There are several things in this communication that cause me deep concern. In Figure 1, it claims that there were 583 breast cancers newly diagnosed during the study period. Out of these, 272 were deemed to be advanced breast cancer or unsuitable for surgery. This is hugely outside any previous publication in this country or for any other country. In general terms, we would have no more than 5% of patients presenting with disease unsuitable for surgical intervention. This invalidates anything that has been said further in this paper.
Furthermore, while the authors are keen to say they are exploring NICE guidelines, they themselves are not adhering to these guidelines. First, NICE explicitly recommends that if sentinel node biopsy is carried out it should be carried out using the triple technique of scintigram, gamma probe and blue dye rather than blue dye alone. Using just blue dye is inaccurate and should not be recommended.
Furthermore, NICE guidance also states that frozen section is not recommended. There are indeed many publications in the literature that show that frozen section of the sentinel node has utility; in other words, when it is positive it is accurate on the whole. However, it has a significant high false-negative rate and therefore has been superseded by other methods such as preliminary change reaction molecular analysis.