Most medical students discover that their research idea has already been done about six weeks into the project. They have read 40 papers, drafted a protocol, and emailed three potential supervisors when someone casually points out the 2023 meta-analysis that answers the exact question. The next six weeks go to figuring out a new angle, restarting the protocol, and rebuilding momentum that was already thin. This is the most common kind of research waste and it is almost always preventable with a 20-minute pre-flight check.

Here is the check.

Step 1: Search PubMed for systematic reviews and meta-analyses (5 minutes)

Open PubMed. Type your concept terms with the publication type filter for systematic reviews and meta-analyses. The fastest version:

(your concept terms) AND ("systematic review"[pt] OR "meta-analysis"[pt])

If you get a recent (within 2 years), highly cited SR or MA that answers your exact question with positive findings, your project is dead on arrival unless you can clearly articulate a different angle — a new population, an updated time window, a head-to-head comparison the existing review skipped. If you get older reviews only, you may have a temporal-gap opportunity (see our piece on identifying research gaps).

Step 2: Search the Cochrane Library (3 minutes)

Cochrane reviews are the gold standard. If your topic has a Cochrane review from the last 3 to 5 years with a "no conclusive evidence" verdict, you might still have an opportunity. If the Cochrane review concludes the evidence is sufficient, you almost certainly do not.

Search at cochranelibrary.com directly, or in PubMed via ("Cochrane Database Syst Rev"[Journal]). Cochrane reviews are also indexed in Europe PMC and PubMed.

Step 3: Check PROSPERO for in-progress reviews (5 minutes)

PROSPERO is the international prospective register of systematic reviews. People register protocols before they publish, so PROSPERO is where you find out that three teams are already six months into a review on your exact question. If you find one, you have hard choices: pivot, race them (almost always loses), or find a meaningfully different angle. See our PROSPERO walkthrough for how to search and interpret results.

Step 4: Check ClinicalTrials.gov for ongoing trials (3 minutes)

If your question is about whether a treatment works and there is a large ongoing trial scheduled to read out next year, your meta-analysis on that question may be obsolete before it publishes. Search clinicaltrials.gov for your intervention and population, filter by Recruiting / Active / Not Yet Recruiting status, and look at primary completion dates.

Step 5: Look for preprints on bioRxiv and medRxiv (3 minutes)

A wave of preprints on your topic in the last 6 months is an early warning that the field is heating up and someone is about to publish what you were planning. Europe PMC indexes both preprint servers and is the easiest place to search them together. You can also browse medrxiv.org and biorxiv.org directly.

Step 6: Use an automated novelty check (optional, 1 minute)

The five steps above are the manual version. You can also run an automated novelty check that does all five queries in parallel, deduplicates the results, scores novelty against the actual evidence found, and flags concurrent work. The output is structured the same way as the manual checks, so you can verify any specific number it reports.

How to interpret what you find

You find a recent SR/MA that answers your question

This is the most common outcome. Look at the review's "future research recommendations" section — reviews almost always end with a list of things still unanswered. That is a free roadmap to a more publishable angle.

You find older reviews and recent original studies but no recent SR

Good sign. The literature has moved since the last synthesis. An updated review is publishable and useful.

You find scattered original studies and no review at all

Great sign for a first review. Confirm the literature is large enough (typically 8+ studies) to support synthesis, and check that the studies are similar enough to combine.

You find almost nothing

Two possibilities: the field is genuinely novel, or your search terms are too narrow. Try synonyms and broaden the population before concluding the topic is virgin territory. If it really is novel, consider a scoping review or a primary study rather than a systematic review.

You find a PROSPERO registration on your exact question

Check the registration date and status. If they registered last month and have not published, you have at least 9 to 18 months before they appear. You can still proceed if your angle is different. If they are 80% done, pivot.

The 20-minute version saves the 8-week version

None of this requires institutional access, a librarian appointment, or anything beyond an internet connection. The 20 minutes you spend on this check is the most asymmetric use of time in your entire research project. You either confirm the idea is viable, or you save yourself two months. There is no scenario where skipping this check is the right call.