Phenytoin Correction Calculator
Correct measured total phenytoin levels for hypoalbuminaemia and renal impairment using the Winter-Tozer equation. Estimates the total phenytoin level the patient would have if their albumin were normal, enabling accurate therapeutic drug monitoring.
Calculate Corrected Phenytoin Level
Enter the measured total phenytoin level, serum albumin, and the patient’s renal function status. The calculator will apply the appropriate Winter-Tozer correction to estimate the adjusted total phenytoin level and the expected free (unbound) fraction.
Use this correction whenever the patient’s serum albumin is below 3.5 g/dL or when renal impairment (CrCl < 20 mL/min) or dialysis is present. In these patients, the measured total phenytoin level underestimates the pharmacologically active free fraction, and the corrected level provides a more accurate estimate of the effective drug exposure. When available, a directly measured free (unbound) phenytoin level is preferred and does not require correction.
Understanding Phenytoin Protein Binding
Phenytoin is approximately 90% bound to plasma proteins, primarily albumin. Only the unbound (free) fraction — normally about 10% of the total — is pharmacologically active, crosses the blood-brain barrier, and produces both therapeutic and toxic effects. Standard therapeutic drug monitoring measures the total phenytoin concentration (bound + free). This total level reliably predicts clinical effect only when protein binding is normal.
When albumin is low (hypoalbuminaemia), there are fewer binding sites available. A larger proportion of phenytoin circulates as free drug. The total level drops (because there is less bound drug), but the free level — the fraction that actually determines clinical effect — may remain the same or even increase. Without correction, the clinician may interpret a low total level as sub-therapeutic and inappropriately increase the dose, risking toxicity.
Standard Winter-Tozer Equation
Corrected PHT = Measured PHT ÷ (0.2 × Albumin + 0.1)
Use when: Albumin is low but renal function is normal (CrCl > 20 mL/min).
Example: Measured PHT 8 mg/L, Albumin 2.0 g/dL → 8 ÷ (0.2 × 2.0 + 0.1) = 8 ÷ 0.5 = 16.0 mg/L
Modified Equation — Renal Impairment
Corrected PHT = Measured PHT ÷ (0.1 × Albumin + 0.1)
Use when: CrCl < 20 mL/min or on haemodialysis. Uraemic toxins further displace phenytoin from albumin.
Example: Measured PHT 6 mg/L, Albumin 2.5 g/dL, on dialysis → 6 ÷ (0.1 × 2.5 + 0.1) = 6 ÷ 0.35 = 17.1 mg/L
Why does the coefficient change in renal failure? In uraemia, accumulated endogenous substances (indoxyl sulphate, hippuric acid, and other organic acids) compete with phenytoin for albumin binding sites, displacing it and further increasing the free fraction from the normal ~10% to as high as 20–25%. The modified equation (0.1 instead of 0.2) accounts for this additional displacement, reflecting the fact that each gram of albumin provides fewer effective phenytoin binding sites in the uraemic patient.
Therapeutic Ranges & Clinical Interpretation
The following table provides the standard therapeutic and toxic ranges for both total and free phenytoin concentrations. The corrected total level should be interpreted against the same thresholds as a directly measured total level in a patient with normal albumin.
| Category | Total Phenytoin (mg/L) | Total (µmol/L) | Free Phenytoin (mg/L) | Free (µmol/L) |
|---|---|---|---|---|
| Sub-therapeutic | < 10 | < 40 | < 1.0 | < 4.0 |
| Therapeutic | 10 – 20 | 40 – 80 | 1.0 – 2.0 | 4.0 – 8.0 |
| High Therapeutic / Early Toxicity | 20 – 30 | 80 – 120 | 2.0 – 3.0 | 8.0 – 12.0 |
| Toxic | 30 – 40 | 120 – 160 | 3.0 – 5.0 | 12.0 – 20.0 |
| Severely Toxic | > 40 | > 160 | > 5.0 | > 20.0 |
Phenytoin toxicity is concentration-dependent and presents in a classic progression. Nystagmus (especially lateral gaze) typically appears first at levels around 20 mg/L. Ataxia and dysarthria emerge at 30 mg/L. Lethargy and confusion develop at 40 mg/L. Above 50 mg/L, seizures may paradoxically worsen, and coma can occur. Cardiac toxicity (arrhythmias, hypotension) is primarily a concern with intravenous administration, especially if the infusion rate exceeds 50 mg/min.
Corrected Level vs Free Level — When to Prefer Which?
Factors Affecting Phenytoin Protein Binding
Beyond albumin and renal function, several additional factors alter phenytoin protein binding and may affect the accuracy of the Winter-Tozer correction. Understanding these factors helps clinicians interpret levels more accurately.
Several commonly used drugs compete with phenytoin for albumin binding sites, displacing it and increasing the free fraction. The most clinically significant displacing drugs include valproic acid (sodium valproate), which is the most potent displacer and can increase phenytoin free fraction from 10% to 15–20% at therapeutic valproate concentrations. Other important displacers include aspirin (high-dose, > 2 g/day), sulphonamides, tolbutamide, and phenylbutazone.
When valproate and phenytoin are co-prescribed, the total phenytoin level may appear deceptively low (because more drug is unbound and available for tissue distribution and metabolism), while the free level may be therapeutic or even toxic. The Winter-Tozer equation does not account for drug-displacement interactions — in these cases, a direct free phenytoin level is essential. Similarly, bilirubin displaces phenytoin at high concentrations (> 170 µmol/L), which is relevant in neonates and patients with severe hepatic jaundice.
- Valproic acid — most clinically important displacer
- High-dose aspirin (> 2 g/day)
- Sulphonamide antibiotics (sulfamethoxazole, sulfisoxazole)
- Hyperbilirubinaemia (> 170 µmol/L)
- Free fatty acids (elevated in critical illness, heparin infusion)
Phenytoin is one of the few drugs in clinical use that exhibits saturation kinetics (Michaelis-Menten pharmacokinetics) at therapeutic doses. Unlike most drugs where steady-state concentrations increase proportionally with dose, phenytoin’s hepatic metabolism becomes saturated at therapeutic levels. This means that small dose increases (e.g., from 300 to 350 mg/day) can produce disproportionately large increases in serum concentration, sometimes pushing levels from therapeutic into the toxic range.
This non-linearity has critical implications for dosing adjustments based on corrected levels. If the corrected total level is sub-therapeutic and a dose increase is needed, increments should be conservative — typically 30–50 mg/day at a time, with steady-state levels rechecked after 7–14 days. The time to reach a new steady state is also dose-dependent and can be prolonged (up to 2–4 weeks) at higher doses, compared to 5–7 days at lower doses. Never double the dose based on a sub-therapeutic level.
Hypoalbuminaemia is common in the patient populations most likely to require phenytoin correction. Key causes include chronic liver disease (reduced albumin synthesis), nephrotic syndrome (urinary albumin loss — often combined with renal impairment), critical illness (capillary leak, haemodilution, reduced synthesis), malnutrition (inadequate amino acid substrate), inflammatory states (albumin is a negative acute-phase reactant — inflammation shifts hepatic protein synthesis away from albumin toward CRP, fibrinogen, and other acute-phase proteins), and major burns or trauma (protein loss through damaged tissues).
In acutely ill patients, albumin levels can change rapidly. A patient admitted with a normal albumin may develop hypoalbuminaemia within 24–48 hours of a septic episode due to capillary leak and reduced synthesis. This means a phenytoin level that was correctly interpreted on admission may need re-correction as the clinical picture evolves. Always use the most recent albumin value when applying the Winter-Tozer equation.
Phenytoin levels should ideally be drawn as trough levels — immediately before the next scheduled dose. For oral phenytoin at steady state, the trough occurs 12–24 hours after the last dose (depending on formulation). For IV phenytoin or fosphenytoin, levels should be drawn at least 2 hours after the end of an infusion (4 hours for fosphenytoin, to allow complete conversion). Peak levels are not routinely used for phenytoin monitoring.
Steady state takes approximately 7–14 days at low-to-moderate doses, but can take up to 4 weeks at high doses due to saturation kinetics. Levels drawn before steady state will underestimate the eventual concentration and may lead to premature dose escalation. After a loading dose, a level drawn 2 hours post-infusion reflects the immediate post-load distribution but not the maintenance steady state.
The Winter-Tozer equation was derived and validated in ambulatory patients. In the ICU, multiple factors converge to make the correction less accurate: rapidly changing albumin levels, altered volume of distribution (oedema, third-spacing), concurrent displacing drugs (heparin releases free fatty acids, which displace phenytoin), renal and hepatic impairment, and hypothermia or hyperthermia (which alter protein binding affinity).
Studies in critically ill patients have shown that the Winter-Tozer equation overestimates the corrected phenytoin level in some cohorts and underestimates it in others, with prediction errors that can exceed 30–50%. For ICU patients, the consensus recommendation is to measure free (unbound) phenytoin directly whenever possible, rather than relying on the corrected total level. If a free level is unavailable, the corrected level should be interpreted cautiously as an approximation.
If the corrected phenytoin level does not match the clinical picture — for example, the corrected level is “therapeutic” but the patient has signs of toxicity (nystagmus, ataxia) — trust the clinical findings and obtain a direct free phenytoin level. The correction is an estimate; the patient’s clinical examination is the gold standard.
Special Populations
Phenytoin protein binding and the applicability of the Winter-Tozer correction vary across several key patient populations. In some groups, direct free phenytoin measurement is strongly preferred.
When to always measure free phenytoin directly: Pregnancy, neonates, concurrent valproate therapy, ICU patients, severe liver or renal disease, albumin < 2.0 g/dL, or whenever the corrected level seems discordant with clinical findings.
Common Pitfalls & Limitations
The most common error is interpreting a raw total phenytoin level without checking the albumin. A patient with an albumin of 2.0 g/dL and a total phenytoin of 8 mg/L appears sub-therapeutic — but the corrected level is 16 mg/L (well within the therapeutic range). Increasing the dose based on the uncorrected level would risk toxicity. This scenario is extremely common in hospitalised patients, many of whom have hypoalbuminaemia due to inflammation, malnutrition, or liver disease.
How to avoid: Always check the albumin before interpreting a total phenytoin level. If the albumin is below 3.5 g/dL, apply the Winter-Tozer correction. Better yet, request a free phenytoin level directly if your laboratory offers it.
The standard equation (0.2 × albumin + 0.1) and the renal-impairment equation (0.1 × albumin + 0.1) produce substantially different results. Using the standard equation in a dialysis patient with albumin 2.0 g/dL would yield a corrected level of 16 mg/L for a measured level of 8 mg/L. Using the correct renal equation gives 26.7 mg/L — potentially indicating toxicity rather than therapeutic levels. Misidentifying the patient’s renal status can lead to dangerously wrong corrections.
How to avoid: Always determine whether the patient’s CrCl is < 20 mL/min or whether they are on haemodialysis before selecting the equation. When in doubt, calculate with both equations and interpret the range, or obtain a free level.
The Winter-Tozer equation accounts for reduced albumin concentration but does not account for competitive displacement of phenytoin from albumin binding sites by other drugs. In a patient on valproate (the most common scenario), phenytoin’s free fraction is increased beyond what hypoalbuminaemia alone would explain. Applying the Winter-Tozer correction in this setting will underestimate the true free fraction, because the equation assumes that each gram of albumin binds phenytoin normally — but valproate is occupying some of those binding sites.
How to avoid: In any patient co-prescribed valproate and phenytoin, measure free phenytoin directly. The corrected total level is unreliable. This also applies to patients receiving high-dose aspirin or other known albumin-binding displacers.
Multiple validation studies have shown that the Winter-Tozer equation performs poorly in critically ill patients, with prediction errors exceeding 30% in some cohorts. The dynamic physiological derangements of critical illness — fluctuating albumin, third-spacing, heparin-induced free fatty acid release, variable renal and hepatic function — create conditions that the equation was not designed to handle. Several studies have found that the corrected level can be both higher and lower than the true equivalent, offering inconsistent guidance.
How to avoid: In the ICU, treat the corrected level as a rough estimate. If clinical decisions depend on the result (e.g., whether to load, increase, or hold phenytoin), obtain a direct free phenytoin level. If a free level is not available, interpret the corrected level in the context of clinical signs (presence or absence of nystagmus, sedation, seizure control) rather than relying on the number alone.
Even when the corrected phenytoin level is accurate, phenytoin’s saturation (Michaelis-Menten) kinetics mean that dose adjustments must be conservative. A corrected level of 7 mg/L (sub-therapeutic) does not mean the dose should be doubled — because the relationship between dose and concentration is not linear, a doubling of the dose could push the level well into the toxic range. Similarly, if the corrected level is 22 mg/L (mildly supra-therapeutic), reducing the dose by 50% could drop the level to sub-therapeutic concentrations.
How to avoid: Adjust phenytoin doses in increments of 30–50 mg/day (or one capsule, typically 30 or 100 mg). Recheck the level (with albumin) after at least 7–14 days. For urgent situations requiring rapid level change, use a supplemental partial loading dose calculated from the volume of distribution (approximately 0.7 L/kg).
Quick Reference Summary
Total PHT (mg/L)
Free PHT (mg/L)
(Normal Albumin)
(g/dL)
| Clinical Scenario | Which Equation / Action |
|---|---|
| Low albumin, normal renal function | Standard: PHT ÷ (0.2 × Alb + 0.1) |
| Low albumin + CrCl < 20 or dialysis | Renal: PHT ÷ (0.1 × Alb + 0.1) |
| Normal albumin, normal renal function | No correction needed — interpret total level directly |
| Concurrent valproate therapy | Correction unreliable — measure free phenytoin directly |
| ICU / critically ill patient | Correction is approximation only — prefer free level |
| Pregnancy (2nd/3rd trimester) | Correction unreliable — monitor free phenytoin |
The Golden Rule: The Winter-Tozer correction estimates what the total phenytoin would be if albumin were normal. When available, a directly measured free (unbound) phenytoin level is always more accurate. When in doubt between the corrected level and the clinical examination, trust the patient, not the equation.
Disclaimer & References
For Educational Purposes Only. This calculator and the accompanying clinical information are intended as educational tools for healthcare professionals. They do not replace clinical judgement. Results should be interpreted in the full clinical context. Lab reference ranges vary by institution — verify with your own laboratory. Drug dosages should be confirmed against current prescribing information.
References
- Winter ME, Tozer TN. Phenytoin. In: Evans WE, Schentag JJ, Jusko WJ, eds. Applied Pharmacokinetics: Principles of Therapeutic Drug Monitoring. 3rd ed. Applied Therapeutics; 1992:25.1-25.44.
- Anderson GD, Pak C, Doane KW, et al. Revised Winter-Tozer equation for normalized phenytoin concentrations in trauma and elderly patients with hypoalbuminemia. Ann Pharmacother. 1997;31(3):279-284. DOI: 10.1177/106002809703100301
- Bauer LA. Applied Clinical Pharmacokinetics. 3rd ed. McGraw-Hill; 2015. Chapter 10: Phenytoin.
- Patsalos PN, Berry DJ, Bourgeois BF, et al. Antiepileptic drugs — best practice guidelines for therapeutic drug monitoring. Epilepsia. 2008;49(7):1239-1276. DOI: 10.1111/j.1528-1167.2008.01561.x
- Wolf GK, McClain CD, Zurakowski D, Dodson B, McManus ML. Total phenytoin concentrations do not accurately predict free phenytoin concentrations in critically ill children. Pediatr Crit Care Med. 2006;7(5):434-439. DOI: 10.1097/01.PCC.0000228290.31032.1C
- Dager WE, Inciardi JF, Howe TL. Estimating phenytoin concentrations by the Sheiner-Tozer method in adults with acknowledged hypoalbuminemia. Ann Pharmacother. 1995;29(7-8):667-670. DOI: 10.1177/106002809502900701
- Kiang TK, Ensom MH. A qualitative review on the pharmacokinetics of phenytoin in pregnancy. Pharmacotherapy. 2014;34(7):724-732. DOI: 10.1002/phar.1419
- Kane SP, Bress AP, Gueorguieva I. Phenytoin binding and Sheiner-Tozer correction in critically ill patients. Ann Pharmacother. 2013;47(2):181-190. DOI: 10.1345/aph.1R516