Testing-testing: do we give verbiage a bad name?
Like many people, I dislike the word “verbiage” , but just testing the waters for catchy blog titles. Directly influenced by my PGY year (4) I have been interested in report construction and the use of language to convey diagnostic findings, diagnostic uncertainty, and the interpretation of newer tests such as new molecular tests which look for genetic variations as discussed with PathDoc15 in this post.
Construction of pathology reports using diagnostically accurate and meaningful language has been a focus of the major pathology organizations, especially the CAP, undoubtedly influenced by the IOM 1999 report and fueled by the relatively frequently cited article “Clinicians are from Mars and Pathologists are from Venus” published in the Archives in 20001. This article reported that surgical pathology reports were “misunderstood” 30% of the time.1 The CAP, the Association of Directors of Anatomic and Surgical Pathology, and other organizations such as the Association for Molecular Pathology publish recommendations for pathology reporting, including the minimal requirements for specific body sites, tumor types and testing modalities. Some institutions have implemented synoptic templates for cancer reporting to ensure that all of the required elements and important features of a patient’s tumor are present in the report and to display these elements in a list-type manner so that features such as tumor size or vascular invasion are readily identifiable.
The CAP in June 2007 published guidelines for reporting of molecular tests.2 The Joint Commission also issued New and Revised Molecular testing standards Jan 2007, and these standards are similar to the minimal criteria proposed by the CAP.3 Minimal criteria include (besides name of lab, patient name, ID date of collection, source, ordering doc, etc):
- List results by test name–clearly stated and obvious
- Reference range or normal range and abnormal range
- type of procedure (PCR, Southern Blot, etc)
- defined target (gene, defect, locus)
Additional recommendations include (not limited to):
- interpretation of the result
- the significance of the results in relation to the patient
- pertinent assay performance characteristics (if control has been degraded and lower detection limit is altered)
- control results
- relevant peer-reviewed literature citations
- details of procedure (kit versus home brew, non-FDA-approved test, etc)
- recommendations on genetic counseling
- residual risk of disease, or risk of disease not assessed by study
Each aspect of the above additional recommendations is a good point of discussion, and all are bound in the attempt to convey useful information to other physicians (including pathologists) for the purposes of patient care–the implementation of which is probably easier said then done. What is certain is that everyone benefits from clear reporting (i.e. the result is properly interpreted by a treating physician) and treating physicians want recommendations about clinical decision making and implications for other family member/genetic counseling bringing to light the importance of the pathologist as consultant in the “current ” testing tech-splosion4-7 (ok getting too bloggy with blogspeak here).
Standardized acceptable terminology is another issue for accurate reporting in molecular pathology as there is a lack of a strict nomenclature.2 Especially important, since a minimal requirement is to write or transcribe the target gene/protein/cDNA/mRNA/rRNA/etc. One simple example written in the recommendations is the common use of abl or bcr-abl, (all lower case) where the correct term is BCR/ABL1.2 There is also no consensus on how to report what is being studied. Should we say BCR/ABL1, BCR/ABL1 transcript, ABL1/BCR fusion product? The standardization of molecular reporting would also benefit long term data storage and easier retrieval for research whether intra-institutional or collaborative between small and large groups.
Another major point of interest is the preservation or study of patient safety in molecular testing, an area which hopefully pathology informatics may support and bolster. Pre-analytical errors such as cross contamination or specimen mix up would render false results and impact patient safety tremendously. (ex= Yes this patient has a homozygous CF mutation, no this patient does not have the PML-RARA fusion transcript). A nice idea would be an internal control on new diagnoses to ensure the specimen of study is of the correct patient. This would be helpful becuase a series of microsatellite SNPs or RFLPs could be studied and filed in a database for future comparison–however, the collection and storage of reference patient DNA for future comparison for specimen result accuracy may bring up patient privacy issues.
1. Arch Pathol Lab Med 2000; 124:1040-1046
2. Arch Pathol Lab Med 2007; 131:852-863
3. Joint Commission Perspectives 2007;1
4. CAP Today 2003; 17:1-82
5. CAP Today 2004; 18:58-66
6. Genet Med 2003;5:166-171
7. NEJM 348:2526-2534
5 Comments