Reducing mortality in patients with chronic heart failure
Ducharme A et al. Am Heart J 2006; 152: 86-92
This paper attracted my attention because of current interest in AF and in additional benefits of the angiotensin-receptor blockers (ARBs) beyond lowering BP and treating heart failure.
AF, especially in patients with heart failure, is undesirable, leading to interest in preventing its development. These results are from a substudy of the CHARM trial for 7,601 patients with symptomatic heart failure. In the full study, patients with reduced or preserved left ventricular systolic dysfunction were randomised to candesartan with a target dose of 32mg and mean dose of 24mg, or placebo.
This was in addition to standard treatment for heart failure, with 55 per cent on a beta-blocker at baseline and almost two-thirds at the end. At baseline, 6,379 patients (83.9 per cent) did not have AF, giving a baseline prevalence of 16.1 per cent.
During the median follow-up of 37.7 months, 5.5 per cent developed AF in those treated with candesartan and 6.74 per cent in the placebo group. Patients who developed AF were older, and more often male and diabetic. They had more severe heart failure and lower ejection fraction (<30 per cent). The reduction in development of AF occurred in all subgroups.
This is thought to be the first study to show the benefit of an ARB. There is a suggestion that this effect occurs by reducing the progression of the severity of heart failure; another is the reduction of afterload or the inhibition of neurohumoral activity.
- Dr Kathryn Griffith is a GPSI in cardiology in York
Cardiovascular calcium in CKD
Santo Dellegrottaglie et al. Am J Cardiol 2006; 98: 571-6
Patients with end-stage renal failure have many cardiovascular events, associated with the presence of cardiovascular calcium. This is the first significant study of risk factors for intravascular calcium in pre-dialysis patients with an eGFR <50ml/min/1.73m2, as calculated using the MDRD equation.
A total of 106 patients were recruited to the study with multidetector tomography. Coronary calcium was scored by a blinded investigator. The majority of subjects were stage 4 CKD (73 per cent); 20 per cent were stage 3 and 7 per cent, stage 5. The aetiology was thought to be hypertension in 58 per cent and diabetes in 25 per cent, the remainder being polycystic kidney disease, interstitial nephritis and glomerulonephritis, a much higher percentage than we would see in primary care. Overall hypertension was present in almost 60 per cent.
The study demonstrated that 70 per cent of the subjects had some evidence of coronary artery calcium and 28 per cent had scores >400, considered to be associated with a high probability of obstructive coronary artery disease.
Overall age, male gender, hypertension and diabetes were the best predictors of high scores, confirming the role of traditional risk factors. Those patients with high scores were more likely to have low HDL, although high total cholesterol levels did not correlate with high scores.
The paper discusses the aetiology of the development of early coronary artery disease in patients with CKD and the possibility of ‘renal specific factors’. In atherosclerosis, calcium is deposited in the intimal layer of the vessel wall. An additional process of medial vascular calcification is found to be associated with diabetes, advanced age and renal failure. There was, however, no relationship between abnormal calcium or phosphate levels and high scores, and no difference between subjects with stable or deteriorating renal function or different haemoglobin levels. KG
Mediterranean diet improves cardiovascular risk factors
Estruch R, Martinez-Gonzalez MA, Corella D et al. Ann Intern Med 2006; 145: 1-11
This substudy of a randomised trial of primary prevention of cardiovascular disease recruited 772 high-risk people to assess the effects of Mediterranean diets (with virgin olive oil or mixed nuts) on surrogate markers of cardiovascular risk.
Participants were men aged 55–80 and women aged 60–80. They had either or both type-2 diabetes, or three or more risk factors: smoking, hypertension, BP >140/90mmHg or being treated, LDL M4.14 mmol/L or being treated with lipid-lowering drug, BMI >25 kg/m2, or family history of premature CHD.
Subjects were randomly allocated with specified subgroups of age and sex. Baseline questionnaires included diet and clinical assessment, and fasting bloods at baseline and three months. All subjects had a 30-minute consultation with a dietitian. Those in the Mediterranean diet group had further sessions and 1 litre olive oil per week, or sachets of walnuts, hazelnuts and almonds.
At three-month follow-up, only three subjects were not evaluable, although 48 were excluded as having unrealistic energy intakes. Analysis was on intention to treat basis and the sample size was determined by the detection in a difference of 0.13mmol/L of LDL.
All participants increased intake of fruit, vegetables and fish, and decreased meat, sweets and dairy. The Mediterranean diet with nuts increased intake of fibre, total fat, monounsaturated fats and polyunsaturated fats. Adverse events were only reported in the nut group, due to problems chewing whole nuts.
The body weight was slightly reduced, with no difference between groups. Both Mediterranean diets were associated with a reduction in BP (4.8/2.5mmHg and 6.5/3.6mmHg with nuts), blood glucose, cholesterol-HDL ratio and increases in HDL (0.62mmol/L with olive oil). Reductions in inflammatory markers were also seen in those on the Mediterranean diets. KG
Carotid disease in the elderly
Fairhead JF, Rothwell PM. BMJ 2006; 333: 525-7
Elderly patients may be reluctant to pursue what they perceive as aggressive treatment, particularly involving surgery. In this study, investigation and treatment patterns were compared between two populations.
One was the participants in the OXVASC study, in which patients with transient ischaemic attack (TIA) and stroke were offered investigation and treatment according to recommended guidelines. The other population was a sample managed as part of routine care in the same region.
The results demonstrated a significantly reduced level of intervention for symptomatic carotid stenosis in patients over
80 in the routine care population. Patient choice or legitimate contraindications were not the reason for undertreatment, because the OXVASC population involved similar decisions discussed with similar types of patients contemporaneously in the same region, but presumably with more consistent information provision and in a more formal research setting.
There is no reason to suppose that this finding is specific to the region studied. This is an important reminder that clinicians (in primary or secondary care) should not make assumptions about patient preferences without informed discussion, and that the risk/benefit ratio for many procedures may be in favour of intervention, even in the very elderly.
- Dr Tim Holt is a GP and clinical lecturer at Warwick Medical School
CHD mortality and obesity
Romero-Corral A, Montori VM, Somers VK et al. Lancet 2006; 368: 666-78
Despite the known association between excess body fat and cardiovascular disease, this systematic review involving 40 studies and 250,152 patients with CHD found no association between mild obesity and cardiovascular or all-cause mortality.
Those who were overweight, but not obese, had lower relative risk of both cardiovascular death or death from any cause, compared to the normal BMI group. The severely obese (BMI >35kg/m2) are at higher risk of cardiovascular death, but not for all-cause mortality. Interestingly, the underweight group was at higher risk of cardiovascular and all-cause death than the normal BMI, the overweight, or the obese patients.
However, the majority of studies included in the review did not take into account the confounding effects of drug treatment or lifestyle intervention in obese patients, who are more likely to be receiving therapy for raised cholesterol or hypertension. Those with low BMI tend to have low muscle mass and may be less likely to exercise effectively. These factors make the results very difficult to interpret.
The major question identified in this paper is not that excess body fat is unimportant or harmless, but that BMI is inadequate on its own in identifying patients with excess adiposity. Raised BMI may be seen in those with a higher volume of lean mass due to regular training, for example, and based on this measure alone, it is impossible to distinguish these patients from those of equal BMI who are unfit and have a higher waist circumference.
This lack of discriminatory power is a problem for general practice teams who are currently building obesity registers under the quality and outcomes framework, registers that are defined solely on the basis of BMI.
As this study suggests, the inclusion of waist circumference, or better still, waist to thigh ratio, would significantly improve the usefulness of these data for targeting at-risk individuals. In the meantime, in our attempts to help obese patients to lose weight, we should not forget the significance of low BMI for patients with established CHD. TH
Cardiology for practitioners with a special interest
Burwell C. Br J Cardiol 2006; 13: 293-6
This study evaluated a postgraduate diploma in cardiology being run by Bradford City Teaching PCT. The course trains local practitioners in collaboration with local cardiologists and supports the development of practitioners with a special interest (PSI).
Designed to tailor learning needs to local service provision, the evaluation found an increase in confidence in all clinical areas among the seven respondents who took part.
This evaluation was small and not independent, and did not include participants who did not return a second questionnaire at course completion, so its conclusions were limited, but it gives encouraging signs for other PCTs interested in developing this approach to cardiology. The Bradford City Teaching PCT course is available to practitioners nationwide.
Inclusion of the views of cardiologists and patients might have strengthened the study, because a major question concerning the PSI role is acceptability to patients who might traditionally have been referred to secondary care. TH
Guidelines on statin treatment for preventing CHD deaths
Manuel DG, Kwong K, Tanuseputro P et al. BMJ 2006; 332: 1419
How do you distinguish a good guideline from a bad one? One measure is the efficiency with which recommendations perform their intended function. This is closely linked to the question of cost-effectiveness, which tells us how much is required to achieve the health gains the guideline seeks to deliver.
These authors chose six guidelines from Canada, Australia, New Zealand the US, the UK and Europe, and compared their effectiveness. Data from the Canadian Heart Health Survey, which includes 6,760 men and women aged 25–74, were used as the base population for their modelling study. In a similar way to the Health Survey for England, this is a weighted sample held to be reflective of the entire population.
The effects of implementing the guidelines was evaluated within this sample, with respect to the number of CHD deaths avoided and the number of patients needed to treat to prevent one CHD death with five years of treatment, assuming full implementation of the respective guideline. Not surprisingly, the guidelines that require screening the largest proportion of people are the most effective in preventing the most CHD deaths. These are the Australian, UK and US guidelines, which, the authors calculate, could avoid more than 15,000 deaths each.
The real question is which guideline is the most efficient, preventing the greatest number of CHD deaths for any given unit of GP or nurse time that a practice can allocate to CHD.
The New Zealand guideline is potentially the most efficient, delivering the benefit of preventing one cardiac death for every 108 people treated with statins. The UK guidelines, despite preventing a large number of CHD deaths, are considerably less efficient and require 139 people to be treated to prevent one cardiac death – thus requiring 30 per cent more people to be screened than the New Zealand guidelines for the same benefit.
In population terms, the New Zealand guidelines require screening of 12.9 per cent of the population, while the UK guidelines require screening of about 17 per cent.
There are three main lessons here. First, recommendations should be based on modelling exercises such as these, rather than on assertion. Second, the JBS2 guidelines are inferior to the New Zealand guidelines and will add to workload without improving efficiency. Third, the numbers needed to treat are surprisingly large and give some reason for pause and reflection.
- Dr Rubin Minhas is a GPSI in CHD and Medway PCT CHD lead
LDL cholesterol reduction for CHD and metabolic syndrome
Deedwania P, Barter P, Carmena R. Lancet 2006; 368: 919-28
Lipid goals are a hot topic in cardiovascular medicine, with an argument raging over whether lower lipids, generally attainable with the more potent, non-generic statins, confer benefit that is more worthwhile for patients than that of the tried and tested, cheaper, generic statins.
This retrospective cohort study looks at the Treating to New Targets (TNT) studies and asks whether identifying patients with CHD and features of the controversial metabolic syndrome selects patients who might derive more benefit than those people without features of metabolic syndrome.
The original TNT study was a prospective randomised double-blind trial that examined outcomes among 10,001 patients receiving atorvastatin 10mg, versus those randomised to receiving atorvastatin 80mg.
Given the large population of the original study, this subgroup analysis of 5,584 patients allows a reasonably well-balanced comparison between patients with and without either metabolic syndrome or diabetes, who were randomised to the low and high doses of atorvastatin.
From an overall baseline of 2.5mmol/L, atorvastatin 80mg reduced LDL to a mean of 1.9mmol/L after three months, compared to atorvastatin 10mg.
After a median follow-up period of 4.9 years, 262 patients (9.5 per cent) with metabolic syndrome who received atorvastatin 80mg had a primary event, in contrast to 367 (13 per cent) receiving atorvastatin 10mg who had a primary major cardiovascular event. These are the headline figures from the study and the number needed to treat is quoted as 29 for this group. The results were the same whether BMI or waist circumference were used.
Consistent with the original trial, there was no increase in total mortality, so no difference in the overall death rate from any cause for patients randomised to atorvastatin 80mg.
The authors suggest their study provides evidence that patients with CHD and metabolic syndrome are candidates for more intensive lipid lowering. However, this study is hypothesis generating and a prospectively designed study is required. The benefits of more aggressive lipid lowering remain moot. RM