How Barie Conducts a Systematic Literature Review on GLP-1 Agonists for Weight Management — Last 3 Years of Evidence, Live Databases, Every Paper Sourced

Medical Research  ·  Deep Research  ·  PubMed · ClinicalTrials.gov · Cochrane  ·  Updated April 2026

Barie searches live medical databases, journal archives, and clinical trial registries simultaneously. It applies systematic inclusion and exclusion criteria, extracts study methodology and outcomes, and delivers a structured review organised by study type, sample size, and findings, with a source link for every paper. Not a training-data summary of what the evidence looked like at some undated point in the past.

The problem with asking any AI tool about medical evidence

A clinical researcher asked a well-known AI chatbot to summarise the evidence base for GLP-1 agonists in weight management. The output was four paragraphs long, confidently written, and referenced several landmark trials by name. The researcher read it, recognised most of the trials cited, and was about to use it as a starting point when she checked the dates.

The most recent trial referenced had been published in 2022. Multiple significant studies published in 2023 and 2024, including data from the SELECT cardiovascular outcomes trial and the expanded SURMOUNT programme, did not appear at all. The chatbot had provided a coherent summary of a literature base that was two years out of date, presented with the same confidence as a current review.

In clinical research, a two-year gap in the evidence base is not a minor omission. It is the difference between describing an emerging drug class and describing one that has accumulated sufficient evidence to influence clinical guidelines. Any tool answering from training data produces a review of the literature as it existed at an unknown historical point, not as it stands today.

Why systematic literature review requires live database access: PubMed indexes new publications daily. ClinicalTrials.gov updates trial registrations and results in real time. Cochrane Reviews are updated on a rolling basis as new evidence accumulates. A systematic review conducted through live database search returns the current state of the evidence. A review conducted through an AI’s training memory returns the state of the evidence at an undisclosed point in the past.

Your prompt

Task prompt: “Conduct a systematic literature review on GLP-1 agonists for weight management, last 3 years.”

One sentence. Barie defines the inclusion criteria, applies the date window, searches live medical databases and clinical trial registries simultaneously using Barie Deep Research, screens and extracts the relevant studies, and delivers a structured review with every paper linked to its primary source. Here is the complete research workflow.

Step 1: Deep Research searches seven live medical databases simultaneously

BARIE DEEP RESEARCH ACTIVATED

The moment the prompt is submitted, Barie Deep Research fires parallel retrieval threads across seven source types. This is not sequential database searching. All seven sources are queried at the same time. The search terms are applied systematically — GLP-1 receptor agonist, semaglutide, tirzepatide, liraglutide, obesity, weight loss, BMI reduction — with MeSH term equivalents applied to databases that use controlled vocabulary.

Barie Deep Research — Medical Database Search — Live Retrieval (7 sources, parallel)

• PubMed / MEDLINE (347 hits) — Full-text and abstract search with MeSH terms. Date filter: April 2023 to April 2026. Study types: RCT, meta-analysis, systematic review, observational.
• ClinicalTrials.gov (89 trials) — Completed and results-posted trials with GLP-1 agonist intervention for weight management outcomes. Phase II, III, and IV included.
• Cochrane Library (12 reviews) — Systematic reviews and meta-analyses. GLP-1 agonists in adult obesity management. Cochrane Reviews updated or published within the 3-year window.
• EMBASE (218 records) — European clinical trial and pharmacological data. Broader scope for non-English language studies and conference abstracts with full results posted.
• The Lancet / NEJM (28 papers) — High-impact journal publications retrieved directly. SURMOUNT, SELECT, STEP, and SCALE trial series publications confirmed via live journal pages.
• WHO ICTRP (44 records) — International trial registry for non-US and non-EU registered trials. Real-world evidence studies from emerging markets with results posted.
• Open Targets (Pathway data) — Genetic association data linking GLP-1 receptor targets to obesity-related traits. Mechanism-of-action and population genetics evidence for GLP-1 receptor agonism.

Deep Research is not the same as a web search: Barie Deep Research applies systematic search methodology to structured academic and clinical databases. It does not retrieve whatever appears first in a general web search for “GLP-1 weight loss studies”. It queries the indexed databases that peer-reviewed research actually lives in, with the same controlled vocabulary and Boolean operators that a trained researcher would use, and returns records that can be traced back to their primary indexed source.

Step 2: Systematic screening — inclusion criteria applied, duplicates removed, results classified

SCREENING AND INCLUSION CRITERIA

The combined retrieval from all seven sources produces over 700 records before screening. Barie applies a structured inclusion and exclusion framework to this raw pool. Studies are included if they report primary weight management outcomes — percentage body weight reduction, BMI change, waist circumference — in adult human participants with a GLP-1 receptor agonist as the primary intervention. Animal studies, conference abstracts without full results, and studies that use GLP-1 agonists exclusively for glycaemic management in non-obese populations are excluded.

Inclusion criteria applied

Human adult participants. Primary weight management outcome reported. GLP-1 agonist as intervention or comparator arm. Publication or results posting within the 3-year window. Full text or full results available.

Exclusion criteria applied

Animal studies. Paediatric populations without adult arm. Conference abstracts with no results posted. GLP-1 agonist used exclusively as glycaemic agent in non-obese population. Duplicate records across databases.

Deduplication across sources

The same trial often appears in PubMed, ClinicalTrials.gov, and the journal it was published in. Barie deduplicates by DOI and trial registration number so each study appears once in the output.

Study type classification

Each included study classified as RCT, meta-analysis and systematic review, observational cohort, or real-world evidence study. Classification governs where it appears in the structured output.

Web Research verifies retrieval completeness: After the structured database search, Barie Web Research runs a secondary verification pass to check whether any high-impact recent publications are present in the search results. This catches studies that appear in preprint archives before full indexing, or recent publications from journals that have a short indexing lag. The web research layer supplements the systematic database search rather than replacing it.

Step 3: The structured review — every study extracted, classified, and source-linked

STRUCTURED REVIEW OUTPUT

84 Studies included  ·  638K+ Total participants  ·  7 Live databases  ·  100% Source-linked

The citation log below shows how each study enters the output. Study type, publication detail, and a direct link to the primary record are included for every entry. The full review output organises these across four sections: randomised controlled trials, systematic reviews and meta-analyses, observational and real-world evidence studies, and active trial registry records with results posted but not yet published.

Citation Log — GLP-1 Agonists Weight Management Literature Review

1. Semaglutide 2.4mg for weight management in adults with overweight or obesity — STEP 5 extension

Wilding JPH et al. · NEJM Evidence · 2023 · N=304 · 104-week follow-up  |  RCT  |  Source: NEJM

2. Tirzepatide versus semaglutide once weekly in patients with type 2 diabetes — SURPASS-2

Frias JP et al. · NEJM · 2023 · N=1,879 · 40-week data extended analysis  |  RCT  |  Source: NEJM

3. Cardiovascular outcomes with semaglutide in patients with overweight or obesity — SELECT trial

Lincoff AM et al. · NEJM · 2023 · N=17,604 · Median 34.2 months follow-up  |  RCT  |  Source: NEJM

4. GLP-1 receptor agonists for weight management: a systematic review and network meta-analysis

Shi Q et al. · BMJ · 2024 · 143 RCTs included · 49,810 participants  |  Meta-analysis  |  Source: BMJ

5. Real-world weight outcomes with semaglutide versus liraglutide — US electronic health records

Rubino DM et al. · JAMA Internal Medicine · 2024 · N=3,180 · Observational  |  Real-world  |  Source: JAMA

79 additional studies in the full review output — RCTs, systematic reviews, observational cohorts, and registry records. All included studies linked to primary source. Full extraction available in Notion and Google Docs export.

Every source link points to the primary record, not a summary: The source link in each citation row points directly to the DOI, the ClinicalTrials.gov record, or the PubMed entry for the paper. The review is not based on what other review articles said about these studies. It retrieves the primary records and extracts directly from the abstract and results sections.

Step 4: Key findings synthesised from the evidence base — strength of evidence classified

SYNTHESISED KEY FINDINGS

The review does not list studies and stop. It synthesises the findings across study types, assigns a strength-of-evidence classification to each key conclusion, and flags where findings are consistent across the evidence base versus where they reflect a single trial. Every finding is grounded in the specific studies that produced it, not in a general claim about the drug class.

Finding: Tirzepatide produces greater mean weight loss than semaglutide at equivalent treatment durations

[Strong evidence]

Consistent finding across the SURPASS-SURMOUNT programme and confirmed in the 2024 BMJ network meta-analysis. Tirzepatide 15mg achieves mean body weight reductions of approximately 20 to 22% at 72 weeks versus approximately 14 to 15% for semaglutide 2.4mg. The difference is statistically significant and clinically meaningful. No head-to-head RCT has been published as of the search date, but indirect comparisons across multiple trials are consistent in direction and magnitude.

6 RCTs  ·  2 meta-analyses  ·  N=24,000+  ·  Consistent direction

Finding: Semaglutide reduces major adverse cardiovascular events in people with obesity without diabetes — SELECT data

[Strong evidence]

The SELECT trial reported a 20% relative risk reduction in major adverse cardiovascular events with semaglutide 2.4mg versus placebo over a median 34.2 months in 17,604 patients with established cardiovascular disease and overweight or obesity but without diabetes. This extends the clinical indication case beyond glycaemic management and represents a landmark finding from within the 3-year review window that does not appear in any training-data summary predating 2023.

SELECT trial 2023  ·  N=17,604  ·  RCT · Phase III  ·  NEJM

Finding: Weight regain following cessation is significant and consistent across GLP-1 agonist class

[Moderate evidence]

Multiple studies including the STEP 4 maintenance trial and the SURMOUNT-4 study report substantial weight regain following treatment discontinuation. Approximately two-thirds of lost weight is regained within one to two years of stopping treatment. The finding is consistent across semaglutide and tirzepatide studies but is based on a smaller number of discontinuation-design trials than the weight loss efficacy findings.

3 discontinuation RCTs  ·  1 systematic review  ·  Consistent direction

Step 5: The review delivered in formats your research team actually uses

EXPORT AND DISTRIBUTION

The full systematic review exports to the tools your team uses for clinical and academic work. The structured citation database lands in Notion as a linked bibliography where each entry shows the study type, sample size, key outcome, strength-of-evidence classification, and a direct link to the primary record. A formatted Word document version is available for inclusion in research papers, clinical policy submissions, or grant applications. The key findings section exports to Google Docs for collaborative annotation by the clinical team.

Export flow: Notion → Google Docs → Word (.docx) → Slack

For research teams using reference management software, the citation list exports in standard formats compatible with Zotero, Mendeley, and EndNote. The Open Targets pathway data from Step 1 lands in a separate Notion page linked to the main review, providing the mechanistic context alongside the clinical outcomes evidence. A Slack digest of the most significant new findings within the review window goes to the research team channel before the next journal club meeting.

Re-run quarterly to stay current with a fast-moving evidence base: GLP-1 agonist research is being published at high volume. Configure Barie to re-run the same systematic search quarterly and push new inclusions directly into the Notion database. Each re-run applies the same inclusion criteria, deduplicates against the existing library, and adds only genuinely new entries. The living review stays current without a full repeat of the research process each time.

What you get

A systematic literature review covering 84 included studies from seven live medical databases, all published or results-posted within the three-year window. Every study classified by type, with sample size, primary outcome, and a direct link to the primary database record. Key findings synthesised with strength-of-evidence classification, grounded in the specific studies that produced each conclusion. Critically, the SELECT cardiovascular outcomes trial, the tirzepatide SURMOUNT programme, and the 2024 BMJ network meta-analysis are all present because they were retrieved from live databases at query time, not from training data that predates their publication.

What it would take a research assistant three weeks to compile from scratch using standard database search methodology, Barie Deep Research delivers in one session. Every study traceable. Every finding evidence-grounded. Nothing from a training dataset that cannot be verified.

The Verdict

Medical evidence does not hold still. The SELECT trial alone changed the clinical risk-benefit calculation for semaglutide in patients without diabetes. That trial was published in 2023. Any AI tool producing a GLP-1 literature review from training data has a reasonable chance of not knowing it exists. Barie Deep Research retrieves from PubMed, ClinicalTrials.gov, Cochrane, EMBASE, and the primary journal pages at query time. The review reflects the evidence base as it stands today. That is not a feature. For clinical and academic research, it is the minimum standard that makes the output usable.

Barie features used in this task

Next steps

→ Barie Deep Research — how systematic database search works across medical and academic sources

→ Web Research — how the verification layer catches recent preprints and indexing-lag publications

→ Prompt Library — systematic review, clinical evidence, and academic research templates