Arthritis & Rheumatism, Volume 63,
November 2011 Abstract Supplement

Abstracts of the American College of
Rheumatology/Association of Rheumatology Health Professionals
Annual Scientific Meeting
Chicago, Illinois November 4-9, 2011.

Comparison of Rheumatoid Arthritis Patient Characteristics From Randomized Controlled Trials to a Registry Designed for Rheumatoid Arthritis Comparative Effectiveness Research.

Patel1,  Aarat M., Amity2,  Christine L., Frydrych2,  Lynne M., Sippel2,  Derek, Jones2,  Donald, Goudeau2,  Danielle, Eng3,  Heather

Univ of Pittsburgh Med Ctr / Children's Hospital of Pittsburgh of UPMC, Pittsburgh, PA
Univ of Pittsburgh Med Ctr, Pittsburgh, PA
Univ of Pittsburgh, Pittsburgh, PA
Brigham & Womens Hospital, Boston, MA
University of Pittsburgh, Pittsburgh, PA


The US federal government and several medical societies have recently endorsed more research targeted to comparative effectiveness studies of existing medical therapies as a way to optimize medical care, to lower health care costs and to provide research results that are more generalizable to a broader population than those typically included in randomized controlled trials (RCTs). We developed a Rheumatoid Arthritis Comparative Effectiveness Research (RACER) registry at the University of Pittsburgh Medical Center (UPMC) to provide a platform for comparative effectiveness study designs that incorporate electronic health record information, improve patient care and incorporate aspects of both randomized studies and real world observation. We determined whether there were significant differences between the characteristics of RACER subjects versus subjects from published RCTs.


RACER was established in February 2010 and consists of patients >=18 years with a rheumatologist diagnosis of RA (n=820). At each clinic visit, self-administered questionnaires including a routine assessment of patient index-3 (RAPID3), short-form-12 (SF12), work productivity and activity impairment (WPAI), DAS28, and physician and patient visual analogue scales (VAS) for arthritis-related pain and global health are collected from subjects. Blood samples for determination of C-reactive protein (CRP), RF and anti-cyclic citrullinated peptide (CCP) levels are collected. We compared RACER descriptive statistics to similar data from all RA patients seen in the UPMC health care system in 2010 (n=2,610) and to published biological therapy RCTs identified by PubMed searches. We also compared the descriptive statistics of RACER subjects initiating new therapies to subjects in RCTs.


RACER subjects had characteristics similar to RA patients evaluated throughout the UPMC system in 2010. RACER subjects that initiated new biologic or DMARD therapies (n=185) had somewhat higher disease activity scores, more work impairment and a lower physical quality of life but were otherwise highly comparable to other RACER subjects (Table 1). When compared to subjects in RCTs of biologic therapies, RACER subjects were older, had longer disease duration and lower disease activity. However, RACER subjects and subjects in RCTs had several similar characteristics including sex, race, RF positivity, functional ability (HAQ or mdHAQ) and quality of life (SF-12 or SF-36).

Table 1. Demographics of RACER Subjects (n=820)

 All RACER subjects (n=820)RACER subjects that began a new DMARD/Biologic (n=185)
Age, yrs59.8 ± 13.558.6 ± 13.4
Race Caucasian88%88%
  African Am.10%11%
Duration, yrs14.3 ± 12.414.2 ± 13.0
RA criteria Both 1987 and 201088%94%
  1987 ACR only5%4%
  2010 ACR/EULAR only4%2%
  Neither 1987 or 20102%0%
ACR/EULAR Remission17%14%
Clinical Setting  
RF, IU median6838
  % Positive77%80%
Anti-CCP, IU median4660
  % Positive76%79%
CRP, mg/dl0.450.51
DAS283.2 ± 1.33.6 ± 1.5
SDAI13.7 ± 11.817.9 ± 14.3
CDAI12.9 ± 11.417.2 ± 13.9
RAPID33.5 ± 2.24.0 ± 2.2
WPAI Activity imp.40.50%48.60%
  Time missed6.20%9.60%
  Imp. working27.60%36.10%
  Overall imp.30.60%39.30%
SF-12 MCS36.1 ± 10.745.3 ± 11.7
  PCS46.7 ± 11.134.7 ± 9.8
Medications DMARD47%40%
  DMARD + Biologic22%29%
  No DMARD or Biologic8%16%
All values are mean ± standard deviation unless otherwise written as median or percent. RACER: Rheumatoid Arthritis Comparative Effectiveness Research; ACR: American College of Rheumatology; EULAR: European League Against Rheumatism; RF: Rheumatoid factor; DAS28: disease activity score-28 joint count; SDAI: simplified disease activity index; CDAI: clinical disease activity index; RAPID3: routine assessment of patient index data 3; WPAI: Work Productivity and Activity Impairment; SF12: Short Form-12; DMARD: disease modifying anti-rheumatic Drug; MCS: mental component score; PCS: physical component score


The RACER cohort is representative of real world RA patients given the similarities between RACER subjects and non-RACER RA patients followed within the UPMC health care system. RACER subjects were typically older, had a longer disease duration and less disease activity than subjects in RCT's. This suggests that comparative effectiveness studies of different therapies tested in RACER subjects may yield different results compared to results from RCTs.

To cite this abstract, please use the following information:
Patel, Aarat M., Amity, Christine L., Frydrych, Lynne M., Sippel, Derek, Jones, Donald, Goudeau, Danielle, et al; Comparison of Rheumatoid Arthritis Patient Characteristics From Randomized Controlled Trials to a Registry Designed for Rheumatoid Arthritis Comparative Effectiveness Research. [abstract]. Arthritis Rheum 2011;63 Suppl 10 :2150

Abstract Supplement

Meeting Menu