Bristol, UK - A team of British scientists has recalibrated Framingham data to generate a new web-based calculator for assessing the risk of heart disease in seven ethnic minorities in the UK [1].

"Until now, ethnic groups in the UK have often been wrongly assessed," lead author Dr Peter Brindle (University of Bristol, UK) told heartwire. "This new tool should more accurately reflect their true risk." Brindle and colleagues report their findings June 8, 2006 online in Heart, and the calculator, dubbed ETHRISK, can be accessed online [2].

"In the UK there are hardly any cohort studies in ethnic minorities, so there are few long-term incidence data, unlike the situation in the US," he notes. "We basically took prevalence data and used it as a reasonably good surrogate for incidence."

Coauthor and original Framingham statistician Dr Ralph D'Agostino Sr (Boston University, MA) explained to heartwire that the British team has taken his 2001 validation of the Framingham scores in US ethnic groups [3] and built on that. "They have assumed that this first step we took works—ie, that the risk factors mean the same thing in different minorities and, using minimal UK data, they have done something practical with it."

Brindle and colleagues used two community-based surveys of more than 8000 individuals to calculate the ethnic- and sex-specific 10-year risks of coronary heart disease (CHD) and cerebrovascular disease (CVD) from the product of the incidence rate in the general population and the prevalence rates for each ethnic group. The work was funded by the British Heart Foundation.

They did not include diabetes or left ventricular hypertrophy (LVH) as risk factors in their model, even though they are included in the standard Framingham tool. D'Agostino told heartwire he agreed with the decision to leave LVH out "because there tends not to be stability about how it is measured across geographic regions. In fact, we ourselves dropped it from our later work." But, he says, he argued with the British "that diabetes should not be left out."

However, Brindle says ETHRISK "is not about screening for diabetes. These patients [diabetics] are high risk anyway and once identified will receive attention to all their CVD risk factors irrespective of predicted risk."

They also did not include any measures of obesity, as they were not included in the original Framingham tool. D'Agostino says this is not a problem because, in reality, "obesity is always reflected in the blood pressure and cholesterol."

Brindle et al found the 10-year risk of CHD and CVD was highest for men of Pakistani and Bangladeshi origin and lowest for Chinese women.

He says this new work is a major advance on existing guidance in the UK. Many primary-care practitioners in the UK currently do not record ethnicity or take it into account when assessing the risk of heart disease or stroke, he notes.

"We have minimal data in the UK on South Asians, and this does not account for heterogeneity between different South Asian groups or between the sexes. We have none at all for black Africans, Caribbeans, and Chinese. This risk score provides a pragmatic solution to include people from diverse ethnic backgrounds in the primary prevention of CVD."

He adds that although the algorithm is specific to the ethnic mix and environment found in the UK, any country could adapt it to generate risks for its minority groups. "Many countries have good prevalence numbers and so could adapt the model for use in their own populations."

In their 2001 paper, D'Agostino and colleagues set out to investigate whether the original Framingham data—which were based mainly on white middle-class individuals—would generalize to other populations.

D'Agostino says while the ATP III guidelines of 2004 recognized his work, it was decided not to try to impose use of the recalibration suggested in ethnic groups in the US. "The feeling was that it might generate confusion if they tried to impose it and stop people from using Framingham risk in ethnic minorities altogether."

However, he feels it shouldn't be hard to progress the concept with one extra factor—ethnicity. "I see it as something quite feasible, it should not be a killer on the physicians." He believes the UK's centralized National Health Service (NHS) may be better equipped than the US to incorporate the ETHRISK calculator into daily use. Most general practitioners in the UK, for example, have desktop computers with medical records instantly accessible.

But Brindle says "the major drawback of ETHRISK" is that it is web-based and therefore requires doctors to enter in extra data. "It is not currently included in the standard general-practitioner database, so we need to design software to integrate it."

He says the current ETHRISK "is not the ultimate solution, but it is an improvement on the existing method." The calculator is due to be appraised soon by the National Institute for Clinical Excellence (NICE).