In this issue of Acta Psychiatrica Scandinavica, Cohen and Loonen  challenge the assumptions of many that extended-release formulations and immediate-release formulations of drugs are not only interchangeable, but that immediate-release drugs should be used preferentially due to lower costs. The authors point out that at least for one agent, carbamazepine, head-to-head studies of extended release versus immediate release show superiority of extended release for both efficacy and tolerability. These authors also argue that when extended-release formulations reduce the frequency of dosing, they enhance compliance, and therefore improve outcomes; thus, Cohen and Loonen submit that the rational weighing of the evidence favours the use of extended-release formulations.
When are immediate-release and extended-release drugs interchangeable?
I reviewed this issue briefly almost 10 years ago  and again recently  and concluded that the value of a given extended-release formulation depended upon the specific drug and that when value over the immediate-release formulation was present, this was based largely upon uncontrolled use in clinical practice and not upon head-to-head randomized controlled clinical trials. Because of the relative lack of ‘gold standard’ head-to-head studies of extended-release formulations with their immediate-release counterparts, many disagree with me and assume that extended-release formulations are just patent extension gimmicks and not worth the extra cost. After all, does not the absence of evidence of head-to-head superiority mean evidence of absence of any advantages?
Perhaps not. Few clinicians or patients would doubt the improved tolerability of venlafaxine in extended release; now that both extended release and immediate release are available as generics, almost no one uses the immediate release for venlafaxine . Yet, there are no head-to-head trials that prove this. Similarly, few clinicians or patients doubt the value of once a day versus three times a day administration of bupropion or of the reduction in seizures the extended-release formulation provides; now that bupropion and its extended-release formulations are all available as generics, this drug is almost never prescribed as the immediate-release formulation . Again, no controlled head-to-head trials exist.
Same drug, dramatically different clinical actions as immediate versus extended administration
The best example of robust clinical differentiation of extended-release versus immediate-release drug delivery is probably the stimulants. Use of stimulants for the treatment of attention deficit hyperactivity disorder (ADHD) especially in children has long been problematic, requiring multiple daily doses and needing to give a dose during the day at school. This caused stigma and embarrassment for the child whose medication need was no longer a secret and also great hassle for the school to handle the responsibility of administering a controlled substance. Once new extended-release technology became available for the stimulants, it eliminated the need for a middle of the day dose of a stimulant, removing significant barriers to the use of stimulants for the treatment of ADHD . Such an advantage is obvious but not demonstrated in any randomized controlled head-to-head trial.
More important than mere convenience is the discovery that stimulants are actually two entirely different drugs, depending upon whether they are given as a rapid, immediate dose or as a delayed and extended dose [5, 6]. The faster a stimulant enters the brain, the stronger are its reinforcing effects, probably because this form of drug delivery triggers phasic DA firing, the type associated with reward and saliency [5, 6]. Immediate release enhances the abuse potential and addiction of stimulants and is why stimulant abusers do not prefer extended-release formulations and instead prefer the fastest way of getting stimulants in the brain, whether by oral immediate release, snorting, intravenous injection or smoking. By contrast, when these same stimulants are given in slow onset, long-duration extended-release oral formulations, they enhance tonic dopamine firing and improve cognition with much lessened reward or abuse potential [5, 6]. Such findings have been demonstrated in small trials, often open label, uncontrolled and using a crossover design, but not in large randomized head-to-head studies.
The tyranny of the majority
Evidence-based medicine is population-based medicine, defining what is good for the median patient on the basis of randomized controlled trials. It has resulted in decision trees so plentiful it is sometimes difficult to see the individual within the forest. That is, the individual can become invisible within the forest of evidence, especially if that individual is not a median patient. In evidence-based medicine, the majority rules. What is right for the median patient is what all patients receive. When there is no consideration for the individual either because of the lack of evidence for this outlier, or cost considerations, that individual patient is to be treated just like everyone else. Cohen and Loonen also point this out, as they note in the case of extended-release formulations, sometimes the individual patient notes differences and does not fit into the well-documented group characteristics for this drug. That is not an uncommon situation in clinical practice. But what should we do with this ‘evidence?’ Should we ignore it?
The pendulum has swung very far towards population-based medicine and against ‘narrative-based’ medicine from observations of individual clinicians on individual patients for fear of returning to ‘eminence-based medicine’ . On the one hand, epidemiologists and researchers practice in the realm of populations; on the other hand, clinicians practice in the realm of individuals including both median patients and outliers, all of whom belong to populations. We are amidst a major paradigm shift in medicine, away from population-based medicine and the tyranny of treating everyone including outliers as a median evidence-based patient, and towards personalized medicine, where the individual has treatments customized based upon biomarkers as well as family history and past treatment successes and failures .
So, the question may not be whether immediate- and extended-release drugs are interchangeable for everyone. Rather, the question seems to be whether any advantages of extended-release drug are worth the premium that they demand when they are more costly, and for whom this can be best applied.
We eagerly await the development of standards for personalized medicine where the tyranny of the majority no longer rules all without exceptions. In the meantime, humble clinicians continue to be informed by ‘evidence’ wherever it exists, including empiric trial-and-error treatments of individuals, while carefully evaluating patient preference, risk/benefit ratio, past response, availability and cost.