No single, universal “he” (or entity) maintains a definitive, centralized record of psychotherapy outcomes across all cases globally—it’s not how mental health research works. Instead, outcomes are tracked through studies, meta-analyses, and clinical trials, often with varying methodologies. Here’s what the data generally shows:

1. Psychotherapy Alone:

• Studies suggest about 75% of people who undergo psychotherapy (e.g., CBT, psychodynamic therapy) experience some benefit, per the American Psychological Association. A landmark 1994 Consumer Reports survey found 90% of respondents rated it effective, with 54% saying it “helped a great deal.” More rigorous meta-analyses (e.g., Cuijpers et al., 2010) report moderate to large effect sizes (around 0.8) for depression and anxiety, meaning significant symptom reduction for most.
• Non-responders or partial responders: Roughly 25-40% don’t improve significantly or only see partial benefits, depending on the condition, therapy type, and study criteria. For example, in PTSD studies, remission rates hover around 50% with CBT, leaving the rest with partial or no response.

1. Psychotherapy Plus Medication:

• Combined treatment often outperforms either alone. A 2020 network meta-analysis (Cuijpers et al.) found combined therapy had a 27% higher response rate than psychotherapy alone and 25% higher than medication alone for depression (response defined as 50% symptom reduction). Effect sizes for combined treatment are small but significant (e.g., 0.32-0.39 over single treatments).
• Non-response or partial benefit: Even with combined treatment, 20-30% of patients don’t achieve full response, especially in chronic or severe cases like treatment-resistant depression (TRD), where meta-analyses show a moderate effect (d=0.42) for psychotherapy added to usual care.

1. Percentage Not Responding or Partially Benefiting:

• Across studies, 5-10% of patients may worsen during psychotherapy (even in controlled settings), 35-40% see no meaningful change, and 25-50% get partial benefits without full remission. Combined treatment reduces these rates slightly but doesn’t eliminate them. For example, the STAR*D trial (2006) showed 30-40% of depression patients didn’t respond fully even after multiple treatment steps, including combined approaches.

How Conclusions Are Drawn

No one “arrives” at psychotherapy alone being the “best choice” as a universal truth—it’s context-dependent. Researchers and clinicians draw conclusions based on:

• Efficacy Data: Meta-analyses compare effect sizes (e.g., Hedges’ g) across treatments. Psychotherapy alone often matches medication for mild-to-moderate cases and is preferred for fewer side effects and higher acceptability (e.g., 17% more patients stick with it than medication alone, per Cuijpers et al., 2020).
• Patient Preference: Surveys show many favor psychotherapy over drugs (McHugh et al., 2013), influencing recommendations where compliance matters.
• Specific Conditions: For anxiety or mild depression, psychotherapy alone (especially CBT) often suffices, while severe depression or psychosis leans toward combined treatment.
• Statistical Modeling: Logistic regression or backward regression in studies (e.g., on bouldering psychotherapy) identifies predictors of response, like quality of life or interpersonal factors, to argue when psychotherapy alone might suffice.
• Ethical and Practical Factors: Psychotherapy avoids medication side effects (e.g., bradycardia from donepezil) and is viable when drugs aren’t tolerated.

Why Psychotherapy Alone Might Be Favored

If someone claims psychotherapy alone is “best,” they might point to:

• Comparable efficacy to medication in many cases (RR=0.99 vs. meds alone in Cuijpers’ meta-analysis).
• Higher patient retention and satisfaction.
• Long-term skill-building (e.g., Gita-like self-awareness or CBT’s cognitive restructuring) versus medication’s temporary symptom relief.

But this isn’t absolute. Combined treatment consistently edges out solo approaches in moderate-to-severe depression, and non-response rates suggest neither is a silver bullet. The “best choice” hinges on individual factors—severity, chronicity, patient goals—not a one-size-fits-all decree.