Page 123 - The value of total hip and knee arthroplasties for patients

Return to work after total hip and knee arthroplasty: a systematic review
Step 3: data extraction
With respect to the study characteristics, the following data were systematically extracted from the selected full-text papers: title, first author, year of publication, journal title, study design (retrospective, prospective or crosssectional) and country where the study was conducted. Regarding the patient characteristics, the number of subjects in the study, the recruitment or selection criteria, sociodemographic characteristics of the subjects (age, sex) and type of operation (THA and/or TKA) were recorded. In addition, the following outcomes were extracted: work status of patients prior to surgery and work status of patients after surgery, including the observation time (duration of follow-up), with work status before or after surgery comprising the number of patients working full time (yes/no), working part time (yes/no), being on early retirement, unemployed, on sick leave and/or being permanently work disabled (partial or full disability pension). Moreover, the number of hours working per week and/or time to return to work (weeks) were registered, as well as any factors described as having an impact on return to work, including the employment of univariate or multivariate analyses.
Step 4: assessment of methodological quality 7 To assess the methodological quality of the included studies, a quality checklist was
developed based on items described in a review of tools for quality assessment7
and on a review of the quality of prognostic studies in systematic reviews.8 This
quality checklist was employed in a previous systematic review on work disability in a rheumatic disease by our own group.9 Two authors independently assessed the quality of each study by scoring 23 items (see Supplementary data, available at Rheumatology Online), divided into three categories: (i) selection bias (items 1-6), (ii) information bias (items 7-18) and (iii) statistical analysis of potential determinants of work status (items 19-23). Bias was considered present if the majority of the items within a category pointed in this direction.The quality of the study was rated as high if there was no evidence for selection bias, information bias or analysis bias. The quality of the study was rated as moderate if there was evidence of bias in one of the two categories in descriptive studies (statistical analysis of factors associated with return to work not applicable) or two of three categories in studies comprising an analysis of associations between various factors on the one side and work status on the other.The quality of the study was rated as low if there was evidence of bias in two categories in descriptive studies and all three categories in the other studies.
121