Background/Purpose

Atherosclerotic disease increases the morbidity and mortality in Rheumatoid Arthritis (RA). NSAIDs are associated with Cardiovascular (CV) risk in the general population. Data from RA cohorts are limited, but some evidence suggests that the association with CV risk may be less marked when NSAIDs are used in RA. We aimed to assess the extent of prescribing NSAIDs amongst a contemporary RA cohort and to evaluate the relationship between NSAIDs use and hospitalized cardiac events.

Methods

We collected data using the electronic medical record (EMR) at King’s College Hospital, United Kingdom, for routine clinical care which contains detailed information on disease as well as drug exposure.

Patients with a consultant diagnosis of RA were classified into regular NSAID users and non-users. Low dose aspirin was not considered an NSAID. Cox-II inhibitors were combined with NSAIDs for this analysis.

Disease activity was analyzed using the mean DAS over the duration of follow up. Patients entered follow up from either March 2008 (when the EMR current coding system was established) or date of diagnosis. Follow up was censored at date of CV event or October 2013 whichever occurs first. CV events (hospitalization with myocardial infarction, angina, heart failure, TIA or stroke) were identified using the hospital electronic coding system and subsequently validated by a physician (AG).

Cox proportional hazard modeling was used to compare rates. The potential influence of confounding due to age, disease severity or existing IHD was explored using multivariate regression.

Results

The clinical characteristics of the population are presented in table 1. In total, 1089 patients with RA were under follow up during the time period available for analysis. The characteristics are typical of a UK cohort with RA, with a mean age 55 years, a female predominance (78%) and mean DAS28 score of 4.0 and a moderate level of disability as measured by HAQ (median 1.625). 28% of the cohort were treated for hypertension. The median follow up time in whole cohort was 3.7 years.

 In our practice NSAIDs use is not associated with a significantly increased rate of CV events.  Inclusion of PRN users did not influence estimates.

Exclusion of naproxen raised adj HR to 1.2 (0.54, 2.45). Naproxen use accounted for 24% of NSAIDs prescriptions.

Conclusion

These data are reassuring about the use of NSAIDs in RA in our practice, as these drugs remain an important adjunct to therapy. This approach to analysis using the EMR opens doors to future research, highlighting the power of routine captured data. Important limitations include unmeasured confounding, challenges of potential misclassification as well as left and right censorship. Awareness of these issues can help drive innovative solutions to make the data more robust for future analyses.

 

Table 1

 

Characteristic	All patients n=1089	No NSAIDs n=866	NSAIDs n=223	p value
Age, mean (SD)	55 (16)	55 (16)	53 (15)	0.044
Gender, n (%) female	847 (78)	676 (78)	171 (77)	0.659
Average DAS, mean (SD)	4.0 (1.3)	3.9 (1.3)	4.4 (1.3)	<0.001
Average HAQ, median (IQR)*	1.625 (0.875, 2.125)	n/a	n/a	n/a
Disease duration in years, median (IQR)	7 (4, 11)	6.9 (4, 11)	7.3 (4, 11)	0.874
Methotrexate use, n (%)	614 (56)	491 (57)	123 (55)	0.679
Biologic use, n (%)	132 (12)	107 (12)	25 (11)	0.640
Treated hypertension, n (%)	301 (28)	249 (29)	52 (23)	0.106
Exposure time (person years)	4633	3751	882
Number of CV events	63	53	10
Incidence CV events / 1000 person years (95% CI)	12.1 (9.5, 15.5)	12.6 (9.5, 16.5)	10.0 (4.8, 18.5)
Unadjusted hazard ratio (95% CI)	n/a	ref	0.77 (0.39, 1.51)
Adjusted hazard ratio (95% CI)*	n/a	ref	0.78 (0.36, 1.68)

*Missing data for HAQ present preclude meaningful analysis across cohort

---

« Back to 2014 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/cardiovascular-risk-with-nsaids-in-rheumatoid-arthritis-an-analysis-using-routinely-collected-data/