Build a systematic review protocol

TASK
Turn the research question and methodological constraints below into a reproducible systematic-review protocol. Do not invent studies, results, database coverage or search yields.

INPUTS
- Research question: {{research_question}}
- Population/problem: {{population}}
- Intervention/exposure: {{intervention_or_exposure}}
- Comparator, if relevant: {{comparator}}
- Outcomes: {{outcomes}}
- Date range: {{date_range}}
- Databases/sources to search: {{databases}}
- Languages: {{languages}}
- Eligible study designs: {{study_designs}}
- Known inclusion/exclusion constraints: {{eligibility_constraints}}
- Review objective: {{review_objective}}

METHOD RULES
1. If the question does not fit PICO, choose a more appropriate explicit framework (for example PECO, SPIDER or a concept-based structure) and state why in one sentence.
2. Separate concepts from synonyms/search terms. Do not claim a term is a controlled-vocabulary heading unless that status is known from the supplied information or can actually be checked.
3. Draft database-specific search strings only where syntax is known. Otherwise provide a platform-neutral Boolean strategy and label database adaptation as required.
4. Keep screening criteria operational: a second reviewer should be able to apply them consistently.
5. Predefine reasons for exclusion at full-text stage. Do not create exclusion reasons after seeing results.
6. Define an extraction schema before synthesis. Separate reported data from reviewer inference.
7. Do not fabricate a registration number, PRISMA counts, risk-of-bias judgments or certainty rating.
8. If essential scope information is missing or contradictory, list the decisions required before finalizing the protocol.

OUTPUT
A. Review objective and review question
B. Question framework and concept map
C. Eligibility criteria: inclusion and exclusion
D. Information sources and coverage limits
E. Search strategy: concepts, synonyms and draft Boolean logic; database-specific adaptations only when supportable
F. Deduplication and screening workflow
G. Full-text exclusion-reason taxonomy
H. Data-extraction schema
I. Risk-of-bias/quality-assessment plan appropriate to the included study designs
J. Synthesis plan, including when quantitative pooling would not be justified
K. Reproducibility log fields: search date, source, exact query, filters, result count, deduplication count and reviewer decisions
L. Decisions still required before protocol lock

Return a protocol ready for human methodological review, not fictional review results.