Article type
Year
Abstract
Background: To provide an easy approach in identifying Cochrane neurological reviews the Neurological Field created a browse list organised as an index in a neurological textbook; since 2000 two Italian neurologists have blindly selected new and updated reviews they feel are of neurological interest from the Cochrane Library, attributing chosen titles to specific diseases, included in a 27 item list (Fig. 1); there were discrepancies in choices made, this was in contrast with the purpose that the browse list be representative for the largest number of users over the globe.
Objectives: To understand how wide the variability is in selecting a review for a person with neurological interests, creating a practical and broad virtual database of systematic neurological reviews.
Methods: Five neurologists representative of different geographical areas: Nigeria, Brazil, Germany and Italy, wereasked to independently select reviews of neurological interest from those published in the Cochrane Library in 2008, classifying them in the neurological disease index. Inter-observational agreements were analyzed with the Latent Class Model for Rater Agreement¹.
Results: Three hundred and fourty-four titles of new/updated systematic Reviews were extracted. The model with five latent classes was selected (Table 1), as the best support to the data in terms of a lower AIC index. There was a wide range of variability: from highest agreement in Latent Class 1 (122 Tiles, 32.6%) to lowest in Latent Class 5, only 1 of the 5 neurologists felt that 54 Titles (15.7%) carried a neurological interest (Table 2).
Conclusion: To enhance global participation it is essential to be aware that ‘neurological interest’ is different in different countries, possibly due to educational background², socio-economic resources and working expertise; considering these variability could help develop appropriate criteria to elaborate a comprehensive browse list of reviews, address tagging objectives and provide an easier and quicker way for the Cochrane to reach new audiences.
References
1. Agresti, A. (2002) Categorical Data Analysis 2nd ed. New York: Wiley.
2. Journal of the Neurological Sciences Volume 246, Issue 1, Pages 59–64, 15 July 2006. Neurology training programs worldwide: A world federation of neurology survey.
Objectives: To understand how wide the variability is in selecting a review for a person with neurological interests, creating a practical and broad virtual database of systematic neurological reviews.
Methods: Five neurologists representative of different geographical areas: Nigeria, Brazil, Germany and Italy, wereasked to independently select reviews of neurological interest from those published in the Cochrane Library in 2008, classifying them in the neurological disease index. Inter-observational agreements were analyzed with the Latent Class Model for Rater Agreement¹.
Results: Three hundred and fourty-four titles of new/updated systematic Reviews were extracted. The model with five latent classes was selected (Table 1), as the best support to the data in terms of a lower AIC index. There was a wide range of variability: from highest agreement in Latent Class 1 (122 Tiles, 32.6%) to lowest in Latent Class 5, only 1 of the 5 neurologists felt that 54 Titles (15.7%) carried a neurological interest (Table 2).
Conclusion: To enhance global participation it is essential to be aware that ‘neurological interest’ is different in different countries, possibly due to educational background², socio-economic resources and working expertise; considering these variability could help develop appropriate criteria to elaborate a comprehensive browse list of reviews, address tagging objectives and provide an easier and quicker way for the Cochrane to reach new audiences.
References
1. Agresti, A. (2002) Categorical Data Analysis 2nd ed. New York: Wiley.
2. Journal of the Neurological Sciences Volume 246, Issue 1, Pages 59–64, 15 July 2006. Neurology training programs worldwide: A world federation of neurology survey.
Images
