Irstly, their simulation model utilizes information taken from the Utrecht and Nijmegen programmes, which started inside the mid1970s. The relevance for breast screening in the United kingdom in the late 1990s is unclear, offered the considerable variation in fundamental screening variables in between programmes, notably the interval cancer rates and screening detection prices of modest invasive cancers.two Secondly, their model shows a poor match together with the benefits of the second screening round in the north west region. It predicts that in the second screening test at 3 years over 60 of invasive cancers will probably be 10 mm in diameter and 9 > 20 mm; the corresponding observed frequencies are 40 and 19 . The impact on predicted mortality of poorly modelling the stage distribution of cancers detected at the second or later screen is most likely to become substantial, and even greater using a two year interval than a three year interval considering that cancers detected on screening are of reasonably higher value. Their estimates in the UNC-926 marginal effect of decreasing the screening interval will then be unreliable to an unknown extent. Thirdly, no uncertainty is attached for the numerous estimates. The authors claim that the marginal cost per life year gained of shortening the screening interval from 3 to two years is 545–a spuriously precise figure. This estimate is most likely to become extremely misleading. With uncertainties more than the data and the poor match in the model, sensitivity analyses are vital. Option data could produce marginal charges several occasions higher than the quoted estimate. The BMJ typically insists on uncertainty estimates, often as self-assurance intervals. For an article intended to influence policy, PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/20030894 omission of uncertainty bounds renders it almost valueless. Future policy choices for the national breast screening programme ought to be primarily based on evidence directly connected towards the United Kingdom’s programme itself. Such proof will shortly be offered in the benefits from the multicentre randomised trial of differentBMJ VOLUME 318 six FEBRUARY 1999 www.bmj.comMoney may be improved spent on symptomatic girls Editor–All the hype, promotional material, and leaflets inviting women to be screened carry the false guarantee of a 25 reduction in mortality from breast cancer. Boer et al explode this myth with out explicitly stating so.1 Their laptop or computer simulation study suggests that the current programme may possibly accomplish a 12.8 reduction in mortality, half that promised by the NHS breast screening programme when it was initiated. Even extending the age up to 69 or lowering the three year interval to two would come nowhere near matching the promises on which the whole infrastructure of this programme was based. To extend the age to 69 or shorten the interval would expense an further 0m a year. Having said that, it is actually the human sources which are most valuable. A current report from the Royal College of Radiologists described the parlous state on the radiological assistance for the current programme.two Morale is at an all time low, recruitment of radiologists to supply the existing service cannot be sustained, and for that reason any extension towards the programme at present is completely impractical. But to perform nothing is just not an solution, as Werneke and McPherson point out.3 However, they do not go far adequate. We must contemplate the resource implications and potential opportunity expenses applying not simply to an expansion from the programme but for the continuation in the programme since it is. Even though no doubt politically unaccept.