Not to make things all awkward but.. wow.

The JAMA authors don't make any explicit recommendations but their findings actually kind of support breast cancer screening and preventive treatments. Especially for black women and women younger than 35. They've simply exposed opportunities for improvement.

Kolata has taken just one of the study findings (that adjuvant radiotherapy did not significantly reduce 10-yr breast cancer-specific mortality among lumpectomy recipients) and concluded somehow that preventive treatments don't impact breast cancer outcomes.. which is not what the JAMA authors suggest at all lol. She also concludes that women with DCIS diagnosis die at the same rate as women in the general population (wrong), misquotes the age risk factor as being under 40 (wrong), fails to mention that adjuvant radiotherapy reduced the likelihood of invasive cancer development by almost half, and seems to have missed the overall purpose of the study.

A hazard ratio (HR) is a 'number of events per unit time' comparison between groups. In this study the events are deaths, so the HR are a comparison of mortality rates (i.e. 1 = both groups have same # deaths per unit time and >1 = one group has more deaths per unit time.) In the results, the JAMA authors clearly state that Black women had a mortality rate 2.5x HIGHER than non-hispanic white women in the cohort, even after adjusting for other factors like income, treatment and tumor features. And the small percentage of women diagnosed with DCIS before age 35 had a mortality rate 2.1x higher than women aged 35+, and approximately 17x HIGHER than similarly aged women in the general population. So where did Kolata get this 'same likelihood of dying' point from??

Imho the focus of the JAMA paper isn't really treatment, or even detection. It's efficacy. The current standard for breast cancer screening is to look at DCIS as a cancer precursor. The JAMA analysis sought to update 10-year and 20-year mortality rate estimates for women diagnosed with DCIS, and characterize additional factors that increase/decrease breast cancer-specific mortality risk. They identified variations in mortality according to age, race, and receptor sensitivity, as well as an unexpectedly high mortality rate among DCIS diagnosed women who didn't even go on to develop invasive cancer (which screening ultimately aims to prevent). This led the JAMA authors to conclude that there are nuances with respect to DCIS prognostics, treatments and mortality risk that have not been integrated into practice. Kolata... seems to have read a different paper or something lol.


Anyway, glad I read the JAMA article before the NYT one. I like this take from the same JAMA issue, much better commentary:

http://oncology.jamanetwork.com/article.aspx?articleid=2427488

And in other news.. I just learned today that, apparently, Kolata has reported erroneously on medical matters before and the NYT had to issue a lengthy correction lol:

http://ahrp.org/ny-times-corrects-gina-kolata-re-alzheimers/