Article Text

AB0195 To develop a regression model for predicting damage-related haq: a nationwide study based on the ninja (national database of rheumatic diseases by ir-net in japan) 2011
  1. S. Nishiyama1,
  2. K. Ohashi1,
  3. T. Aita1,
  4. Y. Yoshinaga1,
  5. S. Miyawaki1,
  6. J. Nishino2,
  7. S. Tohma3,
  8. NinJa group
  1. 1Rheumatic Disease Center, Kurashiki Medical Center, Kurashiki
  2. 2Orthopaedics and Rheumatology, Nishino Clinic, Tokyo
  3. 3Department of Rheumatology, Clinical Research Center for Allergy and Rheumatology, Sagamihara National Hospital, National Hospital Organization, Sagamihara, Japan


Background HAQ is one of evaluation methods for physical function and is thought to consist of activity HAQ (ACT_HAQ) and damage-related HAQ (DAM_HAQ)1.

Objectives To search for important factors of ΔDAM_HAQ and to develop a regression model for predicting DAM_HAQ in patients with rheumatoid arthritis (RA).

Methods Data of RA patients excluding surgical case were extracted from NinJa, an annually updated nationwide database in Japan. Correlation between ΔHAQ and ΔSDAI was calculated using 177 early RA patients (stage I and disease duration ≤1 yr) who were registered both 2010 and 2011, where ΔHAQ was approximately equal to ΔACT_HAQ. The result was ΔACT_HAQ/ΔSDAI≈ΔHAQ/ΔSDAI=0.0235. According to previous report1, we calculated ACT_HAQ=ΔACT_HAQ/ΔSDAI×SDAI=0.0235×SDAI, and DAM_HAQ=HAQ-ACT_HAQ using data of 2193 patients. Multiple regression analysis was performed to obtain a regression model for predicting ΔDAM_HAQ[2010,2011] (=DAM_HAQ2011-DAM_HAQ2010). The model was validated using 844 sequentially registered data from 2006 to 2011.

Results ΔDAM_HAQ did not differ between male and female patients. Significant high ΔDAM_HAQ was observed in patients with older, longer disease duration, progressive stage and class, higher TJC28 and SJC28, larger VAS of physician and patient assessment, and high level of CRP. MTX users and boolean remission achievers had significant low ΔDAM_HAQ. Multiple regression analysis revealed that age, stage, class, TJC28, SJC28, patient assessment, CRP, MTX user or not, and boolean remission state were effective predictors for ΔDAM_HAQ. While DAM_HAQ2006+Σpredicted ΔDAM_HAQ[t, t+1] where t=2006 to 2010 (A) was overestimated DAM_HAQ2011, DAM_HAQ2006+Σ10% predicted ΔDAM_HAQ[t, t+1] (B) showed close correlation with DAM_HAQ2011 (figure).

Conclusions ΔDAM_HAQ was affected by age, stage, class, TJC28, SJC28, patient assessment, CRP, MTX user or not, and boolean remission. One tenth of predicted ΔDAM_HAQ may be irreversible and accumulated on the former DAM_HAQ.

  1. Smolen JS, et al. Estimation of a numerical value for joint damage-related physical disability in rheumatoid arthritis clinical trials. ARD 2010;69:1058-64.

Disclosure of Interest None Declared

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.