Ionized Calcium vs Corrected Calcium
Ionized calcium is measured; corrected calcium is estimated. When the two disagree, which to trust, and the situations where ionized calcium is needed.
Ionized Calcium vs Corrected Calcium: Which Should You Trust?
You have a lab result that shows a total calcium of 2.10 mmol/L and a corrected calcium of 2.28 mmol/L. Your patient is hypoalbuminemic. Do you order the expensive ionized calcium test, or do you trust the free corrected value? Trust the ionized calcium. The corrected calcium is a calculated guess from 1973. It misclassifies the true ionized calcium status in up to 20-30% of sick patients. For any patient with an acid-base disturbance, critical illness, or abnormal protein profile, the corrected number is dangerously misleading.
Total, Ionized and Corrected Calcium: What Each Is
Total calcium is the measured serum concentration of all calcium fractions: protein-bound calcium (about 40%, mostly bound to albumin), complexed calcium (about 10%, bound to citrate, phosphate, and bicarbonate), and the free, ionized calcium fraction (about 50%). Ionized calcium (Ca²⁺) is the physiologically active form that controls muscle contraction, nerve conduction, and coagulation. It is measured directly by an ion-selective electrode, following strict pre-analytical protocols specified in CLSI C31-A2 (2001) and its 2017 edition, CLSI C31-Ed3.
Corrected calcium (also called adjusted calcium) is a calculated estimate of total serum calcium, mathematically adjusted for the concentration of serum albumin, intended to approximate the ionized calcium fraction. The honest version is that this correction is a crude, population-level statistical patch, not a measurement. The original Payne formula from 1973, corrected Ca = total Ca + 0.02 × (40 − albumin in g/L), was derived from a single UK hospital cohort of patients with abnormal serum proteins. It only adjusts for one cause of protein binding variation and fails precisely in the sickest patients.
Ionized Calcium Reference Range and Sample Handling
The reference range for ionized calcium is narrow: 1.15-1.30 mmol/L (UK Pathology Harmony range). For corrected calcium, the standard is 2.20-2.60 mmol/L. These are not interchangeable. A corrected calcium of 2.30 mmol/L can correspond to an ionized calcium anywhere from 1.00 to 1.40 mmol/L in a sick patient.
Sample Handling Errors That Destroy Reliability
Ionized calcium measurement requires an anaerobic sample drawn without a tourniquet, or with a tourniquet applied for less than one minute. Prolonged tourniquet use causes venous stasis, which lowers pH and falsely increases ionized calcium. A pH change of 0.1 unit changes ionized calcium by about 0.05 mmol/L. The sample must be collected in a heparinized syringe (not EDTA, which chelates calcium) and kept capped to prevent CO₂ loss. If the sample is exposed to air, CO₂ diffuses out, pH rises, and ionized calcium falls. CLSI C31 guidance requires analysis within 30 minutes at room temperature or 2 hours if stored on ice. Failure to follow these rules invalidates the result.
How Well Corrected Calcium Predicts Ionized Calcium: The Evidence
The evidence is clear: the Payne equation works reasonably well in patients with normal albumin and no acid-base disturbance, but it fails in the populations that need it most. Lian & Asberg (2018, Scandinavian Journal of Clinical and Laboratory Investigation) studied renal patients and found the bias between corrected and ionized calcium was −0.06 mmol/L, but the limits of agreement were wide: −0.35 to +0.23 mmol/L. That means a corrected value of 2.30 mmol/L could correspond to an actual ionized calcium anywhere from 1.95 to 2.53 mmol/L, a range that spans both hypocalcemia and normal.
Ladenson et al. (1978, Clinical Chemistry) studied ICU patients and reported that the correction failed to identify true hypocalcemia in 30% of cases. Smith et al. (ICU cohort) and Kenny et al. (CKD cohort) both replicated these findings: the misclassification rate in critically ill and renal patients is consistently 20-30%. The KDIGO 2017 CKD-MBD guideline explicitly advises against using albumin-adjusted calcium as the primary measure in CKD. The KDOQI US commentary (2018) reinforces this caution. The Royal College of Pathologists (2019) also warns that the formula does not account for pH, citrate, phosphate, or protein binding changes.
| Attribute | Corrected Calcium | Ionized Calcium |
|---|---|---|
| Agreement with true ionized calcium | Poor in ICU/CKD (bias up to 0.2 mmol/L) | Gold standard, direct measurement |
| Sensitivity for hypercalcemia | Reasonable in mild hypoalbuminemia, falls in critical illness | High, method-dependent |
| Specificity for hypocalcemia | False positives common in hypoalbuminemia | High, with proper sample handling |
| Albumin range of validity | Narrow, 30–50 g/L; error increases outside | Not affected by albumin level |
| Time from draw to analysis | Stable for hours (total calcium) | Unstable, must be analyzed within 30 min |
| Cost | Free (calculated) | Extra charge, typically $15–50 |
| Turnaround time | Immediate | 30–60 min |
| Robustness to hemolysis | Falsely lowers total calcium | Interference common |
| Clinical actionability | Misleading in sick patients | Directly actionable |
| Evidence base for use | Decades of data showing failure | Extensive validation |
When to Measure Ionized Calcium
Order ionized calcium directly in these specific clinical scenarios. For every one of these, the corrected calcium is unreliable enough to change management.
ICU and Critically Ill Patients
Acid-base disturbances are the rule, not the exception, in the ICU. A patient on a ventilator with a pH of 7.25 will have a higher ionized calcium than a pH 7.45 patient with the same total calcium. The corrected calcium cannot account for this. Ladenson et al. found that the correction missed hypocalcemia in 30% of ICU patients. If you are managing a trauma or sepsis patient, order the direct measurement.
Chronic Kidney Disease (CKD)
KDIGO 2017 explicitly recommends against albumin-adjusted calcium in CKD. The Payne equation assumes a normal albumin binding constant, which is altered by the dysproteinemia and metabolic acidosis of renal failure. Lian & Asberg showed that in renal patients, the corrected value misclassifies calcium status in up to 25% of cases. For dosing calcimimetics or bisphosphonates, trust only the ionized calcium.
After Blood Transfusion or Albumin Infusion
Citrate in transfused blood chelates calcium, and albumin infusion adds binding sites. Both events change the relationship between total and ionized calcium unpredictably. The corrected calcium will show a normal value while the patient is actually hypocalcemic. Measure ionized calcium 15-30 minutes after the infusion ends.
Very Low or Very High Albumin
The Payne equation is only validated for albumin between roughly 30 and 50 g/L.For a patient with nephrotic syndrome or liver failure, the corrected number is fiction.
Acid-Base Disturbances
Why pH changes ionized calcium: hydrogen ions compete with calcium for binding sites on albumin. In acidosis, more calcium is displaced into the free, ionized form. In alkalosis, more calcium binds to albumin, reducing the free fraction. A pH change of 0.1 unit shifts ionized calcium by about 0.05 mmol/L, enough to misclassify a borderline result. The corrected calcium has no mechanism to adjust for pH.
Who Should Use Ionized Calcium and Who Should Not
Ionized calcium suits clinicians who need to make treatment decisions, ordering bisphosphonates, calcimimetics, or vitamin D analogs, in patients with abnormal protein profiles or acid-base disturbances. It suits pharmacists who need to interpret calcium results in patients on albumin or with CKD. It suits lab scientists who need to guide clinicians away from the deceptive security of the calculated value.
Ionized calcium does not suit researchers using corrected calcium as a primary outcome in populations with normal protein profiles. It does not suit veterinarians, because the Payne equation is calibrated for human albumin. And it does not suit anyone with a known monoclonal gammopathy or paraproteinemia, because the correction is invalid for them entirely. The single thing that most often goes wrong here: a clinician trusts the corrected calcium in an ICU patient, the number comes back normal, the patient is actually hypocalcemic, and no one acts in time.
Common Questions
What is the normal range for ionized calcium?
The UK Pathology Harmony ionized calcium reference range is 1.15-1.30 mmol/L. This is the physiologically active fraction. For corrected calcium, the normal range is 2.20-2.60 mmol/L. These ranges are method- and lab-specific, so always verify with your laboratory's reported values.
Why is it called free calcium?
Free calcium is another term for ionized calcium. It refers to the calcium that is not bound to proteins or other molecules. About 50% of total serum calcium exists in this free form, which is the only fraction that is biologically active. Total calcium includes the free, protein-bound, and complexed fractions.
When should I order ionized calcium instead of total or corrected calcium?
Order ionized calcium directly in ICU patients, patients with CKD (per KDIGO 2017), after blood transfusion or albumin infusion, in patients with very low or high albumin (below 20 g/L or above 55 g/L), and in any patient with an acid-base disturbance. In these populations, the corrected calcium misclassifies calcium status in 20-30% of cases.
Is total vs ionized calcium a simple comparison?
No. Total calcium includes all fractions, while ionized calcium measures only the active form. In a patient with normal albumin and pH, total calcium times 0.5 gives a rough estimate of ionized calcium. But in any sick patient, that ratio is unreliable. Only direct ionized calcium measurement is trustworthy.
Can I use the corrected calcium formula if my lab reports albumin in g/dL?
Yes, but you must convert. The Payne equation uses g/L. If albumin is 4.0 g/dL, that is 40 g/L. The formula is: corrected Ca (mmol/L) = total Ca (mmol/L) + 0.02 × (40 − albumin in g/L). Using g/dL without conversion introduces an arithmetic error. The same formula rearranged with a normal albumin of 3.5 g/dL is algebraically identical, not a separate method.
Does the KDIGO guideline recommend any calcium test for CKD?
KDIGO 2017 explicitly recommends against using albumin-adjusted calcium as a surrogate for ionized calcium in CKD. For monitoring calcium status, the guideline prefers direct measurement of ionized calcium when available. If ionized calcium is not available, total calcium with clinical context is considered more reliable than the corrected value.
Why does pH change ionized calcium but not corrected calcium?
Hydrogen ions compete with calcium for binding sites on albumin. In acidosis (low pH), more calcium is displaced into the free, ionized form. In alkalosis (high pH), more calcium binds to albumin, reducing ionized calcium. The corrected calcium calculation has no pH term, it assumes normal binding, which fails in any acid-base disturbance. A pH change of 0.1 unit shifts ionized calcium by about 0.05 mmol/L.