Sidelines 18 (1)

Centre for Reviews and Dissemination

University of York

Web: http://www.york.ac.uk/inst/crd

Kastner M, Wilczynski NL, Walker-Dilks C, McKibbon KA, Haynes B.
Age-specific search strategies for Medline. Journal of Medical Internet Research. 2006;8(4):e25 URL: http://www.jmir.org/2006/4/e25/

Finding studies relating to specific age groups in Medline can be problematic. The age group indexing offered by Medline may not match the searcher’s specific age query and/or Medline’s indexing may not be specific enough for helpful retrieval because all age groups included in a study may receive indexing. This paper reports the development of search strategies to identify age groups within Medline records with better precision and specificity.

The authors developed their strategies from keywords and MeSH terms elicited from a wide range of contributors. They tested out these terms’ retrieval performance as single and multiple-term strategies compared to a gold standard set of records. The gold standard was compiled by hand searching and coding all publications from 161 core health care journals for the year 2000. The extra coding added to the records reflected the age bands into which 50% or more of the study participants fell.

The authors offer a range of strategies for age-specific specialty areas: geriatric medicine, adult medicine, pediatric medicine, neonatal medicine and obstetrics. Strategies for sensitive searches, specific searches and best optimization of sensitivity and specificity are offered, so that the searcher can choose the one that is closest to their own focus.

The validation process used for the strategies is unclear from this paper but may be reported in the authors’ earlier detailed methods paper. The authors note that their gold standard represents records from only one year, but are sanguine about the generalisability of their results. The gold standard was developed from core journals, and this may also affect the generalisability of the strategies.

Aberegg SK, Arkes H, Terry PB. Failure to adopt beneficial therapies caused by bias in medical evidence evaluation. Medical Decision Making. 2006;26(6):575-82. 

To address the reasons why many evidence-based therapies are under utilised, the authors of this paper pose the question “is the rate of adoption of beneficial therapies lower than the rate of abandonment of harmful ones?”.

Two case vignettes describing hypothetical RCTs were designed. In version one a treatment in common use was shown to be harmful, in version two the same treatment not in common use was shown to be beneficial. Each version of the vignette was randomly dispensed to two independent samples of critical care and pulmonary practitioners.

The results showed that 76.5% of vignette one respondents would apply the evidence and abandon the harmful treatment. In contrast, only 33.3% of respondents of vignette two would adopt the beneficial treatment.

The conclusions drawn are that physicians are biased when evaluating medical evidence and are less willing to adopt beneficial therapies than to abandon harmful therapies.

Verheijden MW, Jans MP, Hildebrandt VH, Hopman-Rock M. Rates and determinants of repeated participation in a web-based behaviour change program for healthy body weight and healthy lifestyle. Journal of Medical Internet Research. 2007;9(1):e1.

This paper looks at the effectiveness of web-based health promotion programs in reaching the appropriate target audience and in retaining that audience. The study investigated the characteristics of those who participated in a web-based health behaviour change program. Analyses showed that the participants in the program were relatively healthy people and that the program failed to reach the audience for which it was intended. However, the study also identified that obese people used the program more frequently than people of normal body weight. Such findings suggest that web-based behaviour change programs may be more successful for weight management than other areas of health promotion.

Oransky I. Bloggers beware: conflicts of interest and diabetes. The Lancet. 2006;368:1641-2.

This short article reports on the use of blogs, particularly those in the area of diabetes. The paper begins with a description of a number of blogs retrieved via www.technorati.com. The authors reflect on the increasing number of people with diabetes writing blogs (known as diabetes bloggers or d-bloggers).

The most interesting part of the article is the comparison between traditional journalists and bloggers in terms of their attitude towards accepting funding from industry. Oransky suggests that bloggers tend to accept money as long as it comes with ‘no strings attached’, whilst journalists’ ethics preclude them from accepting such funding. There is also a growing amount of fake-blogging (or flogging) where pseudo blogs are created by companies for marketing purposes. The authors call for more safeguards to protect the public and warn against an unquestioning acceptance of the provenance of blogs.

Tang H, Ng JHK. Googling for a diagnosis – use of Google as a diagnostic aid: internet based study. BMJ. 2006;333:1143-45.

One of the major uses of the internet is the search for health information by patients and clinicians. With the advent of sophisticated search engines such as Google, medical information is more easily accessible than ever before. In this paper, the authors report a study that examines the potential use of Google as an aid to diagnosis. Using a sample of 26 case records they entered a selection of search terms based on the patient’s symptoms into Google. In 15 cases Google returned a correct diagnosis and the authors conclude that ‘Google may help doctors to formulate a differential diagnosis in difficult cases’.

Unsurprisingly the BMJ received a large number of rapid responses to this article and the story was picked up by the mass media. Overall this is an interesting article and provides a good starting point for debate. It should be noted, however, that resources other than Google may provide more useful information. For example, one rapid responder replicated the searches on PubMed and obtained a much better success rate.

In addition, any article located on the web should be critically appraised before use.

McKibbon KA, Wilczynski NL, Haynes B. Developing optimal search strategies for retrieving qualitative studies in PsycINFO. Evaluation and the Health Professions 2006(29)4:440-454.

Most search strategies have been developed for MEDLINE to retrieve therapy and review articles. It is refreshing to see here that the authors have chosen a different database, (PsycINFO) and a different area to search for (qualitative research).

Qualitative studies are particularly difficult to search for due to the multiplicity and inconsistency of the use and naming of concepts, the broad array of databases and journals in which studies can be found, the nature of article titles and the frequent absence of any abstract.

The authors use 125 journals articles identified by hand searching 64 journals published in 2000 to derive three search filters; a narrow focussed search strategy, a broad search strategy and a best compromise search strategy.

The Health Information research Unit at McMaster University (HIRU) has used the same methods to develop a wide range of filters for various databases. The strategies from this article have been made available as filters on the Ovid interface to PsycINFO.

Back to Inform 18 (1)

This page was last updated on: 27 July 2007