Ni Tanto que Queme al Santo
Ni Tanto que no lo alumbre
Neither so Much that it Burns the Saint
Nor so Little that it Does not Light it
-Traditional Mexican Saying
The past few years have been trying for those of us prescribing psychotropic medications to children and adolescents. The possible association of sudden death with desipramine or with combined clonidine and stimulant use were two of the hot button controversies in pediatric psychopharmacology during the 1990s. Shortly after those concerns had abated, insofar as the former drug was abandoned, and the latter combination used more judiciously and with additional safeguards, a whole new cast of troubling adverse effects appeared onstage. And no sooner had one of them made a transient exit that another emerged to take its place. Like the Hydra sprouting new heads every time one was lopped off, the more of these villains we addressed, the more we encountered.
With the turn of the century, antidepressants came under attack first, and since 2003 have certainly remained the most contentious. The association between the selective serotonin reuptake inhibitors (SSRIs) and emergent suicidal ideation went from a counterintuitive curiosity to an all-but-proven (if modest) fact (1). The black box warning that now graces the label of all 34 antidepressants is an enduring legacy of these findings. But suicidality is far from the only developmental conundrum surrounding the antidepressants of late: manic induction (2), neonatal withdrawal syndrome (3), and persistent pulmonary hypertension of the newborn (4) have each been raised in turn.
Were it that only antidepressants had taken the heat: far from it. The atypical antipsychotics came next. They proved fallible soon enough, especially among young patients prone to accelerated weight gain and metabolic dysregulation. These remarkably useful compounds all too often led to uncomfortable decision points: efficacy yes, but at what cost? Just how relevant their touted advantages and how favorable the trade-offs between their almost predictable increases in girth and lower rates of tardive dyskinesia when compared to older agents? Mood stabilizers of course were anything but stable when it came to such matters: valproate seemed off-limits for girls liable to polycystic ovary syndrome, lamotrigine too risky for children at heightened risk for Stevens-Johnson syndrome, and topiramate, even when dosed gingerly, too cognitively dulling among school-aged children. Moreover, no large scale studies have been conducted for any of these agents to show that in children they provide benefits beyond placebo. For these and other medications, side effect Jeremiads became the norm, and usually long before efficacy could be tested, let alone demonstrated.
Fortunately for all, the stimulants remained largely untainted. Questions about their propensity for height deceleration and tic exacerbation did come (and mostly go), but stimulant use flourished under the strongest empirical foundations in all of pediatric psychopharmacology. And while other possible treatments for attention deficit hyperactivity disorder (ADHD) emerged, including novel pharmacotherapies and psychotherapies alike, psychostimulants consistently showed major advantages in efficacy. Reassuring news indeed, for these are by far the most widely prescribed psychotropics for children; doubly reassuring, given that stimulants are no new kid on the block: if more than five decades’ worth of safety track record was not sufficient, what timeline could realistically be required of other drugs?
As it turns out, the stimulants have had a rough year. Concerns have ranged from the vitriolic and impassioned though ultimately dismissible accusations by movie icons (5), through the very preliminary suggestion of chromosomal aberrations with uncertain implications (6), to the most biologically plausible and dire possibility of cardiovascular fatalities. Initial reports from Canadian regulatory agencies linking Adderall XR to strokes appeared flawed and were summarily retracted. But subsequent analyses coming out of the Food and Drug Administration were harder to dismiss. Indeed, debate over the need for a black box warning alerting clinicians and patients to the possible risks followed soon enough and may eventually become the law of the land (7).
It is against the backdrop of these black box warnings–-modern day scarlet letters of sorts––and especially against the uncannily timed most recent debate around a possible such warning for the stimulants, that the report by Zuvekas, Vitiello, and Norquist in this issue of the Journal needs to be considered. Their report is not only a welcome contribution, but an overall positive one at that. Using a large, nationally representative database (the Medical Expenditure Panel Survey), their main finding is that stimulant use among children and adolescents between 1997 and 2002–well before the black cloud came out of the black box–had settled into stable rates (to a period prevalence just shy of 3%). The ‘flattening’ of previously fluctuating and steadily increasing usage rates suggests that a degree of equipoise had emerged in clinical practice by century’s end. Further good news includes the narrowing of seemingly ingrained health disparities, most notably across ethnicity and insurance coverage lines (minorities and publicly funded families traditionally getting the short end of the stick, in this instance defined as fewer youths diagnosed with ADHD filling stimulant prescriptions). Finally, the fact that rates of use among children under the age of 6 remained low and steady is a reassuring bonus of this study, coming as it does in the wake of earlier reports suggesting alarming increases in the prescription of psychotropic drugs for preschoolers (8).
What are we as a psychiatric community to do with these findings and these federally imposed warnings, and how are we to respond to life and death safety concerns at the forefront of the public’s and our patients’ minds?
As clinicians, the first thing we would be well advised to do is to do nothing abruptly. Rash decisions made in the heat of the scary-side effect-moment are rarely warranted, and more likely to do harm than good (the sudden discontinuation of antidepressants being an especially concerning case in point). Similarly, we should not become frozen in inaction, awaiting the elusive definitive study to disprove this adverse effect or that: we should be cautious, thoughtful, vigilant and informed, but not withhold indicated treatment. We should take heed of current debates, and even if data are preliminary, incorporate their lessons into our daily practice: in the case of antidepressants, by being decisive–and never casual–about when to use them, and by monitoring explicitly and assertively for emerging suicidality; in the case of the stimulants, by monitoring heart rate, blood pressure and cardiovascular parameters routinely, by having a lower threshold to consult with our colleagues in pediatrics, internal medicine, or cardiology, and, especially in adult patients, by being attentive to medical comorbidity. Finally, we owe it to our patients to discuss with them our current thinking in each of these areas, so that families can become our partners in weighing the risks and benefits of each potential intervention.
Integral to our mission as clinicians is our charge as educators. We must inform the public and our patients about the very rare occurrence of the outcomes in question, about the probabilistic rather than deterministic nature of our science’s impressive figures, and about the similarity between the fine print of our drugs and all others in medicine (pain killers, anesthetics and antibiotics can also have dire consequences). We should provide reassurance that we are minding the shop and that certified bad apples are routinely taken out (liver-damaging pemoline and nefazodone come to mind). We should defuse, rather than stoke, our patients’ understandable ambivalence around psychotropics. In that capacity, all of our discussions should factor in one of the greatest risks: that of sitting by the sidelines doing nothing. And while not medicating is far from ‘doing nothing’ (as there is plenty more that we have to offer), withholding pharmacotherapy comes close to negligence for certain conditions awash in efficacy and safety data. The past decade clearly has shown potential problems associated with psychotropic use, but it has also provided research that demonstrates the tremendous burden of developmental psychopathology and the availability of effective interventions.
Although the type and quality of data that the community of pediatric psychopharmacology researchers has amassed over the past decade is impressive, it remains limited. Some of its shortcomings have been well articulated before, including by one of this report’s authors (9, 10). As one example, the agenda in pediatric psychopharmacology remains reactive (among other things, to rare side effects and, dare I say, to black box warnings) rather than proactive (by advancing potentially novel mechanisms and compounds). As another, the priorities across academic centers, industry, and federal dependencies have not been sufficiently aligned to maximize impact in the service of suffering children (11).
To a community still struggling to make sense of heated controversies directly relevant to its widespread clinical practice, I propose even further items to add to its busy research agenda and stretched funding’s reach: As policy makers, informers and consumers, we need to better understand the impact of well-intended black box warnings and similar top-down advisories. Could they for example, perversely and at odds with their intended spirit, cause more harm than good by scaring a prescribing community away from antidepressants, and as a downstream consequence tip suicide rates upward once again? Pharmacoepidemiologists, psychopharmacologists and policy stakeholders could converge to provide empirically informed answers to these pressing questions–-and ideally before the next black box warning comes, as it surely will.
One wishes that the elegant pharmacoepidemiology showcased by Zuvekas et al. may be pushed even further in the coming years: not only by describing rates, but by addressing qualitative outcomes and consistency (or discrepancy) with evidence-based practices; in short, real-world outcomes. If such an agenda is overly ambitious at this time, one does at least wish that adult rates of usage could be addressed the next time around, as ADHD and stimulant use among grown ups appear to be here to stay.
Stimulating times we live in. May debates and controversies such as these help us do better by our patients. May we be wise and informed enough to light their paths without burning them along the way.