Despite treatment in health care settings, patients with bipolar disorder are well only half of the time; the other 50% of the time is spent experiencing syndromal or sub-syndromal mood symptoms which in addition to increasing the risk of suicide can also significantly impact functioning and quality of life. This presentation will review opportunities and challenges for intervening early during clinical high risk stages and in the early stages of bipolar disorder and examine if such strategies can improve outcomes.

Several challenges exist for intervening during “high risk” stages. First, high risk stages for BD can be defined only clinically at present, and no biomarkers exist that can increase the prediction of conversion to BD. Indeed, current data suggest that only upto 20% of patients defined as high risk for BD convert to BD I or BD II. Further, most studies that assessed the efficacy of interventions in clinically high risk BD to date focussed on the impact of interventions mostly on improving sub-syndromal symptoms. However, a small RCT that assessed the efficacy of multifamily focussed therapy reported lower rates of conversion to BD compared with the wait list control offering some promise for intervention strategies in high risk stages to improve outcomes. Since BD onset is often associated with increase in inflammation and oxidative stress markers, strategies targeting these areas might be beneficial in preventing or delaying the onset of BD.

A firm diagnosis of BD can only be made after a patient has had their first manic or hypomanic episode. Optimizing treatment at the first manic episode may also improve outcomes. This presentation will review outcome data from a first episode mania cohort and present results of clinical, cognitive and functional outcomes in first episode cohorts.

L. Yatham Grant / Research support from: Received grant support from Sumitomo and Abbvie in the last 3 years, Consultant of: Sumitomo Pharam, Alkermes, LivaNova, Puretech, Neurotorium,

Impairments in mentalizing, or theory of mind, occur across psychiatric disorders. Static illustrations are widely used to assess mentalizing due to their simplicity, and they allow assessment of specific cognitive processes. However, systematic comparisons of impairments between psychiatric disorders, neurodevelopmental disorders, and at-risk groups in mentalizing tasks with static illustrations are currently lacking.

A systematic review with pairwise and network meta-analyses (NMA) was conducted to evaluate mentalizing impairments using tasks with static illustrations across psychiatric disorders compared to healthy controls (HCs) and between groups. Subgroup analyses examined specific mentalizing domains (false belief, humor, and intentionality), and meta-regression analyses explored potential moderators. The ceiling effects of specific tasks were also examined.

Eighty-nine studies were included, involving 9,038 participants and 11 psychiatric conditions. Significant mentalizing deficits were observed across all conditions versus HCs, except for the familial risk for bipolar disorder group. NMA demonstrated that schizophrenia (g = −0.960) and early schizophrenia (g = −0.785) exhibited the most pronounced impairments, followed by borderline personality disorder (g = −0.612) and obsessive-compulsive disorder (g = −0.613). Particularly, schizophrenia showed significantly greater deficits than autism, bipolar disorder, clinical and familial high risk for schizophrenia, and depression. Domain-specific analyses highlighted differential impairment patterns. The presence of prominent ceiling effects suggests major limitations of tasks with static illustrations.

This review provides detailed insights into transdiagnostic and disorder-specific patterns of mentalizing impairments with tasks using static illustrations. Findings highlight the importance of domain-specific approaches, examining interindividual variability, refining assessment tools, and implementing targeted interventions.

This presentation reviews current evidence on the relationship between cannabis use and psychosis with special emphasis on cannabis induced psychosis, schizophrenia and the transition between the two. Cannabis has become increasingly potent over time, with THC concentrations rising dramatically, and this change is accompanied by a growing body of evidence linking cannabis to a range of psychiatric outcomes. Although cannabis is generally considered less harmful than alcohol, it nonetheless contributes to significant mental health risks. Epidemiological data consistently show associations between cannabis use and depression, anxiety, cognitive impairment, and, the primary focus of this lecture, psychotic disorders. Both heavy use and early onset substantially increase these risks, especially among vulnerable individuals.

Acute cannabis-induced psychosis is well documented and typically presents with confusion, hallucinations, emotional instability, depersonalization, and paranoid ideation. The condition can resemble acute schizophrenic psychosis, and symptom patterns are quite similar. While most cases are resolved within days, an acute psychotic episode represents an important clinical warning sign. Some debate persists about whether cannabis can cause persistent psychosis, but accumulating findings indicate that cannabis so consistently triggers schizophrenia in vulnerable individuals, that we mostly look at it as a separate causal factor. Large cohort and genetic studies suggest that the association reflects both shared vulnerability and a direct causal contribution of cannabis. Dose–response relationships have been observed for frequency of use, potency, and age at initiation.

Several candidate genes—including COMT, DRD2, and FAAH—appear to interact with cannabis to increase psychosis risk, but findings remain inconsistent. Individuals with schizophrenia who use cannabis experience poorer clinical outcomes, including reduced medication adherence, more frequent relapses, and increased hospitalizations. Studies show rising treatment-seeking for cannabis-induced psychosis paralleling increased potency of cannabis products.

The transition from cannabis-induced psychosis to schizophrenia is a central clinical concern. Recent meta-analytic evidence indicates that about 20% of people with cannabis-induced psychosis later develop a schizophrenia-spectrum disorder, underscoring the long-term implications. Diagnostic and legal frameworks complicate assessment, as time-based criteria in DSM and ICD attempt to separate substance-induced from primary psychosis, despite substantial clinical overlap. These distinctions have direct consequences for treatment, follow-up, and legal responsibility.

Overall, the evidence strongly supports cannabis as a risk factor for psychosis—particularly for early, frequent, and high-potency use—and highlights the importance of identifying vulnerable individuals and providing early, appropriate intervention.

None Declared

Across medicine, accurate prognosis can inform clinical decision-making about management and resource allocation. Several such models have been developed and validated in psychiatry, including those estimating risk of psychosis onset and relapse, and other key adverse outcomes, such as suicide and violence. Reliable risk estimation for these outcomes could inform follow-up intensity, continuity of care, collaborative safety planning, and risk management.

This presentation will provide an overview of prognostic models for violence in people with mental illness. I will outline the predictive performance, acceptability, and impact of new models, which have been translated into clinical risk assessment tools. I will show promising findings for these new models, which could be included in clinical guidelines. Further research is required to study whether and how they supplement and augment decision-making in different clinical settings.

None Declared

Tobacco smoking is highly prevalent in patients with psychosis, who also often experience negative affect (NA) and stress. The relationship between these factors remains unclear in this population. We aimed to investigate everyday life associations in 158 patients with psychosis, 136 unaffected siblings, and 117 controls from the Genetic Risk and Outcome of Psychosis (GROUP) study with Experience Sampling Method measurements.

Generalized linear mixed models were used to evaluate across time and within-subject associations. Across time analyses investigated the relationship between smoking status and overall NA and stress. Within-subject analyses assessed whether smoking between two measurements (t−1 and t0) was associated with changes in NA and stress at the measurement after smoking a cigarette (t0) and at the subsequent measurement (t+1).

Across assessments, smoking status was initially associated with NA in patients (B=0.26, p=0.036), but this association disappeared after controlling for psychotic symptoms and cannabis use. Within-subject analyses in smokers showed a decrease in NA in patients after smoking (t0: −0.23, p=0.016), which remained significant after correcting for confounders (t0: −0.20, p=0.015). Siblings showed a decrease in NA (t0: −0.22, p=0.009), also after controlling for confounders (t0: −0.14, p=0.018). No time-lagged effect was found at t+1 after correction for subsequent smoking.

Overall smoking behavior was not associated with NA in patients with psychosis. In the short term, smoking in the daily life context is associated with a reduction in NA in people vulnerable to psychosis, possibly due to alleviation of withdrawal symptoms, which may complicate smoking cessation.

Exposure to stress is a ubiquitous aspect of human life, but prolonged or traumatic stress is associated with an increased risk for a wide range of psychiatric disorders. However, not all individuals exposed to such stress develop psychiatric disorders, suggesting that additional factors must mediate risk. Indeed, a growing body of evidence suggests that stress exposure and genetic predisposition interact in complex gene–environment interactions. Much of what is known about these interactions in psychiatric disorders comes from epidemiological studies or from investigations focusing on interactions involving a single gene or a small set of genes. Methodological advances in recent years have made it possible to explore gene–environment interactions using novel approaches that provide insight into their polygenic nature as well as into potential mechanisms through which these interactions act, including epigenetic processes. This session will examine the current state of the field, identify existing limitations, and discuss technological advancements that facilitate focused mechanistic investigations into the interplay between stress and genetic risk.

Specifically, an example will be presented illustrating how novel molecular genetic tools can be used to identify gene–environment interactions and map them onto stress physiology and stress-related neuroimaging phenotypes in a human cohort. This case will serve as an example of how multimodal data—ranging from molecular genetic readouts to human phenotypic measures—can be integrated to better understand stress physiology across multiple levels.

None Declared

Peripartum depression (PPD) is one of the most common pregnancy complications; nevertheless, it often goes underdiagnosed. Pinpointing important correlates is crucial for early risk identification and pathophysiology understanding. We aimed to investigate the association between self-reported mood swings during oral contraceptive (OC) use and peripartum depressive symptoms (PPDS).

We used data from the Swedish longitudinal cohort study Mom2B. 3829 women who had reported previous usage of OCs were included. Self-reported mood swings during OC use were assessed through a single question, and PPDS were evaluated using established cut-offs of the Edinburgh Postnatal Depression Scale (EPDS).

Self-reported mood swings during OC use were associated with PPDS at gestational weeks 12–22 (OR = 1.30, 95% CI, 1.01–1.66), 24–34 (OR = 1.37, 95% CI, 1.10–1.71) and 36–42 (OR = 1.39, 95% CI, 1.05–1.82) as well as at postpartum weeks 6–13 (OR = 1.46, 95% CI, 1.12–1.92) and 24–35 (OR = 2.07, 95% CI, 1.43–2.99). Interestingly, self-reported mood swings during OC use were associated with higher odds for newly developed PPDS in early postpartum (OR for weeks 6–13 = 1.92, 95% CI: 1.19–3.08).

Women with self-reported mood swings during OC use have higher risk of experiencing depressive symptoms across the peripartum period and twice the risk of newly developed PPDS during the early postpartum, adding to current evidence of a hormonal sensitive subgroup of women and the opportunity to use this simple question in future predictive efforts.

The last decade has seen a remarkable resurgence of interest in psychedelic drugs such as psilocybin (from magic mushrooms) LSD and DMT (dimethyl tryptamine – the active ingredient of ayahuasca). This has been driven by the discovery that these psychedelics all act agonists of 5-HT2A receptors plus human imaging studies that reveal this action leads to profound alterations in brain measures of activity particularly in terms of increased entropy of EEG MEG and fMRI signals and reduced within-network, but increased between-network, connectivity. In addition they all can increase synaptic growth and brain plasticity. These findings not only explain the subjective nature of the psychedelic experience but also have implications for the treatment of internalising disorders such as depression addiction anorexia and OCD. Subsequent trials, particularly of psilocybin, in these disorders has revealed significant clinical benefits from even just a single administration. My talk will explore these brain mechanisms and clinical data and discuss the potential place of psychedelic medicine in the future of psychiatry.

D. Nutt Shareolder of: chief scientific officer Solvonis Therapeutics ltd, Grant / Research support from: Support for psychedelic research from Compasspathways, Beckley Psytec, Usona, Filament, Small Pharma, Employee of: chief scientific officer Solvonis Therapeutics ltd.

Standard treatments for eating disorders (EDs) typically include nutritional rehabilitation, psychotherapy, and pharmacotherapy, often resulting in modest effectiveness. Neuromodulation has been proposed as a potential add-on strategy. This review aims to critically evaluate its effects on EDs.

The PICO framework was used to conduct a search according to the PRISMA guidelines for systematic reviews (PROSPERO Registration no. CRD42024559700). Descriptive and clinical data for each study, including comorbidity and safety, were tabulated by disorder. Quality assessment was performed using the RoB2 and ROBINS-I tools.

Fifty-eight studies examining rTMS, tDCS, DBS, ECT, iTBS, and dTMS in AN, BN, and BED met inclusion criteria. In AN, the effects of rTMS and tDCS on weight and BMI were modest, with notable benefits for adolescents. Conversely, in refractory AN, DBS produced significant and lasting increases in BMI and reductions in comorbid psychopathology. For BN, single-session rTMS reduced cravings acutely, while multisession results were mixed; tDCS enhanced self-control. In BED, right-DLPFC tDCS decreased cravings, food intake, and binge frequency, showing a dose-dependent effect, whereas early data on NAcc DBS were promising but limited.

Noninvasive stimulation methods effectively reduced cravings and impulsivity, showing more consistent improvements in BN and BED than in weight restoration for AN. Overall, this pattern replicates recent review reports. However, variations in study designs, stimulation protocols, and outcome measures may undermine the reliability of these conclusions and complicate comparisons across studies. Future studies should be adequately powered, multisession, with durability endpoints, and target and dose refinement.

Hundreds of randomized controlled trials have examined the effects of psychotherapies on mental disorders. In my presentation I will give an overview of the current knowledge about these effects, based on the Metapsy initiative, a large project with living systematic reviews of psychological treatments of mental disorders. Large meta-analyses of more than 1,000 randomized controlled trials show that psychotherapies are effective for many mental disorders, with substantial symptom reduction, meaningful response rates across age groups and diagnoses, and long-term effects. Cognitive behavioral therapy has been best examined, but there is no evidence in most disorders that it is more effective than other therapies. Although treatments work, population prevalence has not declined, and evidence-based care can only reduce the overall disease burden by roughly 40%. Innovations are therefore necessary, and many are currently developed. Four main domains of innovation will be highlighted: digital interventions, personalized treatments, development of new or improved therapies, and simplification and wider dissemination of existing therapies. However, there are no “silver bullets” that will dramatically change outcomes, and progress can only come from multiple small, incremental advances.

None Declared

As GLP-1 receptor agonists (GLP-1 RAs) are increasingly used worldwide, concerns about their association with mood disorders have grown. Yet real-world observational studies have produced conflicting findings. This study aims to fully examine the link between GLP-1 RAs and emotional/behavioral outcomes.

Disproportionality analysis of GLP-1 RA adverse events was conducted using FAERS data. Mendelian randomization (MR) employed GLP1R cis-eQTLs as instrumental variables to assess links with mood/behavior-related disorders. Summary-data MR (SMR) was then performed using GLP1R cis-eQTL data.

275,718 adverse events (AEs) associated with GLP-1 RAs were retrieved and analyzed. A mild signal for suicide-related AEs was observed only in the obesity indication subgroup (ROR:1.65, 95% CI: 1.28–2.12). Genetic evidence showed that GLP-1 RAs were likely associated with reduced risks of anxiety, depression, emotional lability, bipolar disorder, and suicide. Mediational analysis indicated that weight loss partially mediated the causal effects of GLP-1 RAs on depression and emotional lability, accounting for 18.28% (95% CI: 9.46–27.10%, P = 0.038) and 7.65% (95% CI: 5.66–9.64%, P < 0.001) of the total effects, respectively. SMR analysis showed that genetically predicted GLP1R expression was negatively associated with anxiety (OR: 0.79, 95% CI: 0.64–0.98, P = 0.031), with no significant associations for other emotional or behavioral outcomes.

Both observational and MR analyses showed that patients treated with GLP-1 RAs may have no increased risk of emotional and behavioral disorders. Instead, genetic proxy activation of GLP-1 RAs may reduce the risk of anxiety, depression, and emotional lability.

Severe mental illness (SMI), including schizophrenia and bipolar disorder, is associated with markedly reduced life expectancy, with premature mortality driven predominantly by preventable somatic comorbidities. Epidemiological data indicate that up to 60% of excess deaths in psychiatric populations are attributable to cardiometabolic disease, including obesity, metabolic syndrome, and type 2 diabetes mellitus (T2DM). Patients with schizophrenia have more than a twofold increased risk of developing T2DM compared with the general population, a vulnerability that emerges even in early, antipsychotic-naïve stages, suggesting shared biological mechanisms beyond lifestyle factors.

Psychotropic treatment, particularly second-generation antipsychotics, may further exacerbate metabolic dysregulation through weight gain, insulin resistance, and direct effects on glucose homeostasis. Clozapine and olanzapine are among the most metabolically burdensome agents, whereas partial dopamine agonists such as aripiprazole, brexpiprazole, and cariprazine demonstrate a more favorable metabolic profile. Despite guideline recommendations (APA, NICE, CANMAT), metabolic monitoring and integrated care remain inconsistently implemented.

We present a clinical case of diabetes decompensation in a patient with SMI receiving long-term antipsychotic therapy, illustrating the complex interplay between psychiatric stability, medication-related metabolic risk, and fragmented somatic care. The case highlights delayed recognition of hyperglycemia, insufficient cardiometabolic screening, and barriers to collaboration between psychiatric and medical services.

Emerging evidence supports immunometabolic pathways linking psychiatric disorders with insulin resistance, inflammation (e.g., IL-6), and gut microbiota alterations. Novel interventions, including GLP-1 receptor agonists, show promise in improving glycemic control and weight without worsening psychiatric symptoms.

Metabolic disease in SMI should be considered a core component of psychiatric practice rather than a secondary complication. Integrated dual psychiatric–somatic interventions, systematic screening, and metabolically informed psychopharmacology are essential to reduce preventable morbidity and mortality in this high-risk population.

None Declared

In the ultra-high risk for psychosis (UHR) field, it is unknown whether understanding symptom relationships, beyond symptom severity alone, may hold prognostic value and inform preventive care. In this study, network analysis was performed to examine the interconnections between baseline symptoms in UHR youth who did and did not transition to psychosis over three years.

In a sample selected from the UHR1000+ cohort, positive and basic symptoms were assessed using the Comprehensive Assessment of At-Risk Mental States. Network analyses and network comparison tests were performed.

195 UHR youth transitioned to psychosis within three years and 346 did not. The two groups did not differ in the network structure, global strength (i.e., the overall level of connectivity between symptoms), or centrality of symptoms (i.e., their importance within networks). The transitioned group was characterized by unusual thought content not being connected to other symptoms; however, its centrality between networks was comparable. Across networks, impaired cognitive functioning connected disorganized speech to impaired emotional functioning, motor functioning, and tolerance to normal stress. Impaired bodily sensation connected perceptual abnormalities to other symptoms.

The networks of youth who transitioned and who did not transition were similar, indicating similar baseline symptom relationships. Across groups, unusual thought content, despite being traditionally associated with transition, had little to no interactions with other symptoms. Clinical manifestations that may need attention include impaired cognitive functioning, which connected several symptoms, and impaired bodily sensation. Future research using time series data may support progress toward individualized care.

Anorexia nervosa (AN) often persists for years, resulting in high morbidity and mortality. Hypoglycaemia, typically assessed from a single morning blood sample, is a critical severity indicator. Continuous glucose monitoring (CGM) provides more comprehensive information on glycaemic patterns. This study aimed to characterize glycaemia in patients with AN and identify its potential drivers among metabolic severity (current BMI), clinical severity (Eating Disorder Inventory-2 [EDI-2] score), and illness duration, in a real-world outpatient setting.

This cross-sectional study included female outpatients with restricting subtype AN. Participants underwent CGM for five days in their usual environment. Collected data comprised age, BMI, illness duration, EDI-2 score, and continuous glycaemic measurements. Glycaemic biomarkers (hypoglycaemic area under the curve [AUC], mean and minimum glycaemia, and coefficient of variation) were computed over 24-hour periods.

Three hundred and four female patients were monitored for a mean of 4.8 days. No significant correlations were observed between glycaemic biomarkers and BMI. Illness duration was significantly associated with mean and minimum glycaemia (r = 0.26 and 0.23, respectively, p < 0.001) and with hypoglycaemia AUC (r = −0.25, p < 0.001).

In female patients with restricting subtype AN, illness duration, rather than BMI, appears to significantly influence glycaemic profiles. This may reflect glycaemic adaptations, a hypothesis that warrants further investigation using CGM, a practical tool for exploring metabolic changes and their potential clinical significance in AN.

In 2025, the European Society of Cardiology (ESC) issued a pivotal document: the Clinical Consensus Statement on Mental Health and Cardiovascular Disease, developed by the ESC Task Force on Mental Health and Cardiovascular Disease under the auspices of the ESC Clinical Practice Guidelines Committee. It was endorsed by the European Federation of Psychologists’ Associations (EFPA), the European Psychiatric Association (EPA), and the International Society of Behavioral Medicine (ISBM).

The document highlights the bidirectional relationship between mental and cardiovascular health, emphasizing its impact on cardiovascular risk and prognosis. It offers evidence-based recommendations for assessing and managing mental health conditions in people at risk for, or living with, cardiovascular disease.

Practical guidance includes integrating mental health into routine cardiology through psycho-cardiac teams, with psychiatrists as integral members. Using a stepwise clinical approach, psychiatrists assess suspected mental health disorders, provide comprehensive pharmacological and psychosocial treatment, and guide care for patients with pre-existing mental disorders who develop cardiovascular disease.

Implementing these recommendations structurally could represent a paradigm shift toward a truly whole-person approach in both cardiovascular and psychiatric care, ensuring that mental and cardiovascular health are addressed in tandem.

None Declared

Optimizing depression treatment intensity and duration is crucial, given an overburdened mental healthcare system. However, decision-making is challenged by heterogeneous treatment effects. We aimed to investigate these effects, accounting for confounders and population heterogeneity, in a real-world dataset from specialized mental healthcare.

The study included 36,946 participants from mental healthcare providers in the Northern Netherlands. We measured the effects of treatment duration and intensity on time to depression recurrence, using monthly costs as a proxy for treatment intensity. An accelerated failure time model was used, adjusting for confounding via entropy weighting. Non-linear effects were examined using restricted cubic splines to identify turning points, after which linear analyses were stratified. Population heterogeneity was explored through K-means clustering analyses, followed by cluster-specific analyses.

In the high-intensity group (above €360/month), a €1000/month increase in treatment intensity may reduce time to recurrence by 16% (acceleration factor [AF] 0.84, 95% CI 0.77–0.92). Conversely, the same increase in the low-intensity group might prolong recurrence-free time by 9.6-fold (AF 9.6, 95% CI 2.18–42.31). Extending treatment duration by 6 months may reduce time to recurrence by 7% (AF 0.93, 95% CI 0.89–0.97) in the long-duration group, with no significant effect in the short-duration group. Five clusters emerged, three of which comprised only women, with AFs of 0.67, 0.80, and 0.81, respectively, under high treatment intensity.

Increasing treatment intensity appears worthwhile only in the low-intensity group, though residual confounding remains possible.

In clinical practice and research, the terms psychosis and schizophrenia are used to describe clinical presentations characterized by a constellation of positive, negative, and cognitive symptoms persisting for longer than one month. Despite shared diagnostic criteria, these disorders are highly heterogeneous with regard to etiology, phase-specific clinical manifestations, illness trajectory, treatment response, and overall outcome. This heterogeneity presents substantial challenges for both clinical management and research.

Over the past two decades, extensive research has focused on identifying biomarkers across multiple domains, including genomics, epigenetics, neuroimaging, transcriptomics, and, more recently, artificial intelligence. The overarching goal of these efforts has been to enable risk prediction and early detection, enhance diagnostic accuracy, facilitate individualized treatment strategies, improve monitoring, and predict treatment outcomes.

These advances have contributed to the emergence of precision psychiatry, which—similarly to other medical specialties—aims to support clinical decision-making and ultimately improve outcomes for individuals with psychosis and schizophrenia. However, most of these innovations remain confined to research settings and have yet to be systematically implemented or validated in routine clinical practice.

In this presentation, we will review the existing tools that show the greatest promise for successful clinical implementation from the perspective of practicing psychiatrists, in alignment with the EPA viewpoint on Precision Medicine in Mental Health: Applications, Challenges, and Recommendations developed by the EPA Task Force Precision Psychiatry.

None Declared

Healthcare systems increasingly recognise that lived experience is not an adjunct to evidence, but a core component of it, to increase the value of care for patients and make healthcare systems more people centred. This presentation examines how OECD member countries collect and use patient-reported outcome measures (PROMs) and patient-reported experience measures (PREMs) to assess the performance of healthcare systems in delivering mental health care. Drawing on OECD work on the Patient-Reported Indicator Surveys (PaRIS) and in mental health PROMs and PREMs, the presentation explores how health outcomes and experiences of care can be translated into comparable policy indicators.

The presentation will address the journey from data to evidence through co-production with people with lived experience across indicator development, data collection, and interpretation, rather than confined to one-off consultations. It will also provide an overview of results from the OECD work on the PROMs and PREMs in mental health care and among people living with mental health conditions.

By reflecting on lessons from international measurement initiatives, the presentation will highlight policy and practical implications for policymakers, researchers, practitioners and service users to make healthcare systems more centred to people’s needs.

None Declared

The COVID-19 pandemic presented significant challenges to infectious disease management and mental health services (MHS). Service demand and delivery changed due to fear of infection, economic hardships, and the psychological effects of protective measures. This systematic review with meta-analysis aims to quantify these impacts on different mental health service settings.

Comprehensive searches were conducted in PubMed, Embase, and PsycINFO, focusing on studies published from the initial outbreak of COVID-19, starting in November 2019. Studies were included comparing the utilization of mental health inpatient, emergency department (ED), and outpatient services (including telemedicine and medication prescriptions) before and during the COVID-19 pandemic. A random-effects model was employed to estimate pooled effects, with study quality assessed using a modified Newcastle-Ottawa Scale.

Among 128 studies, significant decreases in utilization were observed during the initial phase of the pandemic for inpatient services (RR: 0.75, 95% CI: 0.67 to 0.85) and ED visits (RR: 0.87, 95% CI: 0.69 to 1.10). Outpatient services showed a similar decline (RR: 0.78, 95% CI: 0.66 to 0.92), while no significant change was found in psychotropic medication prescriptions (RR: 0.90, CI: 0.77 to 1.05). In contrast, telemedicine utilization increased significantly (RR: 7.57, 95% CI: 3.63 to 15.77).

The findings reveal substantial shifts in mental health service utilization during the pandemic, with the largest reductions in inpatient services and significant increases in telemedicine use. These results emphasize the need for flexible healthcare models. Further research is essential to evaluate the consequences of reduced MHS utilization.