Divide a Bazett-corrected QT by a Fridericia-corrected one and the answer depends on heart rate alone. On our arithmetic Bazett reads 12.2% higher than Fridericia at 120 beats a minute, whatever the tracing, and below 60 the gap reverses. The QT interval, the time the heart's ventricles take to fire and recover, shortens as the heart speeds up, so formulas "correct" it to the value expected at 60 beats a minute, where Bazett, Fridericia and Framingham all return the raw QT. A tracing at 110 beats a minute with a raw QT of 360 ms comes out at 487 ms by Bazett, above the lines in the American Heart Association's 2009 and 2010 statements for both sexes. Fridericia gives 441 ms and Framingham 430 ms, under the 2009 limits for either sex. The figures describe formulas and say nothing of any one heart.
Each correction builds on the same anchor of sixty beats a minute
Every formula works through the RR interval, the time between beats: 60 divided by the heart rate, in seconds. At 60 beats a minute RR is one second, so each correction leaves the QT untouched. Bazett divides the QT by the square root of RR, Fridericia by the cube root. Framingham adds a straight-line term, 154 ms multiplied by one minus RR.
Both power formulas date from 1920. Sagie and colleagues noted in 1992 that Bazett's came "from observations in only 39 young subjects." Their linear formula came from 5,018 Framingham Heart Study participants aged 28 to 62, at heart rates of 41 to 120 beats a minute, so beyond that range it is an extrapolation. The 2009 AHA/ACCF/HRS recommendations measure in the lead showing the longest QT, "usually V2 or V3".
Bazett readings hang on heart rate more than Fridericia readings do
Bazett's QTc divided by Fridericia's equals RR to the power of minus one sixth, so nothing about the tracing enters it. The table, our arithmetic, runs a raw QT of 400 ms through all three, rounded to the nearest millisecond, halves up.
| Heart rate (beats a minute) | Bazett ÷ Fridericia | Bazett (ms) | Fridericia (ms) | Framingham (ms) |
|---|---|---|---|---|
| 50 | 0.970 | 365 | 376 | 369 |
| 60 | 1.000 | 400 | 400 | 400 |
| 80 | 1.049 | 462 | 440 | 439 |
| 100 | 1.089 | 516 | 474 | 462 |
| 120 | 1.122 | 566 | 504 | 477 |
Real QTs shorten as the rate climbs, so the table describes no patient. In Vandenberk and colleagues' 2016 study of 6,609 patients at a Leuven hospital, Bazett and Fridericia differed by a mean of 24 ms at 80 to 90 beats a minute; our row for 80 gives 22 ms.
Readings on a fast heart turn on the formula chosen
A sharper test asks what raw QT reaches a line. At 120 beats a minute, the top of Framingham's fitted range, Bazett reaches the 2009 limit for men of 450 ms from a raw QT of 319 ms. Framingham needs 373 ms. Anything between is prolonged for a man by one formula and under the line by the other, a window of about 55 ms on our arithmetic; against the 460 ms line for women it is about 58 ms.
Slow rates are untidy too: at 50 beats a minute Fridericia reads above Bazett, yet Framingham reads below it for any raw QT under about 353 ms. Vandenberk's team ranked Fridericia and Framingham best and Bazett worst, in one Belgian hospital cohort that excluded heart rates above 90.
The QTc calculator on this site prints all three corrections side by side. Until 8 October 2026 it graded results normal, borderline or prolonged, with prolonged above 450 ms for men and 470 ms for women, and credited that scheme to the 2009 recommendations, which have two tiers and a women's line of 460 ms. It warned about heart rate only above 100 beats a minute, gave Framingham values outside 41 to 120 without a flag, and banded the unrounded value, so in 511 of 333,666 input combinations we ran the number shown contradicted its band. A revision committed that day bands the whole-millisecond figure on the 2009 and 2010 lines, without a borderline tier.
Normal and prolonged QTc limits depend on the question each source asked
The 2009 AHA/ACCF/HRS recommendations call a corrected QT of 450 ms or longer in men, and 460 ms or longer in women, prolonged. Drew and colleagues' 2010 AHA statement gives approximate 99th percentiles for "otherwise healthy postpubertal individuals" of 470 ms in men and 480 ms in women. The ICH E14 drug-trial guideline uses lines from 450 to 500 ms for both sexes and finds "no consensus" on the upper limit. The European Society of Cardiology's 2022 guideline uses 480 ms for long QT syndrome, an often inherited condition, but only "in repeated 12-lead ECGs" or with a points score above 3; we read that row as reproduced in a 2023 review by Wang and colleagues, because the guideline would not open for us.
A 99th-percentile limit means one healthy person in a hundred reads above it, so a value past it is unusual and settles nothing on its own. Drew describes where healthy people past puberty end, while E14 asks whether a medicine moves the interval.
The formula moves the healthy limit too. In Vandenberk's healthy subset the 99% upper limit by Bazett was 472 ms in men, 22 ms above the 450 ms "clinical standard" the team tested. In women it was 482 ms, 12 ms above the standard of 470.
Drug-trial guidance leaned on Fridericia after ICH changed course in 2012
In 2005 ICH E14 asked for data corrected by both Bazett and Fridericia in all applications, saying Bazett "overcorrects at elevated heart rates". Question 1.5 of the E14 Q&As, approved on 5 April 2012, dropped Bazett from that requirement unless there was a compelling reason, and called Fridericia "likely to be appropriate in most situations". We read the December 2015 version only.
The 2009 AHA/ACCF/HRS recommendations went another way, preferring linear regression corrections, Framingham's family, to Bazett's. Drew and colleagues wrote in 2010 that "QTc generally refers to the Bazett correction" unless stated otherwise, and forecast "increasing roles" for Fridericia. Irregular rhythms split the two AHA documents: the 2009 recommendations advise against rate correction when RR intervals vary widely, as often in atrial fibrillation, while Drew averages the QTc at the shortest and longest RR intervals.
Even the 500 millisecond line hinges on the formula
The drug-trial guideline and Drew's hospital statement share one figure for men and women alike: 500 ms. Drew calls a QTc above it "highly abnormal" and cites studies of congenital long QT syndrome in which it went with a 2- to 3-fold higher risk of torsade de pointes, a fast, disordered ventricular rhythm; that figure describes people with the inherited condition only. E14 calls it "a threshold of particular concern" in clinical trials. Beside it Drew sets a rise of "at least 60 ms" over the reading taken before a drug.
The line stays put; the tracing that crosses it does not. At 120 beats a minute, by our arithmetic, a raw QT of 354 ms takes Bazett over 500 ms, while Framingham needs 424 ms.
View author profile → · Editorial policy
