How to Use the Research Library: What Has Been Studied, What Was Used Traditionally and How Strong Is the Evidence

The short answer

The Research Library is a place to look things up, not a place that tells you what to take. It brings together records about herbs, ingredients, formulas, conditions, genes and study records, and it shows how they connect. It does not rate whether anything works. To use it well, start with a question, pick the collection that fits, and keep three ideas apart as you read: what has been studied, what was used traditionally, and how strong the evidence is.

What the library is

The library holds imported records: about 6,800 herbs, 44,600 ingredients, 6,700 formulas, 30,500 condition and syndrome terms, 19,900 symptom (phenotype) terms, 15,700 gene records and 25 biomarker records. All of them were imported from public databases and ontologies, and none has been reviewed by a clinician. Most records have names, identifiers and linked study records, but no written summary.

That is useful for lookup and for seeing links between topics. It is not the same as a written, checked profile. For how to read the labels on a record, see How to Read Research Library Records.

Five questions to start with

Why the three ideas must stay apart

Health information often blends these together, which is how a claim can sound more settled than it is. The National Center for Complementary and Integrative Health (NCCIH) warns that there is a lot of misinformation, from anecdotes disguised as evidence to excessive claims by supplement makers. [1]

• Studied means researchers looked at it. The result may be positive, negative or mixed, and the study may involve cells, animals or people.
• Used traditionally means people have used it over time, in a specific culture or tradition. That is a historical record.
• Strong evidence means well-designed human studies, ideally pooled in a systematic review, point in a consistent direction. A systematic review tries to find and weigh all the research that meets set rules, to reduce bias. [2]

The amount of evidence varies a great deal: NCCIH notes there is a lot of information on some supplements and very little on others. [3]

Finding studies in the library

Search supports exact lookup by DOI, PubMed ID, NCT number and CRD number in the evidence collection. You can also filter by record type, research model (human, animal or cell), completed versus not completed, and subject type. The research model filter matters: laboratory and animal work comes before human testing, and FDA describes preclinical research as the step that checks for potential harm before a drug is tested in people. [4] Animal and cell research is not human evidence.

Of the library's 8,558 registered clinical trials, 6,881 have the status Completed. A registration describes planned or ongoing work unless it is completed, and it carries no result in the library. The 6,705 publication records include PubMed IDs and abstracts, and PubMed itself holds citations and abstracts rather than full articles. [5]

Exploring by collection

Use Herbs, Ingredients and Formulas to start from a substance or preparation. Use Symptoms and Conditions and Genes and Biomarkers to start from a health topic or a biological marker. Body Systems groups records by part of the body. The Remedies page explains how the word remedy is used.

Common misunderstandings

• More links means better studied. It does not. A record with many connections is not better studied or more effective.
• A record on this site was checked. Records are imported by default and are not clinician-reviewed.
• A trial in the list means the treatment worked. A registered trial is a plan until results exist.
• Traditional use means safe. Natural does not necessarily mean safe; MedlinePlus notes that some herbs can cause serious harm. [6]

Related pages

Continue with Evidence Types and Study Designs or the Research Library Glossary.