Improving facility-level data synthesis and reporting

As responsibilities for health planning and service delivery increasingly devolve to lower levels, it is imperative to get timely and accurate data in the right place at the right time to empower frontline health workers to inform critical program decisions. While dashboards and visualization technology can help facilitate the analysis and availability of routine health data, national health management information systems typically aggregate and display data at the district level and above. It can be a challenge for health workers at lower levels to regularly access information products that synthesize HMIS data. Furthermore, existing information products are not always designed to facilitate decisions at the facility level on individual patient treatment and management, resource allocation, and program performance. Under the CDC Kenya HMIS-II project, Palladium developed dashboards for electronic medical records systems giving HIV providers a snapshot of their clinic performance and to support decision making on the quality of HIV services at the facility. Participants will learn about the benefits and challenges designing, tailoring and implementing visualization solutions that meet the needs of frontline data users.

Globally, health programs are increasingly replacing manual systems with open source Electronic Medical Record (EMR) Systems. EMRs can provide efficiencies in documentation and retrieval of clinical records for individual patient care. This in turn translates to faster and better quality of care. Additionally, EMRs can simplify aggregation of patient records into routine reports, which health facilities are required to submit to a national Monitoring and Evaluation (M&E) system. In Kenya, KenyaEMR, built on OpenMRS, is the preferred EMR. While the Entity–Attribute–Value (EAV) data model used by the OpenMRS platform is great for clinical care, it is not optimised for retrieval of aggregate reports and custom reports. This presents a challenge to users who need to generate aggregate and custom reports, especially health facilities that handle data for large numbers of patients. The process for generating these reports can be time consuming and frustrating. OpenMRS developers agree that flattening the EAV database can increase the speed of running reports from the EMRs, and literature describes how to manually flatten the OpenMRS database to gain these efficiencies. However, the CDC-funded Kenya HMIS-II project has gone a step further and automated this process. The innovative process automates the Extraction Transformation and Loading (ETL) of data from the EAV database to flat tables that can be used for faster aggregate and custom reporting. This success resulted in greatly increased reporting for the 325 facilities in Kenya that use OpenMRS. This session will highlight the steps that were taken to increase reporting efficiencies in OpenMRS and how this increased uptake and use of the EMR and provided time-savings for frontline heath workers during routine reporting. This approach has been shown to be easily scalable and re-usable by various implementers of OpenMRS in other countries and is available for adoption by developers.

Mortality and cause of death data are an important aspect of national health information. Countries need to know how many people die each year, in order to design effective public health policies (Mahapatra, 2007). In Kenya, the Civil Registration Services (CRS) registers vital statistics. Although there are steps to automate the CRS data collection system, a manual system is still in use. Some gaps include aggregated reporting and use of outdated data tools that do not capture all the causes of death as recommended by WHO. The Ministry of Health (MOH) complements the work of the CRS in reporting vital events. Health facility reported deaths are medically certified per the WHO ICD-10 standards for statistical comparability. The Kenya Health Information System (KHIS) tracker module to capture this individual level data. There are gaps in the module, it addresses many of the data challenges in CRS and can serve as an essential data source for real-time mortality data in Kenya. The project provided technical assistance to departments of health in targeted health facilities. The County ICD-10 resource persons conducted: ICD 10 trainings for certifiers and coders; structured mentorship on data quality and KHIS reporting at health facility level; advocacy to leadership and manager’s and mortality data reviews to create demand for facility mortality data. Over 50% of the targeted health facilities report cause of death data in KHIS, which has increased by 216%, from 442 events (2017) to 1398 events (Dec 2018). Outdated ICT infrastructure, poor internet connectivity and health worker attitude have hindered 100% uptake in all targeted facilities. To institutionalise KHIS event reporting and use, a government led process, regular health worker mentorship, mortality data reviews in hospitals, investment in ICT infrastructure, leadership and stakeholders’ support are pertinent.