approached in the waiting room of the health center. They were given the necessary information about the study and if they showed interest in participating, the informed consent was read and signed to obtain their permission. Subsequently, a questionnaire was applied to obtain socio-demographic data, as well as instruments to assess attitudes, self-efficacy and breastfeeding practices. In the end, they were asked for their main obstacle to breastfeeding from a list of 10, obstacles that were already observed and identified from this population. Two additional options were available: to add a different obstacle and / or to express not having any problem with breastfeeding.
Group assignation was conducted by the medical interns at the study site. Each time a new dyad was recruited, they opened sealed envelopes previously prepared by one member of the research team (DBG) using computer generated random numbers. Participants assigned to the intervention group received interpersonal counseling by the trained nurse, lasting between 5 to 10 min focusing on the main problem with breastfeeding. Participants assigned to the control group received information on vaccination. Both groups were scheduled for a follow-up vaccination service appointment.
Follow-up data was obtained by medical interns at the subsequent visit for vaccination service after the 2 month-period. If the participants had not attended the service, they were contacted by telephone to answer the questionnaires related to attitudes, self-efficacy and breastfeeding practices.
Outcome Assessment
Breastfeeding practices
The primary outcome was EBF 2 months after the intervention. The definition of EBF utilized was: “only breast milk or prescribed medications / vitamin supplements, without the use of water, juices, formula or solid food”. The evaluation was made based on a 24-hour recall. Other measured outcomes were any other type of breastfeeding and the main 3 obstacles for breastfeeding.
Breastfeeding attitudes and self-efficacy
To measure the attitudes and self-efficacy for breastfeeding, two subscales of the “Breastfeeding Attrition Prediction Tool (BAPT)” (14-16) were used: Negative Attitudes towards breastfeeding and lactation control (self-efficacy). The construct and predictive validity of the full scale and of these subscales have been previously determined in their original version in English (14-15) and in one translated into Spanish (16). Internal consistency has been estimated with a Cronbach's alpha of 0.784 for the negative attitude sub-scale and 0.864 for the control sub-scale. The reproducibility of this instrument was evaluated in a sample of women with similar characteristics to the study population and an internal consistency of 0.65 for negative attitudes and 0.78 for control were found.
Co- Variables
To measure social norms, a questionnaire from the Alive & Thrive Breastfeeding impact evaluation program in Vietnam was used (17). The socio-demographic variables collected were: occupation, marital status, level of education, place of birth, length of time living in Tijuana. Data from the mother: Number of children, type of delivery of the last child, and if mothers had received any information about breastfeeding from health care services. Infant data: Date of birth, sex, birth weight, and weeks of gestation at birth.
Data Analysis
Based on the population that attends immunization services at the health center and the inclusion criteria, it was determined to use a sample of 50 mothers for the intervention and 50 for the control group. This sample size provided 80% power to appreciate a change of 20% in EBF prevalence.
We used SPSS v21 for data analysis. Descriptive statistics were used to measure central tendency and dispersion for continuous variables and frequencies for categorical ones. In the case of inferential statistics, continuous variables were analyzed with Student t-test and categorical variables with Chi-square. To evaluate the effectiveness of the intervention, Chi-square was used for EBF and Student t-test for attitude and self-efficacy scores.
To adjust covariates, we used regression models, linear regression for continuous variables (Attitudes and self-efficacy) and logistic regression for EBF. A Difference in Differences (DID) analysis was performed using regression models. The Difference in differences indicates the pre-post difference that occurs in the intervention group minus the difference in the control group.