Should We Prescribe Selective Serotonergic Reuptake Inhibitors/Serotonergic and Noradrenergic Reuptake Inhibitors in the ICU?*
When patients on antidepressants become critically ill, intensivists must decide whether these medications should be continued, tapered, adjusted, or stopped abruptly. Reasons to discontinue include the risk of drug interactions (4, 5), serotonergic excess (6), upper gastrointestinal and perioperative bleeding (7, 8), delirium (9), and other potential problems in the context of other medications and organ failure. That is, drugs that are relatively safe and well tolerated in medically well outpatients might not be so safe in critically ill patients. Nevertheless, abruptly stopping an antidepressant might precipitate an uncomfortable but non-life-threatening withdrawal syndrome. In addition, continuing or restarting an antidepressant may prevent an exacerbation of a psychiatric condition (e.g., depression) that could impede recovery from critical illness (10). Finally, intensivists might consider starting an antidepressant de novo in a nondelirious longer stay patient who has persistent symptoms of depression and/or anxiety.
What do we know about the safety and efficacy of the most commonly prescribed antidepressants, SSRIs and SNRIs, in the context of critical illness? In this issue, we find that the answer is, unfortunately, not much. In this issue of Critical Care Medicine, Kelly et al (11) carefully and systematically reviewed studies reporting on benefits and harms associated with administration of SSRIs and SNRIs to critically ill patients. The authors searched relevant medical databases from 1990 to 2014, ultimately selecting 13 studies representing 20,048 patients—five observational cohort studies, one case series, and seven case reports. Eleven of the studies reported apparent harm associated with SSRI/SNRI use in critically ill patients, whereas only one case report reported possible benefit. Alas, based on information in the reviewed studies, it was often not clear whether apparent harm due to SSRIs or SNRIs reflected lasting effects of use prior to critical illness, confounding by indication (e.g., harm associated with psychiatric illness itself), effects of ongoing use during critical illness, or effects of antidepressant withdrawal. In addition, there is reason for concern about potential reporting bias. For example, though severe serotonin syndrome is striking and crucial to recognize when present, and patients with such severe symptoms are often admitted to ICUs, it would not occur to most clinician investigators to prepare a report on critically ill patients who are on SSRIs or SNRIs and do not have severe adverse effects. Kelly et al (11) cogently argue for further observational research with clearer drug administration reporting to inform these issues.
In the meantime, what should an intensivist do? Kelly et al (11) provide a thoughtful analysis of the issues, including the risk of a recurrence of severe depression in a patient whose antidepressant is discontinued versus the risk of serotonin syndrome or postprocedural bleeding in a patient who is started, restarted, or continued on an antidepressant.