Page 122 - The value of total hip and knee arthroplasties for patients
P. 122

                                Chapter 7
adapted for use in other databases. The following databases were used: PubMed, EMBASE (OVID version),Web of Science, Cochrane Library, CINAHL (Cumulative Index to Nursing & Allied Health Literature; EbscoHost version), PsycINFO (EbscoHost version), Academic Search Premier and ScienceDirect. Restrictions included papers in English, French, German or Dutch and studies of humans. The search was performed on 23 April 2013.
Data collection and analysis
Four steps in the selection of studies and data extraction were defined. All steps were performed by two of the authors independently (C.T. and W.S.) and any discrepancies were resolved by consensus. If consensus between the two authors was not achieved, a final decision was made by a third author (T.P.M.V.V.).
Step 1: screening of titles and abstracts
First, all duplicates were removed. For screening of the remaining titles and abstracts, the following criteria were used: (i) clinical study including a minimum of 10 patients undergoing THA and/or TKA and (ii) reporting on the patients’ work status before and /or at least on one occasion after surgery.
Step 2: selection of full-text papers
Titles and abstracts identified as potentially eligible were selected for full-article review. If an abstract was not available, the full-text paper was requested. For the screening of the full-text papers the abovementioned criteria were again used, with the following specification regarding the reporting of work status: the study reported quantitative information on work status before and after THA and/or TKA, including working full time (yes/no), working part time (yes/no), number of hours working, early retirement, sick leave, unemployment, and/or permanent work disability (partial or full disability pension). Finally, the references of all selected papers and of systematic reviews included in the yield of the search strategy were checked for potentially eligible studies that were not identified with the original search strategy.The titles and abstracts of these references were screened using the abovementioned selection procedure.
120


























































































   120   121   122   123   124