A facility-based study of women’ satisfaction and perceived quality of reproductive and maternal health services in the Kenya output-based approach voucher program

Background: This is a facility-based study designed to assess perceived quality of care and satisfaction of reproductive health services under the output-based approach (OBA) services in Kenya from clients’ perspective. Method: An exit interview was conducted on 254 clients in public health facilities, non-governmental organizations, faith-based organizations and private facilities in Kitui, Kilifi, Kiambu, and Kisumu counties as well as in the Korogocho and Viwandani slums in Nairobi, Kenya using a 23-item scale questionnaire on quality of reproductive health services. Descriptive analysis, exploratory factor analysis, reliability test, and subgroup analysis using linear regression were performed. Results: Clients generally had a positive view on staff conduct and healthcare delivery but were neutral on hospital physical facilities, resources, and access to healthcare services. There was a high overall level of satisfaction among the clients with quick service, good handling of complications, and clean hospital stated as some of the reasons that enhanced satisfaction. The County of residence was shown to impact the perception of quality greatly with other social demographic characteristics showing low impact. Conclusion: Majority of the women perceived the quality of OBA services to be high and were happy with the way healthcare providers were handling birth related complications. The conduct and practice of healthcare workers is an important determinant of client’s perception of quality of reproductive and maternal health services. Findings can be used by health care managers as a guide to evaluate different areas of healthcare delivery and to improve resources and physical facilities that are crucial in elevating clients’ level of satisfaction. The study was conducted in Kitui, Kilifi, Kiambu, and Kisumu counties as well as in the Korogocho and Viwandani slums in Nairobi which are the OBA program sites. The services in OBA sites are provided by public, NGOs, FBOs, and private service providers. All participating sites were offering SMH services (ANC, Delivery, treatment of delivery complications, and post-natal care up to 2 weeks), LTFP methods, and a small number was providing SGBV services. This was a cross-sectional study conducted in OBA sites using a semi-structured interview guide administered through face-to-face in-depth exit interviews. Participants receiving OBA services were asked to describe their perceptions of the quality of services and reasons for satisfaction with the quality of services they had received in their current and previous visits. Perception was measured using a questionaire (Additional file 1) that was developed on the basis of literature review and suited for a healthcare setting [10, 24]. The questionaire consisted of a large number of items that were found to be imperative in measuring quality of and satisfaction with care. Women were specifically asked how they perceived the care they received during SMH visits, LTFP visits, and SGBV visits. Besides, they were also asked about the information they received, the conduct of the healthcare professionals, and adequacy of resources and services. The items were re-grouped into 23 items measuring perception. There were two additional questions; one, on whether the women were completely satisfied with the services and two, on the reasons for satisfaction or dissatisfaction. Perceived quality of services was rated on a five point Likert Scale 1 being “Completely Disagree”, 2 “Disagree”, 3 “Agree”, 4 “Completely Agree”, and 5 “Do Not Know”. In selecting participants, a multistage sampling technique was used to select the facilities offering OBA services. First, all OBA facilities were classified according to type of ownership-public and private and grouped according to County. Classification has been described elsewhere [16]. Within each County, a representative sample of facilities both public, NGOs, FBOs and private facilities was randomly selected. In the second stage, a conservative sample size was calculated to be 313 respondents. In order to determine the sample size the formula developed by Cochran [25] for proportion that are larger: n = z2pq/d2, where n = was the number of clients/respondents, z = is the critical value for standard normal distribution for the 95% confidence interval around the true population (1.96), p = estimated proportion utilising OBA services (which was based on the proportion of women of reproductive age living below the poverty line in Kitui, Kiambu, Nairobi, Kisumu and Kilifi estimated at 28.56% [26]), q = represented 100-p, and d = was the degree of accuracy (5%). The number of clients were equally divided amongst the chosen facilities (5 clients). A simple random technique was used to select the OBA clients who sought SMH, LTFP, and SGBV care at the time of the study. To randomly select the participants at the facility, the researchers used Stat Trek Random numbers generators which have been applied in other cross sectional studies [27]. The method uses statistical algorithm to give random numbers and instructions on how to use it (http://stattrek.com/statistics/random-number-generator.aspx). The researchers hit a calculate button and the number generator gave a random number table with five numbers between 1 and 20. Subsequently, the interviewers then interviewed the participants presented by these numbers on a single basis until the sample size was obtained. After data collection, the questionaires were then retured to the central OBA program management offices in Nairobi after which they were checked for completeness before inclusion into the database. Only fully completed questionaires with all essential details were included in the analysis and “do not know” response in the questionaire was treated as a neutral term for ease of interpretation. The data were analysed using Statistical Packages for Social Scientists (SPSS) version 18. Descriptive statistical analysis was carried out to describe the respondents’ social demographic characteristics and the time taken to reach the facility either by bus or by foot. Additionally, descriptive statistical analysis was conducted on the women’ perceptions of OBA services. Data were then subjected to exploratory factor analysis (EFA) of the 23 items to break down the items into homogonous sub-scales coherent with the quality dimentions as proposed by Donabedian [20]. Principal component analysis with orthogonal varimax rotation was conducted. In addition, the Kaiser-Meyer-Olkin measure (KMO) was done to evaluate the suffiency of data for EFA and Bartlett’s test of sphericity to evaluate the degree of patterned relationship between the items. Additionaly, reliability analysis was performed to test the reliability of the scale and internal consistencies of extracted factors; whereby Cronbach’s alpha coefficient was calculated. The multivariate response model was used to study whether level of education, ante-natal clinic visit, marital status, age, and County of residence were predictors of the factors related to perceived quality of reproductive care (Table 1). The questions on overal satisfaction and reasons for satisfaction were analysed using Microsoft excel 2010 and Pareto chart [28] was obtained for the level of satisfaction. Definition and measurement of variables used in multi linear regression model The authorization to carry out the study was obtained from the Ministry of Health-Kenya as part of routine monitoring of the process (Development of the Health Sector, Health Financing Support and Output Based Approach, Phase III, BMZ-No. KENYA 2010 65853) of the OBA services. The proposal was approved by the health research unit of the Ministry of Health Kenya (MOH/HRD/1/ (32)). Additionally, permission was obtained from the county headquarters and hospital administrators to proceed with the study. Verbal informed consent for the study was obtained from every woman who agreed to participate. The interviewers explained the purpose of the study to the mothers in their local dialect (language) and asked them whether they were willing to participate. For those who agreed, the interviewer indicated a unique patient identifier and the date of the interview on the front page of the questionnaire before proceeding with the interview and data were only used for the study.