Qualitative vs. Quantitative Assessment; Art vs. Science

| August 21, 2008

In this month’s American Journal of Clinical Pathology, there is an article by Dr. Carlos A. Rubio and colleagues from the Karolinska Institute and University Hospital in Stockholm, Sweden.  Dr. Rubio and his colleagues assess different methods for accurately diagnosing collagenous colitis using quantitative and qualitative means and entitle their paper "Quantitative Assessment of the Subepithelial Collagen Band Does Not Increase the Accuracy of Diagnosis of Collagenous Colitis".  The paper is an interesting read.

Here is the abstract:

The thickness of eosinophilic band in collagenous colitis (CC) was assessed by 3 methods: histologic estimates (22 observers), conventional measurements using a calibrated micrometric scale (1 observer), and semiautomatic micrometric measurements (1 observer). By the histologic estimate technique, 7.4% of the results failed to diagnose CC; by calibrated micrometry, the failure was 6% and by semiautomatic micrometry, 6%. The main difficulty in measuring the thickness of the CC band is that the deeper border of the band appears fuzzy and hairy-irregular. CC should be defined not exclusively on the basis of the thickness of the collagen table, but as a microscopic constellation characterized by a distorted superficial cell arrangement, with areas of epithelial denudation and inflammatory cells in the superficial epithelium and the lamina propria. In agreement with Lazenby’s statement: "Focusing solely on the collagen band can result in both over- and underdiagnosis."

There is also an editorial by Dr. Chris Moskaluk, a pathologist at UVA with research interests in the molecular changes of neoplasia.  He makes the appropriate point that there will always be the need for "traditional aspects" of anatomic pathology, even in the "molecular and digital era of medicine." For "medical" biopsies as Dr. Moskaluk writes, "Subtle alterations in the microanatomy of tissue compartments and specific subanatomic distribution of inflammatory cells will be most easily and accurately identified in the foreseeable future by the eyes and brains of skilled and experienced practitioners of histopathology. The acquisition and judicious application of such skills will ensure the maintenance of the traditional aspects of anatomic pathology even in the molecular and digital era of medicine."

All of this is true, and while there will always be a need for the "art" as well as the "science", I gather the "science" can be taught the "art" and require less "eyes and brains" once those "skilled practioners" refine the algorithms to the point that they can be taught, learned and refined without a histopathologist skills to decide upon refinements beyond standard published accepted diagnostic criteria.


Category: Digital Pathology News, Publications

Comments are closed.