Row over NICE statin guidance brings confusion for GPs
By Nigel Praities
GPs are caught in the middle of a huge stand-off over how to assess patients for statins, after leading figures from the Joint British Societies launched a fierce attack on NICE over its decision to abandon the Framingham risk score.
In a move reminiscent of the long-running row over drug choice in hypertension, experts from the British Hypertension Society and Heart UK strongly criticised NICE for failing to evaluate the evidence properly.
They warned that NICE's plans to replace Framingham with the new QRISK score on practice systems would cause massive disruption with no proof of benefit.
One respected figure raised questions over the presence of QRISK researchers on the guideline development group.
Use of the QRISK tool is set to have a huge impact on GPs' statin prescriptions, with reductions in overall statin use, but significant increases among women and in deprived populations. But the row over its provisional approval last week by NICE has cast a shadow over its introduction, and leaves the institute's forthcoming lipid modification guidance mired in confusion.
Professor Paul Durrington, professor of medicine at the University of Manchester and Heart UK representative for the JBS guidelines, said QRISK was far from being a validated CV risk score.
‘Framingham is extremely reliable and we have done prospective, epidemiological studies and have a proper realisation of its accuracy and limitations, but QRISK is based on a register with huge amounts of missing data,' he said.
Dr Mark Davis, a member of the BHS and a GP in Leeds, said: ‘There is a lot of concern about QRISK. For NICE to recommend something, it needs to make sure it is evidence-based. I am not confident this is.'
Both Professor Neil Poulter and Professor Bryan Williams, the BHS authors of the JBS guidelines, told Pulse they were unhappy about changing to QRISK.
Professor Williams said: ‘If the change involves the need for wholescale re-education of the workforce and changes to computer systems, it must be a substantial improvement on what we have. From what I've seen the added value does not justify the change.'
Although the decision to recommend QRISK was based on three independent reviews by three academics, Professor Durrington claimed allowing QRISK researchers to sit on the NICE guideline development panel was an ‘abuse of the principles for which NICE was created'.
QRISK researchers and GPs Dr John Robson and Dr Peter Brindle were members of the guideline development group, with Dr Robson as chair. He refused to comment, but Dr Brindle said: ‘I had no part in the final decision due to my conflicts of interest.'
Professor Julia Hippisley-Cox, professor of general practice at the University of Nottingham and lead researcher for QRISK, defended the data the score is based on. She said: ‘Some 99% of all people in the UK are registered with a GP and so information about their age, sex and cardiovascular risk factors are held on the system.
‘There is probably a bit of resistance from secondary care doctors to something that is developed by GPs.'
NICE said the two QRISK researchers on the guideline development group had left the room during discussions and were ‘not involved in decisions on choice of risk score'.
It added that it had sought the views of three independent experts, including the ren-owned Sir Richard Peto, who had agreed unanimously that QRISK be recommended.QRISK vs Framingham
• Based on data taken from UK primary care, and provisionally recommended in NICE lipid modification guidance
• First described in July 2007 in a population of 1.3 million patients, and separately validated in a further 1.6 million patients
• Risk factors include age, gender, smoking, blood pressure, total cholesterol to HDL-cholesterol ratio, BMI, family history of CVD, deprivation and antihypertensive prescription
• Set to reduce statin use overall, but increase it in women and deprived populations
• Based on a cohort study of 5,573 affluent white patients in Boston, USA
• Numerous prospective studies have examined Framingham data, providing strong evidence on both its strengths and its weaknesses
• Risk factors include age, gender, smoking status,
blood pressure, total cholesterol to HDL-cholesterol ratio and left ventricular hypertrophy
• Overpredicts risk overall but underpredicts risk in wome
and deprived populations
QRISK is based on a register with huge amounts of missing data