Health sentries in Ghana

Health data is a priced commodity in recent times. The COVID-19 pandemic demonstrated the need for governments and policymakers to quickly ascertain data, analyze and act.

The DHIMS software is used by many LMICs, including Ghana, to aggregate and report on important health indices. However, this data comes in monthly and often belatedly, sometimes leading to fatal consequences (https://www.graphic.com.gh/new...). However, there are mechanisms to report "diseases of interest" rather faster than the routine. Again, DHIMS data is opaque and only accessible to healthcare professionals, mostly within the Ministry of Health. DHIMS also does not have a public API where third-party organizations and private individuals can access the vast data that exists and provide other insights other than the one provided by DHIMS. This development in our opinion has led to the stagnation of insights into one of the most important health data in most countries.

Again, in many LIMCs, the concept of One Health is often not followed, and health data is often collected in isolation, ignoring data from the agricultural sector, especially animal health that can impact.   

We intend to undertake a project where we recruit volunteers who would collect community and facility-level data. The community volunteers would collect and submit data daily via a REDCap. REDCap is an internationally recognized tool for data collection, including offline data. it is also suitable for longitudinal data collection, which we intend to undertake. The data to be collected would include a wide area of areas important to health such as:

• Number of malaria and diarrhea cases in children
• Ante-natal visits
• Antimicrobial prescriptions and purchases
• Immunization activities
• Household expenditure on health
• Absenteeism
• Domestic animals morbidity and mortality
• etc.

The volunteers will collect and sync their data daily. Most parts of Ghana have a 3G internet connection which is more than enough in these scenarios. The REDCap infrastructure is more than capable of handling such loads.

A dashboard would be developed to support the REDCap application. The dashboard application would communicate directly with the underlying database to provide better insights into the data. The information would be public to anyone without restrictions (unless prohibited by ethics or legal constraints).

Through the REDCap API, other organizations, including academia can access the data through the impressive API already available on REDCap. National Health Authorities (MOH, NIA, GHS, etc.) would all be given direct access to the data and the dashboard. Health informaticians at the ministries can then further develop their own use of the data.

Our project would decentralize health data with the purpose of improving health analytics in many countries. We hope our initiative will lead the drive in the use of rapid health data in solving community-level issues by academia and even private developers. Private developers would be required to use the data only for open-source projects which should lead to additional use of the data.

The technologies to be used would be

• Virtual Server purchased from a cloud provider (Azure, Google, digital ocean
• REDCap  as the front-end application
• R (Statistical software) to develop dashboard (shiny) application and API (plumber)
• ASP.Net WebAPI to provide additional API support

Our solution serves different people differently.

The first beneficiary is the communities that would be partaking. They would have access to aggregated health data which they can use for decision-making.

The Ministry of Health and other ministries and agencies should also benefit from the project. We intend to cover a shortfall in the current system where there is not much communication between agencies.

Health-related software developers can also tap into the project and provide additional insights that are currently not available.

We are a team made up of young health experts with considerable experience in Primary Health Care in Ghana at remote areas of Ghana. We have all worked with the Ghana Health Service and understand the health system. Through our experience, we came to appreciate the shortcomings of our existing infrastructure and the need for a complementary open-sourced approach to aggregating health data across the country.

We are fairly conversant with open-source technologies and the impact they would have on our project to the benefit of our communities and the larger population.

The project would be community-led starting with selecting volunteers from participating communities. We hope to implement a program that would be self-sustaining even after the initial funding through the community-led approach.

The project is an ambitious one and we are applying specifically for 2 reasons

• Such a project with an international backing is more likely to receive the necessary ethical and political buyout from authorities that would need to provide their assent before the project kicks off
• The project also requires initial capital, one that we hope to secure.

The solution leverages existing technologies to provide more insights into PHC data. The solution is open-ended in the sense that, additional tools can be added rather quickly than the existing infrastructure provides. It is an open-source first solution that should ensure the participation of additional software developers to extend the project to provide better insights.

Local health authorities and the local government would have direct access to data that is affecting their communities in real-time so that they can make better decisions.

By next year, we hope the project would have received mass acceptance and adoption by the health and political authorities in the country.

In the long term, we hope to extend the project to other LMICs. A team member from Sierra-Leone is would be expected to enroll a similar project there

We expect a reduction in negative health indices, e.g. malaria and diarrhea cases among children under 5, etc. Our solution should provide the platform and data for other organizations to leverage their interventions.

Our team should also be able to identify actionable items from the data that would be sourced.

We expect academia to tap into the data to provide better insights through publications from the study.

Actionable and directly sourced data that is openly accessible should lead to changes in policy. Our project can lead to health and political authorities designing tailor-made solutions at the community level rather than the conventional one-fit-all. Communities with surges in diarrhea should have immediate intervention.

Our solution relies on the Web and internet infrastructure to serve as the backend.

We would first acquire a virtual server from a cloud provider (Most likely Azure, as we are more familiar with it). The server would be running the Linux OS, 1 to minimize cost and as a requirement for the dashboard application to run. The R shiny server on runs on Linux.

We would be relying heavily on REDCap to provide the interface to capture data. REDCap is in active development and provides means to sync offline data when there is an internet connection.

We would rely on the open-source R statistical software to design dashboards and APIs through the shiny and plumber packages.

Additional API support may be provided through the open-source ASP.Net.

We currently rely on extracting data from DHIMS through a "hacked" API. We have an in-house application that is able to extract data from DHIMS. But the process is a bit tedious and requires a lot of manual workflows. It also has to be done per element, i.e, each subject area

Our team has respect for everybody. We are made up of people of diverse backgrounds, cultures, and religions. Our tolerance for everybody is what keeps us still.

We are an open-source first team. We believe in democratizing data.

If the project is executed successfully, we hope to get a buy-in from governments to continue the solution. We are mindful however of the potential for conflict of interest and political interference at that stage.

We hope the project would be strictly academic with membership mainly from academia. A governing body would be sourced to ensure the continued independence of the solution.

Another continued funding under discussion is getting people to volunteer information periodically through our APIs. That way, the financial requirements should be minimal.