Risks, Benefits of Using AI to Reach Universal Health Coverage

Machine Learning (ML) and Artificial Intelligence (AI) can boost the achievement of universal health coverage, according to a paper from the WHO. However, the use of these technologies has to be guided by the public interest instead of the specific interests of stakeholders and insurers. 

AI and ML can be leveraged for multiple purposes related to healthcare. “The use of the data obtained by these machines in healthcare can help in decision-making that positively impacts the diagnosis and more personalized treatment of patients,” wrote Eduardo Medeiros, Co-Founder and CEO, Welbe, on MBN.

Through computers and machines, AI mimics the problem-solving and decision-making capabilities of the human mind, according to IBM. Meanwhile, ML is a type of AI based on the use of statistical and mathematical modeling techniques to define and analyze data, according to WHO. “The main added value of ML lies in its enhanced speed and precision or accuracy compared with traditional statistical methods, also due to the fact that ML can be more easily applied to large volumes of data,” reads the paper. 

However, the use of these technologies on health financing has not been fully explored. AI and ML can be leveraged for the prediction of health expenditure, risk scoring, claims management and fraud detection, among other fields. The technologies can have both positive and negative implications in the achievement of universal health coverage. While they can put vulnerable groups in the spotlight, they can also lead to actions that hurt these groups. Although the use of these technologies can help to generate valuable insights and improve service delivery, the generated information can equally be used to increase insurance contributions, reduce distribution and result in the exclusion of vulnerable groups from health coverage. Also, the equitable distribution of resources can be impacted. “If ML enhances risk selection and cost prediction in private commercial insurance schemes, it may contribute to further fragmentation, reduced or inequitable financial protection and inequitable financing,” reads the paper. 

The benefits of AI can be overshadowed by poor algorithms that create biases. For this reason, transparency is essential as algorithms tend to be biased. Regulation is also needed to decrease this risk. “Data privacy and ethics are trending and everyone in the industry should align to these principles. Algorithms must represent a competitive advantage; they should not be biased and results must be impartial,” said Jorge Huerta, Chief Business Development Officer, X-Data to MBN.