Background Despite the importance of the Health Assessment Questionnaire (HAQ) and Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) in assessing patient-reported functional status and disease activity, they have been critiqued for being time-consuming, not convenient on a daily-basis and thus not contributing to decisions in routine care.
Objectives The aim of this analysis was to describe the correlation of individual HAQ and BASDAI questions with patient and physician reported measures used in AS and to examine whether the instruments could be reduced to better reflect routine clinical practice.
Methods BioTRAC is an ongoing prospective registry of patients initiating infliximab or golimumab. Data from AS patients treated in 2005-2014 were used. The correlation of individual HAQ and BASDAI questions with patient (pain, BASDAI, HAQ and BASFI) and physician (MDGA) reported measures was described with the Pearson's correlation coefficient. The impact of each question on the need for help in each HAQ domain was assessed with logistic regression. Factor analysis was used to assess the variability due to each individual question in HAQ and BASDAI.
Results A total of 413 AS patients with 1660 BASDAI and 1654 HAQ assessments were included. HAQ and BASDAI questions correlated at different extents with each AS measure. Questions related to “eating” and “gripping” showed the lowest correlation with patient and physician reported measures. All HAQ questions had higher correlations with patient reported measures than with MDGA. The BASDAI question on “fatigue and tiredness” showed the highest correlation with BASFI, while the question on “other joints pain/swelling” showed the lowest correlation with MDGA. None of the HAQ and BASDAI questions were associated with needing help for eating. All other HAQ individual questions were significantly associated with the need for help within their corresponding category, with the exception of Q5C and Q7A. BASDAI question on level of discomfort was significantly associated with the need for help in all HAQ categories, with the exception of “eating” and “walking”. Q2A and Q7C accounted for 59.6% of the HAQ variance. The level of morning stiffness accounted for 73.8% of the BASDAI variance. When combining the HAQ and BASDAI, Q2A and Q3A from HAQ and Q1 from BASDAI accounted for 63.5% of the variance.
Conclusions Variability exists in the correlation of HAQ and BASDAI questions with patient and physician reported AS measures. The results suggest that “standing up straight from an armless chair” and “turning faucets on/off” are the main drivers of HAQ, while the level of morning stiffness drives the BASDAI. Three questions were found to drive the combined HAQ and BASDAI which may have implications in the design of self-report instruments.
Disclosure of Interest P. Rahman Consultant for: Abbott, AbbVie, Amgen, BMS, Celgene, Janssen, Novartis, Pfizer, Roche, M. Zummer: None declared, W. Olszynski: None declared, M. Khraishi: None declared, D. Sholter: None declared, R. Faraawi Consultant for: Janssen, W. Bensen Consultant for: Janssen, M. Baker: None declared, A. Chow: None declared, J. Vaillancourt Employee of: JSS Medical Research, J. Sampalis Shareholder of: JSS Medical Research, F. Nantel Employee of: Janssen, S. Otawa Employee of: Janssen, A. Lehman Employee of: Janssen, C. Tkaczyk Employee of: Janssen, M. Shawi Employee of: Janssen