How Rinnit Research Studies Are Done
Written and medically reviewed by Dr. Baraa Alnahhal, MD · Last reviewed: November 2026 · How we review
Quick Answer: What is Rinnit Research?
Rinnit Research is the original-studies programme of rinnit.com. Each study follows a written analysis plan published before the data is pulled, has every number calculated in code and recalculated independently, is reviewed and interpreted by a physician, and is released with its data and code under a DOI. The studies are not peer-reviewed.
What kinds of studies does Rinnit Research publish?
Rinnit Research publishes analyses of open public data, and systematic reviews are planned as its second study type. The first study used national estimates published by The DHS Program (Demographic and Health Surveys). Rinnit Research works only with published, open sources. It does not run patient surveys and does not use hospital records.
How is a Rinnit Research study planned?
Every Rinnit Research study starts with a written analysis plan that is published before the data is pulled. The plan states the question, the data source, which records are included, the calculations and the limits. The physician author approves the plan, and it is published on Zenodo with its own DOI, which gives it a public date. Any later change from the plan is listed in the published study.
The plan for study 1 is here: https://doi.org/10.5281/zenodo.24136796
How are the numbers calculated and checked?
Every number in a Rinnit Research study is calculated in code and checked twice before publication. The steps are:
A script pulls the data from the source and records the date of the pull. The raw extract is kept and published with the study.
Claude (Anthropic) runs the analysis code that produces the result tables.
A second AI system, Gemini (Google), recalculates the main tables independently from the same data. The two sets of tables must match before the study goes further.
A script compares every number in the article's text and tables with the result files, so no number is typed by hand.
In study 1, the independent recalculation exposed a rounding inconsistency. It was corrected before publication and changed four summary figures by 0.1.
Who writes and reviews each study?
Each Rinnit Research study is reviewed and interpreted by its author, Dr. Baraa Alnahhal, MD, a practising physician. He approves the plan, the results and the text, and he writes the clinical interpretation himself.
AI systems assist, and each study states their role. Claude runs the analysis code and verifies the draft. Gemini recalculates the tables and writes the first draft from the verified tables only. When the two systems disagree on a claim, the author decides from a written list of the disputed points.
Where are the data and code published?
Each Rinnit Research study is published in two places: as an article on rinnit.com and as a record on Zenodo with a DOI. The Zenodo record holds the full report as a PDF and an archive with the raw data extract, the result tables, the charts and all analysis code, so anyone can rerun the analysis.
The text, charts, code and result tables are released under the Creative Commons Attribution 4.0 licence (CC BY 4.0). Raw data from an outside provider stays under that provider's own terms.
What are the limits of Rinnit Research studies?
Rinnit Research studies are original analyses and are not peer-reviewed, and each one carries the label "Rinnit original analysis, not peer-reviewed". Studies based on published public data are descriptive. They show what the published figures are and cannot show why a result exists or what it causes. National survey estimates also carry sampling error, so small differences between countries are not treated as real differences. In study 1 the threshold was 5 percentage points.
How are corrections handled?
A confirmed error in a Rinnit Research study is corrected in the article with a dated note, and a new version of the record is published on Zenodo. Earlier versions stay available on Zenodo, so readers can see what changed. To report an error, email support@rinnit.com with the article link and the issue, as described in the Editorial Policy.
How should a Rinnit Research study be cited?
A Rinnit Research study is cited by its Zenodo record and DOI. Each article gives its own citation line. Study 1 is cited as: Alnahhal B. The blood-pressure gap in antenatal care: a multi-country analysis of DHS surveys (Rinnit Research study 1). Version 1.0. Rinnit; 2026. https://doi.org/10.5281/zenodo.24142462
Which studies has Rinnit Research published?
Rinnit Research has published one study so far. New studies are added to this table when they are published.
Study | Topic | Published | Article | Data and code |
Study 1 | The blood-pressure gap in antenatal care, 72 countries | November 2026 |
FAQ
Is Rinnit Research peer-reviewed?
No. Rinnit Research studies are original analyses that have not been peer-reviewed, and every study says so. The plan, data and code are published so that readers can check the work themselves.
Does Rinnit Research use AI?
Yes. Claude (Anthropic) runs the analysis code and verifies the draft, and Gemini (Google) recalculates the tables independently and writes the first draft. A physician approves the plan, the results and the text, and each study states the role AI played.
Who is responsible for a Rinnit Research study?
The author, Dr. Baraa Alnahhal, MD, is responsible for each Rinnit Research study. He approves the analysis plan, the results and the final text, and he writes the clinical interpretation himself.
Why is the analysis plan published before the data is pulled?
A plan published first fixes the question and the calculations before any result is known. Its DOI gives it a public date, so readers can compare the plan with the published study.
Can I reuse Rinnit Research charts and results?
Yes. The text, charts, code and result tables are released under CC BY 4.0, so they can be reused with credit to the study and its DOI. Raw data from an outside provider stays under that provider's own terms.
Does Rinnit Research collect patient data?
No. Rinnit Research uses published, open data only. It does not run patient surveys and does not use hospital records.
How can I report an error in a Rinnit Research study?
Email support@rinnit.com with the article link and the issue. Confirmed errors are corrected in the article with a dated note, and a new version of the record is published on Zenodo.
Medical disclaimer
Rinnit Research studies and this page are for information and education. They are not medical advice and do not replace a consultation with a qualified health professional.
Comments