Escrever uma avaliação PREreview

Treatment non-adherence is among the most consequential and modifiable determinants of outcome in psychiatric practice, yet prevailing models—built on insight, medication beliefs, side-effect burden, and access—chiefly explain deliberate omission. They leave unexplained a large residue of doses that patients fully intend to take and then miss. This integrative, mechanism-oriented review advances the proposition that micro-cognitive lapses—transient, seconds-to-minutes failures of executive control—constitute an independent and under-recognised determinant of non-adherence in psychiatric outpatients. Evidence is synthesised on executive function, prospective memory, attentional lapses, goal neglect, mind wandering, intraindividual variability, effort-based decision-making, and metacognition across schizophrenia-spectrum, bipolar, depressive, and attention-deficit/hyperactivity disorders, together with the measurement and intervention literatures. Executive dysfunction is transdiagnostic and tracks real-world functioning more closely than symptom severity. Executive and prospective-memory capacity predicts medication-taking most strongly when doses must be self-initiated amid distraction or routine disruption; forgetting, not reasoned refusal, dominates patients’ accounts of missed doses; and retrospective self-report overestimates adherence relative to electronic monitoring. Ambulatory cognitive testing shows that performance fluctuates within persons with distraction, low effort, and being away from home, establishing lapse propensity as state-dependent and measurable in real time. Because brief interruptions rapidly erode therapeutic drug concentrations, individually trivial lapses carry disproportionate clinical cost. Interventions that reduce reliance on moment-to-moment control—habit formation, implementation intentions, environmental cueing, and long-acting formulations—show the most consistent benefit, whereas purely educational approaches show small effects. Reframing adherence as the output of a state-dependent executive system redirects care towards ecological assessment and automaticity-supporting, lapse-tolerant treatment architectures.