Article type
Year
Abstract
Introduction: Rehabilitation is a broad area of medicine with a diversity of pathologies and types of intervention. Randomized Controlled Trials (RCTs) are often difficult to be designed and conducted in this field, but many authors consider the "cross-over" design a suitable solution. We decided to trace published cross-over studies in order to analyze the clinical settings where this design was used.
Objective: To identify the descriptors and textwords necessary to find all the articles in the rehabilitation field, and to test possible overlaps and how informative Medline and Embase were used as a sources for article retrieval.
Methods: We looked for articles published since 1990 without language restriction, limiting the search to neurologic, orthopedic, respiratory, urologic and rheumalologic rehabilitation. We used all the descriptors pertinent to rehabilitation: "Cross-Over" either as descriptor and text-word, "Rehabilitation" and its narrow terms as descriptors, text-words and subheadings, the descriptors "Physical Therapy" or "Physical Medicine" and its narrow terms, "Prosthesis and Orthoses", "Orthopedic Equipment", "Artificial Limb", "Walkers", "Walking Aid," "Self Help Devices".
Results: We found 191 articles on Medline and 164 on Embase with an overlap of only 59 articles. Only 26% of the articles on Medline and 29% on Embase were relevant to our aims. Of the 59 overlapping articles, 43% were relevant to the search. Overall we thus found only 72 useful articles. We analyzed how the articles published in 1990 and 1995, as sample years, were indexed to see why so many non-relevant studies were retrieved. Causes were a) the use of the text-words "cross-over" and "rehabilitation", b) poor indexing in some cases, c) the impossibility to excluding cardiological rehabilitation in Medline, d) lack of specificity of narrow terms under the descriptors "rehabilitation" and "physical therapy". They are not specific to rehabilitation but are used in many other clinical settings.
Discussion: At least two databases must be used to ensure a comprehensive search. Bibliographic search in the rehabilitation field is particularly complex because of the heterogeneity of the subject matter. Cooperation between an information professional and a clinician is essential to ensure extensive coverage over bibliographical databases and clinical settings.
Objective: To identify the descriptors and textwords necessary to find all the articles in the rehabilitation field, and to test possible overlaps and how informative Medline and Embase were used as a sources for article retrieval.
Methods: We looked for articles published since 1990 without language restriction, limiting the search to neurologic, orthopedic, respiratory, urologic and rheumalologic rehabilitation. We used all the descriptors pertinent to rehabilitation: "Cross-Over" either as descriptor and text-word, "Rehabilitation" and its narrow terms as descriptors, text-words and subheadings, the descriptors "Physical Therapy" or "Physical Medicine" and its narrow terms, "Prosthesis and Orthoses", "Orthopedic Equipment", "Artificial Limb", "Walkers", "Walking Aid," "Self Help Devices".
Results: We found 191 articles on Medline and 164 on Embase with an overlap of only 59 articles. Only 26% of the articles on Medline and 29% on Embase were relevant to our aims. Of the 59 overlapping articles, 43% were relevant to the search. Overall we thus found only 72 useful articles. We analyzed how the articles published in 1990 and 1995, as sample years, were indexed to see why so many non-relevant studies were retrieved. Causes were a) the use of the text-words "cross-over" and "rehabilitation", b) poor indexing in some cases, c) the impossibility to excluding cardiological rehabilitation in Medline, d) lack of specificity of narrow terms under the descriptors "rehabilitation" and "physical therapy". They are not specific to rehabilitation but are used in many other clinical settings.
Discussion: At least two databases must be used to ensure a comprehensive search. Bibliographic search in the rehabilitation field is particularly complex because of the heterogeneity of the subject matter. Cooperation between an information professional and a clinician is essential to ensure extensive coverage over bibliographical databases and clinical settings.