Corrected Phenytoin Calculator

Only unbound drug crosses the blood-brain barrier and blocks voltage-gated sodium channels, yet routine laboratories report the total of bound plus unbound phenytoin. This tool normalises that total back to what it would read if the patient carried a normal albumin, so the number can be compared against the familiar 10 to 20 mcg/mL band that was itself established in patients with normal protein. Phenytoin binds mainly to site II of the single albumin molecule, and the free fraction is about 10 percent when albumin sits in the reference interval of 3.5 to 5.0 g/dL. Halving albumin roughly doubles that fraction. A practical consequence is that in nephrotic syndrome, cirrhosis, burns, sepsis, or late pregnancy a reported total of 6 to 8 mcg/mL frequently represents fully therapeutic unbound drug, and increasing the dose on the raw number is a common route to toxicity. The uremic form of the calculation exists because renal failure alters binding by a second, independent mechanism. Retained organic acids such as hippuric and indoxyl sulphate occupy albumin binding sites, and carbamylation of albumin lowers its affinity, so free fraction rises to roughly 20 to 25 percent even at a normal albumin concentration. Applying the uremic coefficient to a patient with preserved kidney function will overstate the corrected value substantially. Other caveats: albumin measured by bromocresol green tends to read higher than bromocresol purple, especially in renal patients, which propagates directly into the answer; the equation assumes binding is linear across the whole albumin range; and it cannot represent competitive displacement by valproate. It is an estimate, not a substitute for an ultrafiltration free level.

The idea that drug effect follows the unbound rather than the total concentration was worked out in the middle of the twentieth century, as equilibrium dialysis and ultrafiltration made it possible to separate free from protein-bound drug and Scatchard analysis let chemists quantify binding sites and affinities on albumin. Applying that to therapeutics required assays sensitive enough for microgram-per-millilitre concentrations, which arrived with gas-liquid chromatography in the 1960s and with enzyme immunoassays in the following decade. Phenytoin, with a narrow therapeutic window, heavy protein binding, and saturable metabolism, became the prototype drug for the new discipline of therapeutic drug monitoring, and hospital pharmacokinetics services grew up around it during the 1970s. Investigators studying patients with renal failure in that period repeatedly found that acidic, highly bound drugs behaved abnormally in uremic serum, binding far less than predicted from albumin concentration alone. Those observations produced the separate coefficient still used for uremia today. The correction equation reached wide clinical use through Applied Pharmacokinetics, the reference text in which Michael E. Winter and Thomas N. Tozer of the University of California, San Francisco set out the relationship around 1980; the literature also attributes it to Lewis Sheiner, and it appears under both the Winter-Tozer and Sheiner-Tozer names. Since the late 1990s several groups have published revised coefficients for trauma, critical care, and elderly populations, and many centres now bypass the estimate entirely by ordering a directly measured free phenytoin level.