Article Text

Download PDFPDF

Cardiovascular disease assessment in rheumatoid arthritis: a guide to translating knowledge of cardiovascular risk into clinical practice
  1. Anne Grete Semb1,
  2. Silvia Rollefstad1,
  3. Piet van Riel2,
  4. George D Kitas3,4,
  5. Eric L Matteson5,
  6. Sherine E Gabriel5
  1. 1Department of Rheumatology, Preventive Cardio-Rheuma Clinic, Diakonhjemmet Hospital, Oslo, Norway
  2. 2Scientific Institute for Quality of Healthcare, Radboud University Nijmegen, Nijmegen, The Netherlands
  3. 3Department of Rheumatology, Dudley Group NHS Foundation Trust, Dudley, UK
  4. 4Department of Rheumatology, Arthritis Research UK Epidemiology Unit, University of Manchester, Manchester, UK
  5. 5Division of Rheumatology and Department of Health Sciences Research, Mayo Clinic College of Medicine, Rochester, Minnesota, USA
  1. Correspondence to Dr A G Semb, Preventive Cardio-Rheuma Clinic, Department of Rheumatology, Diakonhjemmet Hospital, PO Box 23 Vinderen, Oslo NO-0319, Norway; a-semb{at}, anne.semb{at}


As physicians we like to have evidence for making decisions about interventions to improve health. The evidence vacuum in the field of cardiovascular disease (CVD) prevention and clinical outcome in patients with rheumatoid arthritis (RA) has received vigorous attention in the recent literature. There is broad agreement that a patient with RA fulfilling the criteria established for the general population on CVD risk reduction should receive proven interventions, including smoking cessation, weight reduction, blood pressure control and lipid-lowering therapy. In accordance with these recommendations, and despite all the uncertainties about CVD treatment threshold, targets and outcome results in RA, we firmly advocate that CVD risk should be assessed and acted on in patients with RA as recommended for the general population, even while educational CVD-preventive programmes are being developed and hard CVD end point studies are undertaken in this patient population. The initial strategies for implementing CVD risk evaluation will necessarily be modest at first. There are several possible strategies for collection of data that can be incorporated into the daily routine during rheumatology consultations at outpatient clinics. We recommend starting with these simple procedures:

1. CVD risk factor recording and evaluation using risk calculators available for the general population

2. Referral of patients with high CVD risk to a primary care physician or a cardiologist skilled in this subject for follow-up

3. Providing information about excess CVD risk and how to modify it to the patients as major stakeholders

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.


The increased risk of morbidity and mortality due to cardiovascular disease (CVD) in patients with rheumatoid arthritis (RA) has been known for decades.1 CVD is a major factor in the widening mortality gap between patients with RA and the general population.2 There is no longer doubt that the increased CVD risk in RA equals that of patients with diabetes mellitus.3

Although there is a large body of knowledge on the increased CVD risk in RA, there is a lack of clinical evidence on management of this increased risk. Important and unaddressed questions include: Are the CVD-preventive recommendations for the general population also adequate for patients with RA? Should the traditional risk factors be tailored specifically to patients with RA, with modified treatment thresholds and targets? Are there subgroups of RA patients with excess or lower CVD risk who should be identified?

As physicians we want to act on evidence when making decisions about interventions to improve health. The evidence vacuum in the field of CVD prevention and clinical outcome in patients with RA has been well discussed in a recently published debate article.4 Interestingly, there is a broad agreement that patients with RA fulfilling general population criteria for recommended CVD risk reduction should receive proven interventions such as smoking cessation, weight reduction, blood pressure (BP) control and lipid-lowering (LL) therapy. In accordance with these recommendations, and despite all the uncertainties about CVD treatment thresholds, targets and outcome results in RA, we firmly advocate CVD risk to be assessed and acted on in patients with RA as recommended for the general population.

Several studies have shown that, in daily clinical practice, the CVD risk in patients with RA is underestimated.5–9 In part, this relates to underuse of tools for systematic evaluation of CVD risk and prevention in these patients. Management of CVD risk in patients with RA is an interdisciplinary task involving cardiology, rheumatology and primary care, and there is confusion about which healthcare providers should be responsible for recording and evaluating CVD risk factors, modifying lifestyle-related risk factors, and initiating appropriate medical interventions.

The implementation of theory into clinical practice is a huge task and must necessarily address areas of uncertainty. This viewpoint therefore focuses on the practicalities of CVD risk factor evaluation, including recording, assessment, evaluation and awareness in patients with RA.


RA is a chronic disabling disease, and patients are often mainly in contact with the health system in relation to their joint disease, with frequent consultations with their rheumatologist, but not with other healthcare providers.5 Therefore, rheumatology clinics may be an ideal location in which to record and evaluate CVD risk. However, most rheumatologists may not feel confident in providing such assessments.

The increased CVD risk in RA is partly related to the high frequency of traditional CVD risk factors, such as hypertension, cigarette smoking, insulin resistance/diabetes, physical inactivity and obesity.1012 The CVD risk imposed by RA is both under-recognised and underassessed in primary care.5 ,6 Furthermore, a recent study reported that 30% of RA patients with comorbid CVD, diabetes or hyperlipidaemia consult a primary care physician (PCP) less than once a year, despite frequent visits to other providers such as rheumatologists.5

There is a clear need for the implementation of strategies for CVD risk factor information, gathering and recording. This is a key challenge, particularly in the rheumatology setting. A systematic approach using healthcare professionals other than doctors; for example nurse specialists, is likely to be the most successful way to implement new systems in clinical practice.13 ,14 Indeed, Ikdahl and coworkers argue that a systematic, team-based model for data collection may be necessary to optimise recording of CVD risk factors in patients with RA attending a rheumatology outpatient clinic.7 In an electronic patient journal, rheumatologists, nurses and secretaries were interlinked in CVD risk factor recording. When all traditional CVD risk factors were logged into the electronic journal, a CVD risk score15 was calculated automatically, based on the European Systematic Coronary Risk Evaluation (SCORE) algorithm. This process facilitated CVD risk evaluation for the rheumatologist and assisted decision-making about indications for CVD prevention. Appropriate software could arguably alleviate the monitoring and retrieval of CVD risk information in patients with RA, reduce costs and improve the quality of CVD risk factor recording. We argue that CVD risk factor gathering and recording most likely will and can be carried out successfully by an interdisciplinary team focused on the rheumatology patient.

CVD risk evaluation

In the latest European Society of Cardiology (ESC) guidelines for CVD prevention,16 immunological diseases such as RA were for the first time mentioned and categorised as a CVD risk factor, and have similarly been included in the latest version of the British CVD risk calculator, the QRISK II.17 The diagnosis of RA alone is not an indication for LL treatment, but is recommended to be taken into consideration as part of the CVD risk evaluation. The European League Against Rheumatism (EULAR) recognises that patients with inflammatory joint diseases (IJDs) have increased risk of CVDs, and made recommendations for management of CVD risk factors based largely on expert consensus.18 The EULAR task force group was well aware of the lack of evidence and made a conservative recommendation for a multiplication factor of 1.5 to the calculated CVD risk by SCORE if the RA patient had two or more of the following: disease duration >10 years, rheumatoid factor/anticitrullinated protein antibody positivity, or extra-articular manifestations. Corrales and colleagues report that using the 1.5 multiplication factor reclassifies only 0.03% of the patients into more appropriate CVD risk classes, but even then patients at high risk and those with asymptomatic carotid atherosclerosis were not adequately identified.19 Thus, the multiplication factor of 1.5 does not adjust for the increased CVD risk in RA. Further evaluation of an accurate CVD risk assessment is needed such as, for example, an RA-specific risk calculator, which is currently being developed by A TransAtlantic Cardiovascular Risk Calculator for Rheumatoid Arthritis (ATACC-RA) Consortium.

The SCORE, QRISK II, Framingham and Reynolds CVD risk calculators underestimate future CVD events in patients with RA, except for those in the highest-risk classes.20 ,21 The underestimation of CVD risk may have several reasons related to RA disease-specific factors such as systemic inflammation and the lipid paradox. Those patients with RA who are at highest CVD risk have the lowest lipid levels, which may be low because of the level of inflammatory burden, as reflected by the erythrocyte sedimentation rate and C reactive protein.22 Imputing low lipid levels into a CVD risk calculator will underestimate risk. Although the total cholesterol/high-density lipoprotein cholesterol ratio has been shown to be less susceptible to change with inflammation, its use as a component of the standard risk calculators is still associated with underestimation of CVD events in patients with RA.20 ,21 Another reason for underestimation of CVD risk in RA may be the high frequency of asymptomatic atherosclerosis,23–26 which is easily visualised on ultrasound of the carotid arteries. In addition, carotid artery plaques in patients with RA have been shown to be predictors of a future acute coronary syndrome.27 Moreover, in the recent European guidelines on CVD prevention,16 the presence of carotid artery plaques is considered a CVD equivalent. This suggests that including ultrasound of carotid arteries in the CVD risk evaluation would increase the proportion of correct risk stratification,25 ,28 ,29 which may not be feasible in rheumatology outpatient clinics currently, but will be a great advantage if included as a risk evaluation tool in the future. In conclusion, available CVD risk calculators developed for the general population seriously underestimate CVD risk in patients with RA, but currently represent the best utility for CVD risk evaluation in these patients.

CVD prevention

In both Europe and the USA, patients with RA are significantly less likely to receive CVD-preventive measures compared with the general population.30 Further, despite similar treatment, patients with RA who had a myocardial infarction had worse long-term outcomes than those without RA.31 This deficiency in CVD risk management is amply confirmed in an investigation of CVD risk factor control in 836 patients with RA by Primdahl and colleagues.32 Among the most striking observations is that, in the 644 patients without established CVD or diabetes, inadequate BP and lipid control was documented in 35.8% and 55.4%, respectively. Findings from a preventive cardio-rheumatology clinic revealed that 63.4% of the patients referred were in need of CVD-preventive medical intervention.26

Compounding the uncertainties surrounding management of CVD risk, there is no evidence on CVD-preventive treatment effects in patients with RA. To date, there is a lack of published prospective randomised controlled trials in RA comparing either different primary prevention strategies (LL or antihypertensive medications) or the effects of the various antirheumatic drugs using CVD outcomes as the primary end point. Such trials require a large number of patients monitored for lengthy periods and are costly. Promising results from a post hoc analysis from two large statin trials with CVD end points showed that there were comparable effects of statins on lipid reduction and CVD events in patients with and without IJDs, including RA.33 A large randomised controlled trial on statin versus placebo (Trial of Atorvastatin for the Primary Prevention of CV Events in patients with Rheumatoid Arthritis (TRACE-RA) including >3000 patients with RA) was recently prematurely terminated because of the low incidence of the primary CVD end point and it did not have adequate power to reach a clear conclusion.

An important CVD risk factor in the general population is cholesterol level, which appears to be less important for patients with IJDs.22 However, meta-analyses of clinical trials have shown that patients in the general population benefit from LL treatment regardless of baseline lipid levels.34 Clinical experience of use of CVD-preventive medications for IJDs is scarce. Experience from a preventive cardio-rheumatology clinic for patients with IJDs revealed that lipid targets were attained using fewer than three consultations in 90% of the patients without serious adverse events.26 These results indicate that treatment of RA patients with statins and achieving recommended lipid goals is safe and uncomplicated. Prospective, longitudinal data are needed to demonstrate whether patients with RA need to achieve the same low lipid targets as people without RA, or if a relatively smaller or larger reduction in lipids will result in the same protection against CVD as in persons without RA.

Modern RA therapy and the multiple comorbidities of the disease lead to polypharmacy, thus increasing the possibility of drug interactions with CVD-preventive medications such as statins. Interaction between antirheumatic treatments such as non-steroidal anti-inflammatory drugs (NSAIDs) and synthetic or biological disease-modifying antirheumatic drugs (DMARDs) with CVD-preventive medications is also an area of uncertainty. Inconclusiveness exists regarding the influence of statins on clinical response and B-cell depletion after rituximab treatment in RA.35 ,36 There is a recognised increased CVD risk associated with NSAID use in the general population,37 ,38 but in a recent Danish nationwide study the CVD risk associated with use of NSAIDs in patients with RA was modest and significantly lower than in non-RA individuals.39 Furthermore, NSAIDs interfere with the thrombocyte-inhibitory effect of aspirin, which complicates secondary CVD prevention.40 There are indications from observational studies that use of synthetic and biological DMARDs is associated with reduced CVD risk in RA.41 ,42 It is not known whether this effect is related to the reduction in inflammation opposing the increased levels of atherogenic lipids occurring with biological DMARDs and improvement of high-density lipoprotein cholesterol function.43 Studies to clarify these issues are needed.

CVD prevention in patients with RA is fraught with concerns of drug- and disease-related complexities. Therapies for rheumatic conditions have evolved enormously in the last two decades, as have CVD-prevention strategies. The available literature suggests substantial underuse of CVD-preventive pharmacotherapy in the general population44 ,45 According to the latest DYSIS report, more than half of all patients were not reaching their low-density lipoprotein cholesterol treatment goal in general practice.44 These findings are also supported by the Eurospire III report where clinical implementation of secondary prevention treatments remained suboptimal.45 There are many possible reasons for the underuse of CVD-preventive therapies in the general population, which may also apply to patients with RA. Reasons may be: (1) patient-related—lack of patient education and/or awareness of the need for life-long therapy; (2) physician-related—lack of incentive for physicians; and (3) system-related—lack of large-scale implementation strategies and/or lengthy and too many guidelines. It is our belief that CVD-preventive treatment including pharmacotherapy is best implemented by either a PCP or a cardiologist working collaboratively with a rheumatologist. There are several national initiatives supporting the responsibility of PCPs for undertaking CVD prevention in RA, most notably the Quality and Outcomes Framework in the UK, which has introduced financial incentives for PCPs to do this ( While in many countries PCPs will be responsible for CVD prevention in the general population, the care of patients at high risk of CVD —for example, patients with familial hyperlipidaemia and diabetes—often falls to preventive cardiologists. The excess risk of CVD in patients with IJDs identifies them as another high CVD risk patient group, and thus, ideally, their preventive care is best directed by specialists in preventive cardiology working collaboratively with rheumatologists. One way to ensure efficient delivery of preventive cardiology for patients with RA is to associate a preventive cardiologist within the rheumatology department, as has been implemented in the Preventive Cardio-Rheuma Clinic at the Department of Rheumatology at Diakonhjemmet Hospital in Norway. A further strategy is to have dedicated expert teams of cardiologists and rheumatologists working together in assessing patients with autoimmune rheumatic diseases, as is done in the Cardio-Rheum Clinic at the Mayo Clinic in Minnesota, USA. These scenarios may not be feasible in all countries, thus the issue is to recognise the clinical need for CVD prevention in patients with RA and develop strategies that effectively and efficiently assess CVD comorbidities.

It is our opinion that preventive CVD services should be as aggressive in patients with RA as in other high CVD risk populations such as those with diabetes or hypertension. Currently, there is low awareness of the high CVD risk in patients with RA among cardiologists, which is also reflected by the scarcity of this topic at cardiology meetings and conferences. Therefore, we argue that efforts to address the high CVD risk in patients with RA should be a focus in the cardiology community as well as among rheumatologists, which hopefully will result in more patients with RA being evaluated for CVD risk and preventive measures being instituted when necessary.

Patient awareness of CVD risk

Communicating future risk and the need for preventive treatment is one of the most challenging aspects of the patient–health professional interaction. Patient's awareness about their CVD risk is important for successful implementation of preventive strategies related to lifestyle-modifiable CVD risk factors.46 Although such programmes for assessing CVD risk in RA have been developed, it will take much effort to fully implement them into clinical rheumatology. These strategies include education of healthcare professionals, written material, and online and group education of patients and their families. Even starting with simple information and brochures as a precursor to implementing a full education programme is a step forward.

Patient education is a cornerstone in management of chronic diseases, and for patients with RA today this is aimed at improving arthritis symptoms in obtaining remission, to improve overall outcome. A similar focus and investment is required to achieve reduction of CVD risk, and recent results indicate that obtaining remission in RA also influences CVD risk. For instance, treatment with methotrexate reduces the risk of CVD events by ∼20% in RA.47 Indeed, systemic inflammation contributes substantially to increased CVD risk both through its adverse impact on traditional risk factors and direct effects on the vasculature.48–52 In addition, patients with RA have more vulnerable carotid artery plaques than controls, but lowering disease activity may stabilise this threat.25 Therefore, effective management of CVD risk most likely comprises not only adequate treatment of conventional CVD risk factors but also tight activity control of the RA disease. In other words, remission is the mission not only for the joint disease but also for the CVD. This implies that patient awareness of remission in joint disease may also be of importance in affecting CVD outcomes.

Implementation of recording of CVD risk factors in routine clinical practice

There are several possible strategies for collecting CVD risk data that can be incorporated into the daily routine during rheumatology consultations at outpatient clinics. Examples are as follows:

  1. Patients self-report CVD symptoms, disease activity and medications on a questionnaire, sent to them with the letter of consultation date/time, and bring it filled out to the rheumatology provider appointment. A secretary or nurse records the data in the medical journal. Estimated time: 3–5 min. CVD risk factors and symptoms can also be collected by web-based means or through local electronic data which patients self-enter at home or at the time of their clinic visit. Such systems are used in clinical studies such as in the DANBIO registry53 and have also been implemented in rheumatology outpatient clinics in Norway, called the Go Treat It (GTI) (

  2. BP is measured as part of the routine consultation visit. Lipids are measured as part of the routine laboratory controls. The nurse records both BP and cholesterol concentration in the medical journal. Estimated time: 4 min.

  3. CVD risk is estimated by the rheumatologist using the national recommended guidelines for management of CVD risk. Facilitation of CVD risk evaluation is accomplished in the electronic medical chart (GTI) in Norway, where the CVD risk is automatically calculated. Estimated time: 3–5 min. The rheumatologist refers patients with a high CVD risk to the PCP or a cardiologist, or, as at the Mayo Clinic, to a dedicated cardiology–rheumatology team specifically addressing the needs of patients with systemic rheumatic diseases for medical intervention (eg, antihypertensive and/or LL therapy) and modification of lifestyle-related risk factors by applying the patient's calculated CVD risk in a standardised letter.

The time to register CVD risk factors in a rheumatology outpatient setting has been estimated to be ∼10 min,54 which is in agreement with the time to complete our regular rheumatology outpatient clinic at Diakonhjemmet. We are launching a similar system to the above in 11 rheumatology outpatient centres in Norway from Tromsø in the north to Kristiansand in the south, through a project called the Norwegian Collaborations on Atherosclerosis in Patients with Rheumatic Joint Diseases (the NOCAR project). We acknowledge that, in countries where rheumatologists spend less time with the patient, this strategy will need modification. With adequate ancillary support and efficient data recording and risk calculation, patients with RA can be assessed and management initiated for CVD risk reduction.


We urge that CVD risk be assessed and recorded in all patients with RA in routine clinical practice. It is undoubtedly better to take action than ignore high CVD risk even while educational CVD-preventive programmes are being developed and hard CVD end point studies are undertaken.

At the outset of such an undertaking, it is imperative not to have ambitions that are too high when implementing CVD risk evaluation, because complex and costly strategies may result in failure of initiatives to reduce CVD risk. Methods such as carotid ultrasound and other tools for CVD evaluation can be pursued in the appropriate setting. Therefore we recommend starting with simple procedures in rheumatology outpatient clinics as follows.

  1. CVD risk factor recording and evaluation using the best available risk calculators. At the moment, the latter are those designed for the general population, but we expect that a validated RA-specific calculator will be available in the near future.

  2. Patients with high CVD risk should be referred to a PCP or cardiologist for follow-up and treatment. Priority should be given to patients with the most severe RA disease activity, as this group appears to have the highest CVD risk.

  3. Most importantly, the major stakeholders are the patients, and they should receive information about their excess CVD risk and how to modify it.



  • Handling editor Hans WJ Bijlsma

  • Contributors All authors have (1) contributed substantially to the conception and design and/or analysis and interpretation of the data, (2) drafted and/or revised the manuscript critically for important intellectual content, and (3) given approval of the final version submitted for publication.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.