Guide · 7 min read
How to read a lab report
A printed lab report is one of the densest documents in medicine. A basic metabolic panel fits fourteen numbers, fourteen units, fourteen reference intervals and a column of flags onto half a page, and none of it is explained. Students are expected to absorb the format by osmosis, and patients are handed it with no key at all.
This guide takes a report apart column by column. It is written for nursing and health students learning to read results, and for anyone who wants to understand the document rather than decode their own numbers - what a value means for a particular person is a conversation for the clinician who ordered it, not for a reference site.
The five parts of a result line
Almost every laboratory information system prints the same five fields per analyte, in roughly the same order. Once you can name them, any report from any hospital becomes readable.
The order and the labels vary - some systems put the flag first, some print the reference interval in a separate column headed "Normal", some abbreviate the analyte to a three-letter mnemonic. The content does not vary.
| Field | What it is | What trips people up |
|---|---|---|
| Analyte | The substance measured, often abbreviated (Na, K, Hgb, ALT) | The same substance has several names - see the aliases listed on each test page |
| Result | The measured value | It is a measurement with error bars, not an exact truth; repeat draws differ slightly |
| Unit | The unit the result is expressed in | Conventional and SI units give completely different-looking numbers for the same result |
| Reference interval | The band this laboratory considers expected for its population | It is specific to that laboratory and analyzer, not a universal constant |
| Flag | H, L, or a critical marker when the result falls outside the interval | "Flagged" is not the same as "abnormal for this patient" |
What a reference interval actually is
A reference interval is not a definition of health. It is a statistical description of a reference population. A laboratory measures the analyte in a group of apparently healthy volunteers who meet defined criteria, then publishes the central portion of that distribution - conventionally the middle 95%, cutting 2.5% off each tail.
That construction has a consequence people rarely state out loud: by design, one in twenty healthy people falls outside the interval for any given test. The tails were not removed because those individuals were unwell. They were removed because a band has to end somewhere.
Now multiply. If a panel reports fourteen independent analytes and each has a 95% chance of landing inside its interval, the chance that all fourteen land inside is 0.95 to the fourteenth power - about 49%. In other words, slightly more than half of perfectly healthy people will have at least one flagged value on a fourteen-analyte panel. Real analytes are not fully independent, so the true figure differs, but the direction is right and the lesson is exact: a single flag in isolation carries far less information than its bold typeface suggests.
This is why clinicians look at patterns, trends over time, and the clinical picture rather than at individual flags. It is also why the site's per-test pages list causes of high and low results rather than conclusions.
Flags: H, L, and critical
Most systems use a simple two-tier scheme. A result outside the reference interval gets an H or an L (some print an asterisk, some highlight the row). A result outside a much wider second band gets a critical or panic flag - often HH, LL, a bold C, or an inverse-video row.
The two tiers do very different jobs. The H/L flag says "this is outside the expected band". The critical flag says "this value is associated with immediate risk and someone must be told now". Critical results trigger a callout: the laboratory telephones a licensed caregiver, the call is read back, and the notification is documented. That workflow, not the number itself, is the point of the second tier.
Critical thresholds are set locally. A hospital with a large dialysis population and a paediatric hospital will not use the same potassium callout value, and both are correct for their setting. The commonly taught critical values shown on this site's test pages are teaching values for exam preparation - the number that governs practice is the one in your facility's policy.
- H / L - outside the reference interval; interpreted in context.
- Critical / panic - outside a wider band; triggers immediate notification under local policy.
- No flag - inside the interval, which is not the same as "nothing to see": a value can move substantially within the band and still matter.
- Some analytes are reported with only one boundary (a maximum or a minimum), so only one flag direction is possible.
Units: the same result, two very different numbers
Conventional units are mass-based and dominate reporting in the United States. SI units are amount-of-substance based (moles) and dominate most of the rest of the world, as well as much of the published literature. Because the conversion depends on molecular weight, each analyte has its own factor and the numbers can look unrelated.
Creatinine reported as 1.0 mg/dL is the same result as 88.4 µmol/L. Glucose at 90 mg/dL is 5.0 mmol/L. A reader who has memorised one system and meets the other will misjudge a value badly unless they convert first, which is exactly the mistake that makes international textbooks and journal papers feel wrong to a student trained on a single system.
Two habits prevent this. First, read the unit before the number, every time. Second, when a source uses the other system, convert rather than estimate - the site's unit converter carries the published factor for each analyte, and each test page shows both bands side by side.
Why one range is often not enough
A single printed band hides a lot of legitimate variation. Several analytes have genuinely different reference intervals by sex - haemoglobin, haematocrit, creatinine and ferritin among them - because the underlying physiology differs, not because the measurement does. Where the dataset carries a published subgroup band, this site shows it separately rather than averaging the two into a band that describes nobody.
Age matters as much. Alkaline phosphatase is expected to be high in a growing child and would be reported against a paediatric interval. Many analytes shift in pregnancy. Some shift with posture, with fasting state, or with time of day.
The practical rule is simple and worth internalising early: use the interval printed on the report you are holding, next to the result you are reading. Any range from a textbook, an app, or this site is a teaching value that helps you learn the shape of the data - it is not the band that result was measured against.
What a report cannot tell you
A result is one measurement, at one moment, on one specimen, at one laboratory. It carries analytical imprecision, biological variation within the same person from day to day, and everything that happened to the specimen between the vein and the analyzer.
That is why a trend usually beats a snapshot, why a surprising result is often repeated before it is acted on, and why the interesting question about an unexpected value is frequently "was something wrong with the sample?" before it is "what disease is this?".
None of it substitutes for the clinician who ordered the test. If you are looking at your own results, the useful next step is to ask the person who ordered them what they mean in your context - they have the history, the medication list, and the previous values that the sheet of paper does not.
Frequently asked questions
Does a flagged result mean something is wrong?
Not on its own. A reference interval is built to contain the middle 95% of a healthy reference population, so about one healthy person in twenty falls outside it for any single test - and on a multi-analyte panel the chance of at least one flag is much higher still. A flag means "outside the expected band for this laboratory's population", which is a prompt to look at the whole picture, not a diagnosis.
Why does my lab's normal range differ from the one in my textbook?
Because reference intervals are laboratory-specific. They depend on the analyzer and assay method, on the reference population the laboratory sampled, and on where that laboratory chose to cut the distribution. Two accredited laboratories can publish different intervals for the same analyte and both be correct for their own instruments and patients.
What is the difference between conventional and SI units?
Conventional units are mass-based (mg/dL, g/dL) and are standard in the United States; SI units are amount-based (mmol/L, µmol/L) and are standard in most other countries and in much of the literature. Each analyte has its own conversion factor derived from molecular weight, so the same result can look like 1.0 in one system and 88.4 in the other.
What is a critical or panic value?
A result outside a second, much wider band that a laboratory has defined as requiring immediate notification of a licensed caregiver. It triggers a documented telephone callout with read-back rather than a routine report. The thresholds are set locally, so the number that applies in practice is your facility's, not a textbook's.
Should I look up my own results online?
Reading about what a test measures is genuinely useful preparation for a conversation with your clinician. Trying to interpret your own values from a reference site is not - the interpretation depends on your history, medications, previous results, and the interval your own laboratory used. Bring the questions; leave the conclusions to the person who ordered the test.