Loon Lens™ AI Platform Selected by ICER to Conduct Systematic Literature Reviews Supporting Value Assessment

The Institute for Clinical and Economic Review (ICER) has selected the Loon Lens™ platform to provide:

• AI-powered systematic literature reviews
• Evidence discovery and synthesis
• Data analytics

The purpose is to conduct its value assessments rigorously and transparently. ICER’s value assessments inform coverage, pricing, and benefit-design decisions across the United States, and each one rests on a systematic review of the clinical and economic evidence conducted in-house by the ICER team.

ICER selected the Loon Lens™ platform after an extensive evaluation spanning multiple areas:

• Calibrated confidence: a calibrated confidence score on every AI decision.
• Evidence discovery: calibrated search strategies with measured recall and precision.
• Seed overview and feasibility assessment: a seed report on likely includes, a Bayesian estimate of review size, and estimated sensitivity and specificity for each search strategy.
• Intelligent records management and enrichment: automatic deduplication, enrichment of citations, quality checks, and study classification.
• Screening: deliberative screening at the title/abstract and full-text stages.
• One-shot data schema generation: a tailored, editable data dictionary built for the dataset, with no fixed templates or manual setup.
• Extraction: every datapoint extracted without restrictions, together with its source; uncapped columns, studies or tables.
• PRISMA flowchart: all your changes update the diagram automatically your project is always up-to-date.
• Analysis: analysis of the resulting evidence.

ICER uses the platform with a hybrid AI–human approach.

Loon AI®, the neurosymbolic engine powering the Loon Lens™ platform, is validated in seven studies, including:

• Peer-reviewed: a paper in Value in Health (Janoudi et al., 2025).
• Independent benchmark: the BC Neonatal Cochrane group found 100% data extraction truthfulness, 99.6% data schema coverage, and 98% cumulative time savings compared with traditional reviews.

The full publication record is at loonbio.com/publications.

The Loon Lens™ platform is an end-to-end platform for systematic literature reviews, from calibrated search all the way to data extraction and feasibility analysis:

1. Protocol: eligibility criteria assessment and PICOS optimisation.
2. Search: calibrated search strategies with measured recall and precision.
3. Evidence discovery: including full-text retrieval.
4. Feasibility: an Overview Report for project scoping and feasibility assessment, including EU Joint Clinical Assessment, with an estimate of review size and the likely includes.
5. Records: deduplication and enrichment.
6. Screening: title/abstract and full-text.
7. Study mapping: an overview of the included evidence.
8. Extraction: comprehensive data extraction using schemas tailored to the dataset, with no fixed templates or manual setup.

The PRISMA 2020 flowchart and audit log update automatically at every step.

Every screening decision and every extracted datapoint carries:

• Reasoning that shows how the AI decision was made.
• A calibrated confidence score that shows which decisions deserve further attention.
• A traceable source for every decision and datapoint.

Human review is configurable, and PRISMA flowcharts and a detailed audit log update automatically as the review progresses.

Teams preparing evidence for scrutiny by HTA bodies and regulators can now work with the same platform contracted for ICER’s internal evidence synthesis, supported by transparent methods and configurable human oversight.

Any type of review:

• Review types: systematic, living (LSR®), rapid, scoping, targeted, exploratory, and umbrella reviews.
• Evidence types: clinical efficacy and safety, disease burden, economic burden, utilities and quality of life, cost-effectiveness, resource use and costs, real-world evidence, diagnostic accuracy, prognostic factors, and guidelines and HTA policy.