Study Design and Participants
The present prospective research was carried out for the purpose of investigating the breastfeeding self-efficacy and its relationship with the perceived stress and breastfeeding performance in 171 nursing mothers with late preterm infants born in Alzahra Medical Center of Tabriz. The inclusion criteria of the research are as follows: nursing mothers, the tendency to breastfeed, mothers with a preterm infant, with 34 weeks to 36 weeks and 6 days of gestational age, single pregnancy, Iranian nationality, and hospitalization in Alzahra Medical Center at least for 12 hours after giving birth. The exclusion criteria included the mother’s who were absolutely or partially prohibited from breastfeeding (e.g. mothers with untreated tuberculosis, HIV infection, Cytomegalovirus, etc.), mothers with problems such as Mastitis, Abscess, breast surgery, or breast anomalies, and the ones suffering from metabolic diseases such as Phenylketonuria (PKU), Galactosemia in previous children, suffering from serious systemic diseases or psychological problems on the basis of the patient’s medical records, infants’ congenital disorders and hospitalization in NICU.
The sample size was calculated to be 171, based on the breastfeeding performance variable and considering 0.05 acceptable errors and around the mean (m=3.6), 95% confidence intervaland standard deviation of 1.2.
Sampling and Data Collection
In the present research, after obtaining the ethical code from the Committee for Ethics of Tabriz University of Medical Sciences (IR.TBZMED.REC.1398.128), the sampling was carried out by the simple and continuous method for 5 months. In this way, the researcher has referred to postpartum ward of Al-Zahra Educational and Medical Center in Tabriz for sampling everyday .Afterward, upon mothers’ permission and providing explanations regarding the purposes of the research, sampling was carried out among the mothers’ that met the requirements, were interested in participating, and declared their agreement with participating in the research. Then, upon receiving their written consent, the mothers were interviewed and the questionnaires were completed usingmothers’ demographic information inventory, breastfeeding self-efficacy scale, and the perceived stress scale. The participants were followed for 4 months and the breastfeeding performance questionnaire was completed by the researcher 4 months after the childbirth through a phone interviewwith the mother.
Data Collection Tools
Personal-social characteristic inventory, breastfeeding self-efficacy scale, perceived stress scale, and breastfeeding performance questionnaires were employed for collecting the data.
The personal-social characteristic inventory included closed-ended questions with short answers concerning age, education, mother’s occupation, a history of any illness or disability, family’s economic status, type of childbirth, infant’sbirth weight, gestational age at the time ofchildbirth, mother’s nursing status in previous children, number of previous deliveries, time interval with the previous child, number of prenatal visits, etc.
The breastfeeding self-efficacy scale was employed to measure breastfeeding self-efficacy. This scale included 13 items, all of which were started with the phrase “I can”. These items were designed by Bandura (1977), (in the theory of self-efficacy) as positive sentences in the 5-point Likert scale. Scores for each response were considered as score 1 signifying never or I am not sure, and score 5 signifying always or I am completely sure. The scores of the breastfeeding self-efficacy scale were ranging from 13 to 65, i.e. the highest score indicates the highest breastfeeding self-efficacy. Psychometrics of the Persian version of the Dennis short-term breastfeeding self-efficacy scale (including 13 items) was performed by Araban et al. In 2014 in Iran (36). The results indicated that the scale has scientific validity and its reliability was evaluated with Cronbach's alpha coefficient which was desirable (91%).
The perceived stress was examined by the Perceived Stress Scale (14-item). The perceived stress scale was designed by Cohen et al. (1983) (37). Three versions of this scale are available, including 10, 4, and 14 items, which are used to measure the general perceived stress in the past month, the thoughts and feeling about stressful events, and the actions such as controlling, overcoming and coping withthe situationin case of experiencing the psychological pressures and stresses. This scale was scored according to the 5-point Likert scale, i.e. Never (0), almost never (1), sometimes (2), often (3) and many times (4). The phrases 4-5-6-7-9-10 and 13 were scored using reverse scoring and ranged from never (4) to many times (0). The lowest score obtained was (0) and the highest score was (56). The cutoff score equaled 21.8 and achieving a higher score indicated more perceived stress. The psychometric analysis of the Persian version of the Perceived stress by Cohen (including 14 items) was carried out by Sigari et al. (2014) (38).
The breastfeeding performance was assessed using the breastfeeding performance questionnaire, designed by Agunbiade in 2012(39). This questionnaire included 6 questions regarding the time of starting breastfeeding, frequency, duration, and exclusive breastfeeding. The questions with the correct answers obtained one score. The score of 4 and higher indicated good breastfeeding performance.
The validity of the personal-social characteristic inventory was assessed using face validity and content validity. The reliability of the breastfeeding self-efficacy scale, the perceived stress scale, and breastfeeding performance questionnaire was evaluated by test re-test on 20 cases, determining the intraclass correlation coefficient (ICC), and Cronbach's alpha coefficient (internal consistency). The intraclass correlation coefficient for breastfeeding self-efficacy scale, the perceived stress scale, and breastfeeding performance questionnaire were calculated to be 0.91, 0.85, and 0.93, respectively, and Cronbach's alpha coefficient for the said questionnaires was estimated as 0.85, 0.78, and 0.92, respectively.
The data were analyzed using SPSS -Version 24 software. The normality of the data was determined by Kolmogorov–Smirnov tests. The relationship between the variables was analyzed using analytical-descriptive statistics such as Pearson and Spearman correlation tests, independent T-Test, One-way ANOVA, and multiple linear regression test.