AI-Native Scoping Reviews & Evidence Mapping

The aim of this paper is to systematically review and synthesize the available evidence on the health effects, both beneficial and adverse, of sauna bathing in human populations, identifying key findings, clinical implications, and future research directions. The review utilises 519 original studies with 541510 total participants (topic deduplicated ΣN). Across the mapped evidence, frequent sauna bathing shows a consistent signal of lower major cardiometabolic and neurological risk, including associations such as a 37% lower risk of sudden cardiac death with 4–7 sessions/week versus 1 session/…

E-cigarettes may reduce selected harms when they fully replace combustible tobacco, but this evidence map shows why they cannot be treated as harmless nicotine products. The full SAIMSARA evidence map gives human- and machine-readable access to 123 references, clarifying substitution benefits, dual-use risks, youth uptake, toxicology, vascular and airway effects, pregnancy signals, and regulatory implications.

This review shows where AI is already delivering real value in drug discovery: not just faster predictions, but experimentally validated hits, better ADMET screening, large-scale virtual screening, and even early human translation. It maps which AI advances are truly credible, which claims remain fragile, and where the field is genuinely moving from hype to practical therapeutic impact.

Superior vena cava syndrome is no longer only an oncologic emergency: this evidence map shows how malignancy, catheters, pacemaker leads, dialysis access, thrombosis, pediatric mediastinal tumors, and procedural risks converge in one clinically complex syndrome. The full SAIMSARA evidence map gives human- and machine-readable access to 1,185 original studies, clarifying when to prioritize tissue diagnosis, endovascular stenting, oncologic therapy, airway planning, or benign-disease venous reconstruction.

Prostitution is not a single behavior but a public-health risk context where trauma, violence, substance use, infectious disease, stigma, and legal environment repeatedly intersect across 633 original studies and 3.76M topic-deduplicated participants. The full SAIMSARA evidence map gives both human-readable synthesis and machine-readable JSON for LLMs, tracing where harm concentrates, which interventions show measurable benefit, and where policy, clinical care, and exit-support models still need stronger evidence.

Hantavirus infection is not just a rare rodent-borne fever: this evidence map shows it as a multisystem vascular infection driven by ecology, endothelial dysfunction, thrombocytopenia, and renal or cardiopulmonary injury. The full SAIMSARA paper and machine-readable JSON provide structured, reference-linked evidence across 1,084 original studies, covering diagnosis, exposure risk, Andes-virus transmission, severity markers, One Health surveillance, and the still-unproven therapeutic frontier.

Renal artery stenosis is not a simple “stent or no stent” disease: this map separates broad neutral randomized evidence from the high-risk phenotypes where revascularization may still matter. The full SAIMSARA evidence map gives clinicians and AI agents a structured, reference-linked view of prevalence, imaging accuracy, medical therapy, physiology-guided selection, transplant/pediatric RAS, and remaining gaps.

Stem cell therapy for cardiovascular disease is biologically compelling but clinically heterogeneous: this map separates the strongest human signals from protocol-only, preclinical, safety, and mechanistic evidence. The full SAIMSARA evidence map gives humans and AI agents a structured, reference-linked view of LVEF signals, refractory-angina outcomes, cell retention barriers, HSCT cardiotoxicity, exosome platforms, and the translational gaps still blocking routine clinical use.