A randomized clinical trial involving adolescent mothers, their babies and their mothers (maternal grandmothers) who lived together in the same house and were attended at the maternity clinic of the Hospital de Clínicas de Porto Alegre was conducted. This general university hospital is accredited as a “Baby-Friendly Hospital” and performs between 3,000 and 4,000 births a year. The majority of the population assisted at this hospital is patients of low socioeconomic status who use the public health care system.
To calculate the sample size, the following parameters were adopted: α = 5%; β = 20%; ratio exposed/not exposed = 1:1; prevalence of EBF in the first month in the group not exposed to the intervention = 56% , and difference in the prevalence of EBF in the first month between the exposed group and group not exposed to the intervention = 25 percentage points. Thus, a minimum number of 48 subjects in each group were evaluated, to which 50% more subjects were added to compensate for possible losses and to facilitate the performance of multivariate analyses, totalling approximately 72 participants in each group.
Initially, a pilot study with 20 mothers was conducted to test the data collection instruments. To ensure that the intervention attended to the interests of the mothers and grandmothers, meetings with the adolescent mothers and the maternal grandmothers were conducted before beginning the study; these meetings helped to define the content and the intervention approach.
The adolescent mothers were recruited in the period from May 2006 to January 2008 and the follow-up occurred from June 2006 to February 2009. Daily, including on the weekends, adolescent mothers (younger than 20 years of age) who did and did not live with their own mothers were identified in the obstetric inpatient unit and screened for the fulfilment of the following inclusion criteria: living in the city of Porto Alegre, with healthy non-twin newborn infants, in the rooming-in ward, having started breastfeeding, with infant birth weight greater than or equal to 2,500 g. Pairs who had to be separated due to problems related to the mother or the baby were excluded from the study; additionally, adolescents who lived with their newborns’ paternal grandmother were excluded.
Once identified, the mothers were allocated into two groups: those who lived with their own mothers (the babies’ maternal grandmothers) and those who did not. These two groups were then randomly subdivided into blocks of two subjects for control or intervention groups; that is, if a mother was selected to the control group, the next eligible mother would necessarily be included in the intervention group. At the end, there were a total of four study groups, namely: (1) adolescent mothers not living with their own mothers and not exposed to the study intervention; (2) adolescent mothers not living with their own mothers but exposed to the study intervention; (3) adolescent mothers still living with their own mothers and not exposed to the study intervention; and (4) adolescent mothers still living with their own mothers and exposed to an intervention program directed at both.
Two researchers were specially hired to perform the selection of the mothers and administer the intervention. In the intervention groups, when the mothers and grandmothers lived together, both participated in the intervention. When they did not live together, only the mother received the intervention. The inclusion of this last group sought to evaluate to what extent the impact of intervention could be attributed solely to the adolescent.
Data collection was conducted at distinct periods. In the maternity hospital, after agreeing to participate in the study and signing an informed consent form, the adolescent mothers and the maternal grandmothers responded, through an interview, a structured questionnaire to obtain sociodeomographic data and information related to pre-natal care, the birth, and their previous experience with breastfeeding. Information about how the babies were fed in their first 6 months of life was obtained monthly, in a telephone interview with the mother, using a structured questionnaire containing questions about the child’s feeding. If telephone contact could not be made, in-house visits performed by research assistants who were blinded to the intervention. To verify the quality of the information, 5% of the mothers were randomly selected each month to undergo a second interview with the principal field investigator; this interview focused on some key questions from the follow-up survey.
The intervention sessions occurred at the maternity hospital and at the mothers’ homes. In the hospital, right before the time of discharge, breastfeeding counselling session were given in accordance with WHO guiding principles , that is, mother and professional dialoguing about many aspects of breastfeeding, such as breastfeeding importance and duration, factors that interfere with the production of milk, the technique of breastfeeding, the consequences of using a pacifier and the baby’s crying and communication. Doubts specific to each mother and grandmother were addressed and clarified as well. The mothers were encouraged to breastfeed during the intervention, whenever possible, to take advantage of the opportunity to observe the breastfeeding technique and to instruct as to the positioning of mother/baby and latch on.
The sessions were conducted by one member of the team formed by 2 nurses, a nutritionist and a paediatrician. All members had ample experience with breastfeeding; 3 were International Board Certified Lactation Consultant (IBCLC). These interventions were conducted individually and separately with each mother and grandmother. At home, the counselling sessions were conducted simultaneously.
For support material, flipcharts specially created for the study were used; one flipchart was prepared for the adolescent mother and another for the grandmother. These tools contained information about breastfeeding and about the prevention and management of the most common problems. One booklet with content similar to what was in the flipcharts was distributed at the end of the counselling session in the maternity hospital. In the booklet, there was space to add photos of the baby, which were obtained by the professional during the home visits.
In the home, when the mothers and grandmothers lived together, they received joint counselling sessions when the infants were 7, 15, 30, 60 and 120 days of age. In these sessions, the team members discussed difficulties confronted by the mothers and how they managed them and reinforced the messages given at the counselling sessions in the maternity hospital. Since many mothers in Brazil introduce complementary feeding when the children turn 4 months, this issue was discussed in the visit of the fourth month, emphasizing the importance of EBF in the first six months of life.
The expected outcome was the likelihood of children being exclusively breastfed at 6 months of life. EBF was considered to occur when the infants received breast milk as their only source of nutrition and hydration, without any solid or liquid supplement, including water and tea .
The database for the study was assembled using Microsoft Excel, with double entry of data. Statistical analysis was performed with SPSS 16.0.
The analyses were based upon the “intention to treat” principle. Initially, the characteristics of the control and intervention group participants were compared using the chi-squared test with Yates’ correction. The EBF median duration was calculated, along with their respective confidence intervals of 95%, in days; the Kaplan-Meier survival curves were constructed for EBF in the first 6 months for the 4 groups. To test the difference between the curves, the log-rank test was used. To quantify the impact of the intervention in the abandonment of EBF in the first 6 months, Cox’s regression was used via the hazard-ratio and its respective confidence interval of 95%.
The study was approved by the Ethics and Research Committee of the Hospital de Clínicas de Porto Alegre. The study was registered at ClinicalTrials.gov under the number NCT00910377.