Article Text

SAT0149 Performance of the Framingham Cardiovascular Risk Prediction Model with and without CRP in RA Patients: Analysis of UK Clinical Practice Research Data
  1. H. Cawston1,
  2. E. Alemao2,
  3. F. Bourhis1,
  4. P. Hines3,
  5. T. Le3,
  6. M. Al4,
  7. M. Rutten-van Molken4,
  8. D. Solomon5
  1. 1OptumInsight, Nanterre, France
  2. 2Bristol-Myers Squibb, Princeton
  3. 3Bristol-Myers Squibb, Hopewell, United States
  4. 4Erasmus University, Rotterdam, Netherlands
  5. 5Brigham and Women's Hospital, Boston, United States


Background Epidemiological studies indicate that RA pts have a 1.5–2.0-fold higher risk of acute myocardial infarction and a 1.4–2.7-fold higher risk of stroke. There is growing evidence that the chronic high-grade inflammation present in RA pts contributes to development of premature atherosclerosis and cardiovascular (CV) events.1

Objectives To evaluate the role of CRP and total cholesterol (TC) in predicting CV risk in RA pts and whether adding CRP to risk models enhances prediction.

Methods We conducted a retrospective cohort analysis, using the Clinical Practice Research Datalink GOLD and Hospital Episode Statistics data. The RA population was defined as all pts presenting with at least one RA read code after 1 January 1988 (index code). Pts were required to have at least 12 months of data reported before the first RA code, free of CV events (primary prevention) and to have no psoriatic arthritis-related codes over the entire period. Onset of disease was defined as date of first RA-related code. CV events were defined based on the Framingham Heart Study (FHS) definition for total CV event. Kaplan–Meier (KM) curves of time to first CV event by baseline CRP tertiles (≤6, 6–20, >20 mg/L) and TC levels (<200, 200–240, >240 mg/dL) were tested using log-rank test (LR). Negative binomial models were used to evaluate the risk of CV events by CRP and TC levels. Effect of CV risk factors was evaluated by multivariate Cox models for primary prevention based on the FHS.2 The role of CRP on CV risk was evaluated by assessing the discriminative properties of the model without versus with CRP using the area under the receiver operating characteristic curve (Δ AUC) via C-Index and net reclassification improvement (NRI).

Results 4772 primary prevention RA pts met the inclusion criteria. On average (SD) at baseline, pts were 58.3 (14.4) yrs old, had CRP of 23.5 (33.5) mg/L and follow-up of 3.8 (2.8) yrs (18,281 pt-yrs). In the sample, 71% were female, 40% were hypertensive and 34% dyslipidaemic. The incidence of CV events was 37.47 per 1000 pt-yrs (95%CI: 34.66, 40.28). Based on the KM analysis, probability of a CV event at 5 yrs in pts with low, medium and high CRP levels were 6.58, 7.21 and 12.22%, respectively (LR p<0.001). The probability of a CV event at 5 yrs in pts with low, medium and high TC levels were 9.99, 9.62 and 12.69%, respectively (LR p=0.364). Within each TC category, the risk of a CV event increased with the CRP level (Figure). In the multivariate Cox model with FHS risk factors, the addition of log(CRP) was positively associated with CV events (hazard ratio=1.12; 95%CI 1.02,1.23). The addition of CRP to the FHS improved the model discrimination with Δ AUC based on C-Index of 0.003 and NRI of 3.2%.

Conclusions CRP has a CV risk-modifying effect on TC in RA pts. After controlling for traditional CV risk factors, baseline CRP in primary prevention RA pts is associated with increased risk of CV events. The inclusion of CRP in the FHS equation moderately improves the reclassification of CV risk.


  1. Peters MJ et al. Ann Rheum Dis 2010;69:325–31; 2. Wilson PWF et al. Circ Cardiovasc Qual Outcomes 2008;1:92–97

Disclosure of Interest H. Cawston Employee of: OptumInsight, E. Alemao Shareholder of: BMS, Employee of: BMS, F. Bourhis: None declared, P. Hines Employee of: BMS, T. Le Employee of: BMS, M. Al: None declared, M. Rutten-van Molken: None declared, D. Solomon: None declared

DOI 10.1136/annrheumdis-2014-eular.1736

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.