The fifth data part of the test name specifies the Scale of the measure, and is a required part. The abbreviation of the type of Scale (previously called precision), given in Table 12, should be used in the fully specified name. Note that with the release of Version 1.0K, May 1998, we changed the codes for these from SQ to ORD and from QL to NOM to more accurately identify the meaning.
In Version 2.75, August 2023, the semi-quantitative (SemiQn) scale was re-introduced for non-continuous measurement of numeric values. See the Technical Brief Non-linear Numerical Values “Binned” to Ordinal or Range for further, extensive discussion.
Table 12: Type of Scale
| Scale Type | Abbr. | Description |
|---|---|---|
| Quantitative | Qn | The result of the test is a numeric value that relates to a continuous numeric scale. Reported either as an integer, a ratio, a real number, or a range. The test result value may optionally contain a relational operator from the set {<=, <, >, >=}. Valid values for a quantitative test are of the form “7”, “-7”, “7.4”, “-7.4”, “7.8912”, “0.125”, “<10”, “<10.15”, “>12000”, 1-10, 1:256 |
| Semi-Quantitative | SemiQn | Semi-quantitative results are identified as being within buckets or discrete ranges of possible values usually with lower and upper numeric boundaries. Measurement is not on a continuous numeric scale. Examples include mass or molar concentrations detected by chromogenic changes on a test strip. The changes in color indicate intervals of concentration, such as 2, 4, 8, and 12 mg/dL of urobilinogen. Titers are another example of discrete values reported serially e.g. 1:8, 1:16 but not all values sequentially are reported. The true concentration may have been 1:9, but the measurement of 1:8 implies a true measure between 1:8 and 1:16. |
| Ordinal | Ord | Ordered categorical responses, e.g., 1+, 2+, 3+; positive, negative; reactive, indeterminate, nonreactive. (Previously named SQ) |
| Quantitative or Ordinal | OrdQn | Test can be reported as either Ord or Qn, e.g., an antimicrobial susceptibility that can be reported as resistant, intermediate, susceptible or as the mm diameter of the inhibition zone. (Previously named SQN) We discourage the use of OrdQn in other circumstances. |
| Nominal | Nom | Nominal or categorical responses that do not have a natural ordering. (e.g., names of bacteria, reported as answers, categories of appearance that do not have a natural ordering, such as, yellow, clear, bloody. (Previously named QL) |
| Narrative | Nar | Text narrative, such as the description of a microscopic part of a surgical papule test. |
| “Multi” | Multi | Many separate results structured as one text “glob”, and reported as one observation, with or without imbedded display formatting. |
| Document | Doc | A document that could be in many formats (XML, narrative, etc.) |
| Set | Set | Used for clinical attachments |
2.6.1 Quantitative (Qn)
Identifies Scales that can be tied to some physical quantity through a linear equation. This means that if we have two reports for the same quantity one with a value of 5 and the other a value of 10 we know that the two are related in amount through the linear equation Y = aX +b. When the intercept, b, is non-zero, we have a difference scale. (Fahrenheit temperature is a difference scale.) When it is zero we have a ratio scale (Kelvin temperature is a ratio scale).[[^22]],[[^23]] A Qn value may be reported as a value for a “continuous” Scale, as is the case for serum sodium.
2.6.2 Semi-Quantitative (SemiQn)
The SCALE designation Semi-Quantitative provides a warning that this data should not be presumed to be true “linear” quantities. This SCALE represents ranges, buckets or binning of upper and lower thresholds due to the methods of detection. The SemiQn scale is associated with PROPERTIES such as Titers, NCNCRange, ScoreRange, and also MCNC and SCNC. The latter being invoked with METHODS of test strip detection or RAST Classes, amongst others. Please note urinalysis test strips provide two different SCALE options: the colored test pads may represent units of mg/dL (SemiQn) or 1+,2+,3+ (Ord).
2.6.3 Ordinal (Ord)
Some observations have values that are well ordered, e.g., “present, absent”, “1+, 2+, 3+”, or “negative, intermediate, positive”, but the values have no linear relationship to one another. We do not know that positive is two or three times as much as intermediate, we just know that positive is more than intermediate. Pain scales are an example of this arbitrary measurement. These kinds of observations have an ordinal Scale (Ord). Tests with “yes/no” answers are always ordinal (Ord). Tests reported as negative when less than the detection level but as quantified values otherwise should be regarded as quantitative (Qn).
2.6.4 Quantitative/Ordinal (OrdQn)
Rarely, a result can be reported in either an ordinal or quantitative Scale. The principal examples of this scale are microbiology susceptibilities: Agar diffusion (Kirby Bauer (KB)), Minimum Inhibitory Concentration (MIC) and others. A MIC, which can be reported as either resistant/intermediate/susceptible or by the MIC numeric value. The need for terms with OrdQn as Scale was further obviated by clarification from HL7 that results such as “POS” and “NEG” should go in the OBX-8 field for normalcy status. Thus, LOINC codes with Scale of Qn can be appropriately used in these cases even if the “values” coming back are coded interpretations of the true numeric result value.
2.6.5 Nominal (Nom)
Some observations take on values that have no relative order. Think of the numbers on football jerseys. These simply identify the players; they do not provide quantitative information or rank ordering of the players. We refer to these as nominal (Nom) in Scale. Blood culture results provide a good example. Possible values could be Escherichia coli (or a code for E. coli) or Staphylococcus aureus. Other examples are admission diagnoses and discharge diagnoses. Any test or measure that looks broadly at patient or specimen and reports the name of what it finds is a Nom Scale. The values of nominal scaled observations are assumed to be taken from a predefined list of codes or from a restricted vocabulary (e.g. a menu of choices). These observations would typically be sent in an HL7 message OBX segment with a Coded Element (CE) data type (in earlier HL7 versions) or its superseding Coded with No Exceptions (CNE) and Coded With Exceptions (CWE) variants (later HL7 versions). It is important to note that the CE and CWE data types allow values to be set as codes with their print text or just as their print text alone. These data types and the Nom Scale would not be used for running narrative.
2.6.6 Narrative (Nar)
Some observations are reported as free text narrative. The content is not drawn from a formal vocabulary or code system. A dictated present illness would be an example of a Scale of narrative (Nar). Many clinical LOINC codes will come in two versions: one for the nominal (coded) version and one for a narrative (free text) version.
2.6.7 Multi
We strongly encourage all reporting to be at the most granular level of detail. That is, if three numbers were reported, they would each be reported under a unique LOINC code and transmitted in a separate HL7 OBX segment. Occasionally reporting systems are not able to comply with this dictum. For example, some chromatography instruments can identify chemicals from the entire spectrum of known chemicals (CAS identifies more than 10 million distinct chemicals), and we may not have specific LOINC codes for reporting out these details. We have designated the Scale of Multi to identify results that include many separately structured results as one text “glob” with or without embedded (display formatting). Some laboratories report all of the details of many multiple measure tests under such globs with test names that correspond to their order name. We strongly discourage such reporting. It defeats the very purpose of individual codes to tag content.
Note
Because the individual elements of an Order set/Panel often have different Scales, the Scale for the order set term may be populated by a dash (-).
2.6.8 Document (Doc)
The Scale of Doc represents a collection of information that is either structured or unstructured. Individual LOINC codes are assigned for different collections of information regardless of the format in which they are presented, meaning that the same LOINC code should be used to represent a given document type regardless of whether it is in PDF, text document, JPG, XML, or HTML formats. The difference between Doc and Nar is that Nar represents a single free text result, while Doc is used for collections of results reported together, which may include narrative results.
Note
“Narrative reporting” as required by regulatory agencies such as the Office of the National Coordinator for Health Information Technology in the U.S. may be fulfilled by using LOINC terms with Scale Doc and thus is not only tied to the LOINC Nar Scale.
[[^22]]: Stevens SS. Measurement, statistics, and the schemapiric view. Like the faces of Janus, science looks two ways–toward schematics and empirics. Science 1968;161:849-856. [PubMed: 5667519]
[[^23]]: Tang YW, Procop GW, Persing DH. Molecular diagnostics of infectious diseases. Clin Chem 1997;11:2021-2038. [PubMed: 9365385]