# Build a SAP Handoff Prompt

Use this only with approved source material and a human methods lead. Do not provide patient-level or client-confidential data.

```text
Create a protocol-to-SAP handoff using assets/protocol-to-sap-handoff.md.

Inputs:
- Protocol and version: [attach or summarize]
- Study question(s): [enter]
- Available data/source documentation: [enter]
- Existing analysis constraints: [enter]
- Applicable modules: [claims/EHR, economic, NMA/ITC, RCT, none known]

Instructions:
1. Create one TRACE-ID for each primary and key secondary question.
2. Preserve the protocol’s stated choices; do not invent a population, estimand,
   estimator, effect scale, threshold, grace period, missing-data rule, or code list.
3. For each question, separately state the causal/clinical target, observed data
   inputs, and proposed estimator. Mark absent information as Unresolved.
4. Populate SAP-001 through SAP-008 and link each to input data, algorithm,
   diagnostic, output, and deviation handling.
5. Route economic and cross-study NMA/ITC work to the actual companion plans.
   Request an investigator-supplied trial SAP for randomized work; the supplied
   trial protocol is not an SAP. Do not imply missing attachments already exist.
6. Produce: (a) completed handoff, (b) unresolved-decision register with owner,
   approver, and decision deadline, and (c) implementation handoff for programming.
7. Label all examples as synthetic unless supplied study facts support them.
```