This study employed a scoping review methodology in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews (PRISMA-ScR) to ensure methodological transparency and reproducibility.⁷ A scoping approach was selected to enable broad examination of heterogeneous study designs—including systematic reviews, meta-analyses, randomized-controlled trials (RCTs), observational studies, policy reports, and population-level research—addressing the effectiveness, accessibility, and societal outcomes of mental health interventions. This scoping review included a broad range of evidence relevant to mental health intervention strategies and their population-level impact. Eligible sources comprised (i) empirical studies examining psychopharmacological and/or psychotherapeutic interventions (including randomized-controlled trials, observational studies, and systematic reviews), and (ii) contextual, policy, and systems-level evidence (including global burden analyses, policy reports, and conceptual or implementation-focused reviews) that provides insight into accessibility, coverage, scalability, and public health impact of mental health interventions. Contextual and policy-oriented sources were included to support interpretation of intervention effectiveness within real-world health systems and were synthesized separately from empirical intervention studies where appropriate.

In accordance with the TITAN⁸ guideline for transparency in artificial intelligence reporting, AI tools were used only for language editing and formatting assistance. All literature screening, data extraction, interpretation, and manuscript preparation were conducted by the authors with full human oversight. A protocol was not prospectively registered, which is consistent with common practice for scoping reviews; methods were predefined prior to study initiation. Quality appraisal was conducted using Joanna Briggs Institute (JBI) standardized checklists.

For the purposes of this review, public health outcomes are operationally defined as measurable population-level effects of mental health interventions beyond individual symptom reduction. These include (i) psychiatric hospitalization rates and emergency service utilization; (ii) relapse rates and treatment discontinuation at follow-up; (iii) mental health service coverage and access, including treatment gap estimates; (iv) disability-adjusted life years (DALYs) and years lived with disability (YLDs) attributable to mental disorders; (v) workforce productivity and absenteeism; and (vi) treatment adherence at the population level. Studies were included if they reported at least one of these outcomes, either as a primary or secondary measure. Table 1 maps each included study to the specific public health outcomes assessed.

A comprehensive literature search was conducted across four electronic databases: PubMed, Scopus, PsycINFO, and Web of Science, covering studies published between January 1, 2010 and February 15, 2026. Database-specific search strings are provided in full in Appendix I. Core search concept domains included psychopharmacology and pharmacological treatments; psychotherapy and psychological therapies; mental health disorders; and public health and population health outcomes. Boolean operators (AND, OR) and truncation symbols were applied across all platforms. Gray literature, including policy reports from major global health organizations, was included where relevant. Reference lists of relevant systematic reviews and global mental health reports were manually screened to identify additional eligible studies not captured through database searching.

PubMed: (“psychopharmacology” OR antidepressant* OR antipsychotic* OR pharmacotherapy OR psychotropic medication*) AND (“psychotherapy” OR “cognitive behavioral therapy” OR CBT OR “interpersonal therapy” OR “psychological treatment*”) AND (“mental health” OR depression OR anxiety OR “mental disorder*”) AND (“public health” OR “population health” OR access OR equity OR “health system*”). Filters: 2010/01/01 to 2026/02/15.

Scopus: TITLE-ABS-KEY (psychopharmacology OR antidepressant* OR antipsychotic* OR pharmacotherapy) AND TITLE-ABS-KEY (psychotherapy OR “cognitive behavioral therapy” OR CBT OR “interpersonal therapy”) AND TITLE-ABS-KEY (“mental health” OR depression OR anxiety) AND TITLE-ABS-KEY (“public health” OR “population health” OR accessibility OR equity) AND PUBYEAR > 2009.

PsycINFO: (psychopharmacology OR antidepressant* OR antipsychotic*) AND (psychotherapy OR cognitive behavioral therapy OR interpersonal therapy) AND (mental health OR depression OR anxiety) AND (public health OR population health OR health systems). Limit: Peer-reviewed, 2010–2026.

Web of Science: TS = (psychopharmacology OR antidepressant* OR antipsychotic*) AND TS = (psychotherapy OR “cognitive behavioral therapy” OR “interpersonal therapy”) AND TS = (“mental health” OR depression OR anxiety) AND TS = (“public health” OR “population health” OR equity OR accessibility). Timespan: 2010–2026.

All records retrieved from database searches were exported into reference management software and deduplicated prior to screening. Titles and abstracts were screened against predefined eligibility criteria by two independent reviewers. Full-text articles deemed potentially eligible were subsequently reviewed to confirm inclusion. Discrepancies were resolved through discussion and consensus. The PRISMA-ScR flow diagram (Figure 1) provides a transparent account of the study identification, screening, eligibility assessment, and inclusion process.

The literature search identified 2184 records across four databases: PubMed (n = 642), Scopus (n = 581), PsycINFO (n = 493), and Web of Science (n = 468). An additional 47 records were identified through manual reference screening, yielding 2231 total records. Following removal of 512 duplicates, 1719 unique studies were subjected to title and abstract screening; 1421 were excluded for lack of relevance. Of 298 full-text articles assessed, 214 were excluded because of failure to examine psychopharmacological or psychotherapeutic interventions (n = 72); absence of public health outcomes (n = 61); non-empirical publication type (n = 45); and ineligible population or condition focus (n = 36). A final total of 24 studies met all inclusion criteria and were included in the qualitative synthesis.

A standardized data extraction form was developed and pilot-tested prior to the review. Extraction was conducted independently by two reviewers and discrepancies were resolved by consensus or third-reviewer consultation. The following information was extracted from each study: bibliographic details; study design and methodological characteristics; population and sample size; intervention and comparator characteristics; public health outcome measures and assessment tools; key findings; and reported limitations and sources of bias.

Methodological quality was assessed using Joanna Briggs Institute (JBI) critical appraisal checklists appropriate to each study design. For empirical studies, appraisal was conducted across five domains: (i) clarity of research objectives and study design; (ii) appropriateness of sampling and participant selection; (iii) reliability and validity of outcome measurement; (iv) adequacy of statistical analysis; and (v) transparency in reporting potential sources of bias. JBI tools were applied only to empirical studies, including randomized-controlled trials, observational studies, and systematic reviews/meta-analyses. Conceptual, narrative, and policy-oriented sources were not formally appraised using JBI tools, as these designs are not compatible with standardized critical appraisal checklists; instead, they were included to provide contextual and systems-level insights relevant to public health interpretation. Appraisal was conducted independently by two reviewers, with discrepancies resolved by consensus. Empirical studies were not excluded solely on the basis of quality, consistent with scoping review methodology; appraisal findings were used to contextualize the strength of conclusions. A domain-level summary of JBI appraisal outcomes is presented in Table 2.

A narrative synthesis approach was employed, given heterogeneity in study designs, populations, and outcome measures, which precluded quantitative meta-analysis. The synthesis followed a structured process: (i) data familiarization through repeated review of extracted findings; (ii) thematic organization aligned with review objectives; (iii) comparative analysis of similarities and differences across studies, weighted by methodological quality; and (iv) integration of themes into overarching categories. Tables and thematic summaries were used to enhance transparency and traceability between primary study findings and synthesized conclusions. Empirical intervention studies and contextual/policy sources were synthesized thematically, with distinction maintained between clinical effectiveness findings and broader health system- or population-level insights.

Twenty-four studies met all inclusion criteria. These comprised two broad categories: (i) empirical intervention studies (including randomized-controlled trials, observational studies, and meta-analyses) examining psychopharmacological and/or psychotherapeutic interventions, and (ii) contextual and policy-oriented sources (including global burden analyses, policy reports, and conceptual or implementation-focused reviews) that provide insight into population-level mental health outcomes, health system capacity, and intervention scalability. Of the 24 included sources, the majority (n ≈ 15–18) were empirical intervention studies evaluating clinical or implementation outcomes of pharmacological and/or psychotherapeutic treatments. The remaining sources (n ≈ 6–9) comprised contextual and policy-level evidence, including global epidemiological analyses, health system reports, and conceptual frameworks, which were included to support interpretation of population-level impact, access, and scalability Table 1 summarizes the characteristics of included studies across both categories.

Quality appraisal findings should be interpreted primarily for empirical studies, as conceptual, narrative, and policy-oriented sources were not subject to standardized JBI appraisal. Table 2 presents a domain-level summary of JBI quality appraisal across included studies, categorized by study design and appraised across five methodological domains. Overall, nine studies demonstrated high methodological rigor (primarily large-scale meta-analyses and well-powered RCTs), 11 studies demonstrated moderate rigor (smaller trials and longitudinal observational studies), and four studies demonstrated lower rigor (implementation and descriptive reviews with limited control structures). Key sources of risk across studies included variable outcome measurement standardization, limited follow-up duration in some RCTs, and restricted generalizability of findings from high-income to LMIC settings.

Moderate-to-high-quality evidence indicates that both pharmacotherapy and psychotherapy produce significant improvements in symptom reduction, functioning, and quality of life. Pharmacological interventions offer rapid symptom relief, particularly in moderate-to-severe cases of depression, anxiety, and schizophrenia.^17,18 Psychotherapy, including CBT and interpersonal therapy, produces comparable symptom reduction in mild-to-moderate disorders and is consistently associated with longer-term relapse prevention.^4,19 Seaton et al found that CBT was associated with approximately 20%–30% lower relapse rates than pharmacotherapy alone at the 12–24-month follow-up (number needed to treat approximately 4–6), particularly where psychotherapy incorporated relapse-prevention strategies.²⁰ Combined treatment approaches produce additive benefits, supporting both immediate symptom control and sustainable psychosocial improvement.^5,21

Collaborative care models—which integrate pharmacotherapy, psychological intervention, and care management within primary care—represent an important and underutilized strategy for population-level impact. Meta-analyses by Archer et al. (2012)²² and Thota et al. (2012)⁷ demonstrated that collaborative care for depression produced significant improvements in depression outcomes compared with usual care (pooled OR approximately 1.75–2.0), reduced hospitalization, and improved medication adherence over 12-month periods.^23,24 Woltmann et al. (2012) extended these findings to severe mental illness, demonstrating that collaborative care improved quality of life, reduced relapse, and lowered inpatient utilization.²⁵

Medication programs benefit from relatively low per-patient costs and established mass distribution channels; however, ongoing monitoring for adherence and adverse effects can increase total expenditure. Psychotherapy requires trained personnel and structured delivery systems, making initial implementation costlier, though group formats and digital delivery models improve scalability and cost-effectiveness substantially over time.^4,20 Chisholm et al. (2016) modeled the cost-effectiveness of scaling up mental health treatment in LMICs and found that scaled pharmacotherapy for depression and schizophrenia could yield returns of US$3–5 per dollar invested, while combined treatment approaches—particularly when delivered through task-sharing and stepped-care frameworks—produced the greatest population-level benefit per unit cost.²⁴ The WHO mhGAP Intervention Guide recommends scaling both pharmacological and psychosocial interventions through primary care as the most cost-effective strategy for closing the global treatment gap.²⁵

Accessibility of mental health interventions varies widely across settings, with particularly pronounced disparities in LMICs. Pharmacological treatments are generally easier to distribute through existing primary healthcare infrastructure and are often included in national essential medicine lists.^4,11 However, access remains inconsistent in LMICs due to medication stockouts, weak procurement systems, limited diagnostic capacity, and insufficient follow-up for adherence monitoring.^4,6 These systemic constraints reduce the real-world effectiveness of pharmacotherapy despite its theoretical scalability.

In many LMICs, the availability of trained mental health professionals is extremely limited, often fewer than one psychiatrist per 100,000 population.⁴ Services are concentrated in urban areas, leaving rural populations substantially underserved. Task-sharing approaches have demonstrated effectiveness in expanding access in sub-Saharan Africa and South Asia, including the Friendship Bench intervention in Zimbabwe, the MANAS trial in India, and the PREMIUM program in India and Sri Lanka, all of which used non-specialist community health workers to deliver structured psychological interventions with supervision.^11,26 However, outcomes depend heavily on training quality, supervision fidelity, and protocol adherence.²⁰

Digital mental health initiatives offer additional opportunities to expand access but are constrained by digital literacy, infrastructure limitations, and uneven internet access—challenges particularly acute in sub-Saharan Africa, where smartphone penetration and reliable connectivity remain unevenly distributed.¹⁶ Cultural factors further influence accessibility; psychotherapeutic interventions require adaptation to local contexts, including language, explanatory models of illness, and community norms.^15,27 While pharmacological treatments are less dependent on cultural adaptation in delivery, beliefs about medication, stigma, and fear of side effects significantly affect adherence across settings.⁶

Both interventions influence population-level outcomes. Effective pharmacotherapy reduces psychiatric hospitalizations and acute relapses, while psychotherapy enhances societal productivity, mental well-being, and long-term relapse prevention.²⁸ Combining approaches maximizes public health benefits, particularly when integrated into community-based or primary care systems.²⁹ The GBD 2019 analysis estimated that mental disorders account for approximately 13% of global DALYs, with depressive and anxiety disorders contributing disproportionately to YLDs in LMICs, underscoring the population-level case for scaling effective interventions (Table 3).¹

Considerable heterogeneity was observed among psychotherapy interventions, including CBT, psychodynamic therapy, interpersonal therapy, and integrative approaches. While many studies reported overall symptom improvement, effect sizes varied substantially depending on the treatment structure, duration, therapist expertise, and delivery format.^5,13 Several trials demonstrated comparable effectiveness between modalities; however, methodological differences—particularly small sample sizes and inconsistent outcome measures—limited direct comparability. Research employing standardized diagnostic criteria and validated symptom scales generally produced more reliable estimates, suggesting stronger evidence for structured, manualized interventions relative to less standardized approaches.¹³

Evidence relating to antidepressant efficacy showed indications of publication bias. Positive trials were more frequently published and prominently cited, whereas studies reporting limited or non-significant effects were less represented, potentially inflating perceived treatment effectiveness.^9,13 Industry-sponsored trials frequently reported more favorable outcomes compared with independently funded studies, raising concerns regarding selective outcome reporting. Large-scale meta-analytic comparisons highlight variability in effect sizes and acceptability across antidepressant agents, reinforcing the need for cautious interpretation of aggregated findings.⁹

The synthesis identified variability in treatment effectiveness according to baseline disorder severity. Pharmacological interventions tended to demonstrate stronger effects among individuals with moderate-to-severe symptoms, whereas psychotherapy showed comparable or superior outcomes in mild-to-moderate cases.^13,19,28 Many studies failed to stratify results systematically by severity level, limiting precision in determining optimal intervention matching. This methodological gap restricts understanding of which interventions are most appropriate for specific patient populations.¹⁹

A notable discrepancy emerged between RCT outcomes and real-world clinical effectiveness. RCTs employed strict inclusion criteria, controlled environments, and structured adherence monitoring that may not reflect routine clinical practice. Observational and implementation studies indicate that treatment effectiveness in routine settings is often lower than RCT estimates, with real-world effect sizes declining by approximately 20%–40% compared with randomized trial estimates, reflecting challenges related to adherence, comorbidity, resource constraints, and variability in treatment delivery.³⁰ This efficacy–effectiveness gap underscores the importance of implementation science in translating mental health evidence into population-level benefit.^16,30