11 May 2009

Measuring QT interval: more complex than you may think!

Inaccurate electrocardiographic interpretation of long QT: The majority of physicians cannot recognize a long QT when they see one. Viskin S, Rosovski U, Sands AJ, Chen E, ... Zeltser D. Heart Rhythm 2005 2;6:569-574

Dear Colleagues,

In this well conducted study from Israel 4 cardiograph tracings (2 with long QT syndrome vs. 2 normal controls) were sent to about 1000 doctors in several countries, including Australia for their assessments. Clinicians included QT interval specialists, electro-physiology experts, cardiologists and ‘other physicians’.

Correct classification of all four ECGs was gratifyingly 96% in QT specialists and their results were used as the expected ranges. Only 62% of arrhythmia experts and less than 25% of other physicians (including cardiologists) were correct in all four cases.

More than 80% of arrhythmia experts but less than 50% of regular cardiologists and less than 40% of non-cardiologist physicians calculated the QTc interval correctly in all four trial subject ECGs. Much of the inaccuracy occurred in the correction for rate - clearly many doctors did not know how to do this important step. The most common errors were underestimating the QTc of patients with long QT syndrome and overestimating the QTc of healthy patients.

While there is talk about alleged dangers of heart rhythm disturbances in association with methadone treatment, the QTc interval is often discussed as if it were a constant. Automated calculations are not dealt with in this article but their use is becoming widespread in advanced centres but is apparently rare in the developing world where most dependent individuals live. Even so, abnormal automated results are also subject to certain difficulties, needing the human touch … which from this study would still appear to be far from perfect, even in specialist hands.

This interesting report should remind us that the QT interval issue in methadone treatment needs to be looked at from a practical standpoint related to patient safety and treatment effectiveness. To date few if any young, new or uncomplicated patients treated with standard induction protocols have been reported to develop torsade. And this is despite many such patients being reported to have substantial QT interval prolongation (Wedam found >10% had over 500ms at some stage in the first three months of standard treatment).

In our own practice we have faced numerous challenges in obtaining a confirmed corrected QT (QTc) interval in those who may be at risk of torsade - largely those needing methadone doses in excess of 150mg daily. In New South Wales since 2002 there has been a requirement for a cardiograph with detailed QTc interval before patients are permitted to exceed 200mg daily dose of methadone. Our difficulties have included (1) specified QT request ignored by cardiologist, (2) a bland response: “normal tracing, including QTc”, (3) some approximate figures: eg. “QTc around 0.3ms” and (4) some results which were just wrong when checked by us. We have become reasonably adept at doing these measurements simply because of the variable results we have obtained from cardiology reports.

Of the growing number of torsade reports in the literature, nearly all are of patients with (1) multiple medical illness and/or (2) multiple drug/alcohol use and/or (3) taking very high doses of methadone (>150mg). Fortunately only one death was reported amongst about 80 such cases I found in the literature. See Justo’s review in Addiction for 40 such detailed cases up to 2006: he found virtually all had co-existing contributors over and above standard methadone treatment.

Thus we can define a sub-group of methadone patients in whom torsade may be a credible risk and act accordingly. These would include those prescribed the drug in very high dose, those over 40 years of age, female gender, co-prescribed medications, HIV infected, continued use of illicit drugs and/or those with structural heart disease. The most obvious is the co-prescription of drugs known to prolong the QT interval such as erythromycin, droperidol and cisapride.

Just doing an ECG in such cases on its own has limited if any likelihood of avoiding torsade. Most of the reported cases in the literature had a normal ECG before and/or after the torsade episode where one was available. Thus an ECG tracing in such cases is only a starting point or baseline. At best it would detect most cases of familial long QT syndrome (Smith) should this occur in a methadone patient (some may have died during exposure to illicit drugs such as cocaine or amphetamine).

Torsade has also been reported with normal and shortened QT intervals, so this is by no means a yes or no situation - like most other situations in medicine it is a continuum. This is why diagnosis should always be individual and why clinical guidelines should be reserved for particular public health priorities, and only when they are evidence based and known to do more good than harm.

Fortunately, the majority of methadone dependency patients are not in a risk category and do not need cardiography. On the other hand, most should probably be recommended hepatitis testing since this is a major public health issue and a communicable disease.

Thus, despite talk about supposed dangers of high doses, there are in fact far more dangers by using inadequate doses. This is especially so in high risk individuals such as during pregnancy, those with co-existing mental illness and/or continuing drug use. We should be confident to prescribe higher doses for those who need them, based on clinical factors with no arbitrary maximum cut-off. There are major benefits in using adequate doses as shown by many controlled comparative studies. The side effect profile is relatively low as long as patients are properly assessed. Many well run clinics have mean doses around 100mg daily which is about the same as the original report by Dole and Nyswander in 1965. Most well run clinics also have a small number of patients taking over 200mg daily due to rapid metabolism and/or high tolerance to the drug. Not all patients do well on methadone and in some countries there has been considerable experience with buprenorphine which suits a substantial minority of opioid dependent individuals.

Nonetheless, we need to remember that some patients on opioid maintenance treatments are now in the age groups which are subject to other illnesses. These include osteoporosis, hypertension, heart failure, cirrhosis, dementia, etc. These are best addressed by a well co-ordinated “shared care” model utilising family physicians and appropriate specialists.

Comments by Andrew Byrne ..

References:

Wedam EF, Bigelow GE, Johnson RE, Nuzzo PA, Haigney MCP. QT-Interval Effects of Methadone, Levomethadyl, and Buprenorphine in a Randomized Trial. Arch Intern Med 2007 167;22:2469-2473

Smith WM. Cardiac repolarisation: the long and short of it. MJA 2008 188;12:688-689

Justo D, Gal-Oz A, Paran Y, Goldin Y, Zeltser D. Methadone-associated Torsades de Pointes (polymorphic ventricular tachycardia) in opioid-dependent patients. Addiction. 2006 101:1333-1338

Recommended audio critique of the subject by Dr Gavin Bart: https://umconnect.umn.edu/methadoneqtcscreening/