Arthritis & Rheumatism, Volume 60,
October 2009 Abstract Supplement

The 2009 ACR/ARHP Annual Scientific Meeting
Philadelphia October 16-21, 2009.

Quality Adjusted Life Expectancies in Patients with Rheumatoid ArthritisA Comparison of the EQ-5D, SF-6D and 15D

Linde1,  Louise, Sorensen2,  Jan, Ostergaard3,  Mikkel, Hetland4,  Merete L.

Department of Rheumatology, Hvidovre Hospital, Hvidovre, Denmark
University of Southern Denmark, Odense, Denmark
Department of Rheumatology, Copenhagen University Hospital at Gentofte and Hvidovre. Denmark
DANBIO, Hvidovre Hospital, Hvidovre, Denmark


Cost-effectiveness analyses of health interventions may be based on outcomes derived from patient-reported utility-based health status measures, such as the EQ-5D, SF-6D and 15D instruments. It is unclear to which degree the outcomes vary depending on the instrument used. We explored differences in index scores and quality-adjusted life expectancies (QALexp) obtained by the above-mentioned three instruments in patients with rheumatoid arthritis (RA).


Health status (EQ-5D, SF-6D, 15D), disease activity score based on 28 joint count and C-reactive protein (DAS28) and Health Assessment Questionnaire (HAQ) regarding patients with RA in routine care were collected cross-sectionally. Individuals with complete health status data were included in the analysis. Danish population based survival data were aggregated into five-year age groups and the expected survival from age 20 to 79 was calculated. The QALexp were estimated as the product of the mean index score and the expected population based survival in each age interval. Data were stratified to compare the QALexp in subgroups according to gender, DAS28 and HAQ assuming that these factors did not impact the mortality.


1689 patients were available for analysis (73% women, 76% IgM rheumatoid factor positive, mean (SD) DAS28: 3.06 (1.21), HAQ: 0.67 (0.63)). The mean index scores were: EQ-5D: 0.732, SF-6D: 0.709 and 15D: 0.874 (p<0.01 for pairwise comparisons (paired t-test)), see Figure. The differences in index scores were reflected in the estimated QALexp, see Table. The EQ-5D and SF-6D index scores and estimated QALexp were largely similar, while the 15D values were consistently higher. For all three instruments, the QALexp differed according to gender, DAS28 and HAQ subgroups. However, the magnitude of the differences varied according to the instrument used: The EQ-5D yielded the largest QALexp differences: men vs women: 2.1; DAS28<3.2 vs DAS28>=3.2: 7.8; HAQ<1 vs HAQ=1–2: 7.3; HAQ=1–2 vs HAQ>2: 6.3, while the 15D yielded the smallest differences.


Table. Quality adjusted life expectancies for a 20 year old until 79 years for the EQ-5D, SF-6D and 15D assuming mortality as for the general population.

Whole sample168940.239.048.0
DAS28 <3.270444.942.750.2
DAS28 >=3.298537.136.646.6
HAQ <196540.939.248.2
HAQ = 1–232733.633.444.6
HAQ >27227.330.441.6


We have identified methodological challenges in the interpretation of cost effectiveness analyses in RA studies. In this large study of 1689 patients investigated with both the EQ-5D, SF-6D and 15D, substantial variation in the index scores and quality-adjusted life expectancies was observed depending on the instrument used. Comparison of cost-effectiveness analyses should only be performed between studies that have applied the same instruments.

To cite this abstract, please use the following information:
Linde, Louise, Sorensen, Jan, Ostergaard, Mikkel, Hetland, Merete L.; Quality Adjusted Life Expectancies in Patients with Rheumatoid ArthritisA Comparison of the EQ-5D, SF-6D and 15D [abstract]. Arthritis Rheum 2009;60 Suppl 10 :1381
DOI: 10.1002/art.26455

Abstract Supplement

Meeting Menu