How this instrument works
"Serum osmolality" and "plasma osmolality" describe the identical physical quantity — how strongly the liquid portion of blood pulls water across a membrane — and clinicians use the two terms interchangeably in practice; serum is simply plasma with the clotting factors removed, and the difference doesn't meaningfully change the osmotic reading. This site carries two separate calculators for it anyway, because two genuinely different formulas have been published for estimating that one quantity from routine labs, and each has its own legitimate use rather than one simply being a typo of the other.
Dorwart and Chalmers took a different approach than the quick-arithmetic shortcut used elsewhere. In 1975 they gathered 715 hospital patients' sodium, potassium, glucose, and BUN readings alongside a directly measured osmolality for each, then tested thirteen candidate formulas from the existing literature against that dataset using statistical regression. The combination that came out on top wasn't a clean doubling of sodium — it was 1.86 times sodium, plus glucose divided by 18, plus BUN divided by 2.8, plus a flat constant of 9, coefficients chosen because they minimized the gap against the actual lab measurements rather than because they were easy to compute in one's head.
The practical upshot is a formula that tends to sit closer to a real osmometer reading than the simpler doubled-sodium shortcut, at the cost of being one step harder to do without a calculator. Neither published equation is "wrong" — the simpler one earned its place through decades of emergency-medicine teaching precisely because it's fast, while this regression-fitted version earned its place through Dorwart and Chalmers actually checking the arithmetic against measured blood. Worth showing a reader both, rather than silently picking one and hiding the other.
- Enter Sodium in mEq/L from a basic metabolic panel.
- Enter Glucose in mg/dL from the same panel.
- Enter BUN in mg/dL, also from the same panel.
- Read the resulting osmolality in mOsm/kg, then compare it against a lab-measured value if one is available.
Worked example — Na 130, glucose 180, BUN 28
A patient's basic metabolic panel shows sodium 130 mEq/L, glucose 180 mg/dL, and BUN 28 mg/dL. The sodium term: 1.86 × 130 = 241.8. The glucose term: 180 / 18 = 10. The BUN term: 28 / 2.8 = 10. Add the constant of 9 and the total is 241.8 + 10 + 10 + 9 = 270.8 mOsm/kg.
Run the same three numbers through the simpler doubled-sodium shortcut instead — 2 × 130 + 10 + 10 = 280 mOsm/kg — and the two published formulas land 9.2 mOsm/kg apart on identical inputs. Neither arithmetic mistake explains the gap; the regression coefficient of 1.86 was fitted specifically to track a real osmometer more closely than a flat doubling does, and the constant of 9 exists purely because that's what the 1975 regression required to minimize error across 715 patients' worth of measured blood.
Questions
Is serum osmolality different from plasma osmolality?
Not in any way that matters clinically. Serum is plasma with the clotting proteins removed, and that removal doesn't meaningfully change how strongly the fluid pulls water osmotically. Clinicians and lab reports use the two terms interchangeably; the reason this site hosts two separate calculators is that two distinct published formulas exist for estimating the value, not that the underlying quantity itself differs.
Why does this formula use 1.86 and add 9 instead of simply doubling sodium?
Because those numbers came from statistics rather than convenience. Dorwart and Chalmers tested thirteen candidate formulas against 715 patients' directly measured osmolality in a 1975 study and found this particular combination — 1.86 times sodium, plus the glucose and BUN correction terms, plus a constant of 9 — produced the smallest average error against the real lab measurements. The simpler doubled-sodium version used elsewhere is easier to compute but was not the best statistical fit in that comparison.
Which formula should I actually trust — this one or the simpler doubled-sodium version?
For a rough bedside read in a hurry, the doubled-sodium shortcut is faster and close enough for most triage purposes. When precision matters more than speed — comparing against a lab-measured osmolality to calculate an osmolar gap, for instance — this regression-fitted version has the better empirical track record, since it was built specifically to minimize the gap against real measurements rather than to be easy arithmetic.
What does an abnormal result from this calculation actually indicate?
A high value typically reflects dehydration, poorly controlled blood sugar, or reduced kidney function raising BUN; a low value often points to excess free water intake or a low-sodium state. As with any calculated osmolality, the more clinically powerful step is comparing it against a directly measured lab value — a large gap between the two suggests an unmeasured substance, such as a toxic alcohol, is contributing to the osmotic load.
Does potassium factor into this equation at all?
No, even though Dorwart and Chalmers measured potassium alongside sodium, glucose, and BUN in their original dataset. Potassium's contribution turned out to be small enough that the best-fitting formula from their regression didn't need it as a separate term — sodium alone, doing double duty for its accompanying anions, already captured most of the signal.
References
- Dorwart WV, Chalmers L. Comparison of methods for calculating serum osmolality from chemical concentrations, and the prognostic value of such calculations. Clin Chem. 1975;21(2):190-194 (PubMed)
- StatPearls — Serum Osmolality (NCBI Bookshelf)
Read this first: This instrument computes a screening figure from population formulas — it is not a diagnosis, and it cannot see the whole picture a clinician can. Use it to inform a conversation, not to replace one.