Garvican L, Littlejohns P
Department of Public Health Sciences, St George's Hospital Medical School, London, United Kingdom.
J Med Screen. 1996;3(3):123-8. doi: 10.1177/096914139600300304.
To evaluate the impact of the prevalent round of the NHS breast screening programme on the whole population of women aged 50-64 during the period 1988-92, by comparing the numbers of cancers detected at screening with those diagnosed symptomatically, in one United Kingdom health region (population 3.5 million). To relate this impact to the achievement of national quality standards and the observed sensitivity and specificity of the programme.
The breast screening programme computer systems and the Thames Cancer Registry database were used to diagnose cancers in women aged 50-64 during the period 1988-1992, who were classified into screen detected, interval cases, eligible but not yet invited, non-attenders, and those not registered with the programme.
The programme met all national quality assurance targets for uptake, diagnostic process, and detection, but screen detected cases comprised only 48% of those diagnosed during 1991 and 1992 when the breast screening programme was fully operational. As fewer than 40% of breast cancers registered occur in the eligible age range, this limits the impact of the programme to the possible early diagnosis of only 20% of all cases. A surprisingly high proportion occurred in women unknown to the programme, especially in inner London. The high number of interval cancers led to a programme sensitivity of only 73%, but 99.7% of women screening negative were correctly reassured as a result of screening- and for them the breast screening programme is a success.
Despite achievement of quality standards, the breast screening programme is apparently having a low impact on the overall diagnosis of breast cancer in South East Thames, so a large mortality reduction due to screening alone is unlikely. Population coverage will need to be improved, and it may be necessary to reduce the screening interval, or extend the age range, or both.
通过比较在英国一个健康区域（人口350万）1988 - 1992年期间，筛查发现的癌症病例数与有症状诊断出的病例数，评估国民健康服务体系（NHS）乳腺癌筛查项目普遍开展阶段对50 - 64岁女性总体人群的影响。将此影响与国家质量标准的达成情况以及该项目观察到的敏感度和特异度联系起来。
利用乳腺癌筛查项目计算机系统和泰晤士癌症登记数据库，诊断1988 - 1992年期间50 - 64岁女性的癌症，这些女性被分为筛查发现的病例、筛查间期癌、符合条件但尚未被邀请的女性、未参加筛查者以及未登记该项目的女性。
该项目在接受率、诊断过程和检测方面均达到了所有国家质量保证目标，但在1991年和1992年乳腺癌筛查项目全面运作时，筛查发现的病例仅占诊断病例的48%。由于登记的乳腺癌病例中不到40%发生在符合条件的年龄范围内，这限制了该项目的影响，仅可能对所有病例的20%进行早期诊断。令人惊讶的是，相当高比例的病例发生在该项目未知的女性中，尤其是在伦敦市中心。筛查间期癌数量众多导致该项目的敏感度仅为73%，但99.7%筛查结果为阴性的女性因筛查而得到了正确的安心结论——对她们来说，乳腺癌筛查项目是成功的。
尽管达到了质量标准，但乳腺癌筛查项目对泰晤士河东南部地区乳腺癌的总体诊断影响显然较低，因此仅靠筛查大幅降低死亡率不太可能。需要提高人群覆盖率，可能有必要缩短筛查间隔，或扩大年龄范围，或两者都做。
