Diagnostic test accuracy (DTA) systematic reviews (SR) provide the most reliable form of evidence for decision makers on the accuracy of a diagnostic test.(1) A full SR is extremely resource intensive, often taking months or years to complete. Unlike trials, DTA studies are poorly indexed on major databases and so searches need to be very sensitive to identify all relevant studies. Searches often retrieve several thousand references which need to be screened in duplicate; this has substantial resource implications.(2) “Restricted searches”, searching a limited number of databases using a diagnostic filter, can limit the number of references identified, but at the expense of missing relevant studies.(2) There is little information on whether “missing” these studies would have impacted on review findings.
When conducting a systematic review, stronger conclusions can be derived from studies at low risk of bias (ROB). It has been suggested, but not yet empirically investigated, that “harder to find” studies (those not found by restricted searches) may be smaller and at greater ROB, such that the additional effort taken to find these studies may not be worthwhile. Further, if smaller studies are likely to be at higher ROB, it may be appropriate to suggest a sample size threshold for inclusion of studies in a DTA review; this would save resources in terms of reviewer time in extracting, assessing and synthesising studies.
Meta-epidemiological research provides a framework that allows researchers to look at the association of study characteristics with effect estimates, via collection and statistical analysis of data from a large number of meta-analyses.(3) These methods have been mainly applied to SRs of trials, but have also been used to look at the association of potential sources of bias with estimates of accuracy. Collection and analysis of new meta-epidemiological data relating to test accuracy, including analysis of novel study characteristics, such as sample size and whether studies were detected by restricted searches, would be extremely valuable in informing systematic review guidance in this area.
1. McInnes MDF. et al. Preferred Reporting Items for a Systematic Review and Meta-analysis of Diagnostic Test Accuracy Studies: The PRISMA-DTA Statement. JAMA : the journal of the American Medical Association. 2018;319(4):388-96.
2. Beynon R. et al. Search strategies to identify diagnostic accuracy studies in MEDLINE and EMBASE. Cochrane Database Syst Rev. 2013(9):MR000022.
3. Sterne JA. et al. Statistical methods for assessing the influence of study characteristics on treatment effects in 'meta-epidemiological' research. Statistics in medicine. 2002;21(11):1513-24.
Aim and approach
This PhD project will apply meta-epidemiological methods to determine whether restricted searches and sample size thresholds can be applied in DTA reviews to optimise the review process and reduce resource requirements.
This project is a collaboration between two GW4 HEIs: Population Health Sciences (PHS), University of Bristol (UoB), and the College of Medicine and health, University of Exeter. The student will be based in Population Health Sciences in the Bristol Medical School, an internationally leading centre for research and teaching in health sciences. At Exeter, the student will be closely connected with PenTAG and the Exeter Test Group. These groups are particularly experienced in health technology assessment, test evaluation and input into policy making. The student will gain valuable complimentary experience by collaborating across these departments. They will be invited to attend seminars, journal clubs and statistics group seminars run by Bristol Medical School and Exeter College of Medicine and Health.
Candidate requirements: Applications are welcome from high performing individuals across a wide range of disciplines closely related to health sciences, statistics, biochemistry, mathematics and psychology who have, or are expected to obtain, a 2.1 or higher degree. Applications are particularly welcome from individuals with a relevant Masters degree.
How to apply: Please make an online application for this project here
Contact: Penny Whiting [email protected]
The University of Bristol is offering a 3.5 year full time PhD in research around Population Health to start in 2019. This studentship is funded through GW4 BioMed MRC Doctoral Training Partnership. It consists of full UK/EU tuition fees, as well as a Doctoral Stipend matching UK Research Council National Minimum (£15,009 p.a. for 2019/20, updated each year). Additional research training and support funding of up to £5,000 per annum is also available.
Additional research and training funding is available over the course of the programme. This will cover costs such as research consumables, courses, conferences and travel. Additional competitive funds are available for high-cost training/research. The studentship is based at the Bristol Medical School. For further information please see the website below. View Website