13.07.2015 Views

Evaluating non-randomised intervention studies - NIHR Health ...

Evaluating non-randomised intervention studies - NIHR Health ...

Evaluating non-randomised intervention studies - NIHR Health ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

<strong>Health</strong> Technology Assessment 2003; Vol. 7: No. 27across <strong>non</strong>-<strong>randomised</strong> <strong>studies</strong>, and there arefewer methodological characteristics of interest(randomisation, concealment of allocation, blindingand completeness of follow-up). Also, there areclear reasons to believe that the biases againstwhich these design features protect are likely to actin favour of a particular treatment (theexperimental treatment), leading to systematic bias,such that the problem of increased heterogeneitybetween study designs does not arise.To take account of our concerns about thecomplexity of biases in <strong>non</strong>-<strong>randomised</strong> <strong>studies</strong>,we have first undertaken a review of qualitydomains and assessment methods used to considerthe likelihood of bias in <strong>non</strong>-<strong>randomised</strong> <strong>studies</strong>(Chapters 4 and 5). Second, we have undertakennovel methodological assessments of theimportance of two quality issues which are specificto <strong>non</strong>-<strong>randomised</strong> <strong>studies</strong>: method of allocation(Chapter 6) and use of case-mix adjustment(Chapter 7). Our methodological assessments havebeen designed to overcome the particularproblems of meta-confounding and variability ofdirection and size of bias that plague theinterpretation of these previous reviews.21© Queen’s Printer and Controller of HMSO 2003. All rights reserved.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!