Sentinel node biopsy in melanoma is not a good prognostic marker for individual patientsBMJ 2013; 346 doi: http://dx.doi.org/10.1136/bmj.f651 (Published 06 February 2013) Cite this as: BMJ 2013;346:f651
- Jonathan L Rees, professor of dermatology1
Torjesen deserves credit for her article on sentinel node biopsy (SNB) in melanoma,1 but it is sad that medicine needs a journalist to put its own house in order (excepting people like John Meirion Thomas). So much for evidence based medicine.
Other commentators have pointed out that the evidence that SNB prolongs life is deeply unconvincing. The National Institute for Health and Clinical Excellence called it right on this occasion. However, the idea that SNB is a good (marginal) prognostic marker for individual patient care needs debunking. People keep confusing patient group effects on Kaplan-Meier plots with what can be predicted for individual patients.
Readers should look at the useful publication by Melski on Rhodes’s data.2 3 For those who have troubling understanding the difference between hazard ratios and uncertainty surrounding how long an individual patient might live, a paper by two biostatisticians is worth reading.4 With reference to another cancer (lung), “The main point of the present contribution is to emphasise that in all realistic scenarios we can imagine, the intrinsic statistical variations in life times are so large that predictions based on statistical models and indices are of little use for individual patients.”
Finally, the simplest explanation for the persistence of SNB—in the absence of convincing support from randomised controlled trials—is surely that given by people such as John Wennberg in explaining variation in healthcare provision—money, or at least practice styles that encourage consumption at the margin of rationality.5
Cite this as: BMJ 2013;346:f651
Competing interests: None declared.