Clinical Overview

The QT interval is measured from the onset of the QRS complex to the end of the T wave, and it represents the total time the ventricles spend depolarizing and then repolarizing — roughly, the duration of the ventricular action potential (Al-Akchar and Siddique, StatPearls Long QT Syndrome, 2022; LITFL QT Interval, 2024). “QT Interval Extension” is the source dataset’s own telegraphic label for that interval being too long. The term a clinician writes is QT prolongation, or a prolonged QT interval, which is what this page is titled and scoped to.

The single most important distinction on this page is the one between a finding and a syndrome, because the two are routinely spoken about as if they were the same thing and they are not. QT prolongation is an ECG finding — a number you measure off a tracing, and one that is far more often acquired than inherited, produced by medications, electrolyte disturbance, or a slow heart rate in a structurally ordinary heart. Long QT Syndrome is a diagnosis — in its congenital form an inherited channelopathy, caused by mutations affecting cardiac ion channels, that produces a prolonged QT interval as one of its manifestations and carries a lifelong arrhythmic risk requiring genotype-aware management (Al-Akchar and Siddique, StatPearls, 2022; Farjam et al., Clinical Cardiology, 2026). Every patient with the syndrome has the finding; the large majority of patients with the finding do not have the syndrome. The dataset’s “QT Interval Extension” label denotes the finding, not the syndrome, and nothing about a record carrying it implies an inherited diagnosis. That difference also changes the threshold you would apply — the European Society of Cardiology’s diagnostic criterion for the syndrome is a Bazett-corrected QTc of 480 ms or more regardless of sex, or 460 ms or more in a symptomatic individual, which is a deliberately higher bar than the level at which a monitoring technician would simply report the interval as long (Robyns, ESC Council on Cardiovascular Genomics, 2024).

That scoping decision is reinforced by this project’s own SNOMED cross-check. The check found that the code carried by the QTIE label resolves to the standard “prolonged QT interval” concept and classed the mapping clean — this page is grounded in a correct mapping. The registry does, separately, carry a second label whose name is “Prolonged QT Interval” (LQT), and the same cross-check found that label’s code resolves to an unrelated ventricular rhythm concept rather than to any QT-interval finding, which is why its page is currently blocked pending a human decision. [CLINICAL REVIEW NEEDED: the registry therefore holds two overlapping QT-lengthening labels — one correctly coded and one mis-coded — and the dataset does not document what distinguished them for the original annotators. Until that is resolved, treat this page as the prolonged-QT-interval page and do not read the two labels as describing different entities. The congenital syndrome itself is deliberately left to a separate page rather than folded in here.]

The mechanism is repolarization, not depolarization. Ventricular repolarization depends heavily on outward potassium currents, and the delayed rectifier current carried by the channel encoded by the hERG gene is the one most drugs interfere with. Blocking it slows repolarization, which lengthens the action potential and therefore the QT interval (Farzam and Tivakaran, StatPearls QT Prolonging Drugs, 2023). Congenital long QT syndrome arrives at the same endpoint through inherited channel defects rather than a drug, and the current understanding extends well past the three classic genes to dual-function channel mutations and post-translational defects (Farjam et al., 2026). A repolarization that is both slow and regionally uneven is the substrate for early afterdepolarizations, which can capture and launch torsades de pointes — a polymorphic ventricular tachycardia whose QRS complexes twist around the isoelectric line, classically triggered by a premature ventricular beat landing on the T wave after a short-long-short sequence of R-R intervals (Cohagan and Brandis, StatPearls Torsade de Pointes, 2023).

Torsades is the reason this finding matters, and it is worth being precise about how strong the link actually is. Sources consistently identify a QTc above 500 ms as the level at which concern rises sharply; StatPearls puts the associated increase at roughly two- to three-fold (Cohagan and Brandis, 2023), and LITFL associates a QTc above 500 ms with an increased risk of torsades de pointes (LITFL QT Interval, 2024). What the sources do not support is a bright line. Risk is graded and heavily modified by context — a 2026 integrative review is explicit that a QTc threshold alone is insufficient for stratification, and that T-wave morphology, genotype, electromechanical window dynamics, and multi-factor risk scores add prognostic information a single number does not carry (Farjam et al., 2026). The practical reading for a monitoring role is that 500 ms is an escalation trigger, not a diagnosis, and that a QTc under 500 ms in a patient with several risk factors is not automatically safe.

The finding itself produces no symptoms. A long QT interval is silent until it degenerates into an arrhythmia, at which point the presentation is syncope — the most common symptom, and in the congenital form characteristically provoked by exercise or intense emotion — along with near-syncope, palpitations, seizure-like activity, or cardiac arrest (Al-Akchar and Siddique, 2022; Cohagan and Brandis, 2023). A great many strips carrying this label come from patients who felt nothing at all.

Causes divide into acquired and congenital, and for a monitoring audience the acquired ones are where nearly all of the day-to-day work is. Medications are the largest single group, acting through hERG potassium-channel blockade; recognized offenders span antiarrhythmics (amiodarone, sotalol, dofetilide, procainamide, quinidine), antipsychotics (haloperidol, droperidol, ziprasidone, quetiapine, thioridazine, olanzapine, risperidone), macrolide and fluoroquinolone antibiotics, tricyclic antidepressants and citalopram, methadone, ondansetron, and cisapride (Farzam and Tivakaran, 2023). Risk compounds when two such drugs are combined, or when a CYP450 inhibitor raises the plasma level of one of them (Farzam and Tivakaran, 2023). A 2026 prospective hospital study is a useful reality check on how this looks on a ward: most alerts were driven by combinations of merely low- or moderate-risk drugs — particularly antipsychotics, antidepressants, and antiemetics — and as-needed prescriptions contributed substantially to the accumulated risk even when they had barely been administered (Simona et al., Frontiers in Pharmacology, 2026). Electrolyte disturbance is the second group: hypokalemia, hypomagnesemia, and hypocalcemia (LITFL QT Interval, 2024; Cohagan and Brandis, 2023). Bradycardia is the third, and it is doubly important because it both lengthens the raw interval and supplies the pauses that pause-dependent torsades needs (Cohagan and Brandis, 2023). Other recognized causes include hypothermia, myocardial ischemia, the post-arrest return-of-spontaneous-circulation state, raised intracranial pressure, and congenital long QT syndrome itself (LITFL QT Interval, 2024). Additional risk factors that raise the chance any of the above tips into torsades include age over 65, female sex, structural heart disease, and diuretic therapy (Cohagan and Brandis, 2023); metabolic factors including insulin resistance and adiposity have also been reported as independent contributors (Farjam et al., 2026).

Interpretation Guide

Key Features:

• Rate: not a defining feature in itself, but it is an input to the finding rather than a bystander — the raw QT lengthens as the rate slows and shortens as it rises, which is the entire reason a correction formula exists. Always record the heart rate the measurement was taken at, because a QT reported without its rate cannot be corrected or checked by anyone downstream
• Rhythm: not a defining feature. The interval is measured against whatever rhythm is present, and the label says nothing about where the impulse originates. Where the R-R intervals vary from beat to beat, a single corrected value is unstable, which is one reason the standard advice is to measure several successive beats and take the maximum (LITFL QT Interval, 2024)
• P waves: not part of the QT measurement and not a defining feature of this finding
• PR interval: within normal limits unless a separate, coexisting conduction abnormality is present. It is not part of any QT criterion
• QRS complex: read this before reporting any QT value. The QT is measured from QRS onset, so every millisecond of QRS widening is added directly to the measured QT without representing one millisecond more of repolarization. Conventional measurement therefore overestimates the interval in the presence of left or right bundle branch block, and several dedicated correction formulas exist for that setting — a 2026 cohort of left bundle branch block patients applied five of them (Rautaharju, Yankelson, Wang, Bogossian, Colunga), found prolonged QTc by every one of them associated with all-cause mortality, and reported the Rautaharju formula as the best performer for risk stratification (Xu et al., BMC Cardiovascular Disorders, 2026). A 2021 systematic review reached the same conclusion about the overestimation and catalogued eight such methods, only three of which are applicable in right bundle branch block; that review falls just outside this page’s five-year recency window and is cited here only because the in-window cohort above independently confirms the same point (Fredholm et al., Journal of Clinical Psychopharmacology, 2021). The practical rule is simple — report the QRS duration alongside the QT, and do not apply the ordinary QTc thresholds to a wide-QRS tracing without saying that you did
• ST segment: worth inspecting for where the extra length sits. Hypocalcemia prolongs the QTc primarily by stretching the ST segment while leaving the T wave itself essentially unchanged, which is a distinctive appearance — a long, flat, isoelectric run before a normal-looking T wave (LITFL Hypocalcemia, 2024)
• T waves: the T wave carries the endpoint, so its shape determines how reproducible the measurement is. The tangent method (also called the maximum slope intercept method) defines the end of the T wave as the point where a tangent drawn along the steepest part of the T wave’s downslope crosses the isoelectric baseline; it is easier to teach and more reproducible between readers than eyeballing the return to baseline, though it may slightly underestimate the QTc by cutting off the final phase of repolarization — ESC records no expert consensus on which of the tangent and threshold methods is preferable (LITFL QT Interval, 2024; Robyns, ESC, 2024). Flat, broad, notched, or biphasic T waves make the endpoint genuinely ambiguous; biphasic T waves are included in the measurement (Robyns, ESC, 2024)
• QT interval: the defining feature. Measure from the onset of the QRS complex to the end of the T wave, in lead II or V5 by preference, and where readings differ between leads use the lead giving the longest measurement; measure several successive beats and take the maximum (LITFL QT Interval, 2024; Robyns, ESC, 2024). A quick sanity check taught alongside the formal measurement is that a normal QT is less than half the preceding R-R interval (LITFL QT Interval, 2024). Then correct it, and say which formula you used — the four in common use are Bazett (QTc = QT ÷ √RR), Fridericia (QTc = QT ÷ RR^⅓), Framingham (QTc = QT + 0.154 × (1 − RR)), and Hodges (QTc = QT + 1.75 × (heart rate − 60)), with RR in seconds (LITFL QT Interval, 2024). They are not interchangeable. Bazett over-corrects above 100 bpm and under-corrects below 60 bpm, so Fridericia or Framingham are more accurate outside the 60-100 bpm band (LITFL QT Interval, 2024). The size of the disagreement is not academic — across 19,955 ECGs from 6,881 patients, the Bazett value ran a median of 26.4 ms above Fridericia and 27.8 ms above Framingham, and of the tracings Bazett graded as severe QTc prolongation, 81.0% were graded less severe by Fridericia or Framingham (Richardson et al., JAMA Oncology, 2022). A smaller 2025 retrospective series found the same ordering, with mean QTc of 445 ± 30 ms by Bazett against 426 ± 29 ms by Fridericia, 424 ± 28 ms by Framingham, and 428 ± 29 ms by Hodges, and recommended Fridericia for its lower variability across differing heart rates (Shah et al., Cureus, 2025). As for how long is too long, LITFL reports the interval as prolonged above 440 ms in men and above 460 ms in women, associated with increased torsades risk above 500 ms, and abnormally short below 350 ms (LITFL QT Interval, 2024); the StatPearls Long QT Syndrome chapter uses under 440 ms in men and under 460 ms in women as normal and names Bazett as the formula most commonly applied (Al-Akchar and Siddique, 2022), while the StatPearls Torsade de Pointes chapter gives the male upper limit as 450 ms (Cohagan and Brandis, 2023). [CLINICAL REVIEW NEEDED: the sources read for this page disagree on the male cutoff (440 vs 450 ms) and none of them states in so many words which correction formula their general-purpose thresholds were derived against, though Bazett is the formula each names as the one in common use and the ESC syndrome criteria above are explicitly Bazett-based. Given the 25-30 ms spread Richardson et al. report between Bazett and both Fridericia and Framingham, a threshold quoted without its formula is not interpretable, and a learner should be taught to report the formula every time rather than to memorize one number.]
• Other findings: prominent U waves, best seen in V2-V3, are the classic accompaniment of hypokalemia — and in that setting the apparently long QT is often not a QT at all but a QU interval, formed by a T wave that has flattened away and fused with the U wave (LITFL Hypokalemia, 2024). Short-long-short R-R sequences and premature ventricular beats falling on the T wave are the immediate mechanical trigger for torsades and are worth flagging on a strip that already shows a long QT (Cohagan and Brandis, 2023)

Key Leads

• Lead II – One of the two leads standard guidance names for QT measurement, and the usual default on a monitored patient. Its P and T waves are typically well formed and upright, which makes the start and end of the interval easier to place (LITFL QT Interval, 2024; Robyns, ESC, 2024)
• Lead V5 – The other named default, and the one to reach for when the T wave in lead II is flat or biphasic. V6 serves the same role when the terminal T wave is hard to resolve (LITFL QT Interval, 2024; Robyns, ESC, 2024)
• Whichever lead gives the longest QT – Not a fixed lead but a rule, and the one that matters most when leads disagree. Standard guidance is to use the longest measurement rather than to average across leads, and to repeat over several successive beats and take the maximum (LITFL QT Interval, 2024)
• Leads V2 and V3 – Where U waves are best seen, which is exactly what you need in order to know whether you are measuring a QT or a QU. In hypokalemia the T wave flattens and fuses with a prominent U wave here, manufacturing an apparently prolonged interval that is really a QU interval with an absent T wave (LITFL Hypokalemia, 2024)
• Any lead showing a wide QRS – Not a measurement lead but a disqualifying one. If the QRS is broad, the ordinary thresholds do not apply to the measured interval, because the widening inflates the QT without lengthening repolarization (Xu et al., 2026)

Differential Diagnosis

• U Wave (UW) — the single most common reason a QT is reported as long when it is not. Distinguishing clue: find the end of the T wave, not the end of the last visible deflection. Sources differ on the rule, which is worth knowing rather than glossing over — the ESC guidance is to exclude U waves from the measurement outright (Robyns, ESC, 2024), while LITFL advises including a large U wave (over 1 mm) that is fused to the T wave and excluding smaller or clearly separate ones (LITFL QT Interval, 2024). Where the T wave has flattened away entirely and merged into a prominent U wave — the hypokalemia picture, best seen in V2-V3 — what you are measuring is a QU interval, not a QT (LITFL Hypokalemia, 2024). [CLINICAL REVIEW NEEDED: the include-large-fused-U-waves rule and the exclude-U-waves rule are not reconcilable and will give different numbers on the same strip.]
• Complete Right Bundle Branch Block (CRBBB) — the finding that lengthens the measured QT without lengthening repolarization. Distinguishing clue: measure the QRS first. At or above 120 ms with an RSR’ in V1 and a broad slurred S wave in I and V6, the extra QT duration is bought at the front end by slow conduction, not at the back end by slow repolarization, and conventional measurement overestimates the interval in this setting. Report the QRS duration with the QT, and note that dedicated bundle-branch-block correction formulas exist for exactly this problem (Xu et al., 2026; Fredholm et al., 2021).
• Sinus Bradycardia (SB) — the rate explanation. Distinguishing clue: the raw QT lengthens as the rate slows, so a long absolute QT at 45 bpm may correct to an entirely normal QTc. This is where formula choice bites hardest, because Bazett under-corrects below 60 bpm and can therefore under-report a genuinely prolonged interval in a bradycardic patient; Fridericia or Framingham are more accurate outside the 60-100 bpm band (LITFL QT Interval, 2024). Bradycardia is also a torsades risk factor in its own right, because it supplies the pauses that pause-dependent torsades requires (Cohagan and Brandis, 2023).
• T Wave Change (TWC) — the finding that makes the endpoint unreliable. Distinguishing clue: a flat, broad, notched, or biphasic T wave does not itself prolong repolarization, but it moves where a reader places the end of the T wave and so changes the reported number. The tangent method places that endpoint more reproducibly than eyeballing the return to baseline, though ESC records no expert consensus on the preferred method; where the T wave is genuinely unmeasurable, say so rather than reporting a value with false precision (LITFL QT Interval, 2024; Robyns, ESC, 2024).
• ST Extension (STE) — the label that answers where the length came from. Distinguishing clue: hypocalcemia lengthens the QTc primarily by stretching the ST segment while leaving the T wave morphology essentially unchanged, so a long, flat, isoelectric run followed by a normal-looking T wave points toward calcium rather than toward a drug or a channelopathy (LITFL Hypocalcemia, 2024). Note that despite its name, this project’s own cross-check found the STE label’s code to be the generic “ST interval abnormal” concept, so the label itself does not assert a lengthened ST segment — the reading clue stands on the tracing, not on the label.

Treatment Brief

The interval itself is not treated. What a monitoring technician or nursing student owns here is measuring it defensibly, reporting it in a form the next reader can act on, and knowing which findings on the same strip turn it from a number into an emergency.

• Report four things together, never a bare QTc — the raw QT, the heart rate it was measured at, the correction formula you used, and the QRS duration. Bazett runs roughly 25-30 ms above Fridericia and Framingham on the same tracing, and four out of five tracings that Bazett grades as severely prolonged are graded less severe by either of the other two, so a QTc without its formula is not a comparable number (Richardson et al., 2022); a smaller series found the same ordering between formulas (Shah et al., 2025).
• Measure the T wave endpoint by a stated method — the tangent method is the easier one to apply consistently — in lead II or V5, across several successive beats, and use the longest value (LITFL QT Interval, 2024; Robyns, ESC, 2024).
• Check whether you are measuring a QT or a QU before escalating. Look at V2-V3 for a prominent U wave fused to a flattened T wave — the hypokalemia pattern — because that changes both the number and the correct intervention (LITFL Hypokalemia, 2024).
• Send electrolytes and expect to replace more than one. Potassium, magnesium, and calcium all belong on the panel; hypokalemia frequently travels with hypomagnesemia, and replacing potassium alone leaves the arrhythmic risk in place (LITFL Hypokalemia, 2024; Farzam and Tivakaran, 2023).
• Review the whole medication list, including as-needed orders. Risk accumulates across combinations of individually low- or moderate-risk drugs — antipsychotics, antidepressants, and antiemetics are the usual contributors on a general ward — and as-needed prescriptions inflate the risk picture even when they have barely been given (Simona et al., 2026; Farzam and Tivakaran, 2023).
• Escalate a QTc above 500 ms to the provider, and treat it as a trigger for review rather than as a diagnosis. Concern rises sharply at that level, but a threshold alone is an insufficient basis for stratification and a lower value in a patient with several risk factors is not automatically reassuring (Cohagan and Brandis, 2023; LITFL QT Interval, 2024; Farjam et al., 2026).
• Watch for the mechanical precursors on the strip, not just the number — short-long-short R-R sequences and premature ventricular beats landing on the T wave are what actually launch torsades (Cohagan and Brandis, 2023).
• If torsades occurs, the standard responses are intravenous magnesium as first-line therapy (a 2 g bolus, then a 1-4 g/hour infusion), aggressive electrolyte repletion targeting potassium of 4.5-5 mmol/L and magnesium above 2 mmol/L, withdrawal of the offending drug, and isoproterenol or overdrive pacing at 90-110 bpm for refractory or pause-dependent episodes. A hemodynamically unstable patient gets synchronized cardioversion and a pulseless one gets defibrillation (Cohagan and Brandis, 2023).
• Long-term management of the congenital syndrome is a different conversation from managing this finding, and it belongs with the diagnosis rather than with the tracing — beta-blockers are first-line, with an implantable cardioverter-defibrillator after cardiac arrest or beta-blocker failure, and genotype-targeted mexiletine and left cardiac sympathetic denervation as validated alternatives (Al-Akchar and Siddique, 2022; Farjam et al., 2026).
• Compare against a prior ECG whenever one exists, and use the same formula for both. A stable, long-standing interval in an asymptomatic patient is a very different report from one that has lengthened since the last tracing.