Section 3.28 min read

Searching Before Starting

Core summary

Before investing months in a project, spend 30 minutes searching PubMed and Google Scholar to check if your question has already been answered. This step saves time, prevents duplication, and often improves your question.

Detailed explanation

A preliminary literature search is not a full systematic review. It is a quick, focused scan. Step 1: Translate your PICO into search terms. Take each PICO element and list synonyms. Step 2: Search PubMed with combined terms. Use AND between PICO elements, OR between synonyms. Step 3: Scan the top 20 results by title and abstract. Look for systematic reviews, large RCTs, and recent publications. Step 4: Interpret what you find: - Question fully answered by a high-quality SR → change your question - Question partially answered → your angle may still be novel - Addressed by weak studies only → your study adds value - Nothing found → either truly novel or wrong search terms Step 5: Check Google Scholar for grey literature. This entire process should take 30-60 minutes. It is not optional.

Clinical example

A resident wants to study 'the effect of honey on wound healing.' A 10-minute PubMed search reveals 3 Cochrane reviews and over 200 RCTs. The question has been extensively answered. But no studies exist on Manuka honey specifically in diabetic foot ulcers in the Middle East — that narrower angle may still be novel.

Research example

Chalmers & Glasziou (2009) estimated that 85% of research investment is wasted — a major cause being failure to check existing evidence before starting new studies.

Knowledge check

Q1. If your PubMed search reveals a recent Cochrane systematic review that fully answers your PICO question, you should:

Q2. In a PubMed Boolean search, how do you combine synonyms for the same PICO element?

Q3. If a PubMed search returns zero results for your question, it definitely means your question is novel.