Health data in practice: human-centred science

The availability, scale and depth of data collected in the course of health care, by or about patients - combined with data-driven approaches to its analysis – is creating a paradigm shift in health care and its delivery. Machine learning and other automated methods of analysis will only succeed in providing useful insights for health and care if we understand health data in practice: how data is actually generated, interpreted and used. Human-centred data science operates ‘at the intersection of human-computer interaction, computer-supported cooperative work, human computation, and statistical and computational techniques of data science’ while preserving ‘ the richness associated with traditional methods while utilising the power of large data sets.’ We adapt this concept to the 'health data in practice' doctoral programme with the goal of developing highly skilled future scientific leaders able to apply interdisciplinary perspectives to research and innovations in health data science for the benefit of patients, the public, health care systems and society. Our training programme will introduce students to a wider context for their science, enabling them to draw on a range of concepts, disciplines and sciences underpinning algorithmic designs, human-interactions, evaluation and decision-making, in real-world settings across diverse populations.