Using Open Data Kit (ODK) for Rapid Data Collection, Transport, Visualization and Management of Malaria Data in Siaya County, Kenya
Project Name: Using Open Data Kit (ODK) for Rapid Data Collection, Transport, Visualization and Management of Malaria Data in Siaya County, Kenya
Project Status: Proposed
Point of Contact: Aaron Samuels
Keywords: Data collection, data transportation, data security, real-time, encryption, data access.
Project Description: Malaria burden in Africa is extremely high despite the existence of several interventions including use of insecticide treated nets (ITNs), indoor residual spraying (IRS), community based case management, larviciding, and, provision of sufficient testing and treatment kits in the health facilities, all of which are aimed at curbing mortality and morbidity. Together with the implementation of the various malaria control & elimination strategies, early detection and containment of epidemics constitute the basic elements of malaria control. Due to lack of easy access to care at health facilities, community case management of malaria is being rolled out whereby community health workers (CHWs) help in malaria case detection and management at the community level. In order to monitor the success of these interventions and make evidence-based decisions, malaria control programs in endemic areas have invested heavily to improve on the surveillance tools used to track trends. However, accuracy, quality, cost and timeliness remain a challenge. Therefore, innovative tools for real-time, accurate and efficient data collection are urgently needed. Relay of accurate and timely data from these community units (CUs) is vital in informing policy that will enable proper targeting of interventions. Surveillance data for such interventions is currently collected using a variety of tools including Personal Digital Assistants (PDAs), netbooks, scan-able forms, paper questionnaires and a few mobile phones then physically transported to the data centers for cleaning and analysis. Access to this data is limited to study staff and not very efficient in providing real-time evidence to help in the evaluation of the integrated malaria control interventions.
Malaria indicator surveys conducted by our teams take long before the data can be ready for analysis. The delay is attributed to the number of errors made in the process of data collection and the time taken to clean and validate the data. The devices used in the data collection process also contribute heavily to this, from complications in using devices such as PDAs to the personnel collecting the data. This therefore translates to late action on the indicators shown by the data. We propose a surveillance system that takes advantage of the rapidly growing technological advancements and appreciates the limited resources we have to help obtain data rapidly while ensuring the quality, integrity and cost implications. We intend to bridge this gap by innovatively using an open source application called Open Data Kit (ODK) and extending its capabilities to address the challenges facing data collection, availability and transmission. ODK is an application with very extensive capabilities in terms of data collection and management. The ease of using it makes it the most appropriate application for rural settings with a focus on CHWs who have very basic education. ODK is capable of collecting text, numerals, images and geographical coordinates in one single questionnaire.
Methodology. The study will be conducted in Siaya County, western Kenya. A two pronged approach will be used, facility and community based. Ten health facilities will be randomly selected and 25 CHWs recruited from the villages around the catchment area of the facilities, within the KEMRI/CDC HDSS study area. Malaria surveillance data will be collected using two methods, smartphone/tablet based questionnaire developed using ODK and PDA based questionnaire. Data will be collected simultaneously using the two methods for comparison. Simple smartphones will be used by CHWs while health facility recorders based at the facility will use tablets. Data collected using both the smartphones and tablets will be securely transmitted using 2G/3G/4G internet in real-time via a Virtual Private Network (VPN) to ODK Aggregate servers that will be set up at the KEMRI/CDC data center while data collected using PDAs will be downloaded to computers and physically transported as has been the standard practice. Data collected at household level will include household characteristics, demographic data, malaria diagnosis results, insecticide-treated net (ITNs) ownership and use, health facility utilization, treatment information and geo-mapping data. At the health facility, data on malaria incidence cases by age groups, antimalarial drugs stocks and stock counts of malaria test kit will be captured. Questionnaires will be rapidly developed and remotely deployed onto the tablets and smartphones. PDA based questionnaires will be manually uploaded. Time taken for the programming, setting up, collection and data transmission will be recorded and compared with the time taken for the PDA data collection system. The malaria treatment algorithm will be programmed into both systems and the prescribed treatment will be saved as part of the data. The server will be configured with a Secure Service Layer (SSL) certificate to ensure the data containing personal identifiers are encrypted while on transit and that only certified devices have access to the data. A customized web-based dashboard with extensive features than what ODK Aggregate provides will be developed to display summary and descriptive statistics on the data and also display geographical positions of the data points on Google Earth to help users of the system to get snapshots on the state of the data at particular points in time. Users will be provided with usernames and passwords to be able to access the system from wherever they are. Once the data has been cleaned, it will be uploaded back to the server to allow for access by other partners. This system would be scalable to the various KEMR/CDC disease surveillance activities that involve treatment procedures, require timely monitoring and carry very sensitive participant information. It would also be linked to the Ministry of Health’s District Health Information System (DHIS) for access by county health administration team. To assess the efficiency of the proposed system, research data collected using the two methods will be compared.
Comparative analysis will be done for time taken to setup each system, time taken for collection, time taken to get results, technical challenges experienced, ease of use, time taken for data transportation, precision of geo-mapping data, availability of the data and costs related to implementation.
The project will be considered successful if the results obtained from this study demonstrate that adoption of new technologies would improve ease of data collection, management, transmission, usage and reporting in a timely, efficient, secure and cost effective manner by using smartphones in disease surveillance activities. This novel technology has the potential to reduce data collection costs, availability, improve quality in limited resource settings. There are opportunities of scaling up this tool outside the study area and adoption by the ministry of health.