Screening for abdominal aortic aneurysm

BMJ 2007; 335 doi: https://doi.org/10.1136/bmj.39307.634560.AD (Published 11 October 2007) Cite this as: BMJ 2007;335:732
  1. Roger Greenhalgh, emeritus professor of surgery,
  2. Janet Powell, professor of vascular biology
  1. Vascular Surgical Research Group, Imperial College, London W6 8RP
  1. Correspondence to: R Greenhalgh rmg.pa{at}ic.ac.uk

    Can save lives but only if operative mortality is low

    A recent Cochrane review has updated our knowledge about screening asymptomatic people for abdominal aortic aneurysm, with respect to their mortality, subsequent treatment for the aneurysm, and the cost effectiveness of screening.1 Four completed randomised controlled studies—Chichester, Viborg, Western Australia, and the multicentre aneurysm screening study (MASS)—with 127 891 men and 9342 women (only the Chichester trial included women) aged from 65-83

    years were included to a cut-off date of 26 January 2007. This excluded the more recent seven year follow-up of men in MASS.2 Acceptance rates (of people agreeing to be screened) ranged from 63.1% (Western Australia) to 80.2% (MASS).

    In men aged 65-79 years screening significantly reduced the risk of mortality related to aneurysm (relative risk 0.53 (confidence interval 0.42 to 0.68). This was achieved at the expense of doubling the rate of aneurysm surgery. However, for studies in men, the review reported no significant reduction in all cause mortality. Although the Western Australia trial found a reduction in all cause mortality (from the time of screening and not randomisation), the authors note that the interval between randomisation and screening could have introduced a bias, such that screening did not account for the reduction. However, the recent update from MASS also hints at a possible benefit in all cause mortality in men who were screened (estimated hazards ratio 0.96, 95% confidence interval 0.93 to 1.00),2 so a further update of the Cochrane review may be needed.

    MASS produced a cost effectiveness analysis at four years with 47 fewer deaths from aneurysm equating to £28 400 (€42 000; $58 000) per life year gained and £36 000 per quality adjusted life year. At seven years this had fallen to £12 334 per life year gained2 and is likely to fall even further after 10 years. The Viborg trial derived a very different figure (£620 per life year saved), and the reason for this disparity seems opaque, although health economists should be able to shed some light on the reasons for the disparity.

    The trials have provided evidence to suggest that screening in itself does not impair quality of life,3 4 although this is not covered in the Cochrane review.

    Data are still lacking on the potential benefits or harms and costs for screening women, although at least one group suggests it may be cost effective5 and the screening of high risk women was supported by the current president of the Society for Vascular Surgery in the United States.6

    All these data are supportive of a national screening programme, and at a time when the NHS is considering the cost implication of establishing such a programme for men aged 65 years, the seven year follow-up data from MASS with the lower cost per life year gained are especially timely.

    Correctly, the mood is in favour of aneurysm screening, but the following policy problems still need to be tackled. How can screening uptake be improved in those at highest risk (such as those in the lowest socioeconomic groups)? How can screening be refused to men older than 65 years and women at highest risk (such as smokers and those with a strong family history of aneurysm)? How and where should patients with screen detected aneurysms be managed?

    All the screening trials, as well as other randomised trials of aneurysm treatment, report operative mortality of about 5% for open elective surgery (as used in all the screening trials) for aneurysms ≥5.5 cm in diameter, the general threshold for intervention. Randomised trials show that operative mortality is lower from endovascular repair (<2%) than from open repair (<5%), although endovascular repair costs more.7 8 9 Not only may medical treatment, including statins, further improve operative mortality and life expectancy in those found to have abdominal aortic aneurysms,10 there is now the expectation that statins and other new treatments will slow the growth of small aneurysms found by screening.11 12

    Screening should do no harm. However, a recent evaluation of administrative and clinical databases looking at predictors of risk of death in hospital suggests that in England the in-hospital mortality for non-ruptured abdominal aortic aneurysm repair is 10.2%.13 A systematic review using the same dataset emphasised that although the worst mortality rates were from low volume hospitals, excellent results were achievable in occasional low volume hospitals.14

    These data show that operative results of hospitals are central to whether screening saves or loses lives. We must tackle how acceptable mortality can be achieved across the whole country, perhaps using the protocols that led to such an acceptable mortality in the 41 EVAR trial centres.7 9 Without such safeguards, screening for abdominal aortic aneurysm may not bring the expected results and instead may cause regret about the new screening programme.


    • Competing interests: None declared.

    • Provenance and peer review: Commissioned; not externally peer reviewed.


    View Abstract

    Sign in

    Log in through your institution