Study Setting
The study was conducted in the Southern Highland Zone (Iringa and Njombe) regions. Iringa Region is served by a total of 36 health facilities, of which 13 are hospitals and 23 health centers. All these health facilities provide CTC and PMCTC services. The Iringa Region borders the dry belt of central Tanzania in the north and south by Lake Nyasa. It lies between latitudes 7∘ 05∘ 32 and 12 South and longitude 33∘ 47∘ 32 to 36 East of the meridian. Iringa Region is contiguous with the Dodoma and Singida regions in the North, Mbeya to the west, and Morogoro in the East, and Ruvuma in the South. Lake Nyasa separates Iringa and Malawi in southwestern Tanzania. It has six districts with a population of 941,238 based on the Tanzania national census 2012 [18].
Njombe Region is among the 31 managerial regions of Tanzania. It was officially registered in March 2012, from the Iringa region as an independent region. The 2012 national census shows that the population of Njombe is 702,097[18]. The region is bounded by Mbeya Region South-east about 100 km. The northern end is well marked in 8° 50′ S by an escarpment falling to the Usangu plains, the eastern branch of the East African rift valley, and the basin of the Great Ruaha River. Southwards the range terminates in the deep valley of the Ruhuhu river in 10° 30′ S, the first decided break in the highlands that is reached from the north along the east coast of Lake Malawi.
These regions are estimated to have the highest HIV prevalence whereby16.5% of adults are infected in Njombe and 19.2% in Iringa [18]. These two regions have therefore been selected because of the remarkably high prevalence of HIV.
Study Design
A hospital-based analytical cross-sectional study employed a quantitative approach was used. The study population comprised of HIV-positive lactating mothers attending the PMCT program at Iringa and Njombe Regions.
Inclusion Criteria
All HIV-positive lactating mothers with an infant aged 6 to 12 months who were attending the PMCT program during data collection were included.
Exclusion Criteria
All HIV-positive lactating mothers with very seriously ill children who were not able to concentrate on answering the questions were excluded from the study. Mothers diagnosed with cognitive or psychiatric conditions were also excluded as their level of comprehension would be limited. HIV-positive lactating mothers who were very seriously sick at the time of data collection were excluded as body weakness would result in the inability to not being able to go through all questions comprehensively.
Sample Size
The sample size was estimated by using the Kish Leslie formula (1965).
Formula: n = (Z) 2 P (1-P)/ e2
Where,
n = the required minimum sample size
Z= constant standard normal deviate (1.96% confidence level)
P= estimated prevalence of HIV positive mothers who breastfeed exclusively up to 6 months of infants age which is 46% as per study by [6].
e= margin of error on p (set at 5)
Whereby Z= 1.96
P= 46%=0.46
e =5%= 0.05
n= (1.96)2 X (0.46) (1-0.46)/ (0.05)2 =0.95425344/0.0025=372
Therefore, the actual sample size for this study was 372 HIV lactating mothers.
Sampling Technique
The Census method was used to include regional hospitals that are Iringa and Njombe Regional hospital. In other health facilities simple random procedure was employed whereby in Iringa there are six districts out of these only 3 were selected by lottery replacement method after that one hospital was selected from each district using the lottery replacement method, and two health centers from each district. The same procedure of simple random method by lottery replacement was used in Njombe Region whereby three districts out of six were selected; one hospital was selected from each district using the lottery replacement method, and two health centers from each district. Then, systematic random sampling was employed in which the first round of HIV-positive lactating mothers was identified from the clinic registration. In the second round, calculation of the Kth interval was done using the Kth formula to select mothers who were invited to participate in the study.
Kth = N/n
Whereby,
N = total population units
n= sample size
In this study, the total population of HIV lactating mothers was 963 and the sample size was 372
963/372=2.5 therefore the sampling interval included every second HIV lactating mother.
Data Collection Technique and Tool
The data for this study was gathered through face-to-face interviewer-administered structured questionnaires adapted and modified from [6,19,20]. Data were collected by two trained research assistants and the principal investigator. Standard structured questionnaires were then translated by the language teacher to Kiswahili which the language is spoken by study participants. The Kiswahili version questionnaire was used. Since data collection was done during COVID-19 precaution against protection was ensured all researcher assistants, principal investigator, and the study participants used sanitizer and masks, as well as one meter between the study participants was observed.
Variables and its Measurements
Dependent Variables
Exclusive Breastfeeding status was measured by nominal scale as exclusive breastfeeding practices and non-exclusive breastfeeding practices as continuous breastfeeding since birth such as starting to provide the baby with breast milk within the first hour after delivery, breastfeeding on demand day and night, and continuous breastfeeding alone up to 6 months.
Independent variables
Socio-demographic Characteristic: comprised 19 questions and were measured by nominal scale i.e. residence, education, marital status, occupation, a model of delivery, place of delivery, counseling on EBF whereby age, and income, were measured by an ordinal scale.
Knowledge on exclusive breastfeeding was measured by a nominal scale involving 12 questions, with yes/no answers which were then converted into correct and incorrect. The total score was obtained and computed for mean to categorize it into adequate and inadequate knowledge. The mean score of knowledge on EBF among lactating mothers was 7.66 the maximum score being 12 points while the minimum score was 1 point. A score below the mean was considered inadequate knowledge and above the mean adequate knowledge.
Perceived Benefits of exclusive breastfeeding
Perception towards EBF was measured using 5 points Likert scale. 13 questions were used to determine the perception of mothers. The order of scoring for positive statements was strongly agree = 5, agree = 4, undecided = 3, strongly disagree = 2, disagree = 1 and vice versa for negative statement. The total score was obtained and computed for mean and then categorized into positive and negative perceptions.
The mean score of perception on EBF among lactating mothers was 41.38 the maximum score being 60 points while the minimum score was 22 points. A score above the mean was regarded as a positive perception, while a score below the mean was considered as a negative perception.
Data Analysis
The descriptive statistics used were frequency and percentages in categorized variables like gender, infant HIV serostatus, areas of residence, education level, income level, marital status, and occupation. Also, frequency and percentages were used in determining the prevalence of outcome variables. The mean or median, standard deviations, and range were used to summarize continuous/discrete random variables such as age, Likert scale items. In analyzing Likert scale items on perception and knowledge the mean score was generated. The hypothesis was tested using chi-square to test the proportion of outcomes (EBF practices) across different exposures.
In building the logistic regression model, the process started from simple to complex analysis that led to parsimonious models. The independent variables were added one after another and those with significant results in the univariate were adjusted in the final model. The measure of effects was estimated by the Odds ratio and was tested at a 95% confidence interval and a 5% significance level