The Avoidable Waste in Medical Research, and the potential from automation of systematic reviews – University of Copenhagen

Global Health > Open events > Archive 2016 > The Avoidable Waste in...

The Avoidable Waste in Medical Research, and the potential from automation of systematic reviews

By Professor Paul Glasziou, Faculty of Health Sciences and Medicine, Bond University, Australia

The seminar will be given Wednesday, June 15, 2016 at 3:15 PM, room 5.2.46 at "Gamle Kommunehospital", Øster Farimagsgade 5, 1353 København K. Tea and coffee will be served at 2:45 PM in the library at the Department of Biostatistics (room 5.2.46).


Without accessible and usable reports, research cannot help patients and their clinicians. In a published personal view, a medical researcher with myeloma reflected on the way that the results of four randomised trials relevant to him had still not been published, years after preliminary findings had been presented in meeting abstracts: 'Research results should be easily accessible to people who need to make decisions about their own health…" His experience is not unusual: a recently updated systematic review of 79 follow-up studies of research reported in abstracts estimated the rate of publication of full reports after 9 years to be only 53 per cent.

The world currently invests over US$100 billion every year in supporting biomedical research, and this results in an estimated 1 million research publications annually. This investment in research is laudable. But regardless of who sponsors research, something needs to be done to protect this investment to improve health from the avoidable waste in producing and reporting research. This talk examines the causes and degree of waste occurring at four successive stages: the choice of research questions; the quality of research design and methods; the adequacy of publication practices; and the quality of reports of research. The estimated losses are roughly 50% loss at stages 2, 3, and 4 implying that the dividends from tens of billions of dollars of investment in research are lost each year because of correctable problems. Some of the problems are readily correctable for a minimal investment.

A key source of waste is failure to use existing research to help design new research, but such reviews are inhibited by the current cost in time and effort. Information systems designed to support or automate individual tasks in the systematic review, and in particular systematic reviews of randomized controlled clinical trials, have considerable potential to greatly reduce these costs. Software already exists for many of the steps, but is currently not integrate. The optimized workflow will lead to system in which each systematic review is described as a computer program that automatically retrieves relevant trials, appraises them, extracts and synthesizes data, evaluates the risk of bias.