Algorithm Predicts Antidepressant Response
Researchers at McLean Hospital, Belmont, Massachusetts, have developed a statistical algorithm that identifies which patients with depression are likely to respond to treatment with antidepressants. They reported their development online in Psychological Medicine.
“This model must be tested prospectively before it can be used to inform treatment selection,” they wrote. “However, findings demonstrate the potential to improve individual outcomes through algorithm-guided treatment recommendations.”
The project centered on data gleaned from an 8-week multisite trial of 216 adults with depression randomly assigned to treatment with the antidepressant sertraline or placebo. Focusing on the demographic and clinical characteristics of participants, the research team developed an algorithm that predicted about a third of patients would respond better to sertraline than to placebo.
Overall, “we found relatively little difference in average symptom improvement between those individuals randomly assigned to the medication vs. placebo,” said researcher Christian A. Webb, PhD. However, “for the one-third of individuals predicted to be better suited to antidepressants, they had significantly better outcomes if they happened to be assigned to the medication rather than the placebo.”
Patients who responded to sertraline tended to have higher depression severity and neuroticism, were older, showed less impairment in cognitive control, and were employed.
“These results bring us closer to identifying groups of patients very likely to benefit preferentially from a selective serotonin reuptake inhibitor,” said researcher Madhukar Trivedi, MD, “and could realize the goal of personalizing antidepressant treatment selection.”