The push to standardize digital health

In the past, community health workers in low-and-middle-income countries would go door to door with a paper notebook, gathering health information from people in their community, such as whether a child in the household has been immunized, tally that information, and send it to the government. And while there’s been a large push to digitize health care systems, in many circumstances digitizing health care has created new problems — different donors and organizations come in with a host of mobile applications and different systems for collecting information. This can be confusing and burdensome for health workers who need to juggle between applications and paper records aimed at addressing different issues. A leading discussion at the Global Digital Health Forum last week in Nairobi was the need to make digital health more integrated and interoperable, and to work toward everyone using the same standards and guidelines. The conference was held for the first time in Africa — it’s usually hosted in Washington, D.C. Digital health, especially in Africa, has shown an important impact in some ways, but it’s also failed collectively to transform patient outcomes, Matt Berg, chief executive officer at the Kenya-based technology company Ona, told Devex. “We haven't really achieved its full potential — a lot of this is because of the fragmentation of services,” he said. Vertical programming and disjointed data Health care in low- and middle-income countries is often siloed — with diseases, such as HIV and malaria, tackled in separate programs. Countries have also faced “pilotitis” in the digital health space where many short-lived pilot programs enter into a space and stop when funding runs out. African governments often haven’t been sensitized on how to use data that arises from these programs, said Jean-Philbert Nsengimana, the chief digital adviser at the Africa Centres for Disease Control and Prevention, or Africa CDC. “The fragmentation of the data space results in missed opportunities for better policy insights, better operational insights. It's a problem across the board — across the whole continent,” Nsengimana said. As countries move on their digital transformation journeys, there have been similar issues around duplication that have been seen in paper registries, Natschja Ratanaprayul, a technical officer in the public digital health technologies unit at the World Health Organization said. “Instead of having multiple paper registers — which are still there by the way — now we have multiple applications, multiple devices, with multiple data entry points,” Ratanaprayul said. “When you actually see a health facility with a whole room of health records in paper format, plus all these computers and applications and mobile tablets that don't always work very well … that seems like something is wrong here,” she added. For example, there are places where immunization data from community health workers aren’t connected digitally with facility-based immunizations, Ona’s Berg said. “​​We're going to really address under immunization or zero dose kids … we have to link those two together,” he said. And software developers often define the same things differently. For example, one might use ‘M’ and ‘F’ to determine males and females, whereas, another could use ‘W’ and ‘M’ for women and men. When combined, the data gives an inaccurate representation. ‘Africa on FHIR’ Africa CDC’s Nsengimana told Devex that standardization is key to tackling siloed health care and “piloitis” that’s plagued Africa. Fast Healthcare Interoperability Resources, or FHIR (pronounced FYE-er) is an open-source standard used around the world for exchanging health care information electronically. “We are going to put Africa on FHIR,” Nsengimana told the audience in Nairobi. The European Union, Africa CDC, and WHO have all endorsed FHIR as a standard, Ona’s Berg said. “We all agree on how to speak a common language which is key,” he said. FHIR provides a common model for representing health data. When digital systems use the FHIR standard, it’s like they are speaking the same language, which makes it easier to seamlessly share information, experts said. And WHO is working to make FHIR easier to use. One of WHO’s primary roles in the global health ecosystem is to create guidance, or evidence-based recommendations for health interventions, such as those around managing tuberculosis or diabetes. WHO is making it easier for countries around the world to implement their guidance by producing SMART guidelines, which leverage FHIR to help represent WHO recommendations in a machine-readable format. For example, a software developer with no medical experience might be tasked with creating a digital health application. This would include translating WHO’s guidance for health providers into the way that the application functions. But because this person is not a medical professional, there is room for error. The SMART guidelines aim to fix that by creating a machine-executable intermediary layer so that guideline authors and software implementers have a common interface for collaboration. This includes work to help translate WHO’s health guidelines for safe, effective care into software. These guidelines will also be helpful when clinical recommendations change, WHO’s Ratanaprayul said. It can become expensive and complex to update digital health programs, either because a vendor increases prices, or there are flaws in the data model that prevent updates — this aims to solve these problems, she said. The U.S. Agency for International Development, WHO, UNICEF, the Endless Network, and PATH are partnering around creating a SMART Guideline for Community Health, this includes creating a digital adaptation kit. This work has just begun and was launched at the forum in Nairobi. They are also inviting interested countries, organizations, and partners to be involved in the co-creation of this kit. And there are other tools and guidance documents WHO’s published or will publish, Ratanaprayul said. These efforts are all aimed at improving interoperability — when a patient seeks care, the provider has a full view of the patient's data needed, experts said. The data flows from the community level, to health facilities, to tertiary hospitals, and to the government. For example, a doctor tells a patient they need lab tests to confirm a diagnosis. The lab wants to send the results to the doctor. They could mail the results, but that’s slow. They could make a phone call, but not everyone has time to sit by the phones making these calls, and that method could be error-prone. They could email the test results to the doctor's office, but that doesn't appropriately protect sensitive health data. The most efficient way for the lab to share that information with the doctor is through their digital system. Ideally, those lab results would be shared with a patient’s electronic medical record, so that they're part of their file. However, if the laboratory information system doesn’t speak the same language as the electronic medical record, the doctor might not get the right information. If the digital system that the laboratory uses does not speak the same language as the doctor’s electronic medical records system, the doctor would not receive results as quickly and could struggle to interpret them correctly. Building consensus But it's not just a flip of a switch — work must be done to get everybody using the same standards. “Standardization, to many people, to many policymakers, is still a very technical term. They’re not able to grasp why standardization is a need of the day,” said Mridul Chowdhury, chief executive officer of Bangladeshi company mPower Social Enterprises Limited. He said there’s not enough evidence around its benefits and there’s a need to build up technical expertise. Countries are at different levels of digital maturity — some countries have not digitalized their community health workforce at all. If they seek to equip community health workers with digital tools, they can do so leveraging FHIR, which will help support continuity of care for patients across geographies and different levels of the health care system, experts told Devex. For countries that already have a digital tool supporting those delivering care at the community level, they can choose to adapt existing tools to leverage the FHIR standard directly, or they can use a “facade” that does not change their existing application, but rather acts as a translator between the original application and other systems that use FHIR. For example, two people speak different languages but talk to each other through a translator. Communication takes more effort and time, but with a good translator, they'll understand each other. That’s similar to what a FHIR facade does. Kenya-based tech firm Ona is helping countries build expertise around FHIR, Berg said. For example, they’re working with Uganda’s government to run an application for community health workers using FHIR, with UNICEF funding. “Countries kind of know what FHIR is. They want to use it but it's still kind of a little abstract,” Berg said. “We're making it easier for them to learn about FHIR.” Moving toward FHIR and SMART guidelines adoption can involve high upfront costs in areas such as infrastructure and moving existing data in this direction, Chowdhury said. This is especially true for countries with existing large digital health systems. But other experts said that with the scale of adoption of this standard, when countries are trying to solve a particular problem involving FHIR, it's likely it's already been tackled by others, or other groups might be facing the same issue. This lowers the cost of adoption or gives countries and technologists the opportunity to join forces to tackle problems that would not have been easy to deal with had FHIR not existed. And one of the challenges is donors often fund health care in vertical ways and with specific units within a government — such as a unit managing tuberculosis, rather than the central government, Chowdhury said. These individual units don’t have enough of an incentive to standardize in order to link up with other health units, he said. “In reality, what is often happening is that these donors, and even WHO, are possibly becoming part of the problem that they're trying to solve. It's a very ironic situation,” Chowdhury said. Many countries don’t have a clear national digital health strategy. Moving forward, digital transformation must be government-led so that they lead coordination — which hasn’t always been the case, Nsengimana said. A government can also legislate in a way that it becomes mandatory to use certain standards if a donor, company, or organizations want to access public funds and procurement systems, he said. Development partners have a huge influence and have led investments in digitalizing health in Africa over the past decade, Nsengimana said. But there’s a need moving forward for them to align on health tech investments that are open source and standards-based. Africa CDC launched the Africa HealthTech Marketplace last month — where they’re vetting health tech innovations with standard-based solutions and encouraging governments and partners to invest in them, Nsengimana said.