Workshop
Diagnostic test accuracy systematic reviews and meta-analysis: a 1-day introductory workshop
Penny Whiting, Hayley Jones, Rhiannon Owen
University of Bristol and University of Leicester
NICE Clinical Guidelines Technical Support Unit
and
NIHR Complex Reviews Support Unit
Course Summary
This is a 1-day introductory course on systematic reviews and meta-analysis of diagnostic test accuracy. It covers: defining a diagnostic review question; searching for studies; sources of bias; quality assessment with QUADAS-2; measures of test accuracy; the Receiver Operating Characteristic (ROC) curve; bivariate meta-analysis using Stata; production of forest plots and summary ROC plots in RevMan
Pre-requisites
Participants will need to bring a laptop with Review Manager (RevMan) installed in order to complete the computer practical exercises. No previous experience with RevMan is required. The software can be downloaded here: http://community.cochrane.org/tools/review-production-tools/revman-5
Please also download this file in advance, which you will use in the compute practicals.
Please read the following paper as preparation for the morning session:
Smith, H., Pickering, R.M., Struthers, A., Simpson, I. and Mant, D., 2000. Biochemical diagnosis of ventricular dysfunction in elderly patients in general practice: observational study. BMJ, 320(7239), pp.906-908.
Some previous experience with meta-analysis is strongly advised, including understanding of the difference between fixed and random effects meta-analysis. Participants with no experience of meta-analysis would be advised to read an introductory paper such as:
Borenstein M et al. (2010) ‘A basic introduction to fixed‐effect and random‐effects models for meta‐analysis’. Research Synthesis Methods. 1;1(2):97-111.
What will participants learn?
Participants will learn how to undertake systematic reviews of diagnostic test accuracy, assess study quality, quantify test accuracy and produce forest plots of sensitivity and specificity and summary ROC plots using RevMan. Participants will also learn how to interpret results from a meta-analysis of diagnostic test accuracy using Stata.
Course schedule
| 9:15 – 9:30 | REGISTRATION AND COFFEE |
| PART 1: | SYSTEMATIC REVIEWS |
|
|
Penny Whiting |
| 9:30 – 11:00 | 1. Introduction to diagnostic research |
|
2. Measures of diagnostic accuracy |
|
|
3. Defining the review question |
|
|
4. Principles behind searching for DTA studies |
|
| 11:00 – 11:15 |
TEA, COFFEE |
| 11:15 – 12:45 | 5. Assessing study quality using QUADAS-2 |
| 6. Plain language summary of a DTA review | |
| Practical exercise | |
| 12:45 – 13:30 | LUNCH |
| PART 2: | META-ANALYSIS |
| Hayley Jones, Rhiannon Owen | |
| 13:30 – 15:00 | 7. Meta-analysis of sensitivity and specificity |
| 8. Summary ROC curves | |
| Computer practical 1: meta-analysis of diagnostic test accuracy | |
| 15:00 – 15:15 | TEA, COFFEE |
| 15:15 – 16:45 | 9. Exploring heterogeneity |
| Computer practical 2: Exploring heterogeneity | |
| 10. A brief introduction to more advanced topics | |
| 16:45 – 17:00 | QUESTIONS, DISCUSSION, FEEDBACK |