How Health Systems Can Spot Risk Before a Crisis

Cancer claims 10 million lives each year, the global shortage of health workers runs into the millions, and in some parts of the world, a single mosquito bite can be fatal. These challenges may seem very different. Yet they share a common element in how many health systems operate: risk is often recognised only after it has become illness, outbreak or crisis. By then, a disease may have advanced, transmission of disease may be under way, or an already stretched workforce may be managing consequences that earlier action could have reduced. The opportunity is to move that point of intervention earlier. This is what anticipatory health means in practice: using signals that already exist, at the bedside, in communities, and in routine health data, to help people and health systems act sooner. It is not prediction for its own sake. The value lies in connecting early and everyday signals to a timely, practical response. Finding practical routes to earlier action That principle sits at the heart of the Future Health Impact Challenge: Building Anticipatory Health Systems through Population Sensing, launched by Future Health – A Global Initiative by Abu Dhabi in collaboration with MIT Solve. The Challenge received 393 submissions from 68 countries. ThinkMD, VectorCam and Huna emerged as its three award-winning solutions: ThinkMD received the grand prize, with VectorCam and Huna recognised as distinguished finalists. Future Health Impact Challenge: Trailer The Challenge was designed to identify practical solutions with evidence of impact and potential to scale. Each of the three selected teams demonstrated their approach to using technology to turn an early signal into decisions that support anticipatory actions a health system can take. Frontline care as a population signal ThinkMD starts with the interaction between a frontline health worker and a patient. Its offline mobile platform guides structured assessments aligned with clinical protocols, supporting providers where doctors, connectivity, and diagnostic tools may be scarce. Almost 10,000 frontline health workers across 885 facilities are now using it. Each consultation also creates structured data. Across facilities, that information can reveal changes in symptoms and service demand. In Zambia, analysis of around one million encounters identified symptom patterns several months before a confirmed cholera outbreak. In Kenya, deployment data showed improvements in identifying danger signs in infants and prescribing antibiotics correctly for childhood pneumonia. It is not prediction for its own sake. “The utilisation of AI in healthcare is an absolute game changer. Not only can we increase access for patients, but we can also increase access to information to make correct and appropriate diagnoses and give the right treatment to the right patient at the right time. Future Health has been a really critical platform to be able to accelerate our mission of moving from reactionary to anticipatory care,” said Dr Jackie Rabec of ThinkMD. The award has given ThinkMD support and visibility as it works to expand into new health systems and strengthen how frontline data becomes timely intelligence for decision-makers. That is one measure of the Challenge’s impact: helping proven solutions move towards sustained, wider adoption. Earlier signals from the environment and routine data VectorCam begins with the environment around a community. Using a low-cost smartphone imaging device and offline AI, community health workers can photograph and identify mosquitoes in the field. Each observation generates time- and location-based data that can help authorities see changing vector risk as the first signal of disease patterns, before a spread in local communities and populations. The value lies in connecting early and everyday signals to a timely, practical response. Field evaluations show data completeness rising from approximately 60% to more than 90%. With support from the Future Health Impact Cha