Sidelines 16 (3)
Jo Akers, Julie Glanville, Kate Light, Lindsey
Myers and Kath Wright
Centre for Reviews and Dissemination
University of York
Web:
http://www.york.ac.uk/inst/crd
Munro J, Sampson
F, Nicholl J. The impact of NHS Direct on the demand for
out-of-hours primary and emergency care. British Journal of General
Practice. 2005;55:790-792.
NHS Direct was
developed to try to control the rising demand for out-of-hours primary
and emergency care. This study investigates whether this has been
achieved during the first three years of its operation.
A postal survey was
sent to GP cooperatives, ambulance services and emergency departments
in England, Wales and Scotland to ascertain the number of patient
calls per month between April 1997 and March 2001. Data on calls to
NHS Direct were collected from published sources. Statistical models
were used to estimate the effect of NHS Direct on demand for the GP,
ambulance and emergency services.
The results showed
that during the period of the study, emergency departments and
ambulance services experienced a negligible change in demand while GP
cooperatives had a fall in calls of almost 8%. The impact this had on
overall general practice workload is unclear, because the study was
not designed to collect data which would have shown if the reduction
in calls was linked to a reduction in face–to–face consultations.
Bowling A. Mode of
questionnaire administration can have serious effects on data quality.
Journal of Public Health. 2005;27(3):281-291.
This is a narrative
review of the literature on the effects that different types of
questionnaire can have on the quality of data collected.
Face-to-face
interviews, telephone interviews, postal questionnaires and electronic
questionnaires were compared to asses the effects of the method of
questionnaire administration. Data quality was indicated by the
completeness of the data (percentage of questions answered and the
completeness of individual answers) and the validity of the answers
given. The response rate and the generalisability of the population
covered were also assessed.
Different methods
proved to have different strengths and different sources of potential
bias, and are summarised below:
The reviewers conclude
that all researchers need to be aware of the different kinds of bias
that attach to the different types of questionnaire administration.
Lavin MA, Krieger
MM, Meyer GA, Spasser MA, Cvitan T, Reese CG, Carlson JH, Perry AG,
McNary P. Development and evaluation of evidence-based nursing (EBN)
filters and related databases. Journal of the Medical Library
Association. 2005;93(1):104-15.
The authors note the
need for discipline-specific filters when carrying out evidence-based
nursing (EBN) searches, and describe how they constructed and tested
their EBN filters developed.
The research team
devised an inductive methodology to build and test the filters:
-
Construction of a
‘sleep’ search strategy to which all filters were applied using
PubMed
-
An EBN matrix was
created and used as a framework to categorise the evidence
retrieved. The rows of the matrix recording the levels of evidence
(primary, secondary and tertiary data). Diagnosis, related factors,
diagnostic tests, interventions and outcomes, served as categories
for the columns.
-
A recursive approach
was taken to develop three filters for nursing diagnosis, patient
outcomes and primary data.
-
The filter searches
were performed hourly on PubMed, and the retrieved citations stored
and updated.
-
Filters were
evaluated using sensitivity and specificity analyses, and retrieval
sets were compared.
-
On completion of the
evaluation process, the EBN filters were put in the public domain by
loading onto the web pages of the Network for Languages in Nursing
Knowledge Systems (NLINKS
http://nlinks.org/research_main.phtml).
The paper highlights
indexing issues and differences in the assignment of publication type.
Alpi KM. Expert
searching in public health. Journal of the Medical Library
Association. 2005;93(1):97-103.
The author recognises
an increased demand for evidence-based public health information, and
acknowledges that the teaching of evidence-based searching techniques
to public health practitioners falls behind that given to medical
practitioners. Therefore the expert searcher is crucial to public
health practitioners and students.
The challenges facing
the public health expert searcher are explored: the difficulties of
capturing the language of public health, the inadequacy of medical
subject headings to cover the ‘place’ and ‘population’ terminology
required in public health, and the non-standardised and poor indexing
that can be found in some databases. The disparate and large volume
of journals and databases in the field creates difficulties of access
in terms of cost, time and technology. The importance of grey
literature, meetings and conference information is explored, as are
the known difficulties in locating such information. The problem of
accessing full text information is acknowledged. Finally the
challenge of meeting the needs of the public health information
requestor and the varied backgrounds and knowledge that he/she is
likely to have is explored.
The author offers
suggestions on how to address these challenges. Networking and
maintaining regular contacts with other expert searchers in the field
is recommended. The value of information experience in other fields
and exposure to a range of databases is also mentioned. Another
suggestion is to be proactive in dealing with public health
practitioners, perhaps updating previously requested searches or
setting up selective dissemination of information searches (SDIS).
Expert searchers should be at the forefront of advocating improved
indexing for databases. The author suggests a number of websites that
describe new public health resources, recommending these as good ways
of ensuring the current awareness of the expert searcher.
The paper concludes
that expert searchers have a key role to play in finding
evidence-based information in the field of public health, and
therefore continually updating knowledge and skills in the field is
paramount.
Gomersall A.
Finding the evidence: looking further afield. Evidence & Policy.
2005;1(2):269-85.
This paper introduces
eight lesser-known bibliographic resources from Europe and
Australasia, giving a brief description of the subject, scope, size,
indexing and access methods for each one. There is also a discussion
of database selection for the social sciences.
The author states that
social science databases are more disparate and varied than their pure
and applied science counterparts. There has been a shift from using a
handful of what the author refers to as ‘classical databases’ i.e.
those academic databases containing peer-reviewed literature, to
additionally searching ‘non-classical’ databases which are often
non-academic, research or practitioner-based, containing
non-peer-reviewed literature and grey literature. The author
acknowledges that some of these databases can vary in quality of
content and lack of sophistication in the search interfaces. He also
refers to the barriers that can prevent the use of such databases:
cost of subscriptions; difficulties in obtaining full text documents;
and language, in the case of many European databases. However, he
presents arguments for overcoming these barriers: many databases offer
free trials so their usefulness can be evaluated; document supply
information is often included as part of bibliographic information,
and an increasing number of articles contain hyperlinks to the full
text; and multilingual thesauri and English search facilities exist
for some foreign language databases. The overall argument is that to
avoid bias databases other than the ‘classical’ ones must be searched,
and the author then goes on to list eight databases from Australasia
and Europe that he feels should be considered by anyone contemplating
a review in the fields of social policy and practice.
Cooper N, Coyle D,
Abrams K, Mugford M, Sutton A. Use of evidence in decision models: an
appraisal of health technology assessments in the UK since 1997.
Journal of Health Services Research and Policy. 2005;10(4):245-250.
Decision analytical
modelling is being used more frequently in the health technology
assessment process. The data to populate such models can be identified
from a wide range of sources including randomised controlled trials,
observational studies, case series, expert opinion and meta-analyses.
Search strategies used to obtain the data and the sources used often
go unreported. The economic decision models developed as part of the
NHS R&D HTA Programme 1997 to 2003 were reviewed and the quality of
evidence used was then assessed using a hierarchy of data sources
developed specifically for economic analyses. The authors call for
greater transparency both in the sources of data used to construct
decision models and in the reporting so that policymakers can better
assess their reliability.
Sullivan F, Wyatt
JC. How decision support tools help define clinical problems.
BMJ.
2005;331:831-833.
This is a brief
discussion of how electronic clinical information systems are becoming
integral components of healthcare services and replacing the existing
paper based systems. Some examples are given. These include how
electronic prompts built into electronic clinical systems can be
effective at identifying potential problem areas that doctors can
investigate further. Clinical practice guidelines available
electronically can generate recommendations about diagnosis, screening
or treatment that are patient specific. In the future, information
about individuals’ genetic status may also need to be incorporated
into GP decision making while electronic patient record systems could
also provide reminders to patients when appointments have been missed.
Back to Inform 16 (3) | 
