Firstly, Dr Canavese and colleagues' interest in the guideline
document is much appreciated.
Before responding specifically, it is worth noting that there are two main
diagnostic decisions to make when a patient presents with suspected
chronic idiopathic inflammatory bowel disease (IBD): distinguishing IBD
from other causes of inflammation; and classifying IBD as ulcerative
colitis (UC) or Crohn's disease (CD). Even with full clinical details,
comprehensive sampling, and assessment by a well-informed pathologist
there will be a significant minority of IBD biopsies that cannot be
classified confidently as UC or CD.[1, 2] There will also be a number
that cannot even be categorised as IBD or non-IBD. However, I agree
entirely that several other factors beyond the pathologist's control may
also contribute to diagnostic uncertainty. Indeed, the proven importance
of adequate clinical details and of thorough sampling are often noted in
clinical guidelines.[1, 3-8] For example, limited sampling within and
between sites may be one of the reasons why histopathologists are
apparently less successful at diagnosing CD than diagnosing UC.[7-9]
I also agree that pathologists have a responsibility to convey the level
of diagnostic certainty as clearly as possible to the clinical teams.
Proposed categories for the conclusion of an IBD biopsy report have been
included in the BSG guideline ("PAID" scheme, Table 14).[10] The
suggested terminology represents a consensus view on the best way to
express probability. Accordingly, its adoption is recommended.
Similarly, vague terms such as "in keeping with" are best avoided or used
sparingly (Table 13).[10]
Dr Canavese and colleagues' suggested solutions are helpful. I agree that
the pathologist should insist on a minimum standard of clinical input.
Indeed, provision of the endoscopy report to the pathologist will be
recommended strongly in a forthcoming guideline for clinicians.[11]
Similarly, identifiable deficiencies in the process should be noted by the
pathologist interpreting the biopsies, especially if they interfere with
assessment. Also, a statement in a histology report that repeat endoscopy
and sampling might be informative would oblige the clinician to consider
this option. However, enforcement of minimum clinical standards can be
difficult. Overall, better communication between pathologists and
clinicians, ideally in the setting of a regular clinicopathological
meeting, is an excellent way to enhance diagnostic accuracy.[4, 10, 12]
On a more general note, the scope and quality of IBD services and the
reasons for suboptimal management may vary within and between countries.
Attempts are being made to remedy the inconsistencies. Reassuringly, a UK
services standards document for IBD includes guidance on the use of
histopathology services and is cognisant both of the value of biopsy
assessment and of the importance of interaction between pathologists and
clinicians.[13]
Yours sincerely,
Professor R M Feakins
Conflict of Interest:
None declared
Firstly, Dr Canavese and colleagues' interest in the guideline document is much appreciated. Before responding specifically, it is worth noting that there are two main diagnostic decisions to make when a patient presents with suspected chronic idiopathic inflammatory bowel disease (IBD): distinguishing IBD from other causes of inflammation; and classifying IBD as ulcerative colitis (UC) or Crohn's disease (CD). Even with full clinical details, comprehensive sampling, and assessment by a well-informed pathologist there will be a significant minority of IBD biopsies that cannot be classified confidently as UC or CD.[1, 2] There will also be a number that cannot even be categorised as IBD or non-IBD. However, I agree entirely that several other factors beyond the pathologist's control may also contribute to diagnostic uncertainty. Indeed, the proven importance of adequate clinical details and of thorough sampling are often noted in clinical guidelines.[1, 3-8] For example, limited sampling within and between sites may be one of the reasons why histopathologists are apparently less successful at diagnosing CD than diagnosing UC.[7-9] I also agree that pathologists have a responsibility to convey the level of diagnostic certainty as clearly as possible to the clinical teams. Proposed categories for the conclusion of an IBD biopsy report have been included in the BSG guideline ("PAID" scheme, Table 14).[10] The suggested terminology represents a consensus view on the best way to express probability. Accordingly, its adoption is recommended. Similarly, vague terms such as "in keeping with" are best avoided or used sparingly (Table 13).[10] Dr Canavese and colleagues' suggested solutions are helpful. I agree that the pathologist should insist on a minimum standard of clinical input. Indeed, provision of the endoscopy report to the pathologist will be recommended strongly in a forthcoming guideline for clinicians.[11] Similarly, identifiable deficiencies in the process should be noted by the pathologist interpreting the biopsies, especially if they interfere with assessment. Also, a statement in a histology report that repeat endoscopy and sampling might be informative would oblige the clinician to consider this option. However, enforcement of minimum clinical standards can be difficult. Overall, better communication between pathologists and clinicians, ideally in the setting of a regular clinicopathological meeting, is an excellent way to enhance diagnostic accuracy.[4, 10, 12] On a more general note, the scope and quality of IBD services and the reasons for suboptimal management may vary within and between countries. Attempts are being made to remedy the inconsistencies. Reassuringly, a UK services standards document for IBD includes guidance on the use of histopathology services and is cognisant both of the value of biopsy assessment and of the importance of interaction between pathologists and clinicians.[13]
Yours sincerely,
Professor R M Feakins
Conflict of Interest:
None declared