Session Information
Session Type: Abstract Submissions (ACR)
Background/Purpose: To enable consistent assessment of impact of tapering, withdrawal and dose optimization strategies, there is an increasing need for validated measures of flare in rheumatoid arthritis (RA) clinical studies. Several DAS28 based flare criteria have been described, but none validated
Methods: We used 3 longitudinal observational databases that included treatment withdrawal or change in relation to RA worsening to test 6 previously published DAS28 based flare criteria on fulfilment of 5 hypotheses concerning criterion and construct validity. Published DAS28 based flare criteria were: 1] an increase in DAS28>1.2, or >0.6 if current DAS28>5.1, 2] an increase in DAS28>1.2, or >0.6 if DAS28≥3.2, 3] an increase>0.6 or DAS28>3.2, 4] an increase in DAS28>1.2, 5] DAS28 >3.2, 6] DAS28 >2.6. The 5 hypotheses used to assess validity were: A+B) Sensitivity and specificity >70% compared to patient’s/physician’s judgment of RA worsening assessed with a transition question, and C) difference in proportion with DMARD/corticosteroid initiation/increase >0.2, D) difference in mean CRP change >10mg/L, and E) no statistical difference in SF36 Mental Health (MH) subscale change in patients fulfilling versus not fulfilling the flare criteria. Sensitivity/specificity, Chi square and two sample student’s T test were done
Results: Analyses included 51, 147 and 744 RA patients in the 3 studies. Two studies included patients treated with infliximab and the larger study (NOR-DMARD) included patients initiating a new synthetic or biologic DMARD. Baseline characteristics are described in Table 1. Criterion 2 (an increase in DAS28>1.2, or >0.6 if DAS28≥3.2) fulfilled most predefined hypotheses (4 out of 5, Table 2). Sensitivity and specificity for criterion 2 varied between 63 – 78% and 84 – 92%, respectively. Construct validity was demonstrated with 23% more treatment change, a higher mean DCRP (11.4 mg/L) and a difference in DSF-36 MH of only -5. Criteria 3, 5 and 6 tended to be more sensitive, criteria 1, 2 and 4 more specific
Conclusion: An increase in DAS28>1.2, or >0.6 if DAS28≥3.2 appears most discriminating and valid by our predefined criteria. The differences in sensitivity and specificity between the various DAS28-based flare criteria may be of importance for selection of flare criteria for specific studies. Further assessment, with evaluation of impact relative to levels of worsening, in additional databases may refine criteria
Table 1 Baseline characteristics – mean (SD) unless otherwise noted
|
Database |
1 |
2 |
3 |
|
Number of patients |
51 |
147 |
744 |
|
Age, years |
59 (11.2) |
58 (12) |
56 (13.5) |
|
Female, No (%) |
29 (57) |
101 (69) |
539 (72) |
|
Disease duration in years |
14 (7.5) |
11 (7) |
6.4 (9.5) |
|
RF positive, No (%) |
42 (82) |
117 (81) |
496 (68) |
|
Anti-CCP positive, No (%) |
37 (73) |
95 (69) |
146 (70) |
|
DAS28 at inclusion |
2.5 (0.7) |
3.5 (1.3) |
5.2 (1.1) |
|
DAS28 at 3 months after inclusion |
|
|
3.4 (1.2)* |
|
No. of previous DMARDs, median [p25-p75] |
3 [2-3] |
3 [2-3] |
0 [0-2] |
Table 2 Fulfilment of DAS28 based flare criteria on 5 hypotheses regarding construct and criterion validity
|
|
Criterion validity: databases 1 and 2 |
Construct validity: database 3 |
|||||
|
Flare criteria |
Hypothesis 1 |
Hypothesis 2 |
Hypothesis 3 |
Hypothesis 4 |
Hypothesis 5 |
||
|
In patients classified as a having a flare: |
Sens/spec patient >70% compared to transition scale |
Sens/spec physician >70% compared to transition scale |
Higher proportion of DMARD/corticosteroid change (>0.2) |
Higher CRP (>10 mg/L) |
No change in depression (≤5 points in MH scale) |
||
|
|
Sens* % |
Spec* % |
Sens* % |
Spec* % |
Proportion difference** |
CRP difference mg/L (SE)*** |
MH difference (SE) |
|
1) ΔDAS28 > 1.2 or > 0.6 if DAS28>5.1 |
46/56 |
95/92 |
53/78 |
95/92 |
0.28 |
13.1 (2.2) |
-6.1 (1.4) |
|
2) ΔDAS28> 1.2 or > 0.6 if DAS28>3.2 |
69/63 |
92/84 |
73/78 |
92/86 |
0.23 |
11.4 (1.7) |
-5.0 (1.2) |
|
3) ΔDAS28> 0.6 or a DAS28>3.2 |
98/94 |
70/61 |
100/89 |
67/60 |
0.16 |
3.7 (1.2) |
-2.8 (1.0) |
|
4) ΔDAS28>1.2 |
46/56 |
96/93 |
51/78 |
95/92 |
0.27 |
13.0 (2.3) |
-6.6 (1.4) |
|
5) reaching DAS28>3.2 |
91/88 |
78/67 |
91/89 |
76/68 |
0.18 |
3.1 (1.2) |
-2.6 (1.0) |
|
6) reaching DAS28 > 2.6 |
98/94 |
55/46 |
100/89 |
53/47 |
0.13 |
2.4 (1.0) |
-3.2 (1.1) |
|
*Sens=sensitivity, Spec= specificity. Results from database1 / database2 are represented on sensitivity and specificity. ** All proportion differences are statistically significant, with p<0.0001.*** All differences in CRP are statistically significant, with at least p<0.05 |
|||||||
Disclosure:
A. van der Maas,
None;
E. Lie,
None;
R. Christensen,
None;
E. Choy,
None;
Y. A. de Man,
None;
P. L. C. M. van Riel,
None;
T. G. Woodworth,
None;
A. A. den Broeder,
None.
« Back to 2012 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/construct-and-criterion-validity-of-several-proposed-das28-based-rheumatoid-arthritis-flare-criteria-a-cohort-validation-study/