Chatbots can’t save broken health systems
Digital health technologies could revolutionise healthcare on the continent – but only if governments get the fundamentals right.
Charles Ebikeme
Chatbots now deliver healthcare across the continent. The coronavirus pandemic, when a surge in people’s need for health information coincided with mandated isolation, was a particular driver of chatbots using large language models and artificial intelligence to provide health education. WhatsApp chatbots in South Africa, Rwanda and Senegal provided reliable information and support for rapid Covid-19 testing. In Ghana, a free Covid-19 Telegram app chatbot tackled misinformation on the coronavirus.
In some cases, chatbots were already in use. In others, they have expanded beyond Covid-19 information. NurseBot in South Africa advises pregnant women and new mothers through WhatsApp. Mum’s Companion in Zimbabwe does the same through Facebook Messenger. In Kenya and Ghana the Nuru Facebook Messenger combines classified ads with information on agriculture, finances and health. Medbit in Nigeria provides medical consultations and referrals through WhatsApp.
Many other digital health initiatives are happening across the continent. African healthtech startups have attracted $550-million in the past three years, according to Salient Advisory, a healthcare consulting firm. Online pharmacies, electronic medical records and telehealth services are just some of the innovations garnering interest from investors.
But it’s important to assess digital health innovations beyond investor interest. After all, many debuted to large fanfare in glowing press releases and then quickly subsided. Babylon in Rwanda, ConnectMed in Kenya (2016-2019) and HelloDoctor in South Africa are some of the most notable failures.
When digital health technologies are heralded as potential game changers for a continent with stark health burdens and outcomes, the optimism often ignores the fact that stable public health infrastructure is a prerequisite for any success in health innovations. “We have to get the basics in place before we can work on the fancy stuff,” says Marlien Herselman, a chief researcher at the Council for Scientific and Industrial Research in South Africa.
The optimism “also misses the fact that healthcare is very conservative,” explains Ayomide Owoyemi, a researcher who has built several digital products for Nigeria and other African countries, and whose code for an AI-based health insurance recommendation system in Nigeria is freely available on GitHub. “Healthcare is expensive and does not follow the expected timeline for return on investment.”
The low chances for that “expected timeline for return on investment” makes rolling out investor-led innovations a precarious choice. In 2018, Babylon announced “the first-ever fully digital healthcare service in east Africa using artificial intelligence”. The service would include a chatbot “to take the power of a doctor’s brain and put it on a mobile phone for medical advice and triage”.
In 2022, Babylon reported an operating loss of $369.8-million, despite generating $1.1-billion in revenue. Months later, the company ceased operations in Rwanda, ending what was initially billed as a 10year partnership with the government.
In countries such as Finland and Estonia, success has come from local collaboration between stakeholders including patients, user communities, technology providers, payers, regulators and policymakers. This sort of configuration is rare in African countries.
Health programming has instead historically revolved around time-bound projects that deliver a set of agreed-upon activities. Donors define a project and time frame, identify an implementer to build and deploy a digital solution, and then expect the solution to live on.
Yet, few digital tools are sustained without support from the public system.
The African Development Bank estimates that African governments need $26-billion in annual capital expenditure to meet evolving health needs over the next decade. They currently spend $4.5-billion a year. More investment would help build the shared infrastructure on which digital health innovations can interoperate.
Like car makers need governments to build roads, digital health innovators need governments to build electronic health information systems, internet connectivity networks and so on. Yet, in Africa, only Mali scores highly in public sector investment for digital health, according to the Global Health Digital Monitor.
“When you want to solve a problem, you shouldn’t think of the tech solution first. Think about the fundamentals first,” says Owoyemi.