keynote: Prof. dr. Darrel P. Francis

Darrel Francis Professor of Cardiology
Imperial College London

How to get positive results in cardiology: a liar’s toolkit

In cardiology we suffer from three problems. 1. The heart beats often, so there are multiple alternative values for each measurement we make. 2. Unlike hard sciences, in cardiology the researchers spend much of their life earning money from doing and reporting tests, so are not motivated to reveal the wide error bars. 3. Medical training encourages us to “use the full picture” of clinical information, which unfortunately teaches us bad habits in research. In this lecture I will describe how this has led to disasters in cardiological research, and what you can do to save yourselves from these traps, as researchers and even as readers of published articles.

Short biosketch:
Darrel Francis is an academic cardiologist specialising in quantitative techniques in clinical cardiovascular science. Having qualified in Cambridge University and trained in Oxford, he carried out his thesis in Cardiology at the National Heart and Lung Institute. He is now Professor of Cardiology at Imperial College London.

His research interests include application of mathematics, physics and engineering approaches to practical problems in cardiology, including imaging, heart failure, valve disease and coronary disease.

Recently he has been criticised for calculating that future blinded trials of renal denervation would show a much lower effect size than widely expected: the subsequent trials showed it to be even lower than his predictions.

In another notorious case his paper, indicating how research misconduct can harm large numbers of patients via clinical guidelines, was retracted from the European Heart Journal. It was later published, along with the lessons to learn from the missed opportunities for early action to protect patients, in the BMJ.

He is currently under attack for observing that the more numerical inconsistencies there are in a bone marrow stem cell trial, the more effective the therapy. Paradoxically, the attacks (1, 2) are from the people whose research Francis’s analysis indicates to be the most reliable.

← Previous: Dr. Mark K. Friedberg   Next: Prof. dr. Michiel R. van den Broeke →