Week 1: Your topic, your protocol and Covidence
SKI3011 · Wed 31 Mar
Lecture
- Course overview
- Choosing your topic and research question
- Protocols on protocols.io
- Introduction: Covidence
Slides and materials
Lecture slides and other materials appear here before the lecture.
Tutorial
- Search and find your topic
- Import in Zotero and Covidence
- PRISMA flow chart in Covidence
- Choose your partner (groups of 1-3)
Homework
- Create: protocols.io account and Covidence project
- Watch: Covidence webinar
- Covidence: title/abstract screening
- Install: R and RStudio
Concepts and practice
This week has no quiz: the first quiz (next to week 3) is on R basics. These concepts you need for your protocol (A1).
| Concept | In one sentence |
|---|---|
| Review protocol | The plan of your review written before you start: question, eligibility criteria, search, selection, extraction, risk of bias and analysis. |
| Protocol registration | Publishing the protocol (here on protocols.io) with a date stamp, so readers can check that you did what you planned. |
| PRISMA-P | The reporting checklist for systematic review protocols. |
| Outcome switching | Changing the primary outcome after seeing the results; registration makes this visible. |
| Title/abstract screening | First selection round on titles and abstracts against the eligibility criteria. |
| Full-text screening | Second round: reading the full paper and recording the reason for exclusion. |
| Dual screening | Two reviewers screen independently and resolve conflicts; Covidence handles this for you. |
| Data extraction form | A predefined form for the study characteristics, risk of bias and outcome data you extract from each study. |
Practice questions
1. Why should you register your protocol before you start screening?
So that decisions (question, outcomes, analysis) are fixed before you know the results. Registration protects against selective reporting and outcome switching, and makes your review transparent.
2. Your review has two people in the group. How do you organise screening in Covidence?
Both screen all titles and abstracts independently (dual screening); Covidence lists the conflicts, which you resolve by discussion. Repeat for the full texts, recording a reason for each exclusion.