Toward harmonization of clinical molecular diagnostic reports: findings of an international survey

Laboratory and patient demographics

The laboratory can define the format in which it chooses to report patient results. Laboratories use the following mechanisms to release data: paper/faxed (also known as “hard” copies) copies, electronic reports to the healthcare provider, electronic reports to electronic medical record (EMR), web-based reporting and reports directly to patient. Some laboratories use a combination of the above.

The format of the report should provide identifying information regarding both the patient and the laboratory performing the test. All survey participants stated that the identity of the laboratory, location and address was included in the molecular diagnostic report. This information provides the healthcare provider with the ability to contact the laboratory if necessary. Report design may repeat selected information on every page of the report. For example, ISO 15189 recommends that the patient name and the location of the patient be included on every page. The patient location may be interpreted as that of a clinic or inpatient facility. All surveyed laboratories reported that the patient name was included on every page of the report. Two respondents (13%) stated that their molecular diagnostic report did not include the location of the patient on every page (Table 3). Similarly, ISO 15189 recommends including the name and contact information from the healthcare provider on the report. Compliance was 81% for this recommendation. Inclusion of the patient location (such as the clinic or outpatient facility) on all pages of the report ensures that each page can be delivered to the healthcare provider. For this reason, reports should be numbered and include the total number of pages contained in the report (such as page 1 of 2). So-called “hard copies” or faxed copies of multipage reports are at risk of being incomplete if pages become separated from each other during the reporting process. Even blank pages generated during the process should be included in the report to assure that the complete report has been delivered to the healthcare provider that ordered the test(s). Eighty-seven percent of laboratories state the total number of pages in their report. The IFCC C-MD requested information inquiring whether laboratories included two unique patient identifiers in their reports. Reports including two unique identifiers (such as patient name and date of birth) improve the process that correct results are delivered to patients with the same or similar names, an aspect that is particularly important for studies on related individuals. Eighty-eight percent (n=14) of laboratories included two unique identifiers in their report.

Sample demographics

Quality and interpretation of test results can be affected by auxiliary information such as date and time of specimen collection. This information enables the laboratory to identify potential preexamination issues that could impact reportable results or interpretation [2]. For instance, delays in receiving the specimen could render the specimen unacceptable. Date of collection also permits the healthcare provider to monitor test results that vary with treatment. Compliance with the date of collection was 88%, whereas compliance with the time of collection was 75%. Including the time of collection is an example of an ambiguous recommendation because the specific recommendation states “…time, when available and relevant to patient care”. Including the time of collection does not universally impact the quality of a molecular result or its interpretation. However, the time of collection may be required for test interpretation if treatment is started before or after specimen collection. For instance, molecular levels of analytes such as tumor biomarkers associated circulating cell free DNA (cfDNA) or infectious agents may change following treatments such as chemotherapy, transfusion or antiviral/antibiotic administration. The date that the report was released was reported by 100%, whereas laboratories who also included the time of the report was 81%. The inclusion of the time of the report should be available to the provider if not included on the report according to ISO 15189. For a quality monitoring perspective, tracking the date and time of release of test results could be valuable metrics used to assess sample processing. Interpretation of laboratory results by the healthcare provider may be influenced by the specimen type, and for this reason, specimen type should be included in the report. For instance, a healthcare provider may request the same test on bone marrow and peripheral blood for molecular monitoring of hematologic disorders. Eighty-eight percent of laboratories included the specimen type on reports. Although not included in the ISO 15189 recommendations for reporting, the surveyed laboratories were asked if the specific collection media (or device) was included in the report. A lower number of laboratories (56%) included the specimen collection media or device on the report. Molecular results and interpretation can be affected by the collection method, as exemplified by the unsuitability of heparinized blood samples in some analytical systems.

International diagnostic codes

Another question that was asked by the IFCC C-MD but not required by ISO 15189 inquired if laboratories included international diagnostic codes in the report. A minority of laboratories (38%) used the World Health Organization (WHO) designated International Classification of Diseases [ICD-10] codes and/or Online Mendelian Inheritance in Man (OMIM®) codes as references for the clinical indications/disease in the report [37], [38]. For example, the ICD-10 code for BCR-ABL1 positive Chronic Myeloid Leukemia (CML) is C92.1, whereas the OMIM code is 608232. Clinical classifications in the report support interpretation of the molecular results. In some instances, testing performed on a specimen using a correct diagnostic code provides evidence of appropriate test utilization and can aid interpretation. In some locales, discrepancies between the diagnostic code and the test result may prevent reimbursement for the test.

Molecular diagnostic test methodology

Molecular methods differ in their analytic and clinical performance characteristics that can negatively impact commutability of results. Describing the molecular method in the report provides information, aids interpretation of test results and enables identification of samples which may require a retrospective reanalysis where methods are ultimately deemed to be unreliable (e.g. based on proficiency testing). The ISO 15189 recommendation for including the measurement/examination method in the report is ambiguous in that the phrase “where appropriate” accompanies the recommendation. Although this recommendation is open to interpretation, 94% (n=15) of laboratories identified the methodology in the report. Laboratories reporting the method in the body of the report were 88% [n=14]. Other locations for methods description included the disclaimer section or a link to an outside source. Although not a recommendation by ISO 15189, some agencies require that the laboratory declare if the method has been validated by a regulatory agency, which may be particularly relevant for laboratory developed tests. Laboratories (specifically 38%) reported if the method was cleared or approved by a regulatory agency and included the statement either in the body, disclaimer and/or in a link outside of the report. A lower percentage of laboratories (31%) disclosed if the laboratory-developed molecular tests were validated. Information on the method/examination provides the healthcare provider the opportunity to interpret changes in diagnostic values that may result from technical artifacts rather than changes caused by pathophysiology of the disorder.

Report size and content

Space restrictions on medical report can be an issue. No recommendations on specific font size, report length or reporting mechanism (human readable) are mentioned in ISO 15189. Some healthcare providers prefer short reports. In this survey, reports ranged from one to three pages with most laboratories utilizing a one page format (specifically, 47%) compared to a three page format (specifically, 13%). Laboratories may maximize the information on the report by altering font size or, as previously mentioned, using links to outside information resources such as websites (specifically disclaimers or statement regarding the measurement procedure). Laboratories can elect to emphasize information by varying font size. Sixty-three percent of laboratories used different size fonts, color fonts and/or underlined words to differentiate “detected” or “affected” from “not detected” or “not affected”. By contrast, 67% laboratories used the same font size as the rest of the text in the body to describe the measurement method. Of the laboratories that changed their font size, the font was reduced to size 8 and 6. Thirty percent of laboratories include non-human readable code (for example, Quick response code [QR] or barcodes) on their reports (Figure 2). A balance between insuring adequate information for test interpretation and readability may be a challenge for laboratories seeking to maximize the content of the report while minimizing the number of pages for the report.

Figure 2:

Non-human readable codes.

Quick Response Code (QR) for the IFCC C-MD.

Laboratories in multilingual countries were surveyed to determine the number of languages used to report the result. Ninety-three percent of respondents representing 14 countries used a single language in the report. One country reported results in two languages (specifically Persian and English). Most countries report in the language spoken by the majority of the population. However, some countries used English as the reporting language even if the majority of the population spoke a different language. Reporting in English in these countries occurred prior to the country gaining their sovereign authority (such as when a region was occupied or colonized by an English speaking country, or if English was the language for medical education). ISO 15189 does not have a recommendation for selecting the primary language of the report.

Nomenclature

ISO 15189 recommends that results should be unambiguous and that standardized nomenclature and units be used in reporting results. Standardization of nomenclature, result terminology, units, reference sequences and relevant test statistical performance (such as clinical and analytical) permit the end user to better interpret laboratory results. Half the laboratories reported using standardized nomenclature in their reports. Not all standardized nomenclature is recognizable by the end user; hence, “non-standardized” nomenclature may be added to a report. Two laboratories include both standardized (specifically HUGO) and colloquial names in their reports [39]. Another challenge for reporting is the selection of “positive/negative” vs. “detected/not detected”. CLSI MM03 recommends the use of “detected” vs. “not detected” for infectious organisms rather than “positive” or “negative” [40]. For inherited diseases, laboratories may use “present/absent” for specific gene variants and conclude with a description as to whether the genotype is homozygous, heterozygous, composite heterozygous and so on. However, if the assay is cleared or approved by a regulatory agency, then the package insert recommendations for reporting may be followed. Thirty-three percent of laboratories included specific reporting language for assays cleared or approved by a regulatory agency in their reports.

Units and reference sequences

ISO 15189 recommendations aim to improve standardization results through recommending the inclusion of international system of units (SI), units traceable to SI units or “other appropriate units”. This recommendation raises considerable challenges for molecular diagnostics reporting as SI units are not available for many analytes. Prior to the introduction of the international unit (IU), quantification of hepatitis C viral load exemplified the heterogeneity of results for different testing methods. Although all the results were reported in copies per milliliters, the conversion factors of 0.9, 2.7, 3.4, 3.8 and 5.2 were required to transform copies/mL to IU for each testing methodology [41]. Similarly, the introduction of the International Scale (IS) unit for BCR-ABL1 quantification harmonized reported results across laboratories. Prior to the IS unit’s introduction, the Minimal Residual Disease (MRD) proficiency testing survey by the College of American Pathology (CAP) documented numerous reporting units (such as ratio, % ratio, RNA copies, cell numbers, % Log BCR-ABL1, copies/mg, log reduction, etc.). Figure 3 presents the various units used by the molecular laboratories (50%) reporting quantitative test results. Specific reference sequences may provide a control for testing and interpretation of sequence data. For laboratories performing sequencing or multiple/complex single nucleotide polymorphism (SNP) testing, 69% specified a reference sequence in their reports and 46% reported variants of unknown or undetermined significance (VUS). Relevant analytical and clinical performance of the molecular diagnostic assay was not always included in the reports issued by participating laboratories [42]. For laboratories reporting quantitative values, 46% of laboratories included either the Limit of Detection (LOD) and/or the Limit of Quantification (LOQ) in their reports. Measures of uncertainty (for example, % CV, 95% confidence intervals, etc.) were included in reports by 43% of surveyed laboratories. The low adherence rates for these recommendations among surveyed laboratories may indicate multiple needs in order to standardize reports.

Figure 3:

Units of measure for molecular diagnostics.

The number of laboratories reporting specific units of measure is listed above each category. Laboratories not reporting units did not use quantitative tests.

VUS and secondary findings

Variants of undetermined/unknown significance (VUS) can be identified with nucleic acid sequencing technologies [43], [44]. Such variants receive this classification when the genotype-phenotype correlation is not known or unclear. The clinical significance of these variants may not be known at the time of diagnosis but may become relevant in the future. A minority of laboratories (46%) performing sequence analysis reported VUS. This low percentage may reflect the difference between laboratories performing complex genotyping using technologies such as parallel sequencing vs. laboratories offering basic genotyping (such as SNP testing for Factor 5 Leiden) or targeted infectious disease testing. Additional discussion on interpreting and reporting biocurated VUS will be addressed in the interpretation section.

Secondary findings (SF), also known as, incidental findings (IF), may occur if the laboratory performs a test and identifies an organism (or genetic variant) that was not requested [44], [45]. For instance, in response to a request for Gardnerella vaginalis, a laboratory may erroneously perform Neisseria gonorrhoeae (NG) testing. This secondary finding could have serious implications if the accidental testing identified a medically actionable result. Additionally, the laboratory may perform panels of tests regardless of whether specific analyte(s)/ measurand(s) were requested. The responses for this question were as follows: 29% reported SF/IF, 21% did not report SF/IF and 21% had variable practices depending on the test and test result. One laboratory stated that SF were eliminated when targeted sequencing was performed. Another laboratory identified SF when BRCA1 but not BRCA2 sequencing tests (and vice versa) was ordered. In this scenario, the laboratory routinely performs and reports sequence results from both genes. When a test is accidentally performed due to technical or workflow error and the result could impact patient care, one participant stated that their laboratory would ask the healthcare provider for revised requisition for the specific test that was erroneously performed. After receipt of the revised requisition, the SF/IF was reported. With respect to releasing results from test panels, reporting in toto vs. singly had varied responses among surveyed laboratories. Fifty-four percent of laboratories performing panels reported all results at one time.