Every monograph carries a Key studies table and an Evidence-by-claim table. Together they form a compact argument: what was studied, what was found, what limits the inference, and how the exact claim is graded.
This guide follows Evidence grading and Identity and structure assets. It is an editorial reading order, not guidance for personal decisions.
The Key studies table
One row represents one registry record, report, or analysis—not the entire evidence base. Read the fields in a fixed order before focusing on a large number or a significance marker.
| Field | Reader’s question | Overreading risk | Existing example |
|---|---|---|---|
| Design | Was the study randomized, controlled, blinded, A study without a separate comparison group. Source de la définition: Neutral gloss; usage context: Evidence grading methodology · Glossaire, or retrospective? | Treating every human observation as equally causal | Tirzepatide controlled trials versus the BPC-157 chart review |
| Population and comparator | Who was studied, and against what alternative? | Generalizing beyond eligibility or ignoring an active comparator | Tirzepatide’s program records the population and comparator for each row |
| Exact material | Which sequence, variant, form, and study product was tested? | Transferring results to a similarly named or marketed material | The identity method requires form-level scope |
| Endpoint | Was the measure a patient outcome, surrogate, biomarker, or subjective report? | Converting a surrogate or report into broad clinical benefit | Semaglutide rows name cardiovascular and weight endpoints |
| Result | What changed, in which comparison, and with what uncertainty? | Reading a number without its denominator or comparator | BPC-157’s uncontrolled reports cannot be read like a randomized comparison |
| Limitations | Which design features restrict the conclusion? | Treating caveats as optional detail | Missing controls, subjective follow-up, attrition, or narrow comparators can be decisive |
Design and population
Study design controls what can be inferred. The current tirzepatide monograph includes large prospective randomized trials, while the BPC-157 musculoskeletal row is based on a retrospective chart review with incomplete follow-up and no control group. A reported improvement does not erase that structural difference.
Population and comparator set the boundary. Results in one eligibility group do not automatically apply outside it, and an active-comparator result depends on which comparator was selected.
Exact material studied
The study row should identify the material and product context closely enough to know what the result covers. Findings do not automatically transfer across fragments, analogues, salts, complexes, mixtures, or An entry with frequent visibility in the research-peptide ecosystem. Visibility is not endorsement, and marketed material is not assumed safe, sterile, authentic, or suitable for humans. Source de la définition: Scope and selection methodology · Glossaire products with a familiar name.
Exposure details belong in the cited monograph or study record. This summary deliberately does not reproduce quantities; its task is to check identity, comparability, and whether key facts were reported.
Endpoints and result
An endpoint is what the study measured, not every benefit a reader might imagine. Statistical significance does not automatically establish clinical importance, and a biomarker change is not silently converted into a patient outcome.
Read the result with its comparator, denominator, uncertainty, and time frame. The BPC-157 chart review’s subjective follow-up cannot carry the same inference as a controlled trial, even if both contain prominent numbers.
Important limitations
Limitations belong beside the result. Lack of randomization, missing controls, selective follow-up, sponsor bias, multiplicity, short observation, or a nonrepresentative comparator can narrow or defeat the claim.
The table should make those constraints inspectable rather than leaving them buried in narrative prose.
- design
- population / comparator
- exact material
- endpoint
- result
- limitations
The Evidence-by-claim table
The claim table synthesizes across study rows. It names the indication or outcome, records development stage, assigns A-E or X, and points to the best evidence for that exact claim.
Claim specificity
The same molecule can have different grades. The current tirzepatide monograph assigns A to labeled glycemic-control and weight-management claims and B to its OSA-with-obesity row, while the status table separately records that use as FDA-approved. That combination is an audit prompt: grade and authorization should be checked as independent fields, not silently reconciled by the reader.
The semaglutide monograph similarly keeps its accelerated MASH claim at B while other labeled claims are A. BPC-157 carries C for preliminary human rows and D for Evidence from in vitro or animal studies with no adequate human efficacy evidence — the atlas Grade D lane. Source de la définition: Evidence grading methodology · Glossaire claims.
Grade boundaries
A requires current label scope plus adequate controlled evidence and post-market context. B records moderate human evidence or a material confirmation gap. C is preliminary human evidence. D is preclinical only. E is unsupported anecdotal or marketing material. X records adequate contradictory evidence on a separate lane.
These are claim boundaries, not molecule rankings. Follow the linked grading method rather than reconstructing a grade from one headline result.
Reading cross-table
The Key studies table supplies study facts; the Evidence-by-claim table supplies synthesis. Check that the cited study actually tests the named claim, that the endpoint supports the wording, and that limitations are reflected in the grade.
When the tables appear to disagree, do not average them. Trace the judgment through the current source rows and verification date. That tension is exactly where review is most valuable.
Four-question summary
Before accepting a conclusion, ask four questions: Was the design adequate for the exact claim? Did the population, comparator, and material match the conclusion? Was the endpoint clinically meaningful for that wording? Do the limitations and current claim grade support the same bounded inference?