Background The correlation of components of composite activity scores among different patient populations may provide insights into universal applicability and use of disease activity indices. TRAV (Turkish acronym for “Turkish Rheumatoid Arthritis Registry”) was established in 2010 with the aim of collecting data on RA. patients seen in routine care in Turkey, the first consecutive patients database to do so.
Objectives To compare the correlation of components of composite indices in different RA cohorts from Turkey and US.
Methods Consecutive patients seen at participating centers complete at each visit a MDHAQ which includes scales for physical function, pain, and patient global. Physicians complete global assessment VAS, in addition to tender and swollen joint counts. RAPID3 (routine assessment of patient index data), DAS28 and CDAI are calculated. Demographics, self-reported disease activity measures (Pt global), clinical data and medication usage were abstracted from the last visit of individuals with RA seen at Turkish (TR) sites these were compared with patients seen at a US routine care setting, where data have been collected since 2001 on all patients. Patients seen during the same period were used in this comparison. Agreement analysis of physician and patient-reported clinical outcomes were done. Significant differences in measures were determined using the Kruskal-Wallis test for continuous and ordinal measures and Chi-square test for categorical measures.
Results 424 TR and 176 US patients were studied. Correlation between MD and Pt global, MD global and pain, DAS28 and RAPID3, MD swollen and tender joint count and Pt self report RADAI joint count are shown in the Table. All measures had similar correlations between US and TR RA patients except MD global and Pt global assessments where US measures showed more correlation than TR patients and physicians. The correlation between patient tender joint count and physicians tender joint count were stronger than swollen joint counts but similar in US and TR. DAS28 and RAPID3 were also correlated similarly. There was no difference in disease activity as measured by RAPID3 (TR=10.5, US=11.2). There was significant difference between TR and US patients for years of education (TR=6.9 yr, US=14.5 yr) and disease duration (TR=14 yr, US=7.3yr)
Conclusions The correlation between physician and patient measures were similar among US and Turkish RA patients except physician and patient global assessments of disease activity. The reasons for this is not clear but the difference in years of education and/or disease duration may play a role. Performances of composite indices, regardless of being based on patient-measures only or patient and physician measures were better than the each individual component in these different cohorts of RA patients.
Acknowledgements The authors would like to thanks Cortex for their help with data entry and Bristol-Myers Squibb unrestricted data entry support.
Disclosure of Interest G. Hatemi: None Declared, N. Inanc: None Declared, I. Simsek: None Declared, K. Tascilar: None Declared, S. Ugurlu: None Declared, M. Can: None Declared, S. Pay: None Declared, H. Erdem: None Declared, S. Yilmaz: None Declared, M. Cinar: None Declared, N. Cakir: None Declared, W. McCracken: None Declared, C. Swearingen: None Declared, H. Direskeneli: None Declared, Y. Yazici Grant/Research support from: Abbott, BMS, Celgene, Genentech, Consultant for: Abbott, BMS, Centocor, Celgene, Pfizer, genentech, UCB, Takeda