Meta-Analysis Submission Rules Compared: Word Limits, Abstract Format, PRISMA Uploads and Registration at 20 Medical Journals

Each journal sets its own medical journal requirements, and they are not aligned. Across 559 biomedical journals from 11 publishers, endorsement of major reporting guidelines was inconsistent [1]. We read the instructions for authors at 20 journals on 14 October 2026 and recorded four items for systematic reviews and meta-analyses: the main-text word limit, the abstract format, how the PRISMA files must be uploaded, and the registration policy.

Fifteen journals’ pages could be read in full or in part. Five blocked automated access, so their rows say “check current instructions” rather than repeat unverified figures. “Not stated” means only that the item was not on the page reviewed.

Rules change, so open the live instructions for authors before you format. Where the BMJ suggests a length, CMAJ sets a hard cap.

• Fixed limits. CMAJ allows 3,500 words for systematic reviews; Scientific Reports caps the main text at 4,500.
• Guideline limits. The BMJ sets no fixed limit but suggests about 4,000 words.
• No limit. PLOS Medicine, PLOS ONE, Heliyon and several MDPI journals set no maximum; Frontiers in Medicine allows up to 12,000 words.

Count by each journal’s own definition. Scientific Reports leaves the Methods out of its count, while CMAJ excludes the title page, abstract, figures, tables, appendices and references.

Abstract rules vary more than word limits. Scientific Reports wants no more than 200 unstructured words; MDPI journals and CMAJ ask for about 250 words under fixed headings; The BMJ allows up to 400 for a PRISMA-style abstract with a registration heading; PLOS Medicine follows PRISMA for Abstracts.

The registration line is the item most often lost when an abstract moves between formats. In 62 systematic review abstracts appraised over 12 years, risk of bias, funding and registration were each reported in 1.6% or fewer [3]. Write the structured abstract to the target journal’s own headings.

PRISMA 2020 replaced the 2009 statement with a 27-item checklist in seven sections [2]. Journals differ in how they want compliance shown:

• Checklist with page references: PLOS Medicine and PeerJ.
• PRISMA flow diagram as Figure 1: PLOS ONE and PeerJ.
• File location: supplemental files at The BMJ, main text or supplement at MDPI, a multimedia appendix at JMIR.
• Statement only: CMAJ, Cancers and Frontiers in Medicine ask for PRISMA conformity without naming a format.

In 131 emergency medicine reviews, those declaring PRISMA 2020 met 72.8% of items against 68.4% for the rest; none met every item, and only 9 of 31 PRISMA 2020 users attached the checklist [4].

Registration is where policies diverge most, which makes it a deciding factor in journal selection. PeerJ requires prospective registration for health-related reviews and accepts PROSPERO, INPLASY or OSF registration. MDPI journals strongly encourage registration before data extraction; PLOS journals ask whether a protocol exists; several pages say nothing.

Among 4,750 evidence syntheses, PROSPERO was the most used registry, and adjusted time from registration to publication was 513 days for PROSPERO against 286 for INPLASY. One author founded INPLASY, which the study discloses [5]. Registration does not guarantee publication: 36.9% of 111 South American PROSPERO protocols were published [6]. Search the registry first, because about 13% of 1,054 COVID-19 registrations in PROSPERO were duplicates [7]. See our PROSPERO registration guide.

Treat these items as part of manuscript preparation, not formatting. For the five journals marked unverified, check the live guidelines before you submit.