Research Terms: A Source-First Glossary

Records and access

The distinctions below follow NLM's descriptions of PubMed in About PubMed and the PubMed User Guide.

Abstract

A summary associated with a citation. It can omit methods, data, limitations, and later publication-history changes found elsewhere in the record or full article.

Citation

A bibliographic record identifying a publication through fields such as title, authors, journal, dates, and identifiers. A citation is not an evaluation of the publication.

Full text

The complete publication made available by a journal, PubMed Central, or another provider. PubMed may link to full text but does not itself contain every article's full text.

Study protocol

The written description of a study's objectives, design, and methods. A registry record summarizes protocol information and is not necessarily the complete protocol document.

Study design and record fields

These definitions follow the current ClinicalTrials.gov glossary and its guide to reading a study record.

Observational study

A study in which investigators assess health-related outcomes without assigning participants to a specific intervention.

Interventional study

A study in which participants are prospectively assigned to one or more interventions so specified outcomes can be evaluated. ClinicalTrials.gov also calls this a clinical trial.

Randomized study

An interventional study in which assignment to study groups uses a chance process. Randomization does not by itself describe masking, follow-up, reporting quality, or the appropriateness of the comparison.

Controlled study

A study with a comparison group or condition. The word controlled does not show whether assignment was random, masking was used, or the comparison was appropriate.

Masking

A design strategy in which one or more parties, such as participants or investigators, do not know which interventions were assigned. The record should identify who was masked.

Primary outcome

A planned measure designated as most important for evaluating the study's main question, including what is measured and the time frame.

Secondary outcome

A planned measure of additional interest that is identified separately from the primary outcome.

Enrollment

The number of participants who joined a study or the number researchers expect to enroll. A record may label the value actual or estimated.

Study status

The current recorded recruitment state, such as recruiting, completed, withdrawn, or unknown. Status can change and should be checked on the current record.

Sponsor

The organization or person that initiates a study and has authority or control over it. Sponsorship and funding are not necessarily the same relationship.

Investigator

A person responsible for conducting the study. The principal investigator is responsible for its scientific and technical direction.

Publication and evidence labels

PubMed treats article types as publication-type data, not quality scores. NLM's publication-type scope notes describe the labels, while its linked-citation policy explains status relationships.

Publication type

A label describing the material represented by a citation, such as a review, clinical trial, letter, or retracted publication. A citation can carry more than one type.

Systematic review

A review that uses explicit methods to identify, appraise, and synthesize evidence meeting stated criteria for a defined question. The label does not guarantee that those methods were well executed.

Meta-analysis

A quantitative method for combining results from independent studies. It is not interchangeable with systematic review, and the combined result depends on the included evidence and analytic choices.

Preprint

A complete, public draft of a scientific document that has not yet undergone peer review. The document may change or later appear as a journal publication.

Correction or erratum

A citable notice that corrects or adds information to an earlier publication. NLM groups corrections, corrigenda, addenda, and some partial retractions under errata.

Expression of concern

A notice indicating concern about the integrity of a publication while an investigation or follow-up remains relevant. It is linked to the affected citation.

Retraction notice

The citable notice stating that a publication is retracted or withdrawn in whole. In PubMed, this label belongs to the notice record.

Retracted publication

The publication affected by a retraction. In PubMed, this label belongs to the affected article, which is linked to its retraction notice.

Regulatory-record terms

ClinicalTrials.gov distinguishes interventions under study from regulatory conclusions. FDA's Drugs@FDA glossary defines application and label fields for specific products and records.

Investigational

A term indicating that an intervention is under study and may not be approved for the studied purpose. The word alone is not a complete determination of regulatory status.

FDA-approved label

FDA's official description for a specific approved drug product, including its approved uses and safety information. It does not describe every material with a similar name.

NDA

New Drug Application: a sponsor's submission asking FDA to review a new drug for marketing approval. A submission or application number is not, by itself, proof of approval.

BLA

Biologics License Application: a submission containing specified information for FDA review of a biological product. The current FDA record shows whether the application was approved.