Study design and setting
A hospital-based cross-sectional study was conducted in the Central Zone of Tanzania from April to July 2020. The zone is made up of three regions namely, Singida, Tabora and Dodoma. Each region has one Regional Referral Hospital where there is Neonatal Intensive Care Unit (NICU) and the KMC unit for preterm care.
Inclusion and Exclusion Criteria All post-natal mothers with stable preterm babies of one to four months post-delivery and practicing KMC were included in the study.
The study involved all postnatal-mothers with stablepreterm of one up to four months post-delivery who were practiced KMC and attending at preterm units/clinics from Singida, Tabora and Dodoma Regional Referral Hospitals. Those postnatal mothers with preterm babies who severely ill, and those preterm babies identified to have severe congenital abnormalities were excluded from the study.
Sample Size and Sampling Technique
A Kish Leslie (1965) formula (n=z2P (1-P)/d2) was used to calculate the minimum sample size. The proportion of 38.4% of factors associated with mortality among premature babies conducted in Mwanza, Tanzania [9]. The calculation (1.962x0.384 (1-0.384)/0.052) yielded a total of 363 sample size. The required sample size was selected from three regional referral hospitals available in the central zone (Dodoma, Singida and Tabora). Proportionate sampling technique was employed to recruit representative samples per each regional referral hospital. This was done after the total monthly attendance of preterm babies per clinic from each RRH. The proportional samples was computed by ni=(n/N)Ni;where ni=minimum sample size per facility, n=total sample size, N=total number of preterm babies from all facilities, and Ni = number of premature babies from each RRH. Therefore, the distribution of study representative per facility was; Dodoma RRH138(38%); Singida RRH 113(31.1%); and Tabora RRH112(30.9%).A systematic sampling technique was thereafter used to get participants from Tabora and Dodoma region. However, Singida regional referral hospital because of its small number of preterm babies; therefore all babies found on the day of data collection were included in the study.
Data Collection Procedure
Data regarding postnatal knowledge on KMC practice was collected using a structured questionnaire. The questionnaire was adopted and modified from the Facilitators Guide for Training of Kangaroo Mother Care (2,27,28). Before filling the questionnaire, the researchers created rapport and introduced themselves including the purpose of the research to the participants. The researcher identified a private place within the KMC unit were the interview was conductedto assure confidentiality and promote freeness of respondents to answer questions. All study respondents answered the same questions about KMC. Mothers of premature babies were interviewed to obtain the sociodemographic characteristics and the information of knowledge of KMC.
Validity and reliability
The principal researcher trained research assistants on the tools to ensure accurate interview and filling of the questionnaire. The questionnaire was tested for internal consistency using Cronbach’s alpha test which found 0.7. The tool was also pretested to assess for answerability of the items in the questionnaire.
Measurement of Variables
On testing the level of knowledge to postnatal mothers, 10 multiple choice questions were asked to respondents. Each item had A, B, C, D sub-items. Each correct response scored 1 point, while wrong response scored 0 point. There were 40 points in total in which those who scored 21 points were considered correct responses. The cut of points for knowledge level of respondents was 10.5 mean scores. Those who fell below the mean score were considered to have inadequate knowledge and vice versa [29,30].
Data Analysis
Data was checked for accurate filling before they were entered into the SPSS software v23. Data were then transformed and recoded into analyzable manner. Descriptive statistics was done to analyze the social demographic characteristics of the study participants to determine the frequency and percentages of each variable. Means and standard deviation were also used to present continuous variables. All continuous variables were transformed into categorical variables for binary logistic regression analysis. To determine the association between level of knowledge of postnatal mothers on KMC, and sociodemographic characteristics, Chi square values was presented at 95% CI interval with p-value <0.05 was considered as statistical significant. Both Univariate and Multivariate logistic regression analysis was used to determine the extent of association between knowledge and other significant variables.
Ethical Considerations
Before conducting the study, an ethical clearance letter from the University of Dodoma Ethical Review Committee was granted for the approval of the study. In addition to that, permission to conduct the study was granted by the Regional Administrative Secretary (RAS) of Dodoma, Tabora and Singida as well as regional referral hospitals medical officer in- charges respectively. All participants were given detailed information about the aim of the study. A verbal and written consent was sought from respondents before interview. The respondents were ensured of confidentiality and freedom of participation in the study.