A group of researchers have advocated three theories that focus for data-driven decision-making in healthcare.
According to a study published on QScience.com, ‘Data-driven Decision-making: A Review of Theories and Practices in Healthcare’, these theories can help form well-judged decisions in healthcare and make them applied in various contexts.
The authors of the article, Chloe Ile, South African Medical Research Council; Charlene Ile, Explore AI, Cape Town, South Africa and Christine Ile, Livingstone Hospital, Port Elizabeth, South Africa, categorise Classical Decision Theory; Ottawa Decision Support Framework and Bayesian Decision Theory as three methods to make use of data effectively, highlighting the benefits of such practices for better healthcare decisions.
“ Use of data for healthcare decision-making has numerous benefits, including increasing knowledge of user demographics and needs enabling adequate planning of healthcare resources and services, and providing a roadmap of decisions made to ensure stakeholder accountability,” state, the authors in the article. However, they note that despite these clear benefits, frameworks and theories guiding decision making in healthcare remain under-utilised and therefore they have presented these decision-making theories that focus on data.
Classical Decision Theory- the study of choices to be considered when making a decision- and its modern iterations emphasise the decision-making process and the use of data in this process. It is divided into descriptive and normative decision theories. Normative Decision Theory points to the prerequisites that should be in place to reach rational decision-making. Descriptive Decision Theory seeks to explain and predict how people make decisions. In both, three key elements are described: the decision, the decision-maker, and the decision-making process.
Ottawa Decision Support Framework was developed based on expectancy-value, decisional conflict, and social support theories. It was designed to address health decisions which needed to be made because of a new circumstance or diagnosis, requiring careful deliberation due to unknown benefits and risks and needing more effort during the deliberation and implementation phase. This theory is employed when the decision relates to new diagnoses or treatments or when extensive deliberation is needed in uncertain circumstances.
Bayesian Decision Theory considers existing knowledge and cost functions in decision-making. Uncertainty is inherent in the human experience and can be observed in almost every role, task, and function resulting in the implementation of solutions without a full understanding of the problem or decision. Therefore, it is a statistical decision model and aims to support decisions made in the face of doubt by shedding light on the uncertainties involved in the process.
The study concludes that the healthcare system is a data-rich environment capable of supporting data-driven decision-making at both the individual level for patients or clients and the system level for organisations or national entities. “Data acquisition, storage, and processing are becoming integral pillars within the health sector, especially as fields such as telemedicine, big data analytics, and artificial intelligence continue to grow in popularity. To handle these data volumes and support data interpretation, Big Data analytics must be used with decision theory to move from observation to an informed conclusion or treatment plan,” concludes the study.