TTE Assessment Skill

Phase 0: Triage

Ask the user:

What are you assessing?

• (A) A published TTE study (provide PDF path or paste text)
• (B) A TTE protocol I designed (provide file path or paste)

If option A and a PDF is provided, attempt to read it. If extraction fails or is incomplete, ask the user to paste the relevant sections (abstract, methods, results, any TTE protocol table).

Phase 1: Information Extraction

Read the provided material and extract information for each scoring item. Work through the document systematically:

1. Identify the study context first — this determines which items are applicable:

   • What is the clinical scenario? (Scenario 1, 2, or 3 — see ../tte-foundation/references/scenarios.md)
   • What treatment strategy type? (point, static sustained, dynamic)
   • What comparator type? (active, blank/no-treatment, standard care)
   • Is this a pharmacologic intervention?
   • Could individuals become eligible at multiple time points?

2. Extract evidence for each scoring item — for each item in the rubric, note:

   • Whether it's reported and where (section, page, table)
   • What was stated
   • Any ambiguity

Read the full rubric from ../tte-foundation/references/rubric.md to guide extraction.

Phase 2: Follow-up Questions

For items that couldn't be determined from the text:

• Ask the user for clarification
• Group closely related items together (e.g., all eligibility-related questions in one batch) to avoid excessive back-and-forth
• The user can answer or say "unknown"
• Items where the answer is "unknown" are scored as "Not Met" with a note

Phase 3: Context Determination

Based on extracted information, determine the applicability of each item. Key rules:

See ../tte-foundation/references/rubric.md for the complete applicability mapping.

Phase 4: Scoring

Score each applicable item using the criteria in ../tte-foundation/references/rubric.md. For each item, record:

• Status: Met / Partially Met / Not Met / N/A
• Evidence: Brief quote or reference to where the information was found (or noting its absence)

Phase 5: Output

Generate a two-part report:

Part 1: Scorecard

Present a summary table:

## TTE Methodologic Quality Assessment

**Study:** [title/reference]
**Scenario Type:** [1/2/3]
**Treatment Strategy:** [point/sustained/dynamic]
**Comparator:** [active/blank/standard care]

### Domain A: Design Quality
| # | Item | Status | Evidence |
|---|---|---|---|
| A1 | Causal question defined | Met/Partial/Not Met | ... |
...

### Domain B: Implementation Quality
#### B1. Eligibility Criteria
| # | Item | Status | Evidence |
...
[repeat for B2-B7]

### Summary
| Domain | Met | Partially Met | Not Met | N/A | Total Applicable |
|---|---|---|---|---|---|
| A. Design | x | x | x | x | x |
| B. Implementation | x | x | x | x | x |
| **Overall** | x | x | x | x | x |

Part 2: Narrative Recommendations

For each item scored "Partially Met" or "Not Met", provide:

1. Finding: What was observed (or not observed)
2. Recommendation: What the Ren et al. framework recommends, citing the specific eAppendix 5 recommendation number
3. Action: A specific, actionable suggestion for improvement

Organize recommendations by priority:

• Critical gaps (items that directly threaten causal inference): eligibility using post-baseline info, no randomization emulation, undefined time zero, unmeasured confounding not addressed
• Important gaps (items that weaken but don't invalidate the study): missing design diagram, incomplete sensitivity analyses, undeclared causal contrast
• Reporting gaps (items that affect transparency): missing flowchart, unregistered protocol, no guideline followed

End with:

To design a new TTE protocol with these issues addressed, run /tte-design. To check data feasibility for implementation, run /tte-data.

Multi-Trial Studies

When a study emulates multiple target trials (common when testing different interventions or populations):

• Score each emulated trial separately
• Provide a per-trial scorecard
• Add a summary view highlighting differences across trials

Reference Files

The following files in ../tte-foundation/references/ contain the detailed criteria:

• rubric.md — Complete scoring rubric with met/partially met/not met criteria per item
• components.md — 7 core methodologic components with definitions and common pitfalls
• scenarios.md — 13 clinical scenario options mapped to 3 scenario types

Read rubric.md at the start of every assessment to ensure accurate scoring. Read scenarios.md when classifying the study's scenario type. Read components.md when writing narrative recommendations.