I read with interest Mays and Popes (1) discussion of the issues
surrounding the assessment of qualitative research, citing 'the
proliferation of guidelines for doing and judging qualitative research' as
a manifestation of the concern about assessing quality. I would propose
that many researchers also include 'quality filters' when searching
bibliographic databases in this list.
I have recently completed a survey of qualitative researchers
surrounding their perceptions of the use and relevance of 'quality
filters' when searching electronic databases. One of the main issues
arising from this survey was the tenet that 'quality filters', such as
those developed by the Cochrane Collaboration, McMasters University and
the NHS Centre for Reviews and Dissemination, aim to make assessments on
the basis of the quality of the research study. This is clearly an
impossibility; a misconception bound up with the use of terminology and
those associated with the received evidence hierarchy.
The term 'quality filter' is misleading in two respects. Firstly, it
suggests a filtering out or sifting of potentially irrelevant - perhaps
poor quality - materials. In reality the search strategies aim to be more
inclusive, aiming to identify journal papers more accurately by overcoming
the limitations of database indexing, such as the misapplication or non-
application of appropriate medical subject headings (MESH headings).
Secondly, there is the issue of quality. To date the majority of
published search strategies have been developed to overcome indexing
problems for randomised controlled trials, systematic reviews etc. Given
their position at the top end of the received evidence hierarchy, there is
often an implicit assumption made that these research methods are of a
'gold standard'. Thus, having identified studies of these types, there is
an implication that little appraisal is assumed to be necessary because
studies of a lesser quality, studies of a less 'rigorous' nature, have
been excluded. However, quality assessments remain of paramount
importance in evaluating the appropriateness of study design, the basis of
conclusions etc. wherever the study appears in the evidence hierarchy.
As such, I believe there is a serious need for the reconsideration of
the terminology used in relation to search strategies developed to
overcome deficiencies in database indexing. The wide spread adoption of
the term 'optimal search strategy' would be a more appropriate and more
accurately representative term than ‘quality filter’.
Only then, can the true nature of quality assessments be recognised.
Maria J Grant
Health Care Practice R&D Unit
University of Salford
Humphrey Booth House
Tel: 0161 295 0454
1. Mays N, Pope C. Qualitative research in healthcare: assessing
quality in qualitative research. British Medical Journal 2000; 320: 50-52.
Competing interests: No competing interests