Guide
Does a LOINC code tell you which method measured the result?
For most routine chemistry, no. LOINC can name a method, and method-specific codes exist for some tests, but laboratory test guides commonly map creatinine, albumin, sodium or ALT to a code that leaves the method out, and the same code is used for different analysers and different chemistries. This guide shows what two laboratories’ published codes look like, what the shared creatinine code hides, and what to record alongside the code.
The short answer
A LOINC code tells you what was measured, in what specimen and on what scale. For routine chemistry, the code laboratories send usually does not tell you how. In one hospital laboratory network’s published test guide, read with two open-access studies, 40 of 42 maker-specific versions of a chemistry test sat on a LOINC code with no method, shared with another maker’s version of the same test; in a second network’s guide, 40 of 40 did. None shared a code that names a method. LOINC does define method-specific codes for some of these tests (ALT, AST and albumin among them); the guides we read did not use them. A result that arrives with only a method-less LOINC code cannot be traced to its analyser or its chemistry, so a change of either is invisible downstream.
What a LOINC code records
Each LOINC term is defined along six axes: the component (creatinine), the property (mass or substance concentration), the timing (a single point in time), the system (serum or plasma), the scale (quantitative) and the method [LOINC]. The method axis is optional. Many common chemistry terms leave it empty, and in the two laboratory guides below those method-less terms are the ones the tests are mapped to. A laboratory that changes from one creatinine chemistry to another, or from one maker’s analyser to another’s, keeps the same code.
Method-specific codes do exist for some of these tests. Alanine aminotransferase, for example, has 1743-4 (with pyridoxal phosphate added) and 1744-2 (without it) beside the method-less code [LOINC]. The guides we read used the method-less codes, so part of what is lost is a mapping choice the laboratory makes, not a limit of LOINC itself.
That is a design choice, not an error: LOINC identifies the observation, and a method-less code lets results from comparable methods be pooled. Where LOINC offers a method-specific code, a laboratory can choose it; in the guides below, none did. The cost of the method-less code is that it does not carry the information a reader needs when two methods are not comparable at a decision limit.
What two laboratories publish
Public data
Health New Zealand Waikato’s laboratory test guide gives, for each test, the analyser used at its main laboratory and at its rural laboratories, which run a different maker’s analysers, and one LOINC code [Waikato laboratory test guide]. We read 19 chemistry analytes that more than one maker measures there or at the laboratories described in two open-access studies [Cholan et al., 2022], [Luu et al., 2024]. Of 42 maker-specific versions, 40 sat on a method-less code another maker also used (95% interval 84.2 to 98.7%), and 0 on a shared code that names a method. The two exceptions were one cardiac troponin coded differently at two institutions. Read the figure with its limit: 34 of the 42 come from Waikato, where one test page gives one code for both analysers, so the sharing there is one laboratory’s reporting choice. The other 8 come from the studies.
The same pages name the principle each analyser uses. For five analytes, one code covers two different chemistries:
| Analyte | LOINC code | Main laboratory | Rural laboratories |
|---|---|---|---|
| Creatinine | 14682-9 | Jaffe | enzymatic |
| Albumin | 1751-7 | bromocresol purple | bromocresol green |
| Sodium | 2951-2 | indirect ion-selective electrode | direct ion-selective electrode |
| Potassium | 2823-3 | indirect ion-selective electrode | direct ion-selective electrode |
| Calcium | 2000-8 | NM-BAPTA | Arsenazo III dye |
Source: [Waikato laboratory test guide]. Method names as the guide gives them; LOINC content [LOINC]. LOINC also defines separate albumin codes for bromocresol green (61151-7) and bromocresol purple (61152-5); the guide uses neither.
UW Medicine’s online test guide documents a move of its chemistry testing from one maker’s analysers to another’s: for each test, the old and new laboratory code, a “Trendable?” flag, which the guide ties to whether the old and new tests share one component code in Epic, the laboratory’s record system, and a mean bias between the platforms [UW Medicine test guide]. For 20 chemistry analytes we read the LOINC code on both versions. All 40 sat on a method-less code shared across the two platforms (95% interval 91.2 to 100.0%). This is a second network, read after the first, and the pattern held.
The informative cases are the five analytes the laboratory split into separate tests with new laboratory codes, which its transition table flags as not trendable in its Epic configuration. Each still carries one method-less LOINC code on both the old and the new test, although for ALT and AST LOINC also offers method-specific codes, with and without pyridoxal 5′-phosphate (ALT 1743-4 and 1744-2; AST 30239-8):
| Analyte | LOINC code on both tests | Printed mean bias, new against old |
|---|---|---|
| ALT | 1742-6 | +33.3% |
| AST | 1920-8 | +25.0% |
| Total bilirubin | 1975-2 | −6.3% |
| Total protein | 2885-2 | −5.6% |
| HDL cholesterol | 2085-9 | +5.6% |
Source: [UW Medicine test guide], facts only. LOINC content [LOINC].
A registry or a research dataset that keys on LOINC would join the ALT results from both platforms into one series, across a +33.3% step on a test the laboratory keeps under separate local codes. The tests the guide’s flag marks as trendable are not all close either: of the 76 it marks trendable with a printed bias, 43 differ by 5% or more and 23 by 10% or more, across all test types, not only chemistry.
Neither guide is a sample of laboratories. Together they show that laboratories do publish their chemistry this way; they do not show how common it is elsewhere.
Jaffe and enzymatic creatinine share one code
At Waikato, creatinine by the Jaffe method on the main analyser and by an enzymatic method at the rural laboratories are both reported as 14682-9 [Waikato laboratory test guide]. Three open-access studies ran both creatinine methods on the same specimens and printed how often the two disagree on kidney function stage. In each, both methods ran on one maker’s analyser, so even a record of the analyser would not separate them: only the method does.
| Study and licence | Population | Paired results | Different stage between methods |
|---|---|---|---|
| [Boss et al., 2024], CC BY 4.0 | Kidney transplant outpatients | 12,081 | 1,589 (13.2%), CKD-EPI 2009 |
| [Schmidt et al., 2015, PLoS ONE], CC BY 4.0 | Outpatients | 529 | 29 (5.5%) on either side of eGFR 60, 45, 30 or 15 |
| [Preda et al., 2026], CC BY 4.0 | Hospital patients over 40 | 427 | 10.77% printed, CKD-EPI |
Counts as each paper prints them, checked against its own tables. Sources: [Boss et al., 2024], [Schmidt et al., 2015, PLoS ONE], [Preda et al., 2026].
In the transplant clinic, 1,589 of 12,081 paired results (13.2%, 95% interval 12.6 to 13.8%) fall in a different stage by the two methods, and the Jaffe result reads worse in 1,256 of them [Boss et al., 2024]. Across the line that defines chronic kidney disease, eGFR 60, 647 results (5.4%) are CKD by one method and not the other: 543 by Jaffe only, 104 by enzymatic only. Across eGFR 30, 231 results (1.9%) change side. The stage count is the paper’s; the counts across a line are ours, from its printed stage tables, whose totals match the paper’s own. A transplant clinic is chosen for kidney disease and its patients are measured repeatedly, so the rate is that clinic’s, not a rate for all creatinine results.
Because the direction is not random, a laboratory that moves from Jaffe to enzymatic creatinine under the same code would show up downstream as an apparent improvement in kidney function across its patients. Nothing in the code says the method changed. The guide to Jaffe and enzymatic creatinine covers why the chemistries differ, and the guide to eGFR after a creatinine method change how a small bias moves stages.
Two counts from one study
The 529 outpatient pairs in the table above are the same data this site uses elsewhere for a different count, and the two numbers should not be confused. They count different things.
- 29 of 529 is the paper’s own count: pairs whose eGFR falls on opposite sides of any of the kidney stage limits, 60, 45, 30 or 15 [Schmidt et al., 2015, PLoS ONE].
- 23 of 529 is our re-analysis at two creatinine decision lines, 1.2 and 1.5 mg/dL, set against a baseline from repeat testing alone. 23 results changed category at those lines, against about 12 expected from repeat testing alone. That expected count uses the 1.8% imprecision the study itself measured; at that imprecision the count from repeat testing alone falls between 6 and 17 with 95% probability, so the 23 crossings are about twice the expected count, and by the exact Poisson-binomial test 23 or more from repeat testing alone has probability 0.00017. That 1.8% is also our program's laboratory-analyser default (1.79%, this study's own reported precision), so the default gives about the same count; the program's cautious point-of-care default of 6%, which it uses when a point-of-care device is involved and no precision data are supplied, would give about 41, which overstates repeat-testing noise for these laboratory analysers. At the upper end of the study's own precision estimate, 2.34%, about 15 would be expected and 23 or more has probability 0.014.
The first asks how many results would be staged differently; the second asks whether the crossings at a laboratory’s own lines exceed what imprecision alone would produce. Separating the crossings seen at a decision limit into the part imprecision alone would give and the part due to the change is a published procedure (Schmidt et al., 2015, Clin Chim Acta), which the creatinine study we use applies. What we add is a program that runs it on a laboratory’s own comparison and precision data at its own decision limits, keeps the re-test list in the laboratory and releases only aggregates.
Source for the re-analysis: Schmidt RL, Straseski JA, Raphael KL, Adams AH, Lehman CM (2015). A Risk Assessment of the Jaffe vs Enzymatic Method for Creatinine Measurement in an Outpatient Population. PLoS One. doi:10.1371/journal.pone.0143205 (opens in a new tab); PMC4657986 (opens in a new tab). Licence: CC BY 4.0 (opens in a new tab). Re-analysed by NakedSignal.
What laboratories and data users can do
- Send the method and the analyser with the result, not only the code. HL7 version 2 result messages have fields for the observation method (OBX-17) and the equipment that produced the result (OBX-18); FHIR has Observation.method and Observation.device. A case report on results exchanged between two institutions lists the lack of reporting of analytical technique or instrument platform among the patient safety issues it found, and names the Device Identifier of the Unique Device Identification system as the way to describe test method and device [Luu et al., 2024]. Where both methods run on one analyser, as in the creatinine studies above, the method or reagent has to be recorded too.
- Map with the maker’s LIVD file where one exists. LIVD is the industry format in which a maker publishes the LOINC code its assay’s results should carry, with the device identifier [IICC LIVD], [CDC LIVD page]. It does not add a method to a method-less code, but it ties the code choice to a named assay and its device identifier. In a study of 5 medical centre laboratories, the laboratory’s LOINC code differed from the maker’s for 133 of the 317 non-COVID tests that had a LIVD file [Cholan et al., 2022].
- Follow SHIELD. The FDA-led SHIELD collaboration, now a collaborative community, works on harmonising how in vitro diagnostic results are coded from the device to the health record [FDA SHIELD].
- Date the change. A laboratory that changes method or analyser can publish the date and the tests affected, as UW Medicine’s guide does, and give a test whose results shift a new local code [UW Medicine test guide]. A data user who pools results across laboratories or years can ask each source for that list before trending a LOINC-keyed series.
- Count the patients near the lines, not only the average bias, when a method changes. The guide to comparability across analysers and sites covers how to report it.
Check a change on your own pairs
If you have paired results from an old and a new method, the change check counts the results that crossed each decision line against those repeat testing alone would move. It runs in the browser; your file is not uploaded.
Sources
- Regenstrief Institute. LOINC: Logical Observation Identifiers Names and Codes. LOINC is copyright Regenstrief Institute, Inc. and the LOINC Committee, available at no cost under the licence at loinc.org/license. loinc.org/
- Health New Zealand Waikato (2026). Laboratory test guide. Public test pages; facts used with attribution, no text reproduced. lab.waikatodhb.health.nz/test-guide/
- Cholan RA, Pappas G, Rehwoldt G, et al. (2022). Encoding laboratory testing data: case studies of the national implementation of HHS requirements and related standards in five laboratories. Journal of the American Medical Informatics Association, CC BY 4.0. doi.org/10.1093/jamia/ocac072
- Luu HS, Campbell WS, Cholan RA, et al. (2024). Analysis of laboratory data transmission between two healthcare institutions using a widely used point-to-point health information exchange platform: a case report. JAMIA Open, CC BY 4.0. doi.org/10.1093/jamiaopen/ooae032
- UW Medicine Department of Laboratory Medicine and Pathology (2026). Online test guide. Public test pages and platform transition table; facts only, no text reproduced (University of Washington reserves copyright). testguide.labmed.uw.edu/
- Boss K, Stolpe S, Müller A, et al. (2024). Effect of Difference in Serum Creatinine between Jaffe and Enzymatic Methods in Outpatient Kidney Transplant Recipients. Journal of Clinical Medicine, CC BY 4.0. doi.org/10.3390/jcm13206066
- Schmidt RL, Straseski JA, Raphael KL, Adams AH, Lehman CM (2015). A Risk Assessment of the Jaffe vs Enzymatic Method for Creatinine Measurement in an Outpatient Population. PLoS ONE, CC BY 4.0. doi.org/10.1371/journal.pone.0143205
- Preda EC, Oprea O, Barabas ZA, et al. (2026). Impact of creatinine measurement methods on eGFR and GFR category assignment. The Journal of Critical Care Medicine, CC BY 4.0. doi.org/10.62838/jccm-2026-0007
- Schmidt RL, Kordy MA, Genzen JR, Straseski JA, Greene DN, Lehman CM (2015). A mathematical procedure to estimate the impact of a change in method on discordance or misclassification at a decision limit in laboratory method comparison studies. Clinica Chimica Acta. doi.org/10.1016/j.cca.2014.10.043
- US Centers for Disease Control and Prevention, Division of Laboratory Systems (2024). LOINC In Vitro Diagnostic Test Code Mapping. CDC web page. www.cdc.gov/laboratory-systems/php/livd-test-codemapping/index.html
- IVD Industry Connectivity Consortium. LIVD specification. IICC web page. ivdconnectivity.org/livd-specification/
- US Food and Drug Administration. Systemic Harmonization and Interoperability Enhancement for Laboratory Data (SHIELD). FDA web page. www.fda.gov/medical-devices/diagnostic-data-program/systemic-harmonization-and-interoperability-enhancement-laboratory-data-shield
This material contains content from LOINC (http://loinc.org). LOINC is copyright © Regenstrief Institute, Inc. and the Logical Observation Identifiers Names and Codes (LOINC) Committee and is available at no cost under the license at http://loinc.org/license. LOINC® is a registered United States trademark of Regenstrief Institute, Inc.
The codes are used to identify tests; nothing on this page is advice on how to code a test.