The rise and fall of cardiac rehabilitation in the UK
Algorithms for the assessment of global cardiovascular risk
Medical therapy with or without PCI for stable coronary disease
Beta-blockers used to prevent AF in heart failure
The effect of statins in the reduction of BP
Financial barriers to care and outcomes after acute MI
Specialty cardiac hospitals and coronary revascularisation
This paper reports the results of annual surveys of all UK cardiac rehabilitation programmes between 1998 and 2004. The survey was started by the British Association for Cardiac Rehabilitation to find out how well the therapy was being provided in the UK.
By 2000 it was established that all NHS hospitals with CCU had access to rehabilitation programmes. The number of centres increased from 286 in 1998 to 330 in 2004.
In 2000 the NSF set a target for 85 per cent of eligible patients to receive rehabilitation after AMI, CABG or PCI.
A comparison was made between the numbers treated and those eligible for treatment; overall numbers with acute MI from the British Heart Foundation , CABG patients from the Society of Cardiothoracic Surgeons , and PCI from the British Cardiac Intervention Society . This rather disappointingly showed only about a quarter of those eligible were enrolled for rehabilitation, with nearly three-quarters of CABG patients and only one-fifth of PCI patients. The results also showed that the overall increase in those attending programmes had not kept pace with the increased activity.
Some criticism could be made of the data when only two-thirds of the centres responded and 55 per cent gave workload figures, despite repeated telephone calls from Professor Lewin. The uptake figures are similar, however, to those found by the Healthcare Commission in 111 hospital trusts.
The paper discusses the reasons for the shortfall, which seems to fall mainly on the lack of a national plan and thus central funding. The 2005 assessment of progress towards the NSF goals does suggest that improving rehabilitation should be a major focus of the next phase, but with no new money, it will require the inclusion of cardiac rehabilitation as a quality indicator to encourage trusts to improve services.
This is a field where we still need to establish local champions to drive improvements and, unfortunately, to raise funds to support the developmnent of the service.
- Dr Kathryn Griffith is a GPSI in cardiology in York
This was a study of almost 25,000 women aged 45 years or more who were followed up for at least eight years (median follow-up 10.2 years), run by centres around the Reynolds Centre for Cardiovascular Research , in Boston, Massachusetts.
Participants were obtained from the Women’s Health Study and had baseline plasma samples of which 76 per cent were fasting. The study was set up because traditional risk scores based on the Framingham study were thought to be less accurate for predicting risk in women. Many women with traditional risk factors do not have events and 20 per cent of coronary events occur in the absence of risk factors. Thirty-five factors were assessed in a random two-thirds to develop new algorithms and tested in the other third to predict outcome.
Two models were produced, model A with the best fit, and model B simplified for clinical application. Numerous statistical tests were used to validate the models and comparison made with ATP-III and Framingham Risk scoring systems, although 10-year risk was used.
The best fitting model used nine variables – age, systolic BP, current smoking, apolipoprotein B-100, hsCRP, apolipoprotein A-1, parental history of MI <60 years, HbA1c with diabetes and lipoprotein (a) if apolipoprotein was >99mg/dl. Other variables, such as homocysteine, BMI, menopausal status and creatinine, were excluded. The simplified model was similar, including the same variables, except for the substitution of total and HDL cholesterol for the apolipoproteins.
Comparison was made with ATP-III prediction. Reclassification was small for those at <5 per cent risk but 43 per cent of those at 5 to <20 per cent risk were reclassified, with risk being over- and underestimated in approximately equal proportions.
The algorithm was developed on data from 24,558 women with 766 hard end points, using the combined cardiovascular end points acute MI, ischaemic stroke, revascularisation and cardiovascular mortality.
This research is larger than that in a recent publication from the Framingham study. The sample is, however, largely white, with a narrow socioeconomic range, being composed of female health professionals. A user-friendly calculator can be freely accessed at www.reynoldsriskscore.org. KG
Medical therapy with or without PCI for stable coronary disease
Boden WE, O’Rourke RA, Teo KK et al. N Engl J Med 2007; published at www.nejm.org Mar 26 2007 (10.1056/NEJMoa070829)
This is a must-read paper for those managing patients with ischaemic heart disease.
In this randomised prospective study of 2,287 patients with objective evidence of ischaemia and significant coronary artery disease in 50 North American centres, it was assumed that PCI would be associated with a reduction in three-year events over optimal medical therapy alone.
Patients with Canadian Cardiovascular Society class IV angina and at least 70 per cent stenosis of one coronary artery and objective evidence of myocardial ischaemia were randomised into either resting or exercise ECG changes. Patients with angina that was not improving with medical therapy, heart failure, or revascularisation in the previous six months were excluded.
Patients were randomly allocated to PCI and optimal medical therapy, or medical therapy alone. All patients received aspirin or clopidogrel, a combination of metoprolol, amlodipine and isosorbide, and aggressive lipid-lowering therapy (simvastatin with ezetimibe if required), to achieve an LDL-cholesterol target of <2.20 mmol/L.
Most PCI was performed before the use of drug-eluting stents. Mean follow-up was 4.6 years and 9 per cent were lost to follow-up. The primary outcome of death from any cause and non-fatal MI occurred in 211 patients in the PCI group and 202 patients in the medical therapy group. When stroke was added for the secondary endpoint, the event rate was 20 per cent in the PCI group, versus 19.5 per cent.
At a median follow-up of 4.6 years, 21.1 per cent of those in the PCI group had further revascularisation, compared to 32.6 per cent with medical therapy. This was performed for worsening symptoms or unstable testing and was the only outcome with statistical significance.
The paper concludes that initial therapy with PCI does not reduce major cardiovascular events during a follow-up of 2.5 to seven years. There was better angina relief and reduced need for further procedures. The paper discusses the possibility that this may be due to the fact that the lesions responsible for unstable events are not the same as those for stable angina and that this may be the explanation for the findings.
When obtaining consent for coronary angiography from patients in the Rapid Access Chest Pain Clinic, I must be clear that PCI, if required, improves symptoms, but it has not been proved to save lives in this study. KG
Beta-blockers are known to improve morbidity and mortality in patients with heart failure (HF), but the exact mechanism remains unclear. One possibility is a reduction in incidence of AF, to which patients with HF are prone.
AF in turn exacerbates HF, reducing cardiac output and raising the risk of thrombo-embolism, so there is a strong incentive to prevent the onset of this arrhythmia in HF patients. A number of drugs are known to do this, but most are associated with significant toxicity.
This meta-analysis involved seven randomised controlled trials involving patients with HF treated with beta-blockers, in which the rhythm at baseline was recorded. Most of the participants in both arms of these studies were also receiving ACE inhibitors or angiotensin receptor blockers.
Despite some heterogeneity in the treatment effect, beta-blockers were found to reduce the relative risk of AF by an overall 27 per cent (95 per cent CI 14-38, p = 0.001). The authors excluded the possibility of publication bias towards positive trials by including a robustness analysis.
The effect of non-selective beta-blockers was more easily demonstrated than that of selective beta-blockers. However, this finding was sensitive to the inclusion of one particular study, so the meta-analysis itself could not offer specific advice on the best choice of drug.
One of the limitations of the meta-analysis was that the trials only detected AF if it was picked up as an adverse event during follow-up. The trial investigators were not necessarily looking for it systematically. This might have reduced the detection of AF in study participants, conceivably more so in the intervention arm, where the raised ventricular rates that typically produce the symptoms of AF would tend to be reduced by the beta-blockers.
A definitive study (the AF-CHF trial) is under way that should settle this issue, but in the meantime, it appears that beta-blockers are effective in preventing AF in HF and this may contribute to their beneficial effect on outcomes.
- Dr Tim Holt is a GP and clinical lecturer at Warwick Medical School
We tend to assume that the effects of lipid-lowering drugs on cardiovascular risk are mediated through reduced lipid levels, but this meta-analysis suggests that in the case of statins, some benefits might also result from a reduction in BP.
Perhaps surprisingly, there is a shortage of data from the large statin trials on BP outcomes, and one inclusion requirement for this meta-analysis was that antihypertensive medication, if used, was kept unchanged throughout the trial. This reduced the availability of suitable trials. The 20 studies involved were all quite small, with a total of 828 patients in the meta-analysis, and significance was reported only at the 5 per cent level.
Despite these limitations, a small reduction of systolic BP of 1.94mmHg (95 per cent CI -3.77 to -0.12) was found to be associated with statin use. Diastolic BP showed a trend towards reduction, falling short of statistical significance. Such a small reduction may seem to be of trivial clinical value, but as the authors discuss, cardiovascular outcomes are improved with even small reductions in BP, and reductions of this magnitude have been shown to improve cardiovascular event rates in trials of antihypertensive medication.
The effect of statins on systolic BP was only evident in patients with BP levels >130mmHg at baseline. At lower levels, no effect on BP was observed. It seems unlikely, therefore, that statins will produce additional BP reduction in patients with adequately controlled hypertension, but they might slightly reduce the need for polypharmacy.
However, because the effect is less than that expected from an antihypertensive drug, the therapeutic implications are limited. This study is important, because it encourages us to think more broadly about the mechanisms through which such drug therapies confer their benefits. TH
This US study evaluates the prevalence and consequences of financial barriers to care systems and medication for patients with acute MI. The study involved 2,498 individuals aged 18 years or more, with confirmed acute MI. Sociodemographic, clinical and treatment data were collected within 72 hours of admission and follow-up interviews conducted a year later.
Patient outcomes were assessed using the disease-specific Seattle Angina Questionnaire and the generic SF-36. Other variables of interest included all causes and cardiac-specific rehospitalisations. The prevalence of self-reported financial barriers to care or medication was 18.1 per cent and 12.9 per cent respectively. The proportion with health insurance was about 68 per cent for both groups – greater than the proportion of Americans known not to have health insurance (one-sixth).
After one year, those who reported financial barriers to healthcare were more significantly likely to have lower quality of life scores and increased rates of all-cause rehospitalisation (about 30 per cent), although the 30 per cent increase in cardiac rehospitalisation was of borderline significance. Those reporting financial barriers to medication were more likely to report angina and to have lower quality of life scores and increased rates of all-cause and cardiac-specific rehospitalisation.
This valuable study addresses a significant problem: access to healthcare. The authors conclude that financial barriers to care are a potent risk factor in the acute MI population. They identify a need for approaches to care that mitigate this increased risk and address this barrier to healthcare and medication. - Dr Rubin Minhas is a GPSI in cardiology in Gillingham, Kent, and Medway PCT CHD lead
Although specialty hospitals provide high-quality cost-effective care, financial incentives for physicians could result in them performing more procedures. Here, US researchers looked at whether the opening of cardiac hospitals is associated with increasing rates of coronary revascularisation.
The research involved a population of Medicare beneficiaries from 1995 to 2003. Annual population-based rates were calculated for total revascularisation (CABG plus PCI), CABG and PCI.
Hospital referral regions (HRRs) were used to categorise healthcare markets as those where cardiac hospitals opened (n = 13), where new cardiac programmes opened at general hospitals (n = 142) and where no new programmes opened (n = 151). Outcome measures were rates of change in total revascularisation, CABG and PCI.
Rates of change for total revascularisation were higher in HRRs after cardiac hospitals opened, compared to HRRs where new cardiac programmes opened at general hospitals and those with no new programmes. Four years after opening, the relative increase in adjusted rates was more than twice as high in HRRs where cardiac hospitals opened, compared to HRRs where new cardiac programmes opened at general hospitals and HRRs with no new programmes. These findings were consistent when rates for CABG and PCI were considered separately. For PCI, this growth appeared largely driven by increased use among patients without acute MI.
It would appear that specialty cardiac hospitals are offering revascularisation considerably in excess of the rates seen in general hospitals. Discretionary PCI and coronary stenting seem to account for much of the increase. The roll-out of new cardiac catheter laboratories across the UK should be watched. RM