Arthritis & Rheumatism, Volume 60,
October 2009 Abstract Supplement
The 2009 ACR/ARHP Annual Scientific Meeting
Philadelphia October 16-21, 2009.
A Comparison Between the Simplified Erosion and Narrowing Score (SENS) and the Sharp-Van Der Heijde Score (SHS) in Assessing Radiological Damage in RA Patients During 5 Years of Follow-up
Klarenbeek1, N.B., Guler-Yuksel1, M., van der Heijde2, Désirée M.F.M., Kerstens3, P.J.S.M., Mallee4, C., Westedt5, M.L., Huizinga1, T.W.J.
LUMC, Leiden, Netherlands,
Leiden University Medical Center, Leiden, Netherlands,
JBI, Amsterdam, Netherlands,
Kennemer Hospital, Haarlem, Netherlands,
Bronovo Hospital, The Hague, Netherlands,
VUMC, Amsterdam, Netherlands
The Simplified Erosion and Narrowing Score (SENS) is an easy to learn and quick method to assess the number of joints with erosions and/or joint space narrowing on radiographs of hands and feet in patients with RA. Since SENS does not take severity of damage per joint into account, sensitivity to change may decrease over time.
To compare the reliability and sensitivity to change of SENS and SHS during 5 years of follow-up in a large number of recent onset RA patients.
Annual radiographs of year 0 to 5 of 508 patients participating in the BeSt study were scored in one session per patient by two readers independently using SHS. X-rays were scored in random order, blinded for patient identity. The SENS scores were derived from the SHS scores. Interobserver reliability (intra-class correlation coefficients (ICCs)) was calculated for different time intervals. Sensitivity to change was compared using the ratio of Standardized Response Means (SRMs) of SHS and SENS and Effect Size (ES) ratios. Progression % > smallest detectable change (SDC, calculated each time interval) between the methods were compared. Sensitivity and specificity of SENS were defined, with SHS as gold standard. Non-parametric tests were used to assess the discriminative ability between treatment arms.
Mean baseline scores for SHS and SENS were 7.1 and 3.8. The interobserver ICCs of SENS and SHS were comparable and constant over time (ICC of progression between annual SHS and baseline: 0.91, 0.91, 0.90, 0.91, 0.91 for SENS and 0.93, 0.93, 0.93, 0.91, 0.95 for SHS). Using change scores with time intervals of 1 year, the ICCs of SENS and SHS highly correlated as well, although the ICCs of both methods were lower after year 1, when there was less progression (ICCs for year 15 SHS vs SENS: 0.93 vs 0.91, 0.65 vs 0.57, 0.69 vs 0.59, 0.66 vs 0.51, 0.68 vs 0.62). SRM ratios ranged from 0.73 to 1.18 without a clear pattern over time. ES ratios showed similar results. Percentages with progression >SDC for both methods were comparable over time. Specificity and sensitivity of SENS varied between 92% and 99% for change scores between each follow-up year and baseline. For change scores with a 1 year time interval, specificity was high at all years (9499%), but sensitivity was lower with more variance over time (86%, 84%, 62%, 70%, 81% in year 15 for reader 1, 93%, 90%, 57%, 66%, 89% for reader 2). The ability to discriminate between the treatment arms was comparable in both methods.
The properties of SENS are comparable to SHS. There is no decrease in sensitivity to change over 5 years time. Therefore SENS seems a valuable method for assessing radiological joint damage in daily practice.
To cite this abstract, please use the following information:
Klarenbeek, N.B., Guler-Yuksel, M., van der Heijde, Désirée M.F.M., Kerstens, P.J.S.M., Mallee, C., Westedt, M.L., et al; A Comparison Between the Simplified Erosion and Narrowing Score (SENS) and the Sharp-Van Der Heijde Score (SHS) in Assessing Radiological Damage in RA Patients During 5 Years of Follow-up [abstract]. Arthritis Rheum 2009;60 Suppl 10 :1246