Last weekend, Lown Institute President Dr. Vikas Saini spoke at the Preventing Overdiagnosis Conference in Quebec. In his talk, Saini questioned the paradigm of disease and called for a new framework for identifying and treating medical conditions. Read the full description of his keynote below:
“The problems of overdiagnosis and overtreatment are gaining increasing recognition worldwide, but there is a fundamental limitation in our current paradigm.
Current debates around the proper diagnostic and treatment thresholds revolve around the Bayesian statistics of sensitivity, specificity and pre-test probabilities. While knowledge about the discriminative accuracy of diagnostic tests, and well as knowledge about rates of clinical efficacy of specific treatments may continuously improve as more studies are done, there will always be residual uncertainty about what we know, as well as uncertainty about the right care for the individual patient relative to the mass statistics of RCT’s. Are these so-called grey zones impossible to resolve? Perhaps not.
The starting point of overdiagnosis is defining what diagnosis is – the gnosology or naming of an illness. Historically, such definitions of illness were functional.
Before the advent of modern techniques of measurement and imaging, silent pulmonary embolus was not detected, and effectively, did not exist outside of an autopsy. The elucidation of disease mechanisms and the more recent radical improvements in anatomic and molecular laboratory examinations have created a situation where the impact on human life of conditions called illnesses is harder and harder to distinguish from normal variance. In these grey zones, many sterile debates occur.
What we need is a new science for defining and naming disease, based on the outcomes that matter: all-cause mortality and whatever functional status that makes life worth living. Seeking unity in the midst of such complexity is a daunting challenge, but not insurmountable. We have examples from the world of mathematics, complexity theory and network science to help us. Such an approach might illuminate scale-free networks that simplify our task of identifying and delivering the right care far more often than we currently do.”