The recommendations of the FDA advisory panel that convened on January 27-28, 2011, may have long-lasting effects on the practice of electroconvulsive therapy (ECT) in the United States. The meeting included FDA presentations of the medical literature on ECT, an open public hearing, and panel deliberations. It is hoped that focusing a spotlight on ECT will both reveal the considerable benefit of the treatment and stimulate further research, training, and standardization of the procedure. Classification of ECT devices is a complicated and technical issue, so let us take some time to discuss the backstory behind the panel’s deliberations.
The FDA regulates devices and drugs in the United States. ECT devices come under the purview of the Neurological Devices Panel, and new devices are placed into 1 of 3 risk-based categories. ECT devices, among others, were “grandfathered” into Class 3 (the highest risk-based classification for devices) in 1976 by the Medical Devices Amendments. The FDA never required premarket approval “to affirmatively demonstrate a reasonable assurance of safety and effectiveness”; rather, ECT devices have been regulated through “the premarket notification [510(k)] regulatory pathway, which requires a showing of substantial equivalence to a legally marketed device and is usually reserved for intermediate and low risk devices.”1 In 2009, the Government Accountability Office recommended that the FDA require that all such “grandfathered” devices be either reclassified into Class 1 or Class 2 or undergo premarket approval.
The crux of the matter is that premarket approval usually requires controlled clinical trials be carried out prospectively; such trials cost many millions of dollars and, in the case of drugs, are typically supported by the pharmaceutical industry. For ECT devices, there may not be a funding source for such trials. FDA officials suggested that the existing scientific evidence base for the efficacy and safety of ECT might be sufficient for a determination of device classification. Such an option is referred to as a “paper” premarket approval. That means no new trials would be required.
But if this advisory panel was not convinced by the evidence (see below), why should a future panel come to a different conclusion? With data accumulated over many decades, millions of patients treated worldwide, and more than 10,000 articles in the medical literature, it is curious that the panel concluded that we lack sufficient evidence for the efficacy and safety of ECT. This conundrum can be explained by several factors.
The FDA requires evidence in the form of placebo- or sham-controlled trials. Eleven such trials were conducted starting in the late 1950s; they showed superiority for real ECT only in the acute treatment phase, but there was no difference between the sham and active groups after 1 month. Patients with delusional depression did particularly well. Many of these trials were methodologically primitive by today’s randomized clinical trial standards. An important feature of at least 2 of those trials was that patients were treated “naturalistically” after the initial controlled treatment period of 4 weeks; that is, they received antidepressant treatments (including ECT) “regarded as appropriate by their psychiatrist.”2,3
A balanced review of the safety and efficacy of ECT is needed, which does not mean weighing anecdotal reports of memory loss equally with systematically collected clinical data.
Certainly, much of the subsequent improvement in the sham group (which lessened the difference in outcomes between the groups) can be attributed to this active treatment in the follow-up phase. The methodological shortcomings of the earlier trials could be easily overcome with modern trial designs, but would it be ethical to conduct sham-controlled trials of ECT in the current era? Probably not, and here’s why: the accumulated evidence base for the efficacy of ECT is, in fact, very large, despite the paucity of sham- or placebo-controlled data. In order for it to be ethical to conduct a sham- or placebo-controlled trial, there needs to be “equipoise” between the treatment arms. That is, there must be clinical uncertainty that one treatment is superior to the other.4 Would an institutional review board (IRB) believe that we do not already know whether a melancholic depressed patient would fare better with true ECT than sham? Perhaps a compromise would be to do a trial in which the sham component were very time-limited, for example, a week or less. Even that, however, would probably face opposition from an IRB. This could really be a “catch-22!”5
The advisory panel was confounded by the safety data as well as the efficacy data. By “safety,” I mean cognition and the issue of “brain damage.” The FDA literature review of these two issues was exhaustive, and their written conclusions seemed quite reasonable and reassuring. However, during the course of the panel’s deliberations, the issues somehow became conflated1; a casual observer might have been left thinking that panel members believed that memory loss caused by ECT is indicative of “brain damage.” This is obviously unfounded and represents confusion between functional and structural brain changes. There is abundant evidence to show that ECT is not associated with deleterious structural brain changes or “damage.”6 In fact, recent research points to the damaging effects of prolonged, severe depression, and the neurotrophic (probably beneficial) effects of antidepressant treatments, including ECT.7-10
The advisory panel had the option of recommending that the FDA reclassify ECT devices to Class 2. Class 2 allows for “special controls” to be imposed on the use of the device, as added safety measures. Much of the panel discussion during the 2-day hearing actually focused on what such potential regulatory controls might be. They included restrictions on who could perform ECT (eg, “physicians with specific training and/or experience”) and the requirement to present patients with a checklist “of all known risks of ECT” and requirements for monitoring of cognitive status prior to, and throughout, a course of ECT, as well as other requirements. While some of these requirements might have been onerous to both patients and practitioners, they would ensure the quality and standardization of the practice of ECT. The point to be made here is that reclassification to Class 2 is not a license to perform unrestricted ECT, as some have suggested; in fact, it is eminently reasonable to reclassify ECT devices to Class 2 with appropriate special controls.
“Cleared indications for use for ECT devices” is FDA language for what amounts to “on-label,” as opposed to “off-label” use of ECT. The list of indications for use currently includes the following1:
• Depression (unipolar and bipolar)
• Bipolar manic (and mixed) states
• Schizoaffective disorder
• Schizophreniform disorder
It came as a surprise to me to learn that a device can be classified as Class 2 for a particular indication for use and Class 3 for another. At the hearing, an FDA official gave the example of a contact lens that is in Class 2 for daily use, but in Class 3 for extended wear. The same is true for ECT devices, and the advisory panel gave its advice (in the form of an “unofficial” vote, although no vote was supposed to have been taken) for each of the 6 indications for use. By a slim majority, the panel advised keeping ECT devices in Class 3 for all the indications for use except catatonia, for which it advised Class 2, because of the impossibility of conducting randomized clinical trials with such emergently ill patients. The votes were decidedly along discipline lines: the psychiatrists on the panel (as well as the one anesthesiologist) voted for reclassification to Class 2; all the psychologists, neurologists, and biostatisticians voted to retain Class 3 designation.
So far, the FDA advisory panel hearing on ECT devices has shed more heat than light on this much-vilified antidepressant treatment. But this is a long process and it will be years before the FDA makes a final ruling about the classification of ECT devices and, by extension, about the practice of ECT in the United States. If new clinical trials are required, one can only hope that they are both ethically and financially feasible. The existing evidence base for the safety and efficacy of ECT should be carefully reviewed in a balanced way; balanced does not mean weighing anecdotal reports of memory loss (mostly from decades ago) equally with systematically collected clinical trial data.
Yes, ECT has adverse effects, including memory loss for some recent events, but all medical procedures for life-threatening diseases have adverse effects and risks. Severe depression is every bit as lethal as cancer or heart disease. It is inappropriate to allow public opinion to determine medical practice for a psychiatric illness; this would never happen for an equally serious nonpsychiatric illness. ECT is a proven lifesaver when antidepressants fail.
ECT in the United States should be appropriately and carefully regulated, but not held to a different, higher standard than any other medical procedure. Such appropriate regulation will ensure that ECT remains widely available for those severely ill patients who need it and will allow the practice and science of ECT to move forward.