Comparison of agreement between internet-based registration of patient-reported outcomes and clinic-based paper forms within the Swedish Rheumatology Quality Register.

Objective: The Swedish Rheumatology Quality Register has implemented an internet-based method (PER) for registering patient-recorded outcome measures. The aim of this study was to compare the agreement between visual analogue scales (VASs) reported via PER and clinic-based reporting using paper forms. Methods: In a cross-sectional study (70 patients), the results of 79 registrations of VASs for global health, pain, and fatigue from PER were compared with corresponding clinic-based paper registrations. For patients with polyarthritis, 28-joint count Disease Activity Scores (DAS28) were computed. Patients with axial disease also completed Bath Ankylosing Spondylitis Disease Activity Index and Functional Index (BASDAI and BASFI) questionnaires. Mean differences and intraclass correlation coefficients (ICCs) were calculated. Agreement was visualized using Bland-Altman plots. Results: No statistically significant differences in VASs were found comparing PER and paper forms for VAS Global, VAS Pain, and VAS Fatigue (p = 0.295, 0.463, and 0.288, respectively). ICCs for VAS Global, Pain, and Fatigue ranged from 0.889 to 0.952, indicating excellent agreement. Bland-Altman plots for VAS did not show any proportional bias. The mean difference for DAS28 calculated by VASs from paper vs PER was -0.02 (n = 65, p = 0.660), and the mean difference for BASDAI was 0.04 (n = 11, p = 0.742). ICCs for DAS28 and BASDAI were 0.962 and 0.985, respectively. Of the participating patients, 60% preferred PER. Conclusion: Internet-based reporting for patient-reported outcomes in a clinical setting resulted in similar data for VASs and corresponding disease activity scores to clinic-based reporting on paper forms.

Psychometric Testing of an Instrument About Intimate Partner Violence.

This study used step-by-step exploratory factor analysis in the framework of confirmatory factor analysis (EFA/CFA) to evaluate the psychometric properties of the translated and culturally adapted Swedish version of "The Violence Against Women Health Care Provider Survey." The Swedish version of the instrument was needed to measure district nurses' preparedness to encounter women exposed to intimate partner violence. In the first step of EFA/CFA, the eight-factor model was confirmed. The item-total correlations ranged from .22 to 1.01, and Cronbach's alphas from .68 to .71. After removing four items, the corrected item-total correlations ranged from .40 to .97. On the basis of the analysis, we concluded that the Swedish version of the instrument is valid and reliable for evaluating the preparedness of district nurses in Sweden to encounter women exposed to intimate partner violence.