Comparing a client's results from different labs and units

Some unit pairs are the same number under a different label: ferritin in ng/mL and µg/L, or TSH in mIU/L and µIU/mL. Others need a factor, such as vitamin D, where nmol/L is about 2.5 times the ng/mL figure. And some results can't be lined up at all, because the two labs measured something different or on scales that don't correspond. Before you compare two results, check the unit, the method and the lab's own interval.

I build Sera, a lab-review workspace for practitioners. The factors below come from its published range set, where every converted row is checked against its pair. I've kept the Sera part to the end.

Units that are the same number

These pairs look different on the report but describe the same quantity on the same scale. A ferritin of 42 ng/mL is a ferritin of 42 µg/L.

The rule behind most of them: a nanogram per milliliter is a microgram per liter, because both the mass and the volume have been scaled by a thousand.

Units that need a factor

These are different scales. Multiply the first unit by the factor to get the second.

For vitamin D, the NIH Office of Dietary Supplements pairs 50 nmol/L with 20 ng/mL and 125 nmol/L with 50 ng/mL, which is the 2.5 factor. Worked through: a result of 32 ng/mL is about 80 nmol/L.

HbA1c is a case of its own. The % (NGSP) and mmol/mol (IFCC) figures are linked by an equation, not a single factor, so you can't multiply your way across. The NGSP's master equation gives 5.6% as 38 mmol/mol and 5.4% as 36. Look the value up rather than estimating it.

Two habits help here. Copy the unit exactly as it's printed, and don't convert a result unless you have to. A conversion done by hand is one more place a typo gets in.

Why labs print different intervals for the same marker

Even when two labs use the same unit, the range next to the result can differ. A reference interval belongs to the lab's assay and the population it was built from. MedlinePlus says labs often use different reference ranges for normal results.

TSH is a good example. The American Thyroid Association describes the traditional reference range as 0.5 to 4.5 mIU/L, but laboratories print slightly different limits for their own assays, from about 0.3 to 0.45 at the bottom and 4.2 to 5.3 at the top (TSH sources). A borderline result needs reading against the interval of the lab that produced it.

The method matters too. With DHEA-S, laboratories disagree most for adults under 35, where immunoassays read higher than mass spectrometry. So a young client's lab may print a higher lower limit than a mass-spectrometry interval does. A result that looks low against one can look normal against the other, and the serum hasn't changed.

The thyroid hormones show the same thing: Sera's notes for free T3 and free T4 both say intervals differ between platforms. If the method isn't on the report, the lab's test directory may say.

When two results can't be compared

Sometimes the honest answer is that the two numbers don't belong on the same line.

Different scales from different assays. Lp(a) is reported in nmol/L by some labs and in mg/dL by others. The nmol/L and mass figures are paired cut-offs from different assays, not conversions of one another, so there's no factor to apply.

Antibodies where the manufacturer sets the scale. For TPO antibodies, one manufacturer's assay reports positive from 9 IU/mL, Labcorp's adult range runs to 34, and UK labs on a newer kit use 20. The number isn't comparable between laboratories. What you can compare is whether each result was positive against its own lab's cut-off.

Different things measured under one name. An EPA or DHA result can come from plasma or from red cells. Plasma fatty acids turn over in days to a couple of weeks and move with recent meals, while the red cell value reflects the last two to three months. Same marker name on the report, different measurement. Fatty acid totals have the same problem: which fatty acids a lab adds into the total differs between panels, so the same blood gives different totals at different labs.

Different conditions. A fasting and a non-fasting glucose, or a morning and an afternoon cortisol, aren't the same test, even at the same lab. TSH is more subtle, but it follows a daily rhythm and runs highest overnight and early in the morning, so write down the time of each draw.

A short checklist before you compare

1. Is the unit the same, an identical pair, or a real conversion?
2. Is the method the same, or at least the same kind of method?
3. What interval did each lab print, and does the result sit in the same place in both?
4. Were the conditions the same: fasting, time of day, cycle day, medications?
5. If any answer is no, can you still say something about direction, such as positive against both labs' cut-offs?

When results really can't be compared, the most reliable way to follow a marker is to retest at the same lab, by the same method, under the same conditions.

Where Sera fits

Sera keeps every result in the unit the lab reported and flags a marker that arrives in more than one unit, so you can see the switch rather than have it hidden. It doesn't convert units, and it doesn't store the lab's own reference interval, so keep the original report for that. Each result stays linked to its page in the report. Trends run across years and labs, with up to three markers on one chart, against bands that are published with their sources and an evidence grade. It doesn't diagnose or suggest treatment.

It's free for three clients, with no card. There's a 3-minute tour if you'd rather watch first.

Alina Radu is the founder of Sera and a software engineer. The idea started with her own Hashimoto's tests, scattered across years without the big picture.