Electroconvulsive therapy (ECT) is an effective treatment for severe depressive symptoms, yet more research is needed to examine predictors of treatment response, and factors associated with response in patients not initially improving with treatment. This study reports factors associated with time to response (early vs. late) to ECT in a real-world setting.

This was a retrospective, single-center cohort study of patients endorsing moderate to severe depressive symptoms using the Quick Inventory of Depressive Symptomatology (QIDS; QIDS>10). Response was defined as 50% or greater decrease in QIDS score from baseline. We used logistic regression to predict response at treatment #5 (early response) as well as after treatment #5 (late response) and followed patients through ECT discontinuation or through treatment #20.

Of the 1699 patients included in this study, 555 patients (32.7%) responded to ECT treatment at treatment #5 and 397 (23.4%) responded after treatment #5. Among patients who did not respond by treatment #5, those who switched to brief pulse width ECT from ultrabrief pulse ECT had increased odds of response after treatment #5 compared with patients only receiving ultrabrief pulse (aOR = 1.55, 95% CI: 1.16-2.07). Additionally, patients with less improvement in QIDS from baseline to treatment #5 had decreased odds of response after treatment #5 (aOR = 0.97, 95% CI = 0.97-0.98).

Among depressed patients treated with ECT, response occurred in 56.0% of patients by treatment #20. Patient receiving ultrabrief pulse ECT at baseline and who did not respond by treatment #5 had greater odds of subsequent response if switched to brief pulse ECT than if continued with ultrabrief pulse.

The aim of the present study was to investigate the risk of nonremission following electroconvulsive therapy (ECT), as associated with borderline personality disorder, in individuals with major depression in the context of the contradictory data available in the literature.

We analyzed demographic and clinical data from 210 individuals with major depression who were treated with ECT. Study participants were recruited from the medical records database of the Psychiatry Department at Erasme Hospital. Only individuals with major depression who were in remission, as demonstrated during the systematic psychiatric interview performed at the end of ECT (ie, with a >60% reduction in their 24-item Hamilton Depression Rating Scale score, combined with a score of <10), were included in the "remission" group. Logistic regression analyses were used to determine the risk of nonremission following ECT.

Nonremission following ECT occurred frequently (42.9%) in our sample. Moreover, after adjusting for major confounding factors, multivariate logistic regression analyses demonstrated that borderline personality disorder was a risk factor for nonremission following ECT in individuals with major depression.

We demonstrated that borderline personality disorder was associated with a higher risk of nonremission following ECT in individuals with major depression. This finding seems to justify more systematic screening as well as more adequate management of this personality disorder in individuals with major depression who are treated with ECT to allow for attaining better remission rates in this subpopulation.

Patients from racial-ethnic minority groups undergo disparate electroconvulsive therapy (ECT) treatment compared with Caucasian peers. One leading hypothesis is that clinicians may unknowingly display racial bias when considering ECT for patients of color. Studies have consistently shown that patients of color face numerous racially driven, provider-level interpersonal and perceptual biases that contribute to clinicians incorrectly overdiagnosing them as having a psychotic-spectrum illness rather than correctly diagnosing a severe affective disorder. A patient's diagnosis marks the entry to evidence-based service delivery, and ECT is best indicated for severe affective disorders rather than for psychotic disorders. As a consequence of racially influenced clinician misdiagnosis, patients from racial-ethnic minority groups are underrepresented among those given severe affective diagnoses, which are most indicated for ECT referral. Evidence also suggests that clinicians may use racially biased treatment rationales when considering ECT after they have given a diagnosis of a severe affective or psychotic disorder, thereby producing secondary inequities in ECT referral. Increasing the use of gold-standard treatment algorithms when considering ECT for patients of color is contingent on clinicians transcending the limitations posed by aversive racism to develop culturally unbiased, clinically indicated diagnostic and treatment rationales.

For many years, psychiatrists have tried to understand factors involved in response to medications or psychotherapies, in order to personalize their treatment choices. There is now a broad and growing interest in the idea that we can develop models to personalize treatment decisions using new statistical approaches from the field of machine learning and applying them to larger volumes of data. In this pursuit, there has been a paradigm shift away from experimental studies to confirm or refute specific hypotheses towards a focus on the overall explanatory power of a predictive model when tested on new, unseen datasets. In this paper, we review key studies using machine learning to predict treatment outcomes in psychiatry, ranging from medications and psychotherapies to digital interventions and neurobiological treatments. Next, we focus on some new sources of data that are being used for the development of predictive models based on machine learning, such as electronic health records, smartphone and social media data, and on the potential utility of data from genetics, electrophysiology, neuroimaging and cognitive testing. Finally, we discuss how far the field has come towards implementing prediction tools in real-world clinical practice. Relatively few retrospective studies to-date include appropriate external validation procedures, and there are even fewer prospective studies testing the clinical feasibility and effectiveness of predictive models. Applications of machine learning in psychiatry face some of the same ethical challenges posed by these techniques in other areas of medicine or computer science, which we discuss here. In short, machine learning is a nascent but important approach to improve the effectiveness of mental health care, and several prospective clinical studies suggest that it may be working already.