By Mukul Bhola, Deputy-CEO, The Defeat-NCD Partnership
Around the world, rapid demographic, sociocultural, and economic transitions have driven a surge in non-communicable diseases (NCDs). Increased affluence has been a double-edged sword, bringing people up out of poverty while predisposing them to NCDs like diabetes, cardiovascular disease, chronic lung diseases, and cancers. Rates of these illnesses are skyrocketing, with a staggering 71 percent of global deaths due to NCDs.
There is no magic bullet to solving the challenge of NCDs — but expanding affordable access to healthcare by investing in technology is a crucial first step. By developing and rolling out the use of digital tools, The Defeat-NCD Partnership is bringing innovative technology to each of our four key pillars — national capacity building, scaling up community health, access to affordable essential medicines and supplies, and financing.
Digitization improves service quality
To bring NCD services directly to more people who need them, we are introducing entrepreneurial approaches to earlier screening, self-care interventions, community mobilisation, education, and financing. By using smart digital tools and data, we can upgrade service quality to achieve better health outcomes. For example, machine learning algorithms could be very useful for spotting early warning signs with more accuracy than a human doctor in the field.
There are also opportunities for using machine learning in our work on innovative financing mechanisms that can bring additional monies into the NCD ecosystem. With regard to social impact bonds and multi-country bonds, we could use predictive analytics to look at existing bonds and determine the characteristics that have generated the greatest success.
We could then attempt to leverage that information in our partner countries’ NCD fundraising efforts. By pooling the risk for several countries and ensuring these bonds are de-risked by guarantees from respected institutions such as sovereign funds and multilateral bodies, we could, as an added bonus, bring reasonable returns to investors.
Technology for increased transparency
There is also a technology-building component to our work on increasing the affordability and accessibility of essential NCD supplies. We are doing this by creating a Defeat-NCD Marketplace, a platform where buyers and sellers can get together to trade. The Marketplace provides this space as an easy-to-use online facility specifically designed to reduce barriers and costs. This will ensure quality-assured medicines, diagnostics, and associated equipment get to where they are needed, when they are needed.
Why is a transparent digital platform necessary? Every day, around the world, people suffer and die due to corruption in the health sector. Transparency International estimates that as much as 6 percent, or US 318 billion, is lost to corruption and error. In public medicine procurement alone, estimates of global spending losses due to corruption range from 10 to 25 percent. In the world of NCDs, these spending losses amount to more than just a sum—they amount to the loss of human lives.
Think of the Marketplace as a digital ecosystem with value added services, including an online platform for essential NCD supplies. Its key feature is increased purchasing power for formerly disadvantaged buyers. The way buyers, sellers and products are organised and the transparency of information are the most important factors for a successful Marketplace. Having all necessary data together on the platform will reduce frictions to keep costs as low as possible, with ultimate benefit to the patients.
As you can see, this type of pricing structure is nothing revolutionary with regard to technology. It will be based on a standard pricing model with supply-and-demand market dynamics. What is revolutionary however, is the availability of previously inaccessible yet crucial information on such a broad global scale.
In this case, machine learning opportunities exist much in the same way that major eCommerce companies use machine learning—to aggregate purchasing data for the suggestion of future purchases. Machine learning could also be used to determine how prices will change in the future, whether they will increase or decrease and within what timeframe and parameters. If you could forecast a rise in prices, that would suggest a reduction in supply. This could be instrumental in identifying potential NCD supplies crises before they happen, potentially saving lives.
Man vs machine
AI and machine learning are very a la mode at the moment. But let’s not forget that as helpful as predictive analytics can be in some instances, there are times when it should not be used. As tempting as it may be to use machine learning to determine prices, for example, prices based on supply-and-demand dynamics have a very clear definition and deviations from that necessarily create a bias towards a particular group. This could only happen with consensus between the parties. When misapplied, machine learning has a tendency to entrench existing biases because by definition it requires historical data to make predictions. It is important to know where to draw the line.
As helpful as technology is, it does not solve problems on its own — it is only as good as its implementation. By marrying human values with the specific technological solutions needed, The Defeat-NCD Partnership is designing scalable solutions for tackling NCDs.