Overview
Quick answer
The range printed beside a result on an Australian pathology report is a reference interval. Laboratories usually set it as the central 95% of results from an apparently healthy reference group. That group is not curated for long-term health, so the floor of the interval is a statistical line, not proof that everyone inside it is well.
Some of the numbers people treat as "the range" are not reference intervals at all. Ferritin under 30 µg/L in Australian adults, and HbA1c of 48 mmol/mol for diabetes, are clinical decision limits: college or WHO cut-offs tied to a diagnosis or a risk, not the 2.5th percentile of a local sample.
Optimal range is not a term RCPA, NPAAC, ADS, or the Endocrine Society of Australia use for a well-person blood-test band. Colleges talk about reference intervals, decision limits, and, in disease, treatment targets. A clinic or membership can still choose a tighter preventative band. That choice needs to be labelled as policy, not as consensus.
Overview
Key takeaways
Use the interval printed on the report from the lab that did the analysis. Pathology Tests Explained says intervals can differ between laboratories for the same test.
A two-sided 95% interval will, by construction, leave about 5% of the reference population outside it. An H or L flag is a prompt, not a diagnosis.
Some common chemistry tests now use AACB and RCPA harmonised intervals. "Every lab is different" is too strong. Hormones, haematology, and non-harmonised assays still vary.
Ferritin under 30 µg/L in an adult is an Australian diagnostic cut-off (RCPA, GESA, Lifeblood). WHO still uses under 15 µg/L for an apparently healthy adult. Those are decision limits, not two versions of the same 95% range.
HbA1c of 48 mmol/mol (6.5%) is the Australian diagnostic decision limit for diabetes. It is not the edge of a healthy-population interval.
A typical Australian adult TSH interval is about 0.4 to 4.0 mIU/L and may vary by laboratory. RCPA says there is insufficient evidence to treat non-pregnant adults to under 2.5.
The 2023 Australian cardiovascular guideline treats from five-year absolute risk, not from an LDL target. An in-range LDL can still sit inside a high-risk score.
Hemexa layers a preventative band on the Healius Pathology reference interval for some markers. That band is company policy, not a published medical standard.
Basics
What a reference interval actually is
Pathologists prefer the phrase reference interval. Reference range is the consumer synonym Pathology Tests Explained and RCPA use on public pages. Both name the same column on your report.
A population statistic, usually the central 95%
The IFCC convention, restated by Australian authors (Jones, Koerbin and colleagues, 2014), is to take the central 95% of a reference population: the 2.5th to 97.5th percentiles. CLSI EP28 is the laboratory protocol that implements that convention. It is a statistical habit, not a law of nature. Some analytes use a one-sided interval. Troponin uses the 99th percentile. Glucose and HbA1c often use decision limits instead.
Apparently healthy is not the same as optimal
The reference group is screened with exclusion criteria, then treated as typical. It is not selected for iron repletion, low long-term cardiovascular risk, or a particular performance goal. The leftover 95% therefore includes people who would fail a stricter clinical cut-off. That limitation is real. Naming the leftover band an optimal range is a separate, unofficial step.
About one in twenty healthy results will flag
A two-sided 95% interval leaves roughly 5% of the reference population outside by design. Pathology Tests Explained says an H or L flag draws attention and does not mean you are unwell. Run ten independent tests and the chance of at least one flag is about 40% (Phillips, Medicine Today, 2014). That is why a flag is the start of a conversation, not the end of one.
Accredited Australian laboratories are required to print a reference interval or a clinical decision limit, and to review it when the method changes. That is ISO 15189 as applied by NATA, not a blog claim.
Reading your results
How to read the range column on an Australian printout
An Australian Healius, Australian Clinical Labs, or Sullivan Nicolaides printout is a table, not a verdict. Read the columns in order before you react to a colour or an H.
Test name and units
The first job is to confirm you are looking at the assay you think you are looking at, in Australian units. Haemoglobin is g/L, not US g/dL. HbA1c is mmol/mol first. Ferritin is µg/L. A number copied from an overseas article will not match this column.
Result and flag
The result is the measured value. H or L (sometimes high or low written out) means the number sits outside the interval printed on this report. Pathology Tests Explained: the flag is attention, not a diagnosis. Age and sex often have their own intervals, so a result can flag on one partition and sit inside another.
The interval on this report
Use only the range from the lab that did the analysis. Intervals can differ by method and analyser. Sinclair, Hall and Badrick surveyed Australian haematology laboratories in 2014 and found wide variation, including different intervals reported for the same instruments, often with a source that was hard to trace.
Harmonised chemistry is the exception
RCPA Manual Table 6 publishes harmonised intervals for a first chemistry panel (sodium, potassium and related assays) when methods are traceable. For those tests, many Australian labs now print the same numbers. Hormones, haematology, and immunoassays are still lab-specific.
If you are comparing this year to last year, confirm it is the same laboratory and the same method. A number that looks new can be a transferred interval, not a change in you.
The distinction
Three different numbers, not two
Most consumer pages collapse everything into "in range" versus "optimal." Australian colleges actually use two official ideas, and a third word that they do not define.
Reference interval
A population statistic. Usually the central 95% of an apparently healthy reference group, method-specific, printed by the laboratory. Example: TSH 0.4 to 4.0 mIU/L on many Australian reports, with RCPA noting that it may vary by laboratory.
Clinical decision limit
An outcome- or consensus-based threshold used to diagnose, risk-stratify, or treat. IFCC C-RIDL (Ozarda et al., 2018) is the paper that splits this from a reference interval. Examples: adult ferritin under 30 µg/L; HbA1c of 48 mmol/mol; the 2023 cardiovascular guideline's five-year risk bands.
Optimal or preventative target
Not a recognised RCPA, NPAAC, ADS, or ESA term for a well-person blood-test band. Clinics and memberships use it for a tighter working range. RACGP does say "optimum management" for people who already have type 2 diabetes (a common HbA1c treatment goal of 53 mmol/mol or below). That is a disease-treatment target, not a preventative optimal range for a well adult.
A result can sit inside a reference interval, past a decision limit, or neither. The useful question is which kind of number you are looking at, not whether a website has painted the cell green.
Worked example
Ferritin: the worked example
Ferritin is the cleanest Australian example because two official cut-offs exist, they disagree, and a normal haemoglobin does not settle the question. The full marker walkthrough lives on the ferritin guide. This page only needs the range logic.
WHO: under 15 µg/L in a well adult
The 2020 WHO ferritin guideline still defines iron deficiency in an apparently healthy adult as ferritin under 15 µg/L. WHO rates the certainty behind that adult cut-off as low. It was retained from 1993. With infection or inflammation the same guideline uses under 70 µg/L. Under 12 µg/L is the cut-off for children under five, not for adults.
Australian practice: under 30 µg/L in adults
The RCPA Manual says a serum ferritin under 30 µg/L for an adult is diagnostic of iron deficiency. GESA 2022 and Australian Red Cross Lifeblood use the same adult diagnostic cut-off, with GESA citing roughly 92% sensitivity and 98% specificity at 30. That is a clinical decision limit. It is not "the Lifeblood reference interval." Lifeblood's donor notification intervals still start at 15 µg/L for female donors, which MJA 2024 notes will miss women between 15 and 30.
Stores fall before haemoglobin does
Camaschella (NEJM 2015) calls haemoglobin and MCV late, insensitive markers. GESA and Lifeblood say anaemia occurs late. A ferritin of 22 µg/L with a haemoglobin of 132 g/L can be iron deficiency without anaemia: inside a typical FBC interval, and already past the Australian ferritin decision limit. The full blood count guide is the companion page for that sequence.
MJA 2024, reading the ABS Australian Health Survey Biomedical Results for Nutrients 2011-13, reports 22.3% of Australian women with ferritin under 30 µg/L and 3.5% of men iron deficient. ABS itself did not print those percentages as a national cut-off, and the dataset is now more than a decade old. It remains the last nationally representative biomedical iron survey. A separate 2020 survey of women aged 18 to 39 in three states found 34.8% under 30. Do not merge those figures.
For units, inflammation, and how to read iron studies on an Australian report, see the ferritin and iron studies guide. The companion fact that a normal haemoglobin does not close the iron question lives on the full blood count guide.
Worked example
HbA1c: a decision limit, not a 95% edge
HbA1c is the teaching case CLSI names when it says some analytes use decision limits instead of a classic reference interval.
48 mmol/mol is a diagnostic line
The Australian Diabetes Society (d'Emden et al., MJA 2015, still the May 2023 ADS guidance) diagnoses diabetes at HbA1c of 48 mmol/mol (6.5%) or above, confirmed on a second test in someone without symptoms. That number comes from outcome and consensus work, not from the 97.5th percentile of a local healthy sample.
42 to 47 mmol/mol is a risk band
The same 2015 diagnostic statement says 42 to 47 mmol/mol (6.0% to 6.4%) marks higher risk of later diabetes and cardiovascular disease, and tells clinicians not to label that HbA1c-only result as prediabetes. A later 2020 ADS-endorsed primary-care statement marks the same band "prediabetes likely." Both documents treat it as a decision problem, not as the edge of a 95% interval.
A treatment target is a fourth idea
Once someone has type 2 diabetes, RACGP talks about goals for optimum management, commonly an HbA1c of 53 mmol/mol (7%) or below, individualised. That is therapy in disease. It is not evidence that colleges endorse an optimal range for a well adult.
For mmol/mol versus the older percentage, Medicare items, and what distorts HbA1c, see the HbA1c guide.
The limit
Why in range is not a clean bill of health
Inside the printed interval can still be the wrong place to stop. The reasons that survive a college read are specific. They are not a licence to call every mid-range result suboptimal.
The decision limit sits inside the interval
Ferritin between 15 and 30 µg/L can look ordinary against a WHO well-population line and already meet the Australian diagnostic cut-off. HbA1c of 44 mmol/mol is not diabetes and is also not a 95% "normal." Read the kind of number first.
The interval includes people who are not iron-replete
If the reference group was not required to have ferritin above 30, the statistical floor can sit below the clinical cut-off a college later adopted. That is how a population interval and a decision limit drift apart.
TSH near the top is a debate, not a disease
Walsh (MJA 2016) says the adult TSH interval debate has largely resolved around 0.4 to 4.0 mU/L, and that TSH rises with age. The RCPA thyroid position statement says treatment should generally aim for TSH within the reference interval, and that evidence is insufficient to target the lower half (under 2.5) in non-pregnant adults. 2.5 appears in Australian pregnancy treatment-target language after thyroxine is started. That is a different question.
Lipids are for five-year risk, not an LDL beauty contest
The 2023 Australian cardiovascular guideline, summarised in the MJA in 2024, starts lipid-modifying treatment from five-year absolute risk of 10% or above (consider at 5% to under 10%), not from an LDL target. Australian Prescriber 2024 notes the guideline does not explicitly mention LDL-C treatment targets. An in-range LDL can still sit inside a high-risk score. Total cholesterol above 7.5 mmol/L is treated regardless of the calculator. Secondary prevention is outside that primary-prevention document.
The lipid half of that argument, including why the calculator uses the total-cholesterol-to-HDL ratio, is on the high cholesterol blood tests guide.
Language
What optimal range is not
Search results will sell you an optimal range. Australian colleges do not.
It is not in the college glossary
RCPA, Pathology Tests Explained, NPAAC, ADS, ESA, and the 2023 cardiovascular guideline talk about reference intervals, decision limits or decision points, harmonised intervals, and treatment targets. Optimal range does not appear as a defined pathology term.
Vitamin D is the other common invention
The Australian and New Zealand Bone and Mineral Society holds that 50 nmol/L and above is adequate for most people. Framing 50 to 75 nmol/L as insufficient was the 2011 US Endocrine Society position, not ANZBMS, and even that society moved away from it: its 2024 guideline update does not recommend routine testing or a target range for healthy adults under 75. A clinic band above 75 is a clinic band.
A membership can still choose a tighter band
Preventative practice can watch a result before it crosses a diagnostic line. That is a tracking choice. It has to be attributed as the organisation's policy, built on the laboratory interval, not dressed up as RCPA consensus.
Product
How Hemexa uses a preventative band
Hemexa does use preventative bands on some markers. The honest version of that sentence names the source and the limit.
The lab floor comes from Healius Pathology
Healius Pathology is Hemexa's pathology partner. The watch band on several markers matches the reference interval that laboratory prints. Collection is through that network, not through a single state brand.
Ferritin is the published worked example
Hemexa's own clinical policy targets a higher preventative band than the diagnostic floor: roughly 50 to 120 µg/L for women and 100 to 200 µg/L for men. That is the band already stated on the ferritin guide. It is Hemexa policy, not a universally agreed medical standard.
Conflicting rows are not quoted as a national optimal
Some Hemexa range-policy rows disagree with each other, and several cite a state brand as the reference-interval source. This page will not invent a single Hemexa optimal table from those rows. Where a band is used in copy, it is one already published on a marker guide and labelled as company policy.
FAQ
Frequently asked questions
- If my result is inside the printed range, can it still be a problem?
- Yes, when the number that matters is a decision limit rather than the 95% interval. Ferritin under 30 µg/L with a normal haemoglobin is the usual Australian example. HbA1c of 42 to 47 mmol/mol is a risk band, not hidden diabetes. A result near the top of a TSH interval is interpreted with symptoms, antibodies and age; colleges have not adopted 2.5 as a national adult cut-off.
- Why does my friend's report from another Australian lab show a different range?
- Method, analyser, and how the lab set or transferred the interval. Pathology Tests Explained: use only the range on this report. Sinclair, Hall and Badrick (Pathology, 2014) found Australian haematology intervals varied, including on the same instrument family. Some common chemistry tests are now AACB and RCPA harmonised, so the difference is not universal.
- Is optimal range something RCPA or my GP is supposed to use?
- No. Colleges use reference intervals, clinical decision limits, and, in disease, treatment targets. Optimal range is not a defined Australian college term. A clinic or membership can still choose a tighter preventative band and should say so.
- Is the HbA1c number on my report a reference interval or a diabetes cut-off?
- The 48 mmol/mol (6.5%) line is a diagnostic decision limit from the Australian Diabetes Society. It is not the 97.5th percentile of a healthy sample. Values of 42 to 47 mmol/mol are treated as higher risk, and in the 2020 prediabetes screening statement as prediabetes likely. The 2015 diagnostic statement tells clinicians not to slap the prediabetes label on HbA1c alone.
- Should I treat TSH 2.8 because I read that optimal is under 2.5?
- RCPA: treat toward the reference interval. Evidence is insufficient to use under 2.5 as the adult goal. 2.5 in Australian endocrine writing is mainly a pregnancy treatment target after thyroxine is started, not a national well-adult optimum.
- My LDL is in range. Does that mean I do not need lipid-lowering medicine?
- The 2023 Australian cardiovascular guideline does not treat from an LDL target. It treats from five-year absolute risk, and from a short list of clinically determined high-risk conditions. An in-range LDL can still sit inside a high-risk score. Total cholesterol above 7.5 mmol/L is treated regardless of the calculator.
- If 5% of healthy people fall outside every 95% interval, why flag results at all?
- The flag is a prompt, not a diagnosis. The more independent tests you run, the more statistical flags you should expect. Decision limits exist because a statistical tail is a poor disease definition for HbA1c, ferritin, and lipids.
- Who chose the numbers on my printout: WHO, RCPA, or the lab?
- It depends on the analyte. Harmonised chemistry: the AACB and RCPA table, if the lab adopted it. Haematology and many hormones: the lab's own or transferred interval. HbA1c diabetes: ADS and WHO decision limit. Ferritin deficiency: RCPA, GESA and Lifeblood clinical cut-off of 30 in adults, which may or may not match the printed interval.
How Hemexa fits
How Hemexa can help
The product argument is not a secret third official range. It is a printed Healius interval, a named preventative band where one exists, and a trend across the annual baseline instead of a one-off PDF.
The interval and the band on the same dashboard
Hemexa reports the 76–80 marker annual baseline against the Healius Pathology reference interval, then, for selected markers, against a tighter preventative band. Ferritin is the example already published: 50 to 120 µg/L for women and 100 to 200 µg/L for men, labelled as Hemexa policy.
Marker guides for the decision limits
The ferritin, full blood count, HbA1c, thyroid, and high-cholesterol guides keep the college numbers in one place so this page does not have to become a second copy of each assay.
Collection through Healius Pathology
Blood is collected through Healius Pathology. Membership starts at AU$1,199 a year and includes the annual baseline plus the six-month metabolic retest. Hemexa does not replace your GP, and a preventative band is not a diagnosis.
Sources
References
Solberg, H. E. (1987). Approved recommendation (1987) on the theory of reference values. Part 5. Statistical treatment of collected reference values. Determination of reference limits. Clinica Chimica Acta, 170(2-3), S13-S32. View source ↗
Clinical and Laboratory Standards Institute. EP28-A3c: Defining, Establishing, and Verifying Reference Intervals in the Clinical Laboratory (approved 2008, corrected 2010). View source ↗
Ozarda, Y., Sikaris, K., Streichert, T., & Macri, J. (2018). Distinguishing reference intervals and clinical decision limits. Critical Reviews in Clinical Laboratory Sciences, 55(7), 450-475. View source ↗
Jones, G. R. D., Koerbin, G., et al. (2014). The evidence-based approach to harmonised reference intervals. Clinica Chimica Acta, 432, 35-43. View source ↗
Pathology Tests Explained. Reference intervals explained. View source ↗