Effectiveness of behavior change intervention in enhancing utilization of health data for decision making by community members in Nyando sub-county, Kenya

Abstract

Utilization of health data is key because it enables individuals and communities to make decisions on their health seeking behaviour. However, studies show low utilization of health data for this purpose with data generated in communities under-utilized especially in the Low- and Middle-Income Countries (LMICs). This can be attributed to the ineffective methods of providing feedback to communities resulting in poor health problem identification. In Kenya, majority of health programs provide feedback on health data to communities through conventional methods such as health talks in health facilities, use of mass media, posters and billboards. Despite these, less than 38% of health data is analyzed and used for decision making by communities. This calls for a paradigm shift in the way health data is communicated to communities. This study therefore investigated the effectiveness of a Behavior Change Intervention on enhancing use of health data for decision making among community members in Nyando Sub- County, Kenya. Specifically, the study sought to: establish the health data needs of community members; examine the factors influencing utilization of health data for decision making; determine trends in utilization of health data for decision making during the intervention; and establish the effectiveness of the Behavior Change Intervention in improving use of health data for decision making. This was a longitudinal interventional (pre-post) study that adopted both quantitative and qualitative approaches to data management over a period of 12 months. The study was implemented in five phases namely: Phase 1: Baseline study to identify community health data needs and factors influencing utilization; Phase 2: Curriculum development and training of Community Health Promoters (CHPs) on how to sensitize households in utilization of health data for decision making; Phase 3: Implementation of the Behaviour Change Intervention in households and community forums through dialogue; Phase 4: Longitudinal Monitoring; Phase 5: End-line survey. A total of 440 participants were sampled using Taro Yamane’s formula (1967). Quantitative data for the baseline and end-line surveys was collected using semi-structured questionnaires while qualitative data was collected through Focus Group Discussions and Key Informants Interviews. Quantitative data was analyzed using SPSS version 25 and R, while qualitative data was thematically summarized and analyzed using the NVivo application. Chí-square test was used to determine association between categorical independent and the dependent variables while Cochran’s Q test determined statistical significance of the differences in utilization of health data for decision making at the different observation points during the intervention.McNemar’s test was used to confirm statistical significance of the differences in utilization of health data for decision making between the baseline and end-line. At baseline, 50.2% of the study participants needed health data on HIV/AIDS, 44.5% needed data on prevention of malaria, 52.95% on TB prevention, 34.55% needed data on child immunization, while 54.3% needed data on hygiene and sanitation. Utilization of specific health data for decision making at baseline showed thatuse of prevention of malaria data was at 187(42.5%), TB prevention management was at188 (42.7%), HIV/AIDS prevention was at 210(47.8%), ANC was at 123(28%), Deworming was at 146(33.2%), Child Immunization was at 156(35.5%) and hygiene and sanitation was at 117 (26.6%). Findings from the qualitative survey resonated with these results. Key informants and Focus Group discussants pointed out specific reasons for use and non-use of the health data. The main factors that influenced utilization of health data for decision making were; Education Level, for HIV data use (P=0.01, OR=2.5); Age,for malaria data use (p=0.07, OR=2.05);Education Level, for TB management data use (P=0.00, OR=2.3); Religion, for ANC data use (P=0.02, OR 2.2);and Gender, for child immunization data use (p=0.03, OR=1.7).The study revealed an increased trend in utilization of health data at 4 months(n=235, 53.5%), at 8months (n=282, 64.6%) and at 12months (n=378, 85.9%). Cochran Q test revealed a statistical significance in the increased trends with a cumulative P-value of 0.025.McNemers test revealed statistical significance in the changes in utilization of health data recorded at the baseline and end-line phases of the study (P=0.019) hence confirming that the Behaviour Change Intervention is effective in enhancing utilization of health data for decision making by communities. These findings will therefore inform national and county level efforts to enhance utilization of health data for decision making among communities in order to improve their health seeking behaviour.