3 - Special cases

3.1 Findings viewed as variables or as values

For some complex tests there are two ways to organize the results into a report.

3.1.1 Value

Assume a set “X” is made up of five “results” that can have a scale of (absent, present) or (0, 1). These results could be reported as:

Finding 1 = Present or 1
Finding 2 = Absent or 0
Finding 3 = Present or 1
Finding 4 = Absent or 0
Finding 5 = Absent or 0

Each finding is then considered a binary variable. This is sometimes called a “panel” approach.

3.1.2 Variable (multiple choice) approach

The alternative would be to report this information as a single variable (or multiple-choice question) with many possible values:

Variable X - Finding 1, Finding 3

In this case the findings are the values of a variable called Variable X; only the positive findings are reported as values. Many laboratory tests, e.g., those that test for HLA antigens, red blood cell antigens, or screens for toxic substances, could in theory be presented either way. The microscopic part of the differential count and urinalysis could also be described either way. History and physical findings and (given a real stretch) even culture results could be structured in the panel or multiple choice/multiple answer format.

A single lab may report red blood cell antigens in either way, as a binary panel or a multiple-choice result, depending upon the purpose of the test. The routine cross and type are reported out in the multiple choice pattern format (only positives from a modest fixed set of tested antigens are reported). But if the tests are being used to prove fatherhood, the results are usually reported as a binary panel.

Blood cultures could in theory be regarded as panels:

Test Name Value
Escherichia coli absent
Staphylococcus aureus present
Diphtheroids absent
Streptococcus pneumoniae absent
Pseudomonas aeruginosa present

Although in practice such tests are almost always reported in the multiple choice/multiple answer format, as follows:

Test Name Values
Blood culture P. aeruginosa, S. aureus

We bring up these issues to explain why we use a somewhat different data format for some types of tests, and why we sometimes provide for both reporting methods (e.g., HLA blood cell antigen tests) in the LOINC database. When a binary scale is used, the Scale will be ordinal (Ord) and the kind of Property will usually be presence (PrThr, for results based on the presence/absence of an analyte regardless of whether an internal cut off value is used to determine the ordinal result). When the multiple-choice multiple-answer approach is used, the Scale will be nominal (Nom) and the type of Property will be presence or identification (Prid).

3.2 Blood bank

Red cell antigens will be named in accordance with the American Association of Blood Banking (AABB) naming standards.1 In addition to the antigen or antibody, a modifier would be included in the super-system (the second subfield of the System field), to indicate whether testing was performed on the patient, donor, or blood product. Unless explicitly stated, testing is assumed to have been on a material collected from a patient. Additional information about the person identified in the System, such as the donor's name or relationship to patient, should be placed in other OBX or comment segments of the message.

Blood bank reporting illustrates the need for a method of reporting by panel and by multiple-answer mechanism. The LOINC database provides observation names for both kinds of reporting.

3.2.1 Panel reporting

Each reportable antigen must have its own test, so that each element in a full set of binary tests could be reported as (negative, positive) or (0, 1).

The fully specified names of A, AB, B, and O blood types (as observations) would be as follows:

Measure of serum antibody against type A blood of donor:

A Ab:PrThr:Pt:Ser/Plas^donor:Ord

Presence of A antigen on donor's red blood cells:

A Ag:PrThr:Pt:RBC^donor:Ord

Presence of A antigen on the blood cells in a pack of blood given to the patient:

A Ag:PrThr:Pt:RBC^BPU:Ord

3.2.2 Multiple answer reporting

All blood antibodies found (or not found) can also be reported in one result term:

Antigens absent:Prid:Pt:BBL^BPU:Nom
Antibodies identified:Prid:Pt:Ser/Plas:Nom

The LOINC database provides other “observations” for reporting: the status of each blood pack (e.g., held, given, discarded), and for reporting that information when HIS and medical records systems want it; how much of each type of blood product was given at a moment in time; the type of each pack; any adverse reaction to that pack; and the pack number to accommodate laboratories that send this information as discrete observations.

Blood product disposition:Type:Pt:^BPU:Nom
Blood product type:Type:Pt:^BPU:Nom

3.3 Immunocompetence studies (flow cytometry)

The CD (Cluster of Differentiation) markers in the LOINC database include all of the single markers and the most commonly reported combinations, e.g., CD11C+CD20C+. Most of these are measuring the number or percent of a population of cells that bear the specific marker(s) of interest, e.g., Cells.CDx. In other cases, the percent of a population of abnormal cells, such as blasts, is measured, for instance, Blasts.CD2 or Abnormal blood cells.CD5. LOINC Components for cells with single markers typically do not include a plus or minus sign after the marker because flow cytometry labs do not report single CD marker-negative results, so when the marker is reported, the plus is implied. In other words, the LOINC Component Cells.CD4 represents the presence of CD4, and we do not have a Cells.CD4- Component, because such a result is not reported. However, when multiple markers per cell are reported, the pluses and minuses are always included, for example:

Cells.CD3-CD8-CD57+

Two kinds of measures are of interest, as described below.

3.3.1 Number of cells containing the marker

The “absolute” number of such cells per cubic millimeter is represented as number concentrations, for example:

Cells.CD16C+CD56+:NCnc:Pt:Bld:Qn

3.3.2 Percent of cells containing the marker

Percent of cells containing the named marker per 100 cells of that type is represented as number fraction, for example:

Cells.CD16C+CD56+/100 cells:NFr:Pt:Bld:Qn

3.3.3 Specifying the parent cell population

Historically, the most common cell marker tests were measuring the marker(s) on a population of lymphocytes, such as CD4+ and CD8+ counts, and the LOINC Divisor for the terms representing the number fraction was simply specified as 100 cells, where the fact that the 100 cells are lymphocytes was understood. Over the last several years, cell markers have become more common in the diagnosis and monitoring of a variety of immune deficiencies and hematologic malignancies, and for these tests, the parent cell population may be not necessarily be lymphocytes. In order to fully specify what is being measured, many newer LOINC terms include the cell marker(s) that define the parent population in the Divisor. So for example, the following term is used to report the percent of cells that do not express CD4 or CD8 out of a population of cells that express CD3 and T-cell receptor (TCR) alpha and beta:

80699-2 CD4-CD8-/100 CD3+TCR alpha beta+ cells in Blood

However, in some cases, the parent population may vary depending on the clinical context and/or the laboratory, so we also have newer terms that have 100 cells as the Divisor. For these, we recommend reporting the parent cell population using a separate LOINC term, 32760-1 Cell type in Unspecified specimen.

3.4 Microbiology

3.4.1 Microbiology cultures

The inherent complexity in reporting microbiology culture results presents unique challenges for the goal of standardized observation names.

3.4.1.1 Result status

The result status (Preliminary, Final) should not be reported as a separate observation or as part of the name. It should be reported in the Result Status field (OBR-25) of the HL7 OBR segment.

3.4.1.2 Specimen

The specimen type (Serum, Blood, Urine, etc.) will be indicated in the HL7 OBR segment with the Specimen Source field (OBR-15), but may also be represented in the name.

Details of specimen collection will usually be noted as OBX segments or comment segments that accompany the culture result message. The observation identifier for the OBX segment will have the fully specified name of “Specimen collection description:Find:Pt:*:Nom” and the Observation Sub-ID field will be used to order or group sets of observations. That is, if the material was collected by swabbing a wound of the right upper arm, multiple OBX segments would be created, each with the name “Specimen collection description:Find:Pt:*:Nom” and the Observation Results fields of the OBX segments would contain respectively “Swab,” “Right,” “Arm,” and “Wound.” (The granularity of the actual terms used in the specimen description is at the discretion of the user. Thus, “Right Arm Wound” as the value of a single OBX segment could be used in place of the three codes described in the previous sentence.)

3.4.1.3 Descriptive results

Descriptions of measurement and culture growth will be noted as separate OBX segments that accompany the culture result message. The name of the observation identifier will provide the context of the observation. For instance, the name for a quantitative test of bacteria in a specimen would be:

Colony count:Num:Pt:XXX:Qn:VC

Descriptions of Gram stain findings will be noted as OBX segments that accompany the culture result message. The name of the observation identifier will be:

Microscopic observation:Prid:Pt:XXX:Nom:Gram stain

The result values that could be reported with this test (which is a multiple-choice, multiple answer type or observation) might include one or more of the following:

  • Epithelial cells
  • Gram-positive cocci in chains
  • Many Gram-negative diplococci

3.4.1.4 Culture results

The organisms identified in a culture will be sent as result values in OBX segments. LOINC provides codes to identify the observation, but not for identifying the names of organisms that would appear as result values (i.e. in OBX-5). SNOMED CT is an appropriate source for these organism concepts.

While “Throat Culture” is the source of the culture inoculum, it is also a label that indicates what kind of media was inoculated and the other techniques used in the laboratory. So, it is a short hand for a kind of method and such will be recorded as the Method part of the name. Thus, “Throat Culture”, “Blood Culture”, and “Clostridium difficile Culture” all represent labels for how a culture was performed.

The LOINC naming model for routine cultures is:

Bacteria identified:Prid:Pt::Nom:Culture

Example names include:

Bacteria identified:Prid:Pt:Bld:Nom:Culture
Bacteria identified:Prid:Pt:Burn:Nom:Culture
Bacteria identified:Prid:Pt:Stool:Nom:Culture

The LOINC names for cultures are based on the expected observations they generate, e.g. “bacteria identified” or “virus identified”. For most routine cultures the users are looking for bacteria. However, in some cases a routine culture may also grow out some fungi like yeasts. Such observations can also be reported under the same test code. But, since the culture would not be optimized for growing out fungi we have a bacteria-focused name.

It is worth emphasizing these culture terms are intended for use as both as orders and observations despite having a more result-oriented name. Furthermore, as a matter of good clinical practice, most laboratories perform susceptibility testing on any significant isolate on a reflex basis. So there is no need for codes that say “culture and sensitivity” because it is implied. Thus, an order for:

17928-3 Bacteria identified in Blood by Aerobe culture

would likely trigger several workflow steps, including bottle blood culture, bacteria isolation, identification, and antimicrobial susceptibility testing.

Note
At the current time, a single code for a routine or organism-specific culture is recommended for reporting the results of the bottle culture, bacterial isolation, and identification. Historically, all of these steps have been included in the laboratory workflow under the “culture” umbrella, and individual results for each step have not been reported to or stored in the patient’s record. Susceptibility testing would be reported with separate susceptibility terms as described in Section 3.5.

LOINC names for Methods of staining a sample/material directly (where many descriptive observations are possible) include:

Microscopic observation:Prid:Pt:XXX:Nom:Gram stain
Microscopic observation:Prid:Pt:XXX:Nom:Dry mount
Microscopic observation:Prid:Pt:XXX:Nom:India ink preparation
Microscopic observation:Prid:Pt:XXX:Nom:Trichrome stain
Microscopic observation:Prid:Pt:XXX:Nom:Giemsa stain

Names for results of staining procedures performed on organisms that are growing in culture will use Isolate as the System (sample type). For example:

Fungus identified:Prid:Pt:Isolate:Nom:Fungal subtyping

Names for organism-specific cultures:

Brucella sp identified:Prid:Pt:Bld:Nom:Organism specific culture
Bordetella pertussis:PrThr:Pt:Thrt:Ord:Organism specific culture
Chlamydia sp identified:Prid:Pt:Gen:Nom:Organism specific culture
Legionella sp identified:Prid:Pt:Sputum:Nom:Organism specific culture

Note if a test applies to a specific species of organism, the Component should include the genus AND species (at least). If the measure applies to a series of species in the same family the string “sp” must be included. If it applies to as subgroup of the genus, then that subgroup should be named.

Names for Method for general class of organism:

Fungus identified:Prid:Pt:Wound:Nom:Culture
Bacteria identified:Prid:Pt:CSF:Nom:Culture

Again, the Result Value of these tests would be either organism names or other statements of culture outcome. The table below contains valid values of the culture result from the HL7 OBX segment:

Table 15: Example Culture Results

No growth
Gram-positive cocci
Small Gram negative rod
Escherichia coli
Normal flora
Candida albicans

3.4.2 Properties for culture terms

Presence or Identity (Prid) as a Property should be used when the value of a test can identify one set of alternative infectious agents. If the culture is for herpes virus and the culture can have results of herpes virus 1, herpes virus 2, etc., then Prid is the right Property. If the culture is for herpes virus and the answer is positive/negative or yes/no, then the Property should be presence (PrThr) and the Scale ordinal (Ord).

3.4.3 Microorganism identification based on nucleic acid targets

PCR-based testing for the presence of microorganisms is becoming more common. Early on, we received requests for and created codes with Components such as Acinetobacter baumannii DNA or Influenza virus RNA. More recently, we have received requests for tests that detect bacteria, viruses, and other pathogens based on a specific genetic target. Thus, LOINC now contains terms with more specific Components such as Clostridium difficile toxin A+B (tcdA+tcdB) genes and Zika virus envelope (E) gene.

As genetic testing continues to evolve, distinguishing the specific analyte (i.e. the genetic target) will be important for understanding the differences between tests, interpreting the results, and guiding further testing. Therefore, we will continue to create new LOINC codes that specify the gene targets, and ask requesters of new LOINC codes to supply this information. This specificity is important for bacterial identification, but also for other pathogens.

3.5 Antimicrobial susceptibilities

The drug susceptibility tests are grouped together in the LOINC database under the Class ABXBACT. Various agencies, including the FDA and CLSI in the U.S. and EUCAST in Europe, publish guidelines for determining susceptibility of different organisms to different antimicrobials.

In LOINC, antimicrobial susceptibility tests are named according to the generic name of the drug tested and the methodology used in testing, with Property of susceptibility (Susc), and with Scale of quantitative (Qn), semi-quantitative (SemiQn), ordinal (Ord), and quantitative or ordinal (OrdQn). Thus, appropriate names would be:

Ticarcillin+Clavulanate:Susc:Pt:Isolate:Qn:MLC
Ampicillin:Susc:Pt:Isolate:OrdQn:MIC
Ampicillin:Susc:Pt:Isolate:OrdQn:Agar diffusion

The following table lists Methods in drug-susceptibility tests.

Method Description
Agar diffusion Bacterial sensitivity via agar diffusion (Kirby-Bauer)
MIC Minimum inhibitory concentration
MLC Minimum lethal concentration
SBT Serum bactericidal titer
Gradient strip Susceptible by E-Test or gradient strip method

Methodless codes also exist for each antimicrobial agent.

3.5.1 Susceptibility thresholds based on type of infection ("Breakpoints")

In most cases, susceptibility thresholds are the same for a given organism+antibiotic combination regardless of infection site or route of antibiotic administration. However, for some antibiotic+organism combinations, the threshold (“breakpoint”) varies based on route of administration and/or type of infection, including meningitis, pneumonia, and urinary tract infection. These are two independent parameters: there are cases where one antibiotic could have multiple different breakpoints depending on the combination of route and type of infection, and others where one antibiotic will have a single breakpoint regardless of route and type of infection. For example, depending on whether or not the patient has suspected meningitis there are two different breakpoints for parenteral penicillin. But, there is a single breakpoint for vancomycin. The susceptibility test itself is carried out exactly the same way regardless of the breakpoint. It is only for the final step of assigning a susceptibility result that the specific breakpoint becomes important. These data are published by the same agencies (FDA, CLSI, EUCAST) that publish the common susceptibility thresholds.

Prior to the 2.58 LOINC release, we specified that the susceptibility threshold was based on a meningitis breakpoint in the Component, e.g., Cefepime.meningitis. However, the concept really identifies an antimicrobial susceptibility result for a patient with suspected meningitis rather than a susceptibility result for a meningitis form of the antibiotic. Therefore, as of LOINC 2.58, we moved “meningitis” from the Component to the System: Isolate.meningitis. In upcoming releases, we will also be adding codes with the Systems Isolate.pneumonia and Isolate.UTI.

Note that Isolate.meningitis (or “.pneumonia”, “.UTI”) indicates that the particular susceptibility result being reported is for an antibiotic that has a meningitis breakpoint and that the patient is suspected to have meningitis. This does not necessarily mean that a particular patient has meningitis. For the same patient with suspected meningitis, the result for an antibiotic that does not have a published meningitis breakpoint will be reported using a code with Isolate as the System.

3.5.2 Genotypic resistance testing and predicted susceptibility

Genotypic resistance testing is done to look for genes or gene mutations that confer resistance to a drug or class of drugs. The subtle difference between phenotypic susceptibility and genotypic resistance testing is that susceptibility testing is used to determine whether or not a particular organism’s growth is actually inhibited in the presence of an antimicrobial, while testing for resistance mechanisms or genes is used to determine whether or not such genes or mutations are present and to predict whether the organism will be inhibited in the presence of a particular antimicrobial.

Various Methods are available to test for the genetic information (e.g., specific resistance genes or nucleotide alterations in native genes) that encodes resistance mechanisms, including Probe.amp.tar, Non-probe.amp.tar, and Sequencing. Until 2018, LOINC bacterial genotypic susceptibility testing terms used the specific Methods named above; however, as approved by the Laboratory LOINC Committee in June 2018, moving forward we are recommending using the Molgen Method for all bacterial molecular resistance testing to obviate the need to create different terms for different molecular Methods (Probe.amp.tar, Sequencing, etc.). New terms will be created with the Method Molgen, and terms that were previously created with a Method other than Molgen will be reviewed and either updated to Molgen if not duplicative or the Status will be changed to Discouraged.

Historically, susceptibility testing was always phenotypic; it was always performed on an Isolate, which is a microorganism from a patient specimen that is grown and isolated on culture media. However, with new molecular techniques, genotypic testing can be performed directly on patient specimens without the pre-step of isolating the organism. Rather than create separate terms for every possible patient specimen that can be tested, the Laboratory LOINC Committee approved an approach in June 2018 to create single terms for each resistance gene or resistance-conferring mutation with the System Isolate/Specimen. These terms will represent assays that can be done using a traditional isolated colony or on a direct specimen. This approach does NOT imply that every manufacturer’s genotypic assay is approved for testing on an isolate or any type of specimen.

Existing terms with Isolate or other specific specimens as the System will be reviewed and either updated to the same model if not duplicative or the Status will be changed to Discouraged.

LOINC terms that represent the predicted susceptibility of an organism to a particular antimicrobial based on the presence or absence of resistance genes or resistance-conferring mutations in native genes will contain the drug name or class in the Component, Isolate/Specimen as the System, and Genotyping as the Method. This model was also approved by the Laboratory LOINC Committee in June 2018.

3.6 Cell counts

Quantitative counts of various entities and cells in blood, urine, CSF, and other body fluids may be performed and reported in one of three ways. Cell counts in blood are often reported as absolute counts per unit volume (Property number concentration, NCnc), or percent of a general cell type, e.g., percent eosinophils, (Property number fraction, NFr). Blood cells are usually reported in such a manner, via either a manual or automated count method. Counts on urine and other body fluids can also be done as direct counts and reported as NCnc or NFr. However, they are more often reported as the number of entities or cells per microscopic high power or low power field, e.g., 5-10 cells per high power field. These are really numbers per area (Property Naric). For example, the number of erythrocytes casts per low power field would be reported as:

Erythrocyte casts:Naric:Pt:Urine sed:Qn:Microscopy.light.LPF

Note that even though the values are reported as a range, the Scale is still quantitative (Qn), because the values can be related through a ratio. We use HPF or LPF to identify high power and low power fields respectively. Large entities (such as casts) are usually reported per low power fields, smaller entities per high power fields.

One other way such entities are reported is as a pure ordinal, e.g., none, few, moderate, loaded. These would be specified as arbitrary concentration (ACnc) Properties with ordinal (Ord) Scale, for example:

Erythrocytes:ACnc:Pt:Semen:Ord:Microscopy.light

3.7 Skin tests

These follow the pattern of a challenge test. For a TB skin test it would be:

Tuberculosis reaction wheal^3D post 25 TU ID:Diam:Pt:Skin:Qn

Where TU means tuberculin units, ID means intradermal, Diam indicates a measure of the diameter of the wheal and so on.

3.8 Toxicology - Drug of abuse screening and confirmation

Many kinds of test methods are used in toxicology:

Screening tests include HPLC, EIA, TLC, RIA, GC, and GCMS (rarely).

Confirmation tests are GCMS, LCMS, GC, and HPLC.

The following table indicates Drug of Abuse Methods.

Abbr. Description
HPLC high pressure liquid chromatography
TLC thin layer chromatography
GC gas chromatography
EIA enzyme immunoassay
RIA Radioimmunoassay
GCMS gas chromatography/mass spectrometry
LCMS liquid chromatography/mass spectrometry

Many laboratories use GCMS to signal that the test is a confirmation of a previous screening test, but other methods are also used to confirm, and a given method can be used to screen or to confirm a test. However, it is important that two different methods be used for screen and for confirm and that they both be applied with techniques appropriate to the mode (screen or confirm). So the LOINC committee has determined it is better to distinguish the screening from the confirming procedure by the use of the words screen or confirm, in the Method part of the name, rather than by naming a specific method. Hence LOINC will distinguish toxicology Method by Screen and Confirm but not by particular methods.

Toxicology tests can also be performed on a group of drugs/substances or on individual drugs/metabolites/substances. We will develop LOINC names and codes for both categories: groups of analytes, e.g., “barbiturates”, and individual analytes, e.g., “phenobarbital.”

Group test results are usually reported as ordinal (present /absent) but can also be reported as mass concentrations when the numerator is the total mass of the detectable substances in the group. Group tests at the screening level may also be followed by a confirmation at the group level or by confirms of the individual drug/substance tests at the confirmatory level. Individual drugs/substances may be reported as present/absent (Ord) or as mass (or substance) concentrations (Qn).

When individual drugs/substances are reported ordinally, the reporting threshold (the threshold at which a test level is considered positive) may also be reported as a separate “result.” Thus we have separate LOINC codes to report the cutoff used for defining a positive or negative value.

3.8.1 Toxicology drug groups

General principles: for each “group” of drugs (amphetamines, benzodiazepines, opiates, etc.) we will define the following kinds of LOINC observations:

3.8.1.1 Screen for a group of drugs/ toxic substances

Terms representing screening for the group as a whole:

“X”:PrThr:Pt:Ord:{System}:Screen

Example answers: present/absent

For example,

Amphetamines:PrThr:Pt:Urine:Ord:Screen

Answer: present

Terms to identify the set of drugs/substances screened for by the group test. Here the answer will be a list of discrete drug/substance names or codes:

“X” tested for:Prid:Pt:{System}:Nom:Screen

Example answers: individual drugs that this screening test could detect, from a fixed list

For example,

Amphetamines tested for:Prid:Pt:Urine:Nom:Screen

Answer: amphetamine, methamphetamine, dextroamphetamine, levoamphetamine, pseudoephedrine

3.8.1.2 Identify the drugs or substances screened for (and perhaps other information) as narrative

Terms of this format carry answers that are reported as a “glob” of narrative text. When a screen is reported as negative, confirmatory testing is not performed. When a screening test is reported as positive, the result must be confirmed by an independent testing method.

“X” tested for:Prid:Pt:{System}:Nar:Screen

Example answers: individual drugs that this screening test could detect, reported as a “blob” of text or canned comment.

For example,

Amphetamines tested for:Prid:Pt:Urine:Nar:Screen

Answer: The EMIT urine screen for amphetamines detects amphetamine, methamphetamine, dextroamphetamine, levoamphetamine as indications of methamphetamine abuse. It is also reactive with a component present in over-the-counter nasal decongestant inhalers, and a positive result must be confirmed by a quantitative method that rules out the non-abuse situation.

3.8.1.3 Confirmatory testing for the presence of one or more members of the group represented as a single observation

“X”:PrThr:Pt:{System}:Ord:Confirm

Example answers: present/absent

For example,

Amphetamines:PrThr:Pt:Urine:Ord:Confirm

Answer: present

3.8.1.4 List of the actual drug/substances confirmed

“X” positive:Prid:Pt:{System}:Nom:Confirm

Example answers: a list of analytes detected

For example,

Amphetamines positive:Prid:Pt:Urine:Nom:Confirm

Answer: dextroamphetamine, methamphetamine

3.8.1.5 Confirmatory testing reported as a set of observations (more common)

More commonly, confirmatory testing is reported as a set of observations that each report the presence (or quantitative amount detected) of an analyte in the group.

“X”:PrThr:Pt:{System}:Ord:Confirm

Example answers: present/absent

or

“X”:MCnc:Pt:{System}:Qn:Confirm

Example answers: a quantitative amount

For example,

Amphetamine:PrThr:Pt:Urine:Ord:Confirm
Dextroamphetamine:PrThr:Pt:Urine:Ord:Confirm
Methamphetamine:PrThr:Pt:Urine:Ord:Confirm
Levomethamphetamine:PrThr:Pt:Urine:Ord:Confirm

Example answers: present/absent

3.8.2 Cutoffs

The cutoff levels for screens and confirms of a given substance or group of substances will usually differ. There are three ways to indicate specific cutoffs in LOINC.

3.8.2.1 We provide separate LOINC terms for reporting the cutoff levels of a number of commonly abused substances and substance groups

“X” cutoff:MCnc:Pt:Urine:Qn:Screen
“X” cutoff:MCnc:Pt:Urine:Qn:Confirm

For example,

Amphetamines cutoff:MCnc:Pt:Urine:Qn:Screen

Answer: 1000 ng/ml

Methamphetamine cutoff:MCnc:Pt:Urine:Qn:Confirm

Answer: 500 ng/ml

3.8.2.2 Two general cutoff terms, one for screen and one for confirm, can be applied to any substance whether or not a pre-coordinated term exists

"X" cutoff:MCnc:Pt:{System}:Qn:Screen
"X" cutoff:MCnc:Pt:{System}:Qn:Confirm

3.8.2.3 For commonly used cutoffs, such as those mandated by regulatory agencies, we provide pre-coordinated terms for reporting a “present/absent” result with the cutoff specified in the Method field

“X”:PrThr:Pt:{System}:Ord:Screen>“N”
“X”:PrThr:Pt:{System}:Ord:Confirm>“N”

For example,

Amphetamines:PrThr:Pt:Urine:Ord:Screen>1000 ng/mL

Answer: not detected

3.8.3 Reporting the Method used for screen and confirm

We provide terms for reporting the Method used for screen and confirm tests:

“X” screen method:Prid:Pt:{System}:Nom:*
“X” confirm method:Prid:Pt:{System}:Nom:*

These would normally be reported in conjunction with terms reporting levels and possibly cutoffs, as in the following example:

Amphetamines:PrThr:Pt:Urine:Ord:Confirm

Answer: positive

Amphetamines cutoff:MCnc:Pt:Urine:Qn:Screen

Answer: 1000 ng/ml

Amphetamines screen method:Prid:Pt:Urine:Nom:*

Answer: EIA

Amphetamines positive:Prid:Pt:Urine:Nom:Confirm

Answer: amphetamine, methamphetamine

Amphetamine cutoff:MCnc:Pt:Urine:Qn:Confirm

Answer: 500 ng/ml

Methamphetamine cutoff:MCnc:Pt:Urine:Qn:Confirm

Answer: 500 ng/ml

Amphetamines confirm method:Prid:Pt:Urine:Nom:*

Answer: GC/MS

3.8.4 Individual drug/metabolite test results

Individual substances are typically reported as screens (ordinal), confirms (ordinal) or confirms (quantitative – usually mass or substance concentrations), though some labs also report as screen (quantitative). Ordinal terms for reporting individual substances should always specify a Method of screen or confirm.

Group test screens may be confirmed by group confirms (as described above) or by individual confirms (Either ordinal or quantitative-depending upon the laboratory's preference).

3.8.4.1 Individual test screen (ordinal)

Methamphetamine:PrThr:Pt:Urine:Ord:Screen

Example answer: present

3.8.4.2 Individual test screen (quantitative)

Methadone:MCnc:Pt:Urine:Qn:Screen

Example answer: 150 ng/ml

3.8.4.3 Individual test confirm (ordinal)

Methamphetamine:PrThr:Pt:Urine:Ord:Confirm

Example answer: present

3.8.4.4 Individual test confirm (quantitative)

Methamphetamine:MCnc:Pt:Urine:Qn:Confirm

Example answer: 250 ng/ml

Individual tests may also be reported as simple quantitative (without confirm or screen), as is the case for therapeutic drug level monitoring.

3.8.4.5 Individual substance measured quantitatively; screen/confirm is not relevant

Digoxin:MCnc:Pt:Ser/Plas:Qn

Example answer: 1.2 ng/ml

3.8.5 Naming issues

For confirms, you would always be looking for specific analytes. For example, you would never look for tetrahydrocannabinol, but would look for delta-9-tetrahydrocannabinol, 11-hydroxycannabinol, etc.

3.8.6 Summary

For each “group” LOINC defines the following set of terms:

“Analyte group”:PrThr:Pt:Urine:Ord:Screen
“Analyte group”:PrThr:Pt:Urine:Ord:Confirm
“Analyte group”:MCnc:Pt:Urine:Qn:Confirm
“Analyte group” tested for:Prid:Pt:Urine:Nom:Screen
“Analyte group” tested for:Prid:Pt:Urine:Nar:Screen
“Analyte group” positive:Prid:Pt:Urine:Nom:Confirm
“Analyte group” screen method:Prid:Pt:Urine:Nom:*
“Analyte group” confirm method:Prid:Pt:Urine:Nom:*

For each individual analyte LOINC now defines the following set of terms:

Analyte:PrThr:Pt:Urine:Ord:Screen
Analyte:PrThr:Pt:Urine:Ord:Confirm
Analyte:MCnc:Pt:Urine:Qn:Confirm
Analyte:MCnc:Pt:Urine:Qn
Analyte cutoff:MCnc:Pt:Urine:Qn:Screen
Analyte cutoff:MCnc:Pt:Urine:Qn:Confirm

3.9 Molecular genetics LOINC naming

3.9.1 Introduction

Molecular pathology testing can be used for many purposes. In infectious disease testing to identify organisms and mutations in organisms; in genetic analysis to identify mutations including substitutions, deletions/insertions, frame shifts and trinucleotide repeats; to identify specific chromosomal translocation and clonality in leukemia and lymphomas; to identify various tumor associated genes and gene deletions; in paternity testing to determine the probability that a person is the parent of a child; and in forensic testing to determine the probability that a criminal is associated with genetic material he/she left as evidence.2

LOINC term names follow the recommendations and nomenclature of other standards. As molecular genetic testing expands, we are evolving our naming conventions. Existing terms are being reviewed for compatibility with current naming conventions. An overall description of LOINC’s approach to naming molecular genetics tests can be found in the Deckard et al3 paper.

3.9.2 Brief review of molecular genetics terminology

For clarity, here we provide a brief introduction to some key genetics terminology. DNA consists of a series of nucleotides, nucleotides encode amino acids, and a string of amino acids forms a protein. There are four types of nucleotides (adenine, guanine, cytosine and thymine), and a sequence of three nucleotides that codes for one amino acid is called a codon. Codons are numbered from the first codon participating in the protein (in humans the codon for Methionine) starting with codon number 1. Locus refers to a specific DNA (or RNA) codon or the corresponding amino acid produced by this codon.

The string of DNA that codes for a protein is usually interrupted by DNA segments called introns, which do not contribute to the protein definition. The coding sequences of DNA between the introns are called exons. Linked together, the exons provide the instructions for creating the specific protein. Exons may be numbered e.g., exon 1, exon 2, etc. Exon numbers sometimes appear in the names of DNA mutations, but for a number of reasons, identifying codon locations relative to an exon is unreliable and we will try to avoid such nomenclature when possible in LOINC names.

The term "mutation" is usually applied to a genetic variant that causes a functional change in the gene and results in disease. Genetic changes that occur during the life of the patient such as tumor mutation are called somatic and those that are inherited are referred to as germ line. The nature of the specimen and the testing usually distinguishes these two, so it is not necessary to include this distinction in the test names.

Alleles refer to different forms of a gene and are distinguished at the phenotype level. The term allele is usually applied to a genetic variant that does not cause a disease.

3.9.3 Background on molecular genetics testing methods

The main laboratory methods used are Southern Blot, which applies hybridization to selected DNA “chopped up” by restriction enzymes, Northern Blot, which applies hybridization to all cellular RNA (which comes naturally in smaller segments), and Restriction Fragment Length Polymorphism (RFLP). RFLP depends on the Variable Number of Tandem Repeats (VNTR), which are normal, but specific variants of each person’s DNA. Southern Blot may be combined with RFLP to target mutations whose exact gene molecular chemistry is not known. For completeness sake, we mention Western Blot, which applies an analogous blot method to protein analysis.

In situ hybridization is a method that applies probes to intact tissue. The cellular patterns of the homologies can then be read microscopically. There are a variety of methods for detecting such in situ probes. One popular method is Fluorescent In-Situ Hybridization (FISH). This technique is analogous to an immune stain except that the molecular binding is based on DNA/RNA homologies instead of antigen-antibody binding.

DNA chips provide a radical new way to identify DNA and RNA sequences. In the patented AFYMETRIX® technique, the nucleoside chains are grown using lithography-like methods. Target DNA is tagged with a detector and “washed” over the chip in steps. The locations of the tags on the chip identify the DNA (RNA) in the sample.

Identity testing is used to identify relationships among people and has special complexity. In paternity testing, it can be helpful to have DNA from the child, the putative father and the mother when possible to distinguish the alleles that come from the father.

Forensic testing has special requirements of stringency and often mixes blood antigen testing with RFLP testing. The results are usually reported as a probability.

3.9.4 General molecular genetics naming rules

LOINC’s approach to naming tests makes use of established conventions. For naming genetic tests that target specific genetic variations, LOINC uses the Human Genome Organization (HUGO) Gene Nomenclature Committee's (HGNC) terminology to name the gene(s) and Human Genome Variation Society's (HGVS) syntax to name the variation(s) of interest.

HGVS provides seven types of prefixes to identify the difference types of reference sequences used in naming gene defects as described in the following table. As of LOINC release 2.66, LOINC has terms for the first four (p., c., g., m.).

Designation Explanation
p Protein reference sequence based on amino acid (codon) counting.
c Coding reference sequence based on a transcript that encodes a protein and therefore excludes introns. Since three nucleotides = one codon, these will produce numbers 3x as large as those based on the protein.
g Genomic reference sequence based on counting from the first nucleotide in the DNA for the gene of interest, as it exists natively in the chromosome with introns included.
m Mitochondrial reference sequence based on counting from the first nucleotide in the full mitochondrial DNA sequence.
n Non-coding reference sequence based on a transcript that does not encode a protein.
o Circular genomic reference sequence, for example, for plasmids or certain viruses.
r RNA reference sequence.

The HGVS recommendations have evolved over time. In our early approach, we used the extant protein-based naming convention to get the following name for the commonest mutation causing cystic fibrosis:

CFTR gene.p.F508del:PrThr:Pt:Bld/Tiss:Ord:Molgen

For clarity’s sake, HGVS now prefers the three letter amino acid abbreviations over the single letter amino acid abbreviations we used in the above example.

We embrace the HGVS naming style for naming mutations at the amino acid level, but the world has not been consistently quick to adopt it. The old style, perhaps because of its brevity, persists. As we update existing codes to harmonize with current HGVS guidelines, we will make it easy for users to find our test names by including the old style as synonyms for the new style. Under the current HGVS rules, the above name would be:

CFTR gene.p.Phe508del:PrThr:Pt:Bld/Tiss:Ord:Molgen

Where we still report variants at the amino acid level, we use the nomenclature for human gene mutations proposed by Beaudet4 in the Component or in the LOINC Answer when the mutation is reported as an answer. A list of single and three letter amino acid codes are given in the following table.

Amino Acid 1-Letter Symbol 3-Letter Symbol
Alanine A Ala
Arginine R Arg
Asparagine N Asn
Aspartic acid D Asp
Cysteine C Cys
Glutamic acid E Glu
Glutamine Q Gln
Pyroglutamic acid pQ pGlu
Glycine G Gly
Histidine H His
Hydroxyproline, 4(R)-L- O Hyp
Isoleucine I Ile
Leucine L Leu
Lysine K Lys
Methionine M Met
Phenylalanine F Phe
Proline P Pro
Serine S Ser
Threonine T Thr
Tryptophan W Trp
Tyrosine Y Tyr
Valine V Val
Unknown X

HGVS has made other changes in naming style. Overall they now prefer describing the variant at the coding DNA level as nucleotide changes rather than at the protein level as amino acid changes. The above cystic fibrosis mutation would have the following coding DNA-based name:

CFTR gene.c.1521_1523del:PrThr:Pt:Bld/Tiss:Ord:Molgen

For the nucleotide changes, they have adjusted the syntax for representing a change from one nucleotide string to another as # String1>String2 rather than String1 # String2, where # is an integer representing the nucleotide position of the variation.

Another common cause of cystic fibrosis is a mutation that at the amino acid level would be named p.Gly551Asp with the new protein based naming convention, c.G1652A with the old DNA naming convention and c.1652G>A with the new coding DNA naming convention. (Realize, of course, that more than one nucleotide-level variation name can correspond to a single amino acid variation name.)

If clear guidelines are not in place for a given variant when a LOINC code is created, the familiar or common variant name may be used. LOINC will resort to using the disease name only when the gene has no name and/or the genetic defect is not yet fully specified. We will include the genetic disease name in the RELATEDNAMES2 field of the database when the disease name is not part of the Component. This allows users to easily find the LOINC term by the disease name as well.

3.9.4.1 Specimens in molecular genetics

Blood is the most common specimen for molecular pathology studies. Leukocytes, bone marrow, tumors, products of conception and forensic specimens also contain DNA and are important specimens. The System (specimen) used in the LOINC name for genetic testing will usually be Bld/Tiss since the distinction between these two specimens is rarely important to the result of a molecular pathology test. We have split this further to accommodate fetal specimens (Tiss^fetus). Other specimens include amniotic fluid, CVS, bone marrow, fixed tissue, and CSF.

3.9.4.2 Methods in molecular genetics

In general, we do not create separate variables for each kind of molecular genetics method (e.g., Southern Blot Northern Blot, PCR, restriction fragment length polymorphism (RFLP)) because the different methods do not provide significantly different results. Moreover, there are a plethora of minor method variants, and we would not be able to keep up with their proliferation. Therefore, instead we use a generic Method of Molgen (for molecular genetics method) to indicate that a result of the analysis is based on a molecular genetics method rather than some chemical or antigen method.

However, when results for a given molecular pathology technique are significantly different, we will distinguish the method. Examples of specific molecular pathology Methods in LOINC include fluorescent in situ hybridization (FISH), sequencing, multiplex ligation probe amplification (MLPA), and microarray comparative genomic hybridization (Microarray). For the LOINC 2.68 release, we added a new Method for Mate pair sequencing. It is unclear at this time whether this new Method produces results that are significantly different from older sequencing methods, and we may update the Method for these terms in the future based on new information.

3.9.4.3 Narrative and document-level reporting in molecular genetics

Bulk narrative reports in molecular pathology are often used to provide results for mutation analyses, without reporting any discrete coded answers. We discourage the use of this approach because it is not useful for automated analyses. Since these kinds of results often include some structured elements, we will use the Property Find and Scale Doc beginning with version 2.54. The Scale of Doc represents a collection of information (e.g. results contained in a report) that are either structured or unstructured. To assist with structured reporting, we will attach to these terms specific associated observations that could be used to send data in a structured format. With this approach, labs will still have the flexibility to send text reports. However, we encourage them to also report key data as structured results using the associated observation variables so that it is more easily understood by computers. We plan to use this model for newly created terms for molecular pathology tests and are contemplating the best approach for transitioning existing terms to this format over time.

3.9.4.4 Inferred phenotype and genotype

Beginning in version 2.66, we have LOINC terms for genotype and phenotype results that are derived from their respective phenotype or genotype analysis. For such concepts we include the word "inferred" in the Component. One example is a blood banking panel that is used to report whether or not an RBC antigen is expected to be present in a patient or donor sample based on molecular analysis of the corresponding gene, such as the presence of K Ag based on analysis of the KEL gene, for which the Component is K Ag inferred phenotype. The reverse example is an assay that reports the genotype for the APOL1 gene based on analysis of apolipoprotein L1 isoforms; for this term, the Component is APOL1 inferred genotype.

3.9.5 Infectious diseases

For information about molecular genetics in the field of infectious diseases, see Section 3.4.3 - Microorganism identification based on nucleic acid targets.

3.9.6 Genetic conditions

3.9.6.1 Diagnostic assays for mutation analysis

We currently have at least four different styles for mutation analyses in LOINC:

  1. Single mutation analysis: Diagnostic assays for the detection of a specific mutation or variant
  2. Targeted mutation analysis: Diagnostic assays for a fixed set of the most common or important mutations
  3. Known mutation analysis: Diagnostic assays for one or more mutations that have been previously identified in an affected family member
  4. Full gene mutation analysis: Diagnostic assays usually done by sequencing of the entire coding region of a gene

3.9.6.1.1 Diagnostic assays for the detection of a specific mutation or variant

When looking for one mutation, use LOINC’s single mutation style of gene name followed by the specification of the nomenclature (p, g, c, or m) and the mutation name. A dot will separate the gene name and the mutation identifier:

<gene name> gene.<mutation nomenclature>.<mutation and its location>

For example, Factor V Leiden mutation would be represented as F5 gene.p.R506Q. Where “F5” identifies the gene, “gene” is a fixed part, “p” identifies the kind of mutation nomenclature (protein) and “R506Q” indicates that the amino acid arginine (R) is replaced by glutamine (Q) (see Table 19) at codon #506.

Some examples of fully specified LOINC names for tests of specific mutation are:

F5 gene.p.R506Q:PrThr:Pt:Bld/Tiss:Ord:Molgen

Synonyms: Factor V Leiden, Factor V resistance, APC resistance gene

HFE gene.p.C282Y:PrThr:Pt:Bld/Tiss:Ord:Molgen

Synonyms: HLA-H gene, hemochromatosis gene

CFTR gene.p.F508 del:PrThr:Pt:Bld/Tiss:Ord:Molgen

Synonyms: Cystic Fibrosis Transmembrane Regulator

As mentioned above, HGVS now recommends the three letter amino acid abbreviations over the single letter amino acid abbreviations. As existing codes are updated, we will add synonyms for the old style to make it easy for users to find our test names.

One type of testing procedure identifies a single mutation using two DNA probes: one for the normal locus and the other for the abnormal locus. When only the normal probe reacts, the laboratory reports “no mutation” or “wild type”. When both the normal and mutation probes react, the laboratory reports “heterozygous”. When only the mutation probe reacts it reports “homozygous”. Consequently, such single mutation testing produces one of three ordinal “answers”:

  1. no mutation (wild type)
  2. heterozygous mutation (the mutation found in one gene)
  3. homozygous mutation (the mutation was found in both genes in the gene pair)

Specific testing such as this is only possible when the molecular pathology of the gene is very well known and only one defect is being reported.

Both document-level (Scale is Doc) and ordinal (Scale is Ord) codes are available in the LOINC database for single variant tests. In cases where more than one variant is tested, a nominal (Nom) Scale will be used. As mentioned above, we encourage the use of document-level codes for the order and overall result report, which can be used in conjunction with additional LOINC codes for reporting the discrete results.

Results for specific mutations or variants may also be reported as two separate observations: one observation reports the kind of mutation (allele) found in the first chromosome and another for reporting the kind of mutation for the paired chromosome. In this case, the identity of the allele is reported in the answer. For example:

APOE gene allele 1:Prid:Pt:Bld/Tiss:Nom:Molgen

Answers: E2, E3, or E4

APOE gene allele 2:Prid:Pt:Bld/Tiss:Nom:Molgen

Answers: E2, E3, or E4

3.9.6.1.2 Targeted mutation analysis

LOINC’s approach to represent gene mutation analysis for many genetic variations within one or more genes has had the following form:

<gene name> gene targeted mutation analysis:Find:Pt:Bld/Tiss:Doc:Molgen

For example:

CFTR gene targeted mutation analysis:Find:Pt:Bld/Tiss:Doc:Molgen

Synonyms: Cystic fibrosis transmembrane regulator

BRCA1 gene targeted mutation analysis:Find:Pt:Bld/Tiss:Doc:Molgen

Synonyms: breast cancer risk gene

For each such targeted mutation analysis, we recommend reporting the mutations that were tested using an additional LOINC code, such as LOINC 36908-2 (Gene mutations tested for) or gene-specific companion LOINC observation codes with the words “mutations tested for”:

<Gene name or disease name> gene mutations tested for:Prid:Pt:Bld/Tiss:Nom:Molgen

For example:

CFTR gene mutations tested for:Prid:Pt:Bld/Tiss:Nom:Molgen

Example answers: “Delta F508”, “G542X”, “R553X”, “W1282X”, “N1303K”, etc.

The additional LOINC code is needed for reporting the mutation(s) that could have been found in a given analysis so that clinicians can know what was looked for when no abnormalities were found.

The above gene mutation analysis terms could be used for ordering or reporting the results of a given targeted mutation analysis. Both document-level (Scale is Doc) and nominal (Scale is Nom) codes are available in the LOINC database. As mentioned above, we encourage the use of document-level codes for the order and overall result report, which can be used in conjunction with additional LOINC codes for reporting the discrete results. Nominal results for such an analyses could be: 1) no pathologic mutations found or 2) a list of individual mutations/variations found. When reporting discrete mutations as results, we propose using the HGVS nomenclature and including (in parentheses) the historic versions of the mutation names when such names exist.

3.9.6.1.3 Known mutation analysis

The first two approaches are commonly called “targeted” mutation analyses, or looking for specific mutations within a given gene. Labs may offer testing for both targeted mutation analysis and analysis for known familial mutations. Known familial mutations are those previously identified within an affected family member. To distinguish these testing approaches, LOINC further defines testing for known familial mutations:

<Gene name> gene mutation analysis limited to known familial mutations:Find:Pt:Bld/Tiss:Doc:Molgen

For example:

TNFRSF13B gene mutation analysis limited to known familial mutations:Find:Pt:Bld/Tiss:Doc:Molgen

Since testing may involve more than one known mutation, especially for autosomal recessive conditions, LOINC uses the plural form of mutations in the Component. However, these terms also refer to testing for only one known familial mutation.

3.9.6.1.4 Full gene mutation analysis

To describe mutation analysis by sequencing of the entire coding (and possibly non-coding) region of the gene, we use:

TNFRSF13B gene full mutation analysis:Find:Pt:Bld/Tiss:Doc:Sequencing

Here, we include the Method of sequencing since this is a technique commonly used to identify mutations with the entire coding region of a gene.

3.9.6.2 Diagnostic assays for large deletions and/or duplications

Detection of larger (>50 bp) genomic duplications or deletions is done by various techniques, including multiplex ligation-dependent probe amplifications (MLPA) and array-based comparative genomic hybridization (arrCGH). MLPA and aarCGH techniques detect gene dosage. To describe testing for large deletions and/or duplications (insertions) within a gene, we use:

LDLR gene deletion+duplication:Find:Pt:Bld/Tiss:Doc:MLPA

In this case, we specify the Method of multiplex ligation-dependent probe amplification (MLPA). MLPA is a common technique used to detect gene dosage of genomic deletions and duplications (e.g. one or more entire exons) and determine gene copy number.

3.9.6.3 Trinucleotide repeats

A number of diseases, most of which manifest as neurologic disorders are caused by excessive repeats of specific trinucleotides, and the age of onset of the disease is inversely proportional to the number of excess repeats. Examples of these disorders include:

  • Fragile X syndrome
  • Huntington disease
  • Spinocerebellar ataxia (SCA1)

We name the Component of these terms by the gene when the gene is well defined or the disease, and the name of the trinucleotide that repeats plus the word repeats.

<disease or gene name>.<trinucleotide> repeats

For example, Huntington disease would be represented as follows:

HTT gene.CAG repeats

Examples of some fully specified LOINC names are:

FRAXE gene.CGG repeats:PrThr:Pt:Bld/Tiss:Ord:Molgen

Synonym: Fragile x syndrome

HTT gene.CAG repeats:PrThr:Pt:Bld/Tiss:Ord:Molgen

Synonyms: Huntington Disease, It15, HD, Huntington Chorea

Spinocerebellar ataxia genes.CAG repeats:PrThr:Pt:Bld/Tiss:Ord:Molgen
DMPK gene.CTG repeats:PrThr:Pt:Bld/Tiss:Ord:Molgen

Synonym: Myotonic Dystrophy

These are usually reported “not expanded”, “indeterminate” or “expanded”, so the Scale is Ord.

When the actual number of trinucleotide repeats is reported, the Property is entitic number (EntNum) and the Scale is quantitative (Qn), and separate results are reported for each allele. Examples include:

HTT gene allele 1.CAG repeats:EntNum:Pt:Bld/Tiss:Qn:Molgen
HTT gene allele 2.CAG repeats:EntNum:Pt:Bld/Tiss:Qn:Molgen
DMPK gene allele 1.CTG repeats:EntNum:Pt:Bld/Tiss:Qn:Molgen
DMPK gene allele 2.CTG repeats:EntNum:Pt:Bld/Tiss:Qn:Molgen

3.9.6.4 Hematopathology gene re-arrangement

Immune cells have an innate genetic variability due to rearrangement. The unique rearrangement can be used to identify the development of a clone of one cell type as occurs in many lymph cell tumors (e.g., lymphoma). We use the following format to identify clonal excess.

Immunoglobulin heavy chain gene rearrangements:Arb:Pt:Bld/Tiss:Ord:Molgen
Immunoglobulin kappa light chain gene rearrangements:Arb:Pt:Bld/Tiss:Ord:Molgen
Immunoglobulin lambda light chain gene rearrangements:Arb:Pt:Bld/Tiss:Ord:Molgen
TCRB gene rearrangements:Arb:Pt:Bld/Tiss:Ord:Molgen

Synonym: T cell receptor beta chain

TCRD gene rearrangements:Arb:Pt:Bld/Tiss:Ord:Molgen

Synonym: T cell receptor delta chain

TCRG gene rearrangements:Arb:Pt:Bld/Tiss:Ord:Molgen

Synonym: T cell receptor gamma chain

These would be reported as “clonal” or “not clonal”.

3.9.6.5 Chromosomal alterations: translocations, deletions, and inversions

LOINC nomenclature follows the International System for Human Cytogenetic Nomenclature (ISCN) guidelines when describing chromosomal alterations, including translocations, large deletions and inversions. Tests to detect fused genes or transcripts (RNA, cDNA) due to a chromosomal alteration are designated as follows:

t(<Chromosome of breakpoint gene 1>;<Chromosome of breakpoint gene 2>)(<Location on chromosome 1>; <Location on chromosome 2)(<gene1>,<gene2>) fusion transcript

For example:

t(9;22)(q34.1;q11)(ABL1,BCR) fusion transcript:Arb:Pt:Bld/Tiss:Ord:Molgen

Synonyms: Philadelphia chromosome, BCR1, chronic myeloid leukemia, CML

t(14;18)(q32;q21.3)(IGH,BCL2) fusion transcript:Arb:Pt:Bld/Tiss:Ord:Molgen

Synonyms: Follicular B cell lymphoma, oncogene B-cell leukemia 2, CLL, chronic lymphatic leukemia, follicular lymphoma

t(15;17)(q24.1;q21.1)(PML,RARA) fusion transcript:Arb:Pt:Bld/Tiss:Ord:Molgen

Synonyms: RAR, promyelocytic leukemia, myelogenous, retinoic acid receptor, acute promyelocytic leukemia, APL

For chromosomal deletions and inversion,‘t’ above would be replace with ‘del’ or ‘inv’, respectively:

del(1)(p32p32)(STIL,TAL1) fusion transcript:PrThr:Pt:Bld/Tiss:Ord:Molgen

Synonyms: SCL/TAL1 interrupting locus, T-cell acute lymphoblastic leukemia, ALL, T-ALL, TAL-1 deletions

inv(16)(p13.1;q22.1)(MYH11,CBFB) fusion transcript:Arb:Pt:Bld/Tiss:Ord:Molgen

Synonyms: inversion 16, AML, Acute myeloid leukemia

In some cases, testing may involve comparing the fused transcript to a control transcript and results may be reported as a number ratio (NRto), log number ratio (LnRto), or relative ratio (RelRto):

t(9;22)(q34.1;q11)(ABL1,BCR) fusion transcript/control transcript:NRto:Pt:Bld/Tiss:Ord:Molgen
t(9;22)(q34.1;q11)(ABL1,BCR) fusion transcript/control transcript:LnRto:Pt:Bld/Tiss:Ord:Molgen
t(9;22)(q34.1;q11)(ABL1,BCR) b3a2 fusion transcript/control transcript (International Scale):RelRto:Pt:Bld/Tiss:Ord:Molgen

To specify “major” or “minor” breakpoints, we use:

t(9;22)(q34.1;q11)(ABL1,BCR) fusion transcript major break points:Arb:Pt:Bld/Tiss:Ord:Molgen
t(9;22)(q34.1;q11)(ABL1,BCR) fusion transcript minor break points:Arb:Pt:Bld/Tiss:Ord:Molgen

To specify specific breakpoints, we use:

t(9;22)(q34.1;q11)(ABL1,BCR) b2a2+b3a2 fusion transcript:Arb:Pt:Bld/Tiss:Ord:Molgen

Synonyms: major breakpoints, p210, e13a2, e14a2

t(15;17)(q24.1;q21.1)(PML,RARA) bcr1 fusion transcript:Arb:Pt:Bld/Tiss:Ord:Molgen

Synonyms: breakpoint cluster region 1, long form

Translocation terms can also be expressed as a fraction of cells that have the rearrangement versus total cells of interest:

Cells.t(9;22)(q34.1;q11)(ABL1,BCR)/Cells.total:NFr:Pt:Bld/Tiss:Qn:Molgen

If specific partner genes are not known, we use the pattern:

CCND1 gene rearrangements:Arb:Pt:Bld/Tiss:Ord:Molgen

Synonym: Lymphoma 1

BCL2 gene rearrangements:Arb:Pt:Bld/Tiss:Ord:Molgen

Synonym: Lymphoma 2

3.9.7 Identity testing

The identity testers usually look at 4 genetic loci (each locus is polymorphic enough that any one match has a 10% error of being incorrect). The loci are independent so if all 4 probes match (including all exclusions and inclusions) the probability of an erroneously match is .0001 (one out of 10,000). They may use more than four depending upon the degree of confidence required by the circumstances of the testing. The forensic community chooses from a set of about 20 probes.

We propose two styles for reporting identity testing: atomic and pre-coordinated definitions.

3.9.7.1 Atomic style

This style uses a series of LOINC names to report the kind of index case, the kind of comparison case, the results of the identity testing, and all of the other separate components of the testing. It includes an observation for reporting the actual probes used, and another observation for reporting the population that the probes assume. The Method will be MOLGEN.IDENTITY.TESTING. For example:

DNA probes used:Prid:Pt:Index case^comparison case:Nom:Molgen.identity.testing
Population base:Prid:Pt:Probes:Nom:Molgen.identity.testing
Relationship:Type:Pt:index case:Nom:Molgen.identity.testing

Example Answers: child, victim, suspect

Relationship:Type:Pt:^comparison case:Nom:Molgen.identity.testing

Example Answers: mother, alleged mother, father, alleged father, evidence (external to victim)

Confidence of relationship:likelihood:Pt:Index case^comparison case:QN:Molgen.identity.testing

Comment: gives the statistical confidence in the conclusion

Conclusion:Imp:Pt:index case^comparison case:Nar:Molgen.identity.testing

Comment: gives a summary statement of the conclusion about identity of relatedness

3.9.7.2 Pre-coordinated definitions alternative

Some of the above atomic terms (e.g., DNA probes used) could also be reported with the pre-coordinated results.

Relationship:likelihood:child^alleged mother:Qn:Molgen.identity.testing

Synonym: maternity testing
Comment: gives the likelihood that the alleged mother is the mother of the index child

Relationship:likelihood:child^alleged father:Qn:Molgen.identity.testing

Synonym: paternity testing
Comment: gives the likelihood that the alleged father is the father of the index child

Relationship:likelihood:victim^suspect:Qn:Molgen.identity.testing

Comment: gives the likelihood that the genetic material on the victim is that of the suspect

Relationship:likelihood:suspect^victim:Qn:Molgen.identity.testing

Comment: gives the likelihood that the genetic material on the suspect is that of the victim

Identity:likelihood:evidence^suspect:Qn:Molgen.identity.testing

Comment: gives the likelihood that the genetic material on the evidence is that of the suspect

Identity:likelihood:evidence^victim:Qn:Molgen.identity.testing

Comment: gives the likelihood that the genetic material on the evidence is that of the victim

Using the models described in this section, LOINC has terms for many genetic tests relevant to cancer, including tests related to tumor suppressor genes (e.g. BRCA1, BRCA2, and p53), oncogenes (e.g. HER2), and more.

3.10 HLA allele and antigen nomenclature

Naming conventions for HLA alleles and antigens are specified by the WHO Nomenclature Committee for Factors of the HLA System. Recommendations for both allele and antigen naming have evolved since the original HLA terms were represented in LOINC. For the LOINC 2.54 release (December 2015), we updated nearly all of the antigen and allele terms to reflect the WHO Nomenclature Committee’s latest recommendations. HLA allele and antigen naming conventions are briefly described below; for more details, see http://hla.alleles.org/.

Allele names begin with the “HLA-” prefix and gene name, followed by an asterisk (e.g. “HLA-A* or HLA-DR*”) and up to four sets of numbers separated by colons. The first set of digits following the asterisk describes the allele type, which may correspond to the serological antigen number (e.g., “HLA-A*01 corresponds to the antigen HLA-A1). The second set of digits is the subtype number, which is assigned in the order in which the subtype was identified and sequenced. The third and fourth sets of digits describe the specific sequence variations of the subtype, if applicable.

Antigen names also begin with the “HLA-“ prefix and specific gene, but antigen names are distinguished from allele names in that the gene name is not followed by an asterisk. For HLA-A, -B, -DQ and -DR, the antigen number immediately follows the gene name, and HLA-C, -D and -DP have a lower case w between the gene name and the antigen number (e.g., HLA-Cw3). The lower case w is used to distinguish the C antigens from complement factors and the D and DP antigens as being identified by cellular techniques rather than serological ones. Most antigen names have one set of digits (the antigen number) following the gene name, but in some cases, if an antigen originally identified as a single antigen was later recognized as being two distinct entities, the original, broader antigen number will follow the specific number in parentheses. For example, HLA-B38(16) and HLA-B39(16).

3.11 Allergy testing

The allergy testing industry provides tests for more than 450 different allergens today. Most testing detects IgE antibodies against these allergens. For some allergens testing for IgG and IgA antibodies are available, as well.

For LOINC terms that represent allergen testing, the Component is the allergen name plus the type of the antibody (mostly IgE). Most allergens relate to animals, plants or derivatives of such entities. In the past (prior to LOINC version 2.04), we used the common name, rather than the scientific name to identify the allergen. However, this approach led to some duplicate term definitions, because two different companies would name the same allergen differently. It also led to ambiguity because two different species of animal or plant would sometimes have the same common name. As of version 2.04, we corrected these problems. To help reduce the ambiguity we now use the Latin name of the species of the biologic entity that causes the allergy.

Some background: First, most allergens can also be identified with a special 2-5 character code assigned by Pharmacia5 that most allergy testing companies reference in their catalog of testing. We used these codes to identify duplicate and ambiguous LOINC allergy test terms. These Pharmacia codes are also included in the related names field of the database. Second, allergen tests are often reported in two styles: a quantitative raw measure and an ordinal (0-6) severity rank (RAST class). LOINC defines separate terms for each of these reporting styles. For example, the two LOINC codes for reporting IgE antibodies to Japanese Millet are:

Echinochloa crus-galli Ab.IgE:ACnc:Pt:Ser:Qn
Echinochloa crus-galli Ab.IgE.RAST class:ACnc:Pt:Ser:Ord

The RAST class is a categorization of the raw measurement based on specific allergy criteria. The specific IgE class result values (0, 1, 2, 3, 4, 5, or 6) are an ordered categorical response rather than a continuous numeric scale, therefore “RAST class” terms have an ordinal (ORD) Scale.

Laboratories also test mixtures of allergens to produce one result. An example of how these tests are represented is shown for LOINC 23797-4:

(Acer negundo+Quercus alba+Ulmus americana+Populus deltoides+Carya pecan) Ab.IgE:ACnc:Pt:Ser:Ord:Multidisk

Related name: tx2

There may be more than one type of allergen for each plant. For instance, IgE antibodies can develop towards tree pollen and the fruit of the same tree. Similarly, antibodies exist for grain and for grain pollen. In these cases, the LOINC Component will contain the word POLLEN to distinguish the pollen allergen from the food allergen. For example, the LOINC term for corn (maize) IgE antibody would be:

Zea mays Ab.IgE:ACnc:Pt:Ser:Qn

Related names: f8; cultivated corn; maize

Zea mays pollen Ab.IgE:ACnc:Pt:Ser:Qn

Related names: g202: cultivated corn; maize

3.11.1 Naming rules for allergens

The Component (Analyte) for an allergen consists of the name of the biologic organism that is the source of the allergen. The formal name will use the Latin taxonomic name e.g. Arachis hypogaea. The Long Common Name will use the common name of that entity, e.g., Peanut, if one exists.

In the case of very well specified allergens, the Component will also indicate whether the allergen has a natural source or has been generated via recombinant method by adding the word, native or recombinant, respectively. Different antigens from the same species are distinguished by the addition of a sequence number that roughly reflects when in time they were identified.

So you will see names that contain content like the following:

  • Arachis hypogaea recombinant 1
  • Arachis hypogaea recombinant 3
  • Arachis hypogaea native 1
  • Arachis hypogaea native 3

However, as is true in other classes of LOINC, we sometimes embed synonyms within the allergen Analyte name to help users recognize the term by their naming rules. Using the LOINC-wide convention, synonyms are enclosed in parentheses and immediately follow the word or words which they represent. We insert standard acronym names used by most allergen manufacturers as synonyms. These begin with lower case "n" or "r" to distinguish native from recombinant allergens followed by the first 3 letters of the genus (in Latin), a space and the first letter of the species (in Latin). If the first 3 letters of the genus and the first letter of the species are not enough to distinguish between two allergens, the second letter of the species name is added (e.g., Prunus avium recombinant (rPru av) 1). Because the content in the parentheses represents the antigen acronym and not the Pharmacia code, the antigen sequence number is purposely placed after the closed parenthesis. Of course we also append Ab.IgE to the end of this entity name.

Putting this all together formal and common name for the Component for the allergens listed above become:

LOINC code Formal name of component Long common name of component
58779-0 Arachis hypogaea recombinant (rAra h) 1 Ab.IgE Peanut recombinant (rAra h) 1 IgE Ab
58777-4 Arachis hypogaea recombinant (rAra h) 3 Ab.IgE Peanut recombinant (rAra h) 3 IgE Ab
65769-2 Arachis hypogaea native (nAra h) 1 Ab.IgE Peanut native (nAra h) 1 IgE Ab
65771-8 Arachis hypogaea native (nAra h) 3 Ab.IgE Peanut native (nAra h) 3 IgE Ab

With LOINC version 2.75, August 2023, the RAST class subcomponent was moved from Analyte to Method fields. RASTs are interpreted by upper and lower boundaries of measurements into six classes, according to this source. Accompanying this change was the introduction of the SemiQn scale, which is more appropriate for interpreting 0-6 classes.

3.12 Urinalysis by Test Strips

Manufacturers of urinalysis testing strips provide two means of interpretation: a range of mass or substance concentration (such as 2, 4, 8, and 12 mg/dL of urobilinogen) or an ordinal counterpart of negative, 1+, 2+, 3+. It is rare for a facility to adopt BOTH means of interpretation. The normal range in the local laboratory information system will inform the mapper of which path to follow.

3.12.1 Two pathways of interpreting urinalysis testing strips

3.12.1.1 Semi-quantitative

LOINC Fully-Specified Name
53328-1 Glucose:MCnc:Pt:Urine:Qn:Test strip.automated
53327-3 Bilirubin:MCnc:Pt:Urine:Qn:Test strip.automated
50557-8 Ketones:MCnc:Pt:Urine:Qn:Test strip.automated
50563-6 Urobilinogen:MCnc:Pt:Urine:Qn:Test strip.automated

3.12.1.2 Ordinal

LOINC Fully-Specified Name
50555-2 Glucose:PrThr:Pt:Urine:Ord:Test strip.automated
50551-1 Bilirubin:PrThr:Pt:Urine:Ord:Test strip.automated
57734-6 Ketones:PrThr:Pt:Urine:Ord:Test strip.automated
62487-4 Urobilinogen:PrThr:Pt:Urine:Ord:Test strip.automated

  1. Walker RH. American Association of Blood Banks Technical Manual. 11th ed. Bethesda, MD: American Association of Blood Banks, 1993.

  2. Antonarakis, SE, and the Nomenclature Working Group. Recommendations for a nomenclature system for human gene mutations. Human Mutation 1998;11:1-3. PubMed: 9450896.

  3. Deckard J, McDonald CJ, Vreeman DJ. Supporting interoperability of genetic data with LOINC. J Am Med Inform Assoc. 2015 May;22(3):621-7. PubMed: 25656513.

  4. Beaudet AL, Tsui LC. A suggested nomenclature for designating mutations. Human Mutation 1993;2 (4) :245-248. PubMed: 8401532.

  5. ImmunoCAP Allergens [Internet]. Uppsala(Sweden): ImmunoCAP; c2006 [updated 2007 November 2; cited 2008 June]. Available from: http://www.phadia.com/en-GB/5/Products/ImmunoCAP-Allergens/

Knowledge Base

Users’ Guide

Search Syntax

RELMA

FAQ