BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:“Extreme reviewing”: Use of text-mining to reduce impractical 
 screening workload in extremely large scoping reviews - Ian Shemilt\, Evid
 ence Synthesis Programme\, Behaviour and Health Research Unit
DTSTART:20121102T130000Z
DTEND:20121102T140000Z
UID:TALK41324@talks.cam.ac.uk
CONTACT:Dr Simon Richard White
DESCRIPTION:Background: In scoping reviews of broad evidence bases\, bound
 aries of relevant evidence may be initially fuzzy\, with a refined concept
 ual understanding of interventions and related phenomena of interest an in
 tended output of the process rather than its starting point. Searches are 
 therefore sensitive\, retrieving large record sets that can be impractical
  to screen using conventional methods.\nObjectives: To evaluate use of tex
 t-mining to reduce impractical screening workload in two large-scale scopi
 ng reviews of evidence for impacts of (i) choice architecture intervention
 s (CA) and (ii) changes in the economic environment (EE) on health-related
  behaviours and corollary outcomes. \nMethods: Baseline inclusion rates (B
 IRs) were estimated by screening random samples of records drawn from retr
 ieved sets comprising over 800K (CA) and 1 million (EE) de-duplicated reco
 rds. Text-mining technologies were applied to prioritise records for manua
 l screening.  47\,541 (CA) and 46\,099 (EE) prioritised records were manua
 lly screened and observed inclusion rates (OIRs) recorded. Text-mining per
 formance was measured in terms of OIRs relative to BIRs. Eligible records 
 prioritised using text-mining were compared with those located using paral
 lel snowball searches to assess unique yields and potential biases of each
  approach.\nResults: Overall unadjusted OIRs were 10.1 (CA) and 8.3 (EE) t
 imes higher than BIRs. Text-mining reduced manual screening workload by 90
 % (CA) and 88% (EE) compared with conventional methods (absolute reduction
 s of approximately 430\,000 (CA) and 378\,000 (EE) records)\, to identify 
 85% (CA) and 38% (EE) of remaining eligible records.\nConclusions: This st
 udy expands an emerging corpus of empirical evidence for use of text-minin
 g to support screening\, by demonstrating its feasibility\, strengths and 
 limitations in extremely large-scale scoping reviews. By reducing screenin
 g workload\, text-mining made it possible to assemble\, describe and delim
 it large and complex evidence-bases that crossed research-disciplinary bou
 ndaries. Findings are transferable to other large-scale scoping and system
 atic reviews that incorporate conceptual development or explanatory dimens
 ions.
LOCATION:Large Seminar Room\, 1st Floor\, Institute of Public Health\, Uni
 versity Forvie Site\, Robinson Way\, Cambridge
END:VEVENT
END:VCALENDAR