### PAPER BRIEF ``` PAPER TYPE: Empirical — Randomized Controlled Trial (RCT) DOMAIN: Clinical Psychology / Psychiatry PRIMARY CLAIM: CBT-E produces 30% greater reduction in anxiety symptoms compared to standard CBT in adults with GAD METHOD USED: RCT, two-arm parallel group, 12-week intervention SAMPLE / DATA: 200 adults with GAD diagnosis; GAD-7 at baseline, week 12, and 6-month follow-up READER GOAL: Should CBT-E be recommended in clinic SOPs for GAD treatment? ``` --- ### CLAIM INVENTORY ``` ── CLAIM 01 ────────────────────────────────────────────────────────────── TYPE: PRIMARY CLAIM: CBT-E produces 30% greater symptom reduction than standard CBT METHOD: RCT with GAD-7 measurement at baseline and week 12 THRESHOLD: CAUSAL — requires RCT with adequate power, blinding, ITT analysis, and controlled confounders VERDICT: WEAK FIT NOTE: RCT design is appropriate for causal claims, but the differential dropout (15% vs 5%) raises serious concern about whether the 30% figure reflects treatment efficacy or attrition bias. Handling of missing data not reported. ────────────────────────────────────────────────────────────────────────── ── CLAIM 02 ────────────────────────────────────────────────────────────── TYPE: SUPPORTING CLAIM: Results are statistically significant (p < 0.01) METHOD: Inferential statistics on GAD-7 scores THRESHOLD: STATISTICAL ADEQUACY — p-value alone is insufficient; effect size and confidence intervals are needed VERDICT: WEAK FIT NOTE: p < 0.01 is reported but no effect size (Cohen's d, partial η²) or confidence interval is mentioned. With n=200, the study may be adequately powered, but the clinical magnitude of the 30% difference cannot be independently evaluated from p alone. ────────────────────────────────────────────────────────────────────────── ── CLAIM 03 ────────────────────────────────────────────────────────────── TYPE: SUPPORTING CLAIM: Treatment effects demonstrate long-term durability METHOD: 6-month follow-up GAD-7 assessment THRESHOLD: LONGITUDINAL — requires retention rates, ITT or per-protocol specification, and attrition accounting VERDICT: WEAK FIT NOTE: If the CBT-E group already had 15% dropout by week 12, the 6-month follow-up sample is likely further reduced and potentially unrepresentative. Durability claim rests on a likely-biased subsample. This is the paper's most vulnerable claim. ────────────────────────────────────────────────────────────────────────── ── CLAIM 04 ────────────────────────────────────────────────────────────── TYPE: BACKGROUND CLAIM: GAD-7 is an appropriate outcome measure for GAD METHOD: Not tested — assumed by authors THRESHOLD: CONSTRUCT VALIDITY — not under examination here VERDICT: APPROPRIATE (GAD-7 is a validated, widely used tool; this assumption is defensible) ────────────────────────────────────────────────────────────────────────── ``` --- ### QUALITY SCORECARD ``` DIMENSION SCORE NOTE ────────────────────────────────────────────────────────────────────────── Internal Validity WEAK Differential dropout (15% vs 5%) is a serious threat to group equivalence at outcome. Missing data handling not reported. ITT analysis not confirmed. External Validity CANNOT ASSESS Sample characteristics (severity, comorbidities, demographics) not described. Generalisability to your clinic population unknown. Construct Validity ADEQUATE GAD-7 is a validated measure. However, "30% greater reduction" is not anchored to a minimal clinically important difference (MCID), so clinical vs statistical significance is unresolved. Statistical Adequacy WEAK p < 0.01 reported; no effect size, no confidence intervals, no power calculation mentioned. Cannot assess clinical magnitude of the finding independently. Reproducibility WEAK Dropout handling undisclosed. Therapist training/fidelity protocols not described. No mention of blinding procedures or allocation concealment. ────────────────────────────────────────────────────────────────────────── ``` --- ### FINDING CARDS ``` ── FINDING 01 ──────────────────────────────────────────────────────────── TYPE: WEAK SIGNAL: CBT-E outperforms standard CBT on GAD-7 at week 12 EVIDENCE: RCT design + p < 0.01 on primary outcome CONFIDENCE: MEDIUM (55%) — RCT is appropriate design; p-value supports a directional finding. Confidence capped at MEDIUM because: (a) differential dropout threatens group comparability, (b) no effect size reported, (c) missing data handling unknown. IMPACT: The direction of the finding (CBT-E > standard CBT) is plausible and worth noting, but the 30% magnitude figure cannot be trusted without attrition correction. ────────────────────────────────────────────────────────────────────────── ── FINDING 02 ──────────────────────────────────────────────────────────── TYPE: CONTESTED SIGNAL: Long-term durability of CBT-E effects at 6 months EVIDENCE: 6-month follow-up GAD-7 assessment reported CONFIDENCE: LOW (30%) — Follow-up data almost certainly drawn from a self-selected subsample (completers only, after 15% CBT-E dropout). No retention figures at 6 months provided. Survivorship bias is highly probable. IMPACT: The durability claim is the weakest in the paper. Do not use it to justify long-term SOP commitments without independent corroboration. ────────────────────────────────────────────────────────────────────────── ── FINDING 03 ──────────────────────────────────────────────────────────── TYPE: GAP SIGNAL: Why did CBT-E have 3× the dropout rate of standard CBT? EVIDENCE: Dropout noted (15% vs 5%) but "not heavily addressed" per paper description CONFIDENCE: UNKNOWN — Cause of differential dropout not reported IMPACT: This is clinically critical for SOP decisions. Higher dropout may indicate patient burden, side-effect- equivalent distress, or poor fit for certain profiles. Recommending CBT-E without understanding why patients disengage at 3× the rate introduces real clinical risk. ────────────────────────────────────────────────────────────────────────── ── FINDING 04 ──────────────────────────────────────────────────────────── TYPE: GAP SIGNAL: Clinical significance of "30% greater reduction" not established EVIDENCE: No MCID anchor, no effect size, no CI reported CONFIDENCE: UNKNOWN IMPACT: A statistically significant 30% improvement on GAD-7 could range from barely perceptible to transformative in practice. Cannot determine whether this crosses the threshold for routine clinical implementation. ────────────────────────────────────────────────────────────────────────── ── FINDING 05 ──────────────────────────────────────────────────────────── TYPE: GAP SIGNAL: Therapist competency and treatment fidelity protocols not described EVIDENCE: Not mentioned in paper description CONFIDENCE: UNKNOWN — UNDISCLOSED gap IMPACT: CBT-E results in an RCT setting depend heavily on therapist training. Replicating outcomes in a routine clinic without fidelity standards is uncertain. ────────────────────────────────────────────────────────────────────────── ``` --- ### EVIDENCE INVENTORY ``` EXISTS: → RCT design (appropriate for causal inference) → Validated outcome measure (GAD-7) → Adequate sample size (n=200) → Statistical significance reported (p < 0.01) → Two measurement points post-baseline (week 12 + 6-month follow-up) → Dropout rates disclosed (partial credit — rates given, causes not) MISSING: → Effect size and confidence intervals [UNDISCLOSED] → Missing data / ITT analysis specification [UNDISCLOSED] → Reasons for differential dropout [UNDISCLOSED] → 6-month follow-up retention figures [UNDISCLOSED] → Sample demographic and clinical characteristics [UNDISCLOSED] → Therapist training and fidelity monitoring [UNDISCLOSED] → Blinding and allocation concealment procedures [UNDISCLOSED] → MCID anchor for GAD-7 change scores [ADDRESSABLE] → Adverse effects or patient burden data [UNDISCLOSED] GAP CLASSIFICATION SUMMARY: ADDRESSABLE: 1 (MCID comparison — retrievable from literature) STRUCTURAL: 0 UNDISCLOSED: 7 (all other gaps above — absent from this paper) ────────────────────────────────────────────────────────────────────────── ``` --- ### READER VERDICT ``` GOAL: Should CBT-E be recommended in clinic SOPs for GAD treatment? VERDICT: NO — not on the basis of this paper alone REASONING: The RCT design gives this paper a credible foundation, and the directional finding (CBT-E outperforms standard CBT) is worth taking seriously. However, the 3× higher dropout in the CBT-E arm — left unexplained — is a direct patient safety and feasibility concern for SOP adoption. The durability claim rests on a likely biased follow-up sample. Seven undisclosed data points prevent independent verification of the primary finding. SOPs affect all eligible patients; a single trial with these gaps does not meet the evidentiary threshold for that scope of implementation. CONDITION: Verdict would shift to CONDITIONAL YES if: (1) Authors provide ITT analysis with dropout imputation (2) Differential dropout causes are explained and clinically manageable (3) Effect size exceeds MCID for GAD-7 (typically ≥5 points or ~30% score reduction anchored to functional outcomes) (4) At least one independent replication exists (5) Therapist fidelity requirements are specified and feasible in your clinic context ────────────────────────────────────────────────────────────────────────── ``` --- ### REPRODUCIBILITY FLAG ``` REPRODUCIBILITY RISK: HIGH KEY THREATS: → Differential dropout (15% vs 5%) with no imputation or ITT analysis reported — results may not survive reanalysis → Therapist training protocols undisclosed — the most common reason RCT psychotherapy findings fail to replicate in community settings → Blinding and allocation concealment not described — risk of selection and performance bias unknown → 6-month follow-up sample size and composition not reported — durability finding cannot be independently evaluated → No effect size reported — magnitude of benefit unverifiable WHAT WOULD HELP: → Full CONSORT-compliant trial report (flow diagram, ITT table) → Sensitivity analysis comparing completer-only vs ITT results → Therapist training manual or fidelity checklist → Dropout interview data or stated reasons for discontinuation → Pre-registration record (e.g. ClinicalTrials.gov) to verify outcome measures were not changed post-hoc ────────────────────────────────────────────────────────────────────────── ``` --- **A practical note for your context:** Before any SOP decision, it would be worth checking whether CBT-E has independent replication trials and whether a CONSORT-compliant version of this paper (or its full publication) addresses the gaps flagged above. If you can share the full methods section, the confidence scores on Finding 01 and 02 could shift substantially in either direction.