Erroneous Computerized Interpretation of QTc and Changes to Patient Drug Therapy: Cautionary Example for Pharmacists.

A prolonged QT interval (>500 ms) combined with individual patient risk factors increases the risk of ventricular arrhythmias and torsades de pointes.1 Pharmacists play a critical role in identifying pharmacodynamic and pharmacokinetic drug-drug interactions that may result in QT prolongation. Most 12-lead electrocardiograms (ECGs) provide a corrected QT interval (QTc); however, this value is computer generated and may be miscalculated. We present a case where a computer-calculated QTc was incorrect and would have led to inappropriate withdrawal of drug therapy. A 64-year-old woman with a past medical history of heart failure with reduced ejection fraction, atrial fibrillation, anxiety, and chronic kidney disease was admitted with urosepsis, initially treated with piperacillin/tazobactam. Her medications consisted of amiodarone 200 mg twice daily, bisoprolol 5 mg once daily, spironolactone 25 mg once daily, furosemide 80 mg twice daily, mirtazapine 30 mg at bedtime, venlafaxine 150 mg once daily, and apixaban 2.5 mg twice daily. Her baseline creatinine was 2.6 mg/dL, potassium, 2.9 mEq/L, and magnesium, 1.7 mg/dL. Her baseline ECG (Figure 1) showed sinus rhythm, heart rate of 75 beats per minute, biphasic p waves with T-wave inversion, and a computer calculated QTc of 465 ms. On day 2 of her admission, urine cultures grew Escherichia coli, which was sensitive to ciprofloxacin, piperacillin/tazobactam, and meropenem. Because an oral antibiotic was desired, piperacillin/tazobactam was changed to ciprofloxacin 500 mg twice daily. Given the concern of prolonged QTc with ciprofloxacin and the patient’s other risk factors, including other QT-prolonging drugs (amiodarone, venlafaxine), the pharmacist ordered potassium replacement therapy and a repeat ECG on day 5 to assess the QT interval. The repeat ECG on day 5 showed sinus rhythm, heart rate of 93 beats per minute, and a computer-calculated QTc of 684 ms (Figure 2). The patient’s creatinine was 2.7 mEq/L and potassium 4.2 mEq/L on day 5. Based on the significantly increased QTc, there were significant drug-drug interaction concerns. However, on visual assessment of this ECG by the pharmacist (and later confirmed by a cardiologist), the QT on the inferior leads is 360 ms, and QTc (Bazett equation) is calculated to 448 ms. The computer incorrectly calculated the QTc by interpreting the biphasic p waves as part of the QT interval, thus resulting in a significantly elevated QT and QTc. Based on manual assessment of this ECG and the QTc being similar to baseline, the patient’s drug therapy was not modified. The patient completed her course of oral ciprofloxacin and was safely discharged. A third ECG done at the completion of ciprofloxacin showed a similar pattern of biphasic p waves, T-wave inversion, heart rate 83 beats per minute, and a computer-calculated QTc of 468 ms. This case highlights that computer calculation of QT interval may not always be accurate, in particular, the challenge of recognizing the QRS complex onset and the end of the T wave as illustrated in the above case.2 The QRS complex and end of the T wave can vary in different leads and can be confounded by variations in T-wave morphology, noisy baseline, and presence of U waves.3 The reliability of automated QTc calculations has long been considered limited, and measurement errors remain in abnormal or poorquality ECGs.4 Previous publications have suggested that clinicians should be mindful of the limitations of computerized interpretation of QTc, and manual interpretation is routinely warranted.2,5,6 Pharmacists should not be overly reliant on a computer-calculated QTc, and the QTc should be manually assessed, especially if it leads to changes in drug therapy or if the QT change is beyond what would be expected.3,6,7 Doson Chua, BSc(Pharm), PharmD, FCSHP, BCPS, BCCP Tanveer Brar, BSc, BSc(Pharm), PharmD St Paul’s Hospital, Vancouver, BC, Canada 1049469 AOPXXX10.1177/10600280211049469Annals of PharmacotherapyChua and Brar letter2021