Breast pathology study
I read the letter by Diane Schecter, MD (July 2015), and I respect her right to remind readers that the findings from the breast pathology study published in JAMA (Elmore JG, et al. 2015;313:1122–1132) are similar to results published nearly 25 years ago. I do object, however, to her framing the study as academia versus private practice. Nearly 25 percent of the participants had either a primary academic or an adjunct academic affiliation (see Table 2, page 1126). It is unfortunate that the media misinterpreted the article as an evaluation of overall accuracy in breast pathology. The data presented in the article need to be corrected for population prevalence of the diagnostic categories if a reader’s goal is to estimate overall accuracy. The breast pathology study sampled a broad cross section of pathologists, and the findings should indicate to all clinicians advising women undergoing breast biopsy that there are persistent challenges in reproducibly diagnosing atypical hyperplasia; some lesions may defy our current abilities to classify. We need a more flexible framework for managing these pre-neoplastic lesions while we await reliable adjunct diagnostic tools. Now that we have reaffirmed and quantified these persistent diagnostic challenges, all of us need to work collectively to help patients and clinicians as we grapple with the uncertainties inherent in the art and practice of surgical pathology.
Donald L. Weaver, MD, Professor of Pathology and Laboratory Medicine, University of Vermont, Burlington
Block and slide labeling
“Visuals to the fore in new histology labeling guideline” (June 2015) discusses the CAP/NSH panel’s recommendations on histology labeling in a more nuanced way than the materials on this subject published in Archives of Pathology & Laboratory Medicine. The CAP TODAY article emphasizes difficulties in the guideline’s implementation from financial, logistic, and privacy aspects.
The Eindhoven original root cause analysis model with its human, organizational, and technical categories is most suited to the implementation in practice of the advancements we have seen in histology blocks and slides labeling. Contrary to the investigative root cause analysis, which is more or less an effective reaction to a current event of identification failure, the Eindhoven model, used in chemistry, aviation, and transfusion medicine error prevention, gives a blueprint for comprehensively implementing policies. It requires proactive managerial efforts.
Dr. Brown emphasizes that the panel was not trying to be proscriptive in developing the guideline. As the first step, this might be right, but by limiting recommendations to only patient identification, the guideline lacks a comprehensive approach. I would agree with Vincent Della Speranza’s comment that there are no significant reported data that a lack of labeling standardization has created problems in the histology laboratory. In my experience at the grossing table, rather than a specimen mixup between patients, the more frequent problem is misidentification of and near-miss events related to specimen parts (skin biopsies especially, as well as in prostate, gastroenterology, and gynecology biopsies) during sampling. Although there is a “real estate” issue on the block, the technical difficulties of placing a barcode chip on different specimen parts would be solved by software designers if the problem were addressed. When I only mentioned barcoding in the 2008 article “Root cause analysis of specimen misidentification in surgical pathology accession and grossing,” published in Laboratory Medicine, as the next step in root cause analysis technical category development, I could not have imagined it would now be commonplace. I hope the interval between guideline revisions will be shorter than four years and the recommendations will be made more comprehensive.
Izak B. Dimenstein, MD, PhD, HT(ASCP), Grand Rapids, Mich.