Dear LOINC community,
I am trying to obtain possible results/examples of answers for many tests that have ordinal or semi-quantitative results. For example: antibody titers - are they represented as 1:80, as “reactive”, as both? Other examples include antibiotic susceptibility testing, avidity testing, NAA testing, etc. Beyond the answer lists currently available, is there another way to obtain example answers for these labs? example loins include: 100915-8, 101478-6, 103630-0, 101750-8, 104428-8, among about others on my list.
Hi Caroline! Welcome to the Forum and thanks for posting the question.
One key to predicting example answers is to follow the property and scale attribute descriptions.
ORD scale with PrThr property LOINCs contain example answers of detected/not detected, positive/negative, reactive/non-reactive, opposite word pairings. This includes 100915-8 (Burr Cells by microscopy), and 101750-8 Adeno-associated virus (AAV) serotype Rh74 [Presence] in Serum by Immunoassay
TITER property will always have a SEMI-QN scale, and will always have a x:z ratio format. 103630-0 Monkeypox virus IgM Ab [Titer] in Serum by Immunoassay is your example here.
There are other SEMI-QN scale entries with different properties; they may be numeric ranges in values, or be ranges of log values. The SEMI-QN scale is indicating they are not sequential in number values. Your question includes 104428-8 Measles virus IgG Ab avidity [Ratio] in Serum by Immunoassay would fall into this arena, as it has a RATIO property and is SEMI-QN. The numeric ratio is not a titration value.
You bring up microscopy susceptibilities with 101478-6 Ampicillin+Sulbactam [Susceptibility] by MIC.VOC detection in Positive blood culture. Microbiology is the only class that uses the OrdQn scale, meaning values could be Sensitive, Intermediate, Resistant OR actual numeric MIC/KB/etc values.
If you have more examples/others on your list, please reply here and let’s get some answers for you!
Sincerely,
Pam
With respect to microbiology susceptibilities, there are multiple methods by which they are obtained. For example, MIC.VOC, gradient strip, disk diffusion (KB), by phenotype, method for slow growing mycobacteria, genotype method, phenotype method, bactericidal titer, or not specified (blank). Based on these specifications, would specific answers be expected? ie. is Sensitive, Intermediate, Resistant OR actual numeric MIC/KB/etc value predictable based on these methods?
For loincs such as 60578-2, would you have any insight into the answer key?
What type of setting are you using LOINC codes and which country? These aspects impact responses.
For example, for the antimicrobial susceptibility testing, it is correct that there are different methods and there are differences in how these test results are reported in accord with Clinical Laboratory Standards Institute (CLSI) Best Practices / Protocols as well as CLIA regulations in the US.
CLIA requires laboratories as to validate a test system including following the manufacturer’s package insert. This includes how a test is reported and responses (e.g. detected or not detected vs reactive or non reactive). If they deviate, it becomes a laboratory developed test (LDT).
A word of caution, any LOINC Answers may or may not (in full or partially) correspond to the package inserts. To address your question the LOINC Answers likely will not predict allowed responses for a test.
In addition, there may be some assays that generate a numeric result value, but if the manufacturer provides interpretation guidance for the numbers and indicates a qualitative result value is to be reported, then that is what the laboratory reports. In one lab in which I worked the same vendor test system was utilized for drug testing. One panel had each drug reported out with numeric result values, while the other panel of the exact same drugs performed on the exact same instrument utilized SAMHSA cut-offs for interpretation and reporting of positive or negative. The cut offs varied by each drug in accord with federal law. Even though a patient may have had low numbers of said drug detected by the analysis of their specimen, they may have had a negative result value reported as they did not meet the cut-off/threshold for reporting a positive test value.
For LOINC 60578-2, this would likely be a LDT, where each performing laboratory will determine how they report their results and values. Best practice is to ask the performing lab with any test questions.
