The Avoid Stroke as Soon as Possible (ASAP) general practice stroke audit

Stroke is the third-greatest single cause of death in Australia, after ischaemic heart disease and cancer.1 Each year, more than 40 000 Australians suffer a stroke;2 a third of these people will die within a year,3,4 and half will be disabled.5 The incidence of stroke increases with age, approximately doubling with each decade of life.4 Given the ageing of the Australian population, the number of strokes occurring is likely to dramatically increase unless stroke incidence can be reduced.

Substantial advances in acute stroke management have occurred in recent years, including the use of aspirin, increased access to stroke units and, for selected patients, thrombolysis with tissue plasminogen activator.6 However, the most effective method of reducing the burden of stroke is prevention.7,8 There are two approaches to prevention: the "high-risk" approach, in which individuals with high absolute risk of stroke are identified and interventions initiated, and the "population" approach, in which levels of risk factors are reduced in the population.

Important risk factors for stroke are increasing age, hypertension, smoking, diabetes, hypercholesterolaemia, atrial fibrillation and prior stroke or transient ischaemic attack (TIA). Other factors include family history, male sex, excessive alcohol intake, prothrombotic factors, socioeconomic status and possibly physical inactivity, obesity, infection and stress.9 The prevalence of these risk factors in a typical general practice population is unknown. Our study was primarily designed to obtain this information, to facilitate both the high-risk and population approaches to stroke prevention.

Our aims were to assess among patients consulting general practitioners in Australia (1) the prevalence of important stroke risk factors, and (2) the type and extent of use of pharmacotherapies to modify risk factors.

We conducted a multicentre, observational study of patients attending general practices in Australia.

The study was undertaken during 2000 in Queensland, New South Wales, South Australia, Victoria, Tasmania and the Australian Capital Territory. Western Australia and the Northern Territory were not included for logistical reasons (GPs from WA had participated in an earlier feasibility study). All registered GPs (18 066) were sent a letter seeking expression of interest in the study. Of these, 1850 replied and were eligible for participation. The study area was divided into 22 regions, each serviced by about 600 GPs. From each region, 18 doctors were randomly selected from the 1850 eligible (396 in total). If a doctor was unavailable or refused participation, the geographically closest GP was approached. This process ensured that rural and urban practices and a representative range of socioeconomic settings were included.

Each GP approached 50 consecutive patients aged 30 years or older to participate in the study. Demographic data only were recorded on patients who refused consent.

Management of patients with risk factors was left to the GPs' discretion.

Our study has yielded unique data on the prevalence of stroke risk factors in an Australian general practice population. The data were obtained from general practices that were spread throughout Australia, including metropolitan and rural communities and a range of socioeconomic settings. Thus, the information obtained is likely to be representative of most Australian general practice environments.

Nevertheless, there is likely to be selection bias in the study. Ethnic origin of patients was not recorded; therefore, we cannot be sure our sample is representative of all ethnic groups. As no doctors were from the Northern Territory, people of Aboriginal populations are found in capital cities covered by the study. The age and sex structure of the general population differs from that of the sample. Whether the high proportion of women in the study is due to women being more likely to attend GPs or to chance is uncertain. Of patients aged over 80 years, 10% were taking warfarin, a higher rate than would be expected in the general population. Our definition of hypertension as a past history or a single recording of greater than 140/90 mmHg may include patients who do not have persistent hypertension. This may partly explain why the prevalence of hypertension in our study is higher than in a previous Australian survey13 (although the previous sample consisted of people between 20 and 69 years of age, therefore excluding the age groups with the highest rates of hypertension). The prevalence of hypertension in our study is similar to other published figures.14 The prevalences of hypercholesterolaemia,13 atrial fibrillation,15 diabetes16 and smoking13 are similar to those previously reported.

About 86% of adults in Australia consult a GP at least once a year (Carolynn Fredericks, Department of Health and Ageing, personal communication). Therefore, GPs can play an important role in stroke prevention by opportunistic case-finding of patients with stroke risk factors. The National Health and Medical Research Committee (NHMRC) guidelines on stroke prevention recommend routine screening of all patients over 45 years of age and younger patients who have risk factors.17 This can be achieved at little additional cost to the community if conducted during routine consultations. Systematic screening for risk factors during routine consultations would enable identification of patients at high risk of stroke, particularly those with multiple risk factors. Of course, people who do not attend general practices may have different prevalence of risk factors than those who do, and will not benefit from case-finding by GPs.

Most of the stroke risk factors increased in prevalence with age. Hypercholesterolaemia and diabetes were more common with increasing age, but less prevalent in the oldest age groups. Age is an independent risk factor for stroke9 and this, together with the risk-factor prevalence profiles, explains the exponential increase in stroke risk with age. Importantly, hypertension, hypercholesterolaemia and smoking had a prevalence of more than 10% in men and women aged 30–39 years. This suggests that screening for risk factors in any patient aged over 30 years is worthwhile, so that risk-factor modification may begin before the onset of symptomatic atherosclerosis. In contrast, atrial fibrillation had a very low prevalence below the age of 60 years and it is older patients who require screening for this.

Some of our results are perplexing, such as 10% of patients receiving diuretics for treatment of ischaemic heart disease. Lower proportions of women were receiving each type of pharmacotherapy than men. Although intriguing, reasons for this were not addressed in the study. Surprisingly, TIA had a higher prevalence than stroke. Possible explanations include chance, post-stroke dependency, or misclassification (TIA is difficult to differentiate from minor stroke or non-vascular causes).

The high prevalence of stroke risk factors suggests that many patients could benefit from interventions to prevent stroke. Antihypertensive agents for hypertension,18 antiplatelet agents in people with prior stroke or TIA,19 warfarin for atrial fibrillation20 and the use of statins for hypercholesterolaemia in patients with evidence of coronary artery disease21 are effective in reducing stroke incidence. The evidence in observational studies for the reduction of stroke incidence with cessation of smoking is overwhelming.22 Hypertension may be modified by reducing salt intake,23 and obesity9 and lack of physical exercise9 have been associated with increased risk of stroke. However, whether lifestyle advice delivered by GPs is effective in reducing stroke risk is unproven.

Medical diagnosis and intervention does not necessarily imply adequate control of stroke risk factors. Of patients receiving medications for hypertension, 30% had blood pressures recorded in the moderate to severely hypertensive range. Stroke is associated with undertreated hypertension,24 so ongoing medical supervision is required to ensure therapy is adequate.

Substantial numbers of patients in our study with a history of hypertension, cardiac failure and ischaemic heart disease were not receiving cardiovascular medication. Although reasons for non-treatment (such as contraindication, poor compliance, or stable disease) were not addressed, it is likely that some of these patients are being denied potentially beneficial therapy.

Overall, our study offers unique Australian data on the prevalence of stroke risk factors in a population highly relevant to GPs. The high prevalence of stroke risk factors in patients attending Australian general practice clinics suggests that GPs are ideally placed to undertake opportunistic risk-factor management. The Avoid Stroke as Soon as Possible (ASAP) audit provides baseline data against which progress in risk factor management can be measured.

3: Stroke risk factor prevalence, by age group and sex, in patients aged 30 years and older consulting general practitioners