# Protocol drafting prompt

Use the attached RWEdnesdays protocol kit to help me build a research protocol. First identify the appropriate design template from my actual goal and whether treatment is assigned or observed. Read the selected template, instructions, study brief, available data dictionary, and supplied guidance. Treat the study brief as the decision context, not proof that every proposed method is valid.

My decision and audience: [complete]
My research question and objectives: [complete]
Population, setting, treatment/exposure and comparator if applicable: [complete]
Outcomes and follow-up: [complete]
Available data and known limitations: [complete]
Guidance or jurisdiction requirements: [complete]
Decisions already made and their owners: [complete]
Evidence cutoff and results already viewed: [complete]

Ask for the few missing answers that materially change the design. Explain consequential alternatives and obtain my decision rather than silently choosing. Draft sections that do not depend on those answers. Label every unknown, assumption, proposed choice, and synthetic illustration. Use primary sources for new methodological claims and verify metadata and claim support; list inaccessible sources as unverified.

Produce a versioned protocol draft, an open-decision log, an item-level guidance crosswalk, a claim/source ledger, and an objective-to-analysis-to-output map for the SAP. Distinguish protocol guidance from reporting checklists. Identify where the data cannot support the intended question. Do not invent input fields, parameters, citations, feasibility results, or approval. Do not execute analyses or submit/publish the protocol. Use only authorized non-sensitive inputs.