Secondary Data Analysis Systematic Reviews Associated Databases Prof

Secondary Data Analysis: Systematic Reviews & Associated Databases Prof. Eileen Savage OMCYA Keeping Children Safe Summer School, 12 th – 16 th September 2011, UCC.

What is a Systematic Review? o a concise scientific investigation, with pre-planned methods that summarise, appraise, synthesise and communicate the results of multiple primary studies (Cooke et al, 1997; Centre for Reviews and Dissemination, 2008). ie Research on Research with findings of existing data becoming raw data Differs from a narrative review/ traditional literature review

Traditional Review vs. Systematic Review (Nasseri-Moghaddeam & Malekzadeh 2006)

Growth of Systematic Reviews

Systematic Reviews – Origins of Growth o o o Expanding volume of published & unpublished literature Conflicting findings across studies leading to ‘uncertainty’ Contributes to Evidence Based Movement o

When to Do a Systematic Review o o When there is ‘uncertainty’ about the effectiveness of interventions in practice When key questions remain unanswered e. g. about treatments, interventions, practices, experiences etc… To inform practice/policy with best available research evidence To identify what is known/not known in an area to guide future research (includes research methods) (Petticrew & Roberts 2007)

Where to find a Systematic Review o o o o Peer Reviewed Journals (search bibliography databases) Cochrane Database of Systematic Reviews (Cochrane Library; several Groups) Database of Abstracts of Reviews of Effects (DARE) Health Technology Assessment Database NHS Economic Evaluation Database The Campbell Collaboration Library (Groups e. g. Crime & Justice; Education; Social Welfare; Methods; Communication & internationalization; Users Groups. The Evidence for Policy and Practice Information and Co-ordinating Centre (EPPI -Centre), University of London. (Education; Health Promotion; Employment; Social care; Crime and Justice) Others

Structure of Systematic Review e. g. o o o o Introduction/Background Questions/Aim & Objectives Criteria for considering studies Search Methods of the Review (e. g screening reading papers; quality assessment; data extraction) Results Discussion Conclusions

Criteria for Including/Excluding Studies o Types of Studies (see hierarchy of evidence) o Types of Participants o Types of Outcomes Link to Questions being asked

Hierarchy of Evidence

Search Method o o o Search Terms Databases +/- Hand searching +/- Reference lists +/-Grey Literature +/- Contacting authors (Refer to Booth et al on standards for Reporting Search Methods)

Search Strategy: CINAHL with Full Text (EBSCO) Date of Search Years Covered Complete Strategy Language Restrictions Search Strategy: Embase (Elsevier) 30 th March 2011 Date of Search January 1990 to March Years Covered 2011 1. cystic fibrosis. TX. 2. mucoviscidosis. TX. 3. 1 or 2 4. clinical trial. PT. 5. trial. TX. 6. random*TX. 7. 4 or 5 or 6 8. educat*. TX. 9. program*TX. Complete Strategy 10. 8 or 9 11. self care. TX. 12. self-care. TX. 13. selfcare. TX. 14. self management. TX. 15. self-management. TX. 16. manag*. TX. 17. 11 or 12 or 13 or 14 or 15 or 16 18. 10 or 17 19. 3 and 7 and 18 None 30 th March 2011 January 1990 to March 2011 1. cystic fibrosis. ti, ab, de. 2. mucoviscidosis. ti, ab, de. 3. 1 or 2 4. 'clinical trial'. ti, ab, de. 5. 'trial'. ti, ab, de. 6. random*'. ti, ab, de. 7. 4 or 5 or 6 8. educat*. ti, ab, de. 9. program*. ti, ab, de. 10. 8 or 9 11. self care ti, ab, de. 12. selfcare. ti, ab, de. 13. self management. ti, ab, de 14. manag*. ti, ab, de. 15. 11 or 12 or 13 or 14 16. 10 or 15 17. 3 and 7 and 16

Methods – Screening for Inclusion/ Exclusion o o Titles & Abstracts – In/Unsure/Out Read in full –> in/unsure /out

Quality Assessment of Individual Studies o Quality threshold for inclusion ie exclude poor quality studies Assessment of Quality may be part of SR o How o - No –’gold standard’ exists - Checklists available e. g. CASP - Quality Scales/Criteria

Data Extraction o Use a data extraction form (s) o 2 or more reviewers / cross-check o Extract details relevant to Questions/ Objectives o Present raw data in report – table format

Data Analysis & Synthesis o o o Collation & summary of results Present in tabular form/forest plots/other + descriptive narrative account +/- Meta-analysis, if appropriate (not possible with heterogeneity ie cannot pool apples with oranges for the purpose of statistical analysis as one data set)

Results, Discussion, Implications

Critical Appraisal of Systematic Reviews o The PRISMA Statement: standards for reporting SRs & Meta-analyses o Critical Appraisal Tool Kit for SRs (CASP)

Website on Critical Appraisal Tools (covers a range of methods) At: Division of Health Sciences International Centre for Allied Health Evidence o http: //www. unisa. edu. au/cahe/Resources/CAT/de fault. asp

References o To follow