Observations Yankee Doodling

Screening for early detection of lung cancer

BMJ 2010; 341 doi: http://dx.doi.org/10.1136/bmj.c6544 (Published 17 November 2010) Cite this as: BMJ 2010;341:c6544

This article has a correction. Please see:

  1. Douglas Kamerow, chief scientist, RTI International, and associate editor, BMJ
  1. dkamerow{at}rti.org

At long last, something that makes a difference?

I give an annual lecture to first year medical students on what makes a good screening test. One of the hardest points to get across is that early detection does not necessarily lead to improved outcomes. Why isn’t it the case, they ask, that finding cancer early is not always better than finding it later? It’s so counterintuitive.

To answer I cite the randomised controlled trials (RCTs) done in the 1970s that tested chest x ray pictures and sputum cytology as screening tests for lung cancer. Despite the fact that the screening tests found many asymptomatic lung cancers, none of the trials decreased mortality rates among the smokers who were screened, in comparison with controls. We would never have known this without RCTs, given that early detection nicely increased “survival times” (but really just the length of time that patients knew they had the disease).

RCTs have always been seen as the gold standard in evaluating screening tests because they eliminate many of the biases that taint uncontrolled observational studies. As a result of the RCTs …

View Full Text