Synthesize a stack of clinical papers into a structured internal brief a clinician still has to sign off on

You are producing an internal evidence brief for a clinician or clinical researcher who has gathered a set of papers on one question and needs them synthesized into a structured working document — not a final answer to act on directly.

CLINICAL QUESTION
Does adding structured sleep hygiene counseling improve outcomes for adult patients on first-line treatment for mild insomnia?

PAPERS OR EXCERPTS PROVIDED
Three abstracts pasted in: a 2019 RCT (n=120) showing modest improvement, a 2021 observational study (n=450) showing no significant difference, and a small pilot (n=24) industry-funded study showing large improvement.

INTENDED USE OF THIS BRIEF
Internal discussion document for a clinic considering whether to add a counseling protocol, not a patient-facing document.

AUDIENCE
A small group of primary care physicians and a clinic administrator, all with a research-literacy background.

PROCESS
For each paper provided, extract only what is actually stated in the excerpt: study design, population/sample size if given, the reported direction and magnitude of effect, and any limitation the authors themselves note — never infer a finding that isn't explicitly present in the text given to you. Group papers by whether their findings point the same direction, point in opposing directions, or are too methodologically different to compare directly (e.g. different populations, different outcome measures), and say so plainly rather than averaging conflicting findings into a single smoothed-over conclusion. Write a short synthesis paragraph per question that describes the state of the evidence as provided (strong and consistent, mixed, thin, contradictory) without translating that into a clinical recommendation — a recommendation is a decision for the clinician using this brief, informed by patient-specific factors this brief has no access to. Flag explicitly anywhere the provided papers are old, small, industry-funded (if stated), or otherwise limited in a way that should temper how much weight the synthesis paragraph gives them.

WHAT NOT TO DO
Do not search your own general knowledge to fill in gaps the provided papers don't cover, do not cite a study, statistic, or guideline that was not given to you in the input, and do not phrase the synthesis as though it settles the clinical question — it organizes the evidence someone already collected, nothing more.

MANDATORY DISCLAIMER
Open the brief with a note that this is an evidence-organization aid built only from the material provided, requires independent verification of every source against the original publication, and must be reviewed and interpreted by a qualified clinician or researcher before informing any patient care decision.

OUTPUT FORMAT
1. The disclaimer.
2. A per-paper table: citation as given, design, population, reported effect, noted limitation.
3. A grouping section: agreement / disagreement / not comparable.
4. A synthesis paragraph per question, ending in a plain statement of what remains uncertain.