Introduction: pregnant women need access to skilled attendance at birth and emergency obstetric care (EmOC) to avert maternal deaths. While poor EmOC services may explain the high maternal mortality, inadequate knowledge of providers is also part of the problem. This forms the basis of this paper, in a setting where 50.2% of women deliver in a health facility but maternal mortality remains high at 531/100,000 live births, compared to the national average of 362/100,000 in Kenya. Methods: a facility based cross-sectional survey was conducted in 2018 with a set of knowledge questions extracted from the averting maternal death and disability toolkit. Providers knowledge for maternal and newborn health (MNH) was assessed by interviewing nurses on duty in the maternity units. Data were entered in Ms Access and exported to R version 3.6.2 for descriptive and logistic regression analysis. Ethical clearance was obtained from Kenya Medical Research Unit. Results: a total of 55 nurses were interviewed. Majority (71%) of the respondents were diploma nurses. The overall knowledge score for MNH among the providers was adequate with a score of (64%). Generally, the midwives and higher diploma nurses consistently scored higher than diploma nurses in all the topic areas of MNH. In the mixed linear regression, determinants of knowledge score were seen in provider-level variables. Conclusion: overall, the providers scores were higher on intrapartum and newborn care compared to scores on care for complications. We conclude that in-service training on EmOC to providers is critical to reduction of maternal mortality.
Study setting: this study was conducted in Vihiga County in Kenya. Vihiga County is one of the 47 Counties located in the western region of Kenya and covers an area of 531 km2 [15]. The County has five administrative sub-Counties namely Hamisi, Emuhaya, Luanda, Sabatia and Vihiga. The total population in the County was 600,000 in 2019 [16]. The County has a total of 90 health facilities (public and private) [15]. From these 45 (50%) offer maternity services. For this study, 30 health facilities were selected on the basis of offering maternity services 24 hours a day, 7 days a week and had conducted a total of thirty (30) normal deliveries the previous three months (January – March 2018). Among them, one was the Vihiga County Referral Hospital. There were three level-4 facilities in Sabatia, Hamisi and Emuhaya sub-Counties. All the four facilities provided comprehensive EmOC while the remaining 25 facilities offered basic EmOC. The facility based cross-sectional survey was conducted in the maternity units of the 30 health facilities between April and May 2018. This study was part of a larger study reference number (KEMRI/CPHR/005/07/2015) that assessed the health systems readiness to offer emergency obstetric care (EmOC) in Vihiga County, Kenya. Study population: the participants in this study were the health providers, mainly nurses on duty in the maternity units at the time of the survey. All the 30 maternity units had three or fewer nurses on duty at the time of the survey hence all nurses were invited to partake in the survey. Knowledge questionnaire: a quantitative structured questionnaire was adopted from the Averting Maternal Death and Disability Emergency Obstetric and Newborn Care (EmONC) needs assessment toolkit [10]. The questionnaire determined the knowledge of health providers on routine and emergency care during pregnancy, intrapartum care, essential newborn care and care for sick newborns. A pre-test of the questionnaire was conducted in a facility outside of Vihiga County. Necessary changes to the tool were made and the final tool adopted for the study. The questionnaire was administered by an interviewer who read each question out loud in English to the participant and responded directly into the tool. Answer options were not provided to nurses; instead, nurses provided their own answers and the interviewer marked the multiple choice answers as appropriate. Because all participants were trained in English, they answered in English without a need to translate to the local language. Two interviewers conducted the survey and they were part of the research team members with extensive familiarity and training in the tool; they contributed to the design of the tool and led the piloting of the questionnaire. Data completeness and accuracy was checked on a daily basis by supervisors. Data analysis: data was entered in Ms Access and exported to R version 3.6.2 for descriptive and logistic regression analysis. Based on the questions administered to the health care providers, a knowledge level score was generated. Average summary scores were then calculated for each specific question included in the topic area. Each knowledge question had multiple correct answers; that is, answers that the respondents were expected to offer spontaneously. If a correct answer was not offered, the interviewer coded the response as “not mentioned”. If a spontaneous answer did not appear as one of the multiple choices, it was not taken into consideration for scoring purposes. Respondents were scored on each question by calculating the number of correct responses mentioned out of the total possible and standardizing this to a scale of 100. A mixed-effect linear regression model was used to identify determinants of health providers´ knowledge on maternal and newborn care. The regression coefficient informs us how much the summary knowledge score is expected to increase when the independent variable increases by one, holding all other independent variables constant. In addition to the uni-variate model, we fitted three models. Model I was fitted using provider level variables, model II using facility level variables and model III was fitted on a combination of both provider and facility level variables. Akaike´s information criterion (AIC) was run to measure the model fits and complexity. For the given models fitted on the same data, the model with the smallest value of the information criterion is considered to be the best. The knowledge scores were operationally defined as follows: adequate knowledge: if the health provider answers correctly more than 50% of the questions; inadequate knowledge: if the health provider answers correctly less than 50% of the questions. Ethical approval: ethical approval for this study was granted by the Kenya Medical Research Institute (SSC protocol No: KEMRI/SERU/CPHR/ 003/3277). Written informed consent to conduct this study was obtained from the director of health in Vihiga County prior to conducting the study. Written informed consent was obtained from individual participants before administering the questionnaire.
N/A