Background/Purpose: Validated tools that incorporate evaluator and patient reported outcomes (PROs) are used for guiding therapy in patients with rheumatoid arthritis (RA). PROs used in RA evaluation may be influenced by co-morbidities (i.e. diabetes, coronary artery disease, congestive heart failure). We hypothesized that these co-morbidities lead to discrepancies between the tender and swollen joint counts and between patient- and evaluator-reported global assessments of RA activity.

Methods: We performed a cross-sectional analysis at baseline visit for all patients enrolled in the Veterans Affairs Rheumatoid Arthritis (VARA) registry with linkage to the VA Decision Support System (DSS). Patient global assessments (PGA) and evaluator global assessments (EGA) of disease activity distribution were divided into tertiles. PGA and EGA values in the same tertiles were defined as concordant, whereas the values in different tertiles were defined as discordant. Swollen joint count (SJC) to tender joint count (TJC) ratio of ≥ 0.8 was defined as concordant, whereas ratio < 0.4 was defined as discordant. Demographic characteristics and frequencies of DSS-derived co-morbidities in patients with concordant and discordant PGA and EGA; and TJC and SJC respectively were compared. Additionally, a longitudinal analysis of the subset of patients with at least 2 visits was performed to determine which ratio of SJC/TJC was associated with a decision to switch to a new biologic agent in patients with at least moderate disease activity (DAS28ESR > 3.2).

Results:

1305 unique patients with linkage to DSS were identified. The mean age was 68.2 years and 90.5% were males. PGA and EGA data was available for 529 patients. There were 95 (18%) patients with a discordant PGA worse than EGA, 123 (23.3%) with a discordant EGA worse than PGA, and 311 (58.8%) patients with concordant PGA and EGA. There was no significant difference in socio-demographics and prevalence of co-morbidities in patients with discordant worse PGA subgroup as compared to patients with concordant PGA and EGA. SJC and TJC data was available for all 1305 patients. There were 825 (63.2%) patients with SJC/TJC < 0.4, 326 (25%) patients with SJC/TJC ≥0.8, and 154 (11.8%) patients in-between these categories. Patients with SJC/TJC ratio < 0.4 were older than those with ratio ≥0.8 (68.8 yrs vs 67.2 yrs, p 0.02). There were no other significant differences in socio-demographic variables or prevalence of co-morbidities. Among all VARA patients treated with biologics who were in moderate or high disease activity (DAS28ESR > 3.2) at a VARA visit (n=1365 visits), approximately 7% were switched at that visit to another biologic. The likelihood of switching for patients with a SJC/TJC > 0.4 was 2.5 (1.5 – 4.4) fold greater compared to those with a SJC/TJC ratio ≤ 0.4.

Conclusion: There was no association of co-morbidities with discordantly high PGA or low TJC/SJC ratio. A SJC/TJC ratio < 0.4 appears useful to identify patients whose patient reported outcome data is unduly affected by factors other than RA and might be used to screen for patients not appropriate for traditional treat-to-target strategy cutpoints.

---

« Back to 2012 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/patient-characteristics-associated-with-discrepancies-in-evaluator-reported-and-patient-reported-outcomes-in-u-s-veterans-with-rheumatoid-arthritis/