Baseline Disease Activity Determines the Level of Improvement Needed to Experience a Minimal and Major Response in Rheumatoid Arthritis

Purpose. Patient reported outcomes are important anchors for the definition of improvement in patients with rheumatoid arthritis (RA). To analyze the patient-based minimal clinically important improvement (MCII) and major response (MajR) of disease activity measures in RA in patients with different baseline disease activity.
Methods. We used data from a Norwegian observational database of 1286 patients with RA (73% female; 65% RF+; RA duration of 7.8±9.3 years), who were newly prescribed a disease modifying antirheumatic drug (DMARD). We calculated the changes in the Disease Activity Score 28 (DAS28), the Simplified Disease Activity Index (SDAI), and the Clinical Disease Activity Index (CDAI) from DMARD start to the 3 months follow up visit. Then patients indicated if their condition had considerably improved, had improved, was unchanged, had worsened, or considerably worsened. We used Receiver Operating Characteristic (ROC) curve analyses to determine MCII (discrimination of “unchanged” from “improved or considerably improved”) and MajR (discrimination of “unchanged or improved” from “considerably improved”). We used the bootstrap method to optimize accuracy of cutpoints (100 resamples of 50% of the patients). This analysis was repeated for patients with different baseline disease activity states (low, moderate, and high disease activity; LDA, MDA, HDA) using the specific cutpoints for the three scores.
Results. Disease activity improved significantly during the 3 month period (ACR20/50/70 responses: 38%, 17%, and 6%). When analyzing MCII and MajR, there was a highly significant association of patients’ perception of improvement with baseline disease activity. This was seen for all indices (Table).

Table. Minimal clinically important improvement and major improvement of different scales by baseline disease activity state.

SDAI RESPONSE	All patients (n=1286)	LDA* (n=103)	MDA* (n=531)	HDA* (n=647)
Minimal response**	12.6	1.4	7.6	20.2
Major Response**	18.1	3.3	11.3	25.9
CDAI RESPONSE	All patients (n=1286)	LDA* (n=104)	MDA* (n=468)	HDA* (n=707)
Minimal response**	11.5	1.3	7.5	17.3
Major Response**	16.6	3.2	10.3	22.1
DAS28 RESPONSE	All patients (n=1286)	LDA* (n=71)	MDA* (n=546)	HDA* (n=614)
Minimal response**	1.3	0.6	1.2	1.7
Major Response**	1.9	1.3	1.6	2.5

* Cutpoints for the SDAI were 3.3 < LDA ≤ 11 < MDA ≤ 26 < HDA; for the CDAI they were 2.8 < LDA ≤ 10 < MDA ≤ 22 < HDA; for the DAS28 they were 2.6 < LDA ≤ 3.2 < MDA ≤ 5.1 < HDA.
** Response cutpoints are estimated from repetitive ROC curves each estimating the most sensitive cutpoint that still provides an 80% specificity for the state.

Conclusion. The degree of disease activity improvement needed to experience clinical response changes considerably in relation to baseline disease activity. Thus, single cutpoints for response might not relate to a similar relevant changes in disease activity in all patients.

 D. Aletaha, None; J. Funovits, None; J.S. Smolen, None; T.K. Kvien, None.