Impact of Rheumatoid Arthritis on Recognition of Hypertension in a Medically Homed Population

Background/Purpose: Numerous studies report increased cardiovascular disease (CVD) events and others describe increased arterial stiffness in patients with rheumatoid arthritis (RA). Still, hypertension diagnosis rates in RA reports are often lower than expected by age.   We tested the hypothesis that RA is a risk factor for missed hypertension diagnosis, given the importance of hypertension for CVD risk.

Methods: Using a cohort design we studied all medically homed adult patients from a large multispecialty practice who met Joint National Committee-7 (JNC-7) hypertension diagnostic criteria but lacked baseline diagnosis or treatment to compare new recognition of hypertension in patients with and without RA. “Medically homed” definitions required ≥2 primary care visits over ≥12 months (2009-11), and RA/inflammatory arthritis algorithms required two ICD-9 claims of 714 in 24 months. Cox proportional hazard modeling was used to examine the impact of RA on hypertension recognition.

Results: Among 33,947 medically homed patients with baseline undiagnosed and untreated hypertension, 575 patients had RA codes. After an average of 14 months follow up, 49% of RA patients compared to 42% without RA remained undiagnosed. RA patients had equal annual primary care visits (mean 1.5 v. 1.6), more total provider visits (mean 7.5 v. 4.6) and longer mean observation time, yet were less often diagnosed. In multivariate modeling controlling for socio-demographic factors, comorbidity, and utilization factors, the presence of rheumatoid arthritis decreased the likelihood of hypertension diagnosis or treatment by 21% [Hazard Ratio 0.79, Confidence Interval 0.71-0.89].

        Hypertension recognition was lower in RA than other comorbidities (Table 1), and contrasted with increased hypertension recognition in patients with diabetes or kidney disease.  RA patients were older (mean 62 v. 57 years), more female, weighed less, and were more likely to see an internist for primary care, and in the final adjusted model, younger age, white race, Medicaid, and non-internal medicine primary care provider predicted lower hypertension recognition.

Conclusion: In this sample of medically homed patients meeting JNC-7 criteria for hypertension those with RA were 21% less likely to be diagnosed or treated despite more total visits,  compared to those without RA. Given that both hypertension and RA increase cardiovascular risk, rheumatologists may need to actively help to improve hypertension recognition to modify CVD risk.