Page 119 - THE EVOLUTION OF EARLY ARTHRITIS AND CARDIOVASCULAR RISK Samina A. Turk

Changes in inflammation markers and (components of) the risk scores over four weeks of treatment were analyzed with a paired t-test (normal distributed) or Wilcoxon test (skewed distributed). The relation between the two CV risk scores was determined with a Spearman correlation coefficient, the percentage of agreement, as well as a weighted kappa. Kappa can be interpreted as the percentage of agreement after correcting for chance(<0 indicates no agreement, 0 to 0.2 slight, 0.21 to 0.40 fair, 0.41 to 0.60 moderate, 0.61 to 0.80 substantial and 0.81 to 1.0 as almost perfect agreement)(22;23). The numbers of patients that were reassigned to another CV risk group (low, medium or high) after four weeks of treatment, according to the two risk scores were calculated, and Stuart-Maxwell analyses were applied.
To analyse the association between CV risk score and disease activity, tobit mixed model
analyses were performed. Tobit mixed model analysis can be used when the outcome
is either left- or right censored (like the maximum risk score in de CVD models)(24). The individual components of the CV risk scores were compared with disease activity, by
linear mixed model analyses and excluded patients who used antihypertensive drugs
or statins in analysis which involved blood pressure or cholesterol, respectively. In the
mixed model analyses time and the interaction between time and the independent
variable were added to assess the relationship at the different time points and all
analyses were performed separately for males and females and were adjusted for 6 age and smoking. The tobit and Stuart-Maxwell analyses were performed with Stata
(version 14), all other statistical analyses were performed using SPSS (version 21.0).
RESULTS
In total 153 patients were eligible to participate, of which 104 were included in the current analyses. Reasons not to include patients for analyses were: three patients did not reach week four, 37 patients did not start on methotrexate in combination with prednisolone, four patients dropped out before week four, three patients had no complete data at baseline and two patients did not fulfil the ACR/EULAR 2010 criteria for RA.
The mean age of the included patients was 49 years and 67% was female. The mean DAS44 was 3.5 which decreased after one month of anti-rheumatic treatment to 1.6. A history of CVD was present in seven patients (7%). The following conditions were present: one patient with a myocardial infarction, one patient with a percutaneous coronary intervention, one patient with a coronary artery bypass surgery, three patients with a cerebral vascular disease and one patient with peripheral arterial disease. No patients experienced a CV event during the first four weeks of anti-rheumatic treatment. Twenty-six (25.0%) patients smoked at baseline, and one quit smoking during these four weeks. Five patients used a statin and 16 patients used antihypertensive drugs, which did not change during follow-up (Table 1).
CHANGE IN CARDIOVASCULAR RISK
117