Fluoxetine or Cognitive-Behavioral Therapy First for Intermittent Explosive Disorder
A dental hygienist's escalating workplace outbursts put her job at real, near-term risk. The two treatments with actual trial evidence behind them work on different timelines — and her deadline is shorter than either trial ran.
E.N., a 28-year-old woman, has worked as a dental hygienist at the same practice for six years, and until this week had never had a formal complaint filed against her. Three days ago, after a patient's second same-day no-show, she screamed at a coworker over a misplaced supply cart, shoved it hard enough to send instruments sliding onto the floor, then punched a supply cabinet in the break room with enough force to bruise the side of her hand. Her office manager issued a written warning the same afternoon and told her plainly that a repeat incident would end her employment there.
This wasn't a first. She describes similar blow-ups — screaming matches, a punched wall at home, one road-rage incident that frightened her enough to pull over — roughly two to three times a month since her early twenties, always disproportionate to whatever triggered them, always followed by real embarrassment and a promise to herself that it won't happen again. Screening today found no current depressive episode and no anxiety disorder underneath the pattern; whatever is driving this isn't presenting as a mood problem with aggression attached to it, it's presenting as the aggression itself, on its own research-criteria terms, meeting intermittent explosive disorder's definition of recurrent, disproportionate outbursts occurring at this frequency.
That absence of a comorbid mood disorder turns out to matter directly to what the team can actually recommend: Coccaro and colleagues' 2009 trial of fluoxetine for IED (J Clin Psychiatry) found the drug's antiaggressive effect held independent of whether a patient had current depression or anxiety — this isn't a drug that needs her to have a mood disorder to plausibly help, it's evidence for the aggression phenotype directly. But her deadline is real and close, and the honest question in front of the team isn't just which treatment has evidence — both options do — it's which one can plausibly move fast enough to matter before her next incident costs her the job.
Initial psychiatric consultation, three days after the written warning
Given her timeline, I'd start fluoxetine today. Coccaro, Lee, and Kavoussi's 2009 trial randomized 100 adults with IED to fluoxetine or placebo, and the antiaggressive effect — a real, sustained drop in OAS-M aggression and irritability scores — was apparent as early as week two. Critically, that effect didn't depend on her having a comorbid depression or anxiety disorder, which she doesn't. This is direct evidence for the aggression itself, not something borrowed from an antidepressant effect she isn't a candidate for.
I'd lead with therapy instead, and not as a slower alternative to medication but as a genuinely different kind of outcome. McCloskey and colleagues ran a randomized trial of twelve weeks of group or individual cognitive-behavioral therapy built specifically for IED, forty-five adults, against a wait-list control — both formats reduced aggression, anger, and hostile thinking, with no medication involved at all.
And I'd push back on “a real, sustained drop” as the whole story, because Coccaro's own conclusion doesn't read that way. Full or partial remission of the impulsive aggression occurred in 46% of the fluoxetine-treated patients — full remission in 29%. The authors say plainly that while the antiaggressive effect looks robust, it produces remission in fewer than half the people who take it. If we are betting her job on a bridge, it is worth saying out loud that this bridge carries slightly under half the people who step onto it.
A skill she keeps whether or not she's still taking a prescription five years from now isn't the same kind of gain as a symptom reduction that hasn't been shown to outlast the drug — her job crisis is real, but building something durable matters too, especially given how long she's already lived with this pattern.
You're both citing real, genuinely different evidence, and I don't think either of you is wrong — I think the actual answer is in how the two timelines line up against hers. Fluoxetine's effect shows by week two. CBT's tested protocol runs twelve weeks. Her employer gave her one written warning and a stated one-strike policy, not twelve weeks of patience.
You're right that 46% is not a bridge anyone should oversell, and that number is exactly why I don't want fluoxetine carrying this alone. But a treatment that helps slightly under half of patients and declares itself by week two is still the only thing on this table that can report back before her employer does. Starting fluoxetine now for the faster bridge while beginning CBT in parallel for the durable skill isn't splitting the difference between your two positions — it's what both trials' own week-by-week data actually support once you stop treating ‘medication or therapy’ as the only two options on the table.
Agreed: fluoxetine starts today and IED-specific cognitive-behavioral therapy begins this week in parallel, with a two-week check-in scheduled to match the point where the trial data suggest an early signal should be visible if the medication is working.
Not fully agreed: whether the two-week check-in should also be the point where the team considers stopping fluoxetine if CBT alone appears to be carrying the improvement. The psychologist wants that question asked explicitly rather than assumed away by inertia once a medication is already started; the psychiatrist agreed to ask it, but was clear she wouldn't recommend stopping a treatment that's working just because a second one might eventually be enough on its own.