Background/Purpose: RA patients (pts) often have multiple comorbidities, and with multiple disease activity (DA) metrics available selecting an appropriate metric is vital. It remains an outstanding question as to how much non-inflammatory disease impacts the composite DA metrics, further increasing the need to understand the impact of each metric. A companion study to this suggested a significant difference between between RAPID3(R3) and the swollen joint count (SJC). In this study we compare the DA scores at each DA state for RA patients without Fibromyalgia, re-examined the SJC/R3 relationship, and subsequently factored in damage.

Methods: Pts > 18 years and diagnosed with RA in the JointMan database between 1 Jan 2009 and 4 Mar 2018 were initially included. Pts having a diagnosis of Fibromyalgia were subsequently excluded. Encounters for the Pts having DAS28, CDAI, and RAPID3 scores were selected and further filtered into four cohorts, high, moderate (mod), low, and remission (rem). Each cohort contained only encounters where the patient was in its named DA state using all three metrics. Encounters where a pt was in two or more named DA states were excluded. All DA scores and the SJCs were normalized to Z-Scores on which was performed an analysis of variance. A subsequent ANOVA sensitivity analysis with pain scores and counts of decreased range of motion and deformed joints was also completed. A TukeyHSD post-hoc calculation compared the individual DA results.

Results: 4659 pts over 31,383 encounters were initially included, only 18.2% (5709 encounters) had DA state agreement (1300 high, 1690 moderate, 616 low, 2103 remission). The F values for each DA cohort were: 441.4 high, 1469 mod, 2326 low, 4392 rem and all had a P-value < 10^-16. In the main analysis, for the pair-wise comparisons of each metric to the others, virtually all pairs had significantly different means at each disease activity state. The difference in means between SJC and R3 was 0.96 in high and 0.91 in remission. In the sensitivity analysis the difference in means between SJC and DAS28 was 0.05 in high and 0.42 in remission. Similarly the differences in means between composite scores and joints with deformities or decreased range of motion was also significant (P-value = 0).

Conclusion: When controlling for fibromyalgia patients, each metric and its DA state is consistently different from other metrics at the same DA state. The exception of SJC/DAS28 scores was not surprising as SJC trends closest to DAS28 scores than other metrics. The SJC/DAS28 pair had a high concordance of means in the high DA state. A disparity between SJC/R3 in remission exists when excluding FM and is not resolved when considering the potential RA related joint damage that might adversely affect the function scores of R3. This raises concern that R3 values at remission may unduly represent functional impairment from concordant non-RA musculoskeletal diseases.