Background/Purpose: Clinical assessment of disease activity is a routine procedure when evaluating individual rheumatoid arthritis (RA) patients in daily practice. Swollen joint count (SJC) and tender joint count (TJC) are traditionally performed by the physician (Ph) or the nurse. Patient (Pa) assessed joint count may be an advantage in the busy clinic. The reliability of DAS28 based on patient self-assessment of tender and swollen joints has not previously been reported. The objective of this study was to examine the agreement between patient and physician evaluation of SJC and TJC, to examine the agreement between physician and patient derived DAS28 and to evaluate the reproducibility of these measures when assessed by the physician and the patient, respectively.  

Methods: 30 out-clinic RA patients (mean age 60±15 years) with a disease duration of a least 5 years, who were familiar with physician joint counting and who were considered to have a stable and controlled disease activity were included. 28 TJC and SJC were assessed by patient self-evaluation and by an experienced specialist in rheumatology (the same specialist for all patients) on two different days (median interval 7 days, range 4–10 days) under standardised conditions. Patient global assessment (PaGA) (range 0-100) and CRP (mg/l) were also determined on both occasions. DAS28-CRP with four variables (4V) and three variables (3V) were calculated (1). The association between the duplicate measures was expressed by Pearson’s coefficient of correlation (r). Lower and upper limits of agreement (LOA) were calculated as the mean difference between scores of duplicate measurements (the bias) ± 1.96 x SD of these differences (2).

Results: The mean values for SJC, TJC, DAS(4V) and DAS(3V) based on Ph joint count versus Pa self-evaluation at the first visit was 3.7±2.4 vs. 3.6±3.2 (NS), 4.1±5.2 vs. 4.1±4.7 (NS), 3.5± 1.0 vs. 3.6±1.1 (NS) and 3.4±0.9 vs. 3.5±0.9 (NS). Mean values for CRP and PaGA were 7.9 ±6.5 and 33±28, respectively. Ph-DAS(4V) and Pa-DAS (4V) were highly correlated (r = 0.90, p < 0.0001), and so were corresponding DAS scores based on 3V (r = 0.85, p < 0.0001). The bias for Ph-DAS(4V) vs. Pa-DAS(4V) and associated lower and upper LOA were -0.1 (NS), -0.9 and +0.8, respectively. For DAS(3V) the corresponding results were -0.1 (NS), -1.1 and +0.9. The bias for the duplicate Pa-DAS(4V) assessments was -0.1 (NS) with lower and upper LOA of -0.9 and +0.8; and for Pa-DAS(3V) 0.0 (NS), -0.9 and +0.9. For duplicate Ph-DAS(4V) assessments the bias was 0.0 (NS), and LOA were -1.1 and +1.1. For Ph-DAS(3V) the corresponding figures were: 0.0 (NS), -1.2 and +1.2.

Conclusion: On group level, patient and physician derived DAS scores were in practical identical. Differences between patient and physician derived DAS on the individual level corresponded to the intra-rater agreement for both patients and physicians with LOA approximating ±1. Thus, patient-derived DAS seems to have suitable reliability and may therefore substitute traditional assessment by the physician, at least when the disease activity is stable.

Ref. 1. http://www.das-score.nl/www.das-score.nl/. Accessed 17 June 2011. 2. Altman DG, Bland JM. Lancet 1986; 1: 307-10.

---

« Back to 2012 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/reliability-of-das28-in-rheumatoid-arthritis-when-based-on-patient-self-assessment-of-tender-and-swollen-joints/