Clinical Trial Intelligence

Which Endpoints Are Most Frequently Measured in European Psychiatry Phase 2 & 3 Trials?

5 July 2026

Across 138 unique CTIS-authorized European psychiatry Phase 2 & 3 trial records, safety and adverse-event endpoints are the most common recurring endpoint family, appearing in 66 trials (47.8%). Among standardized clinical scales, functioning/quality-of-life endpoints appear in 54 trials (39.1%), Clinical Global Impression (CGI) in 53 (38.4%), and Montgomery-Åsberg Depression Rating Scale (MADRS) in 41 (29.7%). MADRS is the leading primary endpoint family, while CGI, functioning/quality of life, and safety dominate secondary endpoints.

138
unique CTIS psychiatry Phase 2 & 3 trial records
66
trials measure safety / adverse events
28
trials use MADRS as a primary endpoint family
51
trials use CGI as a secondary endpoint family

Which endpoint families are most common overall?

Safety/adverse-event endpoints appear in 66 of 138 trials (47.8%), followed by heterogeneous disease-specific endpoints in 65 (47.1%), functioning/quality of life in 54 (39.1%), CGI in 53 (38.4%), and MADRS in 41 (29.7%).

Endpoint family frequency, any endpoint role
Safety / adverse events66 / 138 · 47.8%
Heterogeneous disease-specific endpoints65 / 138 · 47.1%
Functioning / quality of life54 / 138 · 39.1%
CGI global impression53 / 138 · 38.4%
MADRS depression score41 / 138 · 29.7%
Response / remission / relapse status37 / 138 · 26.8%
Other depression scales29 / 138 · 21.0%
Biomarkers / imaging / EEG29 / 138 · 21.0%
Percentages are trial-level presence across 138 unique CTIS psychiatry Phase 2 & 3 records.
Interpretation

Psychiatry endpoint strategy is split between universal safety capture and disease-specific symptom measurement. CGI and functioning/quality-of-life endpoints act as cross-disease secondary anchors, while MADRS is the most visible disease-specific standardized scale.

How do primary, secondary, and exploratory endpoints differ?

Primary endpoint selection is led by MADRS in 28 trials (20.3%). Secondary endpoints are broader: CGI appears in 51 trials (37.0%), functioning/quality of life in 50 (36.2%), and safety in 49 (35.5%). Exploratory/other endpoints are visible in only 4 trials (2.9%).

Trial-level endpoint family by role
Endpoint family Primary Secondary Exploratory / other
MADRS depression score28292
Safety / adverse events15493
Response / remission / relapse11311
Substance use / abstinence / craving10120
Biomarkers / imaging / EEG9241
PANSS schizophrenia score890
CGI global impression1512
Functioning / quality of life7501
Counts show number of unique trials where the endpoint family appears in the specified role.
Interpretation

Primary endpoints concentrate around disease-defining symptom scales, especially MADRS in depression and PANSS in schizophrenia. Secondary endpoints carry the broader clinical interpretation layer: CGI, quality of life, safety, response/remission, suicidality, and biomarker signals.

Which diseases drive endpoint selection?

Depression disorders are the largest group with 48 of 138 trials (34.8%), followed by schizophrenia/psychosis with 21 (15.2%), substance use disorders with 13 (9.4%), neurodevelopmental disorders with 10 (7.2%), dementia-related neuropsychiatry with 9 (6.5%), and bipolar disorder with 7 (5.1%).

Disease group distribution
Depression disorders48 · 34.8%
Schizophrenia / psychosis21 · 15.2%
Substance use disorders13 · 9.4%
Neurodevelopmental disorders10 · 7.2%
Dementia-related neuropsychiatry9 · 6.5%
Bipolar disorder7 · 5.1%
Major disease groups are normalized from CTIS trial disease labels.
Interpretation

The European psychiatry Phase 2 & 3 endpoint landscape is primarily shaped by depression and schizophrenia programs. Smaller but analyzable clusters exist for substance use, autism/ADHD, dementia-related symptoms, and bipolar disorder.

Which endpoint families dominate by disease?

Depression is anchored by MADRS in 29 of 48 trials (60.4%). Schizophrenia/psychosis is anchored by CGI in 15 of 21 trials (71.4%) and PANSS in 13 (61.9%). Substance use trials rely on abstinence/craving/use endpoints in 12 of 13 trials (92.3%).

Dominant endpoint family within analyzable disease groups
Disease group Top endpoint family Trials
Depression disordersMADRS depression score29 / 48 · 60.4%
Schizophrenia / psychosisCGI global impression15 / 21 · 71.4%
Substance use disordersUse / abstinence / craving12 / 13 · 92.3%
Neurodevelopmental disordersAutism / irritability / social scales5 / 10 · 50.0%
Dementia-related neuropsychiatryNPI / agitation scales7 / 9 · 77.8%
Bipolar disorderCGI global impression5 / 7 · 71.4%
Only disease groups with sufficient trial counts are interpreted.
Interpretation

Endpoint standardization is strongest where disease-specific scales are established: MADRS for depression, PANSS/CGI for schizophrenia, NPI/agitation scales for dementia-related symptoms, and TLFB/abstinence/craving endpoints for substance use disorders.

Which sub-diseases have enough trials to compare?

The largest sub-disease clusters are major depressive disorder with 33 trials, schizophrenia with 15, treatment-resistant depression with 9, autism spectrum disorder with 8, alcohol use disorder/dependence with 6, and psychosis with 6.

Sub-disease trial counts
33
Major depressive disorder
15
Schizophrenia
9
Treatment-resistant depression
8
Autism spectrum disorder
6
Alcohol use disorder / dependence
6
Psychosis
Sub-disease labels are normalized from CTIS disease strings.
Interpretation

The most benchmarkable sub-diseases for endpoint planning are major depressive disorder, schizophrenia, treatment-resistant depression, autism spectrum disorder, alcohol use disorder, and psychosis. Smaller sub-diseases are better used for case-level endpoint landscaping rather than percentage-based comparisons.

How do Phase 2 and Phase 3 endpoint strategies differ?

The phase comparison includes 68 Phase 2 cohort records and 72 Phase 3 cohort records. Phase 2 is more safety/translational: safety appears in 40 of 68 records (58.8%) and biomarkers/imaging/EEG in 18 (26.5%). Phase 3 is more clinical-scale driven: CGI appears in 34 of 72 records (47.2%) and MADRS in 25 (34.7%).

Endpoint family presence by phase cohort
Endpoint family Phase 2 Phase 3
Safety / adverse events40 / 68 · 58.8%27 / 72 · 37.5%
Functioning / quality of life28 / 68 · 41.2%28 / 72 · 38.9%
CGI global impression20 / 68 · 29.4%34 / 72 · 47.2%
MADRS depression score17 / 68 · 25.0%25 / 72 · 34.7%
Biomarkers / imaging / EEG18 / 68 · 26.5%11 / 72 · 15.3%
Response / remission / relapse19 / 68 · 27.9%18 / 72 · 25.0%
Suicidality / self-harm10 / 68 · 14.7%15 / 72 · 20.8%
Phase counts use CTIS phase-cohort records; two multi-phase protocols contribute to both phase cohorts.
Interpretation

Phase 2 psychiatry trials show more exploratory/translational endpoint architecture, with higher safety and biomarker/imaging use. Phase 3 shifts toward regulator-facing clinical rating outcomes, especially CGI and MADRS.

Do modality factors change endpoint selection?

Small molecules dominate the dataset with 111 of 138 trials (80.4%). Peptide/protein/enzyme interventions account for 8 trials (5.8%). Other modalities are individually small and do not support stable percentage comparisons.

Modality comparison where trial counts allow comparison
Modality Most frequent endpoint families
Small molecule
111 trials
Safety 53 / 111 (47.7%); CGI 47 (42.3%); functioning/QoL 45 (40.5%); MADRS 39 (35.1%)
Peptide/protein/enzyme
8 trials
Autism/irritability/social scales 4 / 8 (50.0%); safety 4 (50.0%); biomarkers/imaging/EEG 3 (37.5%)
Modality values are normalized from CTIS drug modality context variables.
Interpretation

Because small molecules account for four out of five psychiatry Phase 2 & 3 records, overall endpoint trends largely reflect small-molecule development. Peptide/protein/enzyme trials appear more concentrated in autism/neurodevelopmental endpoint patterns.

What does CTIS submission timing add?

The CTIS-authorized psychiatry Phase 2 & 3 records are distributed across authorization years as follows: 59 trials in 2024 (42.8%), 57 in 2025 (41.3%), and 22 in 2026 (15.9%). This makes endpoint benchmarking most robust for the 2024–2025 CTIS period.

Trials by first CTIS authorization year
59
2024 · 42.8%
57
2025 · 41.3%
22
2026 · 15.9%
Authorization year uses the first CTIS authorization date in the trial context variables.
Interpretation

For EU submissions, CTIS endpoint disclosure makes it possible to benchmark whether a proposed psychiatry endpoint package aligns with recent European precedent by disease, phase, and endpoint role before protocol finalization.

Adjacent questions answered from the same data

The same CTIS endpoint data supports three practical planning questions for psychiatry protocol teams.

Which primary endpoints are most defensible by precedent?

MADRS is the leading primary endpoint family with 28 of 138 trials (20.3%), followed by safety/adverse events in 15 (10.9%), response/remission/relapse in 11 (8.0%), and substance-use endpoints in 10 (7.2%).

Which disease areas show the strongest endpoint standardization?

Substance use disorders show the strongest concentration, with use/abstinence/craving endpoints in 12 of 13 trials (92.3%). Dementia-related neuropsychiatry follows with NPI/agitation scales in 7 of 9 trials (77.8%).

Are exploratory endpoints central in CTIS psychiatry records?

No. Exploratory/other endpoint fields are visible in 4 of 138 trials (2.9%), while primary endpoints appear in 126 trials and secondary endpoints in 107 trials.

Definitions

MADRS = Montgomery-Åsberg Depression Rating Scale. CGI = Clinical Global Impression. PANSS = Positive and Negative Syndrome Scale. NPI = Neuropsychiatric Inventory. HAM-A = Hamilton Anxiety Rating Scale. HAMD/HDRS = Hamilton Depression Rating Scale. CTIS = Clinical Trials Information System used for EU clinical trial submissions and authorization records.