De Troeyer Katrien, Silversmit Geert, Rosskamp Michael, Truyen Inge, Van Herck Koen, Goossens Mathijs Michiel, Martens Patrick, Kellen Eliane, Hendrickx Erik, Rummens Elise, De Smet Frank, Broeders Mireille, Verdoodt Freija, De Schutter Harlinde
Belgian Cancer Registry, 1210 Brussels, Belgium; Family Medicine and Population Health, Department of Epidemiology and Social Medicine, University of Antwerp, 2000 Antwerp, Belgium.
Belgian Cancer Registry, 1210 Brussels, Belgium.
Cancer Epidemiol. 2023 Feb;82:102320. doi: 10.1016/j.canep.2022.102320. Epub 2023 Jan 4.
Breast cancer screening programs were introduced in many countries worldwide following randomized controlled trials in the 1980s showing a reduction in breast cancer-specific mortality. However, their effectiveness remains debated and estimates vary. A breast cancer screening program was introduced in 2001 in Flanders, Belgium where high levels of opportunistic screening practices are observed. The effectiveness of this program was estimated by measuring its effect on breast cancer-specific mortality.
We performed a case-referent study to investigate the effect of participation in the Flemish population-based mammography screening program (PMSP) on breast cancer-specific mortality from 2005 to 2017. A multiple logistic regression model assessed the association between breast cancer-specific death and screening program participation status in the four years prior to (pseudo)diagnosis (yes/no), with adjustment for potential confounders (individual socio-economic position and calendar year of diagnosis) and stratified for age. In addition, we performed different sensitivity analyses.
We identified 1571 cases and randomly selected 6284 referents. After adjustment, women who participated in PMSP had a 51 % lower risk of breast cancer-specific mortality compared to those who did not (adjusted odds ratio [aOR] =0.49, 95 % CI: 0.44-0.55). Sensitivity analyses did not markedly change the estimated associations. Correction for self-selection bias reduced the effect size, but the estimate remained significant.
Our results indicate that in a context of high opportunistic screening rates, participation in breast cancer screening program substantially reduces breast cancer-specific mortality. For policy, these results should be balanced against the potential harms of screening, including overdiagnosis and overtreatment.
20世纪80年代的随机对照试验表明乳腺癌筛查项目可降低乳腺癌特异性死亡率，此后全球许多国家都推行了该项目。然而，其有效性仍存在争议，估计结果也各不相同。2001年，比利时弗拉芒地区引入了乳腺癌筛查项目，该地区存在大量机会性筛查行为。通过测量该项目对乳腺癌特异性死亡率的影响来评估其有效性。
我们进行了一项病例对照研究，以调查参与弗拉芒地区基于人群的乳腺钼靶筛查项目（PMSP）对2005年至2017年乳腺癌特异性死亡率的影响。一个多重逻辑回归模型评估了在（假）诊断前四年乳腺癌特异性死亡与筛查项目参与状况（是/否）之间的关联，并对潜在混杂因素（个体社会经济地位和诊断年份）进行了调整，并按年龄分层。此外，我们还进行了不同的敏感性分析。
我们确定了1571例病例，并随机选择了6284名对照。调整后，参与PMSP的女性与未参与者相比，乳腺癌特异性死亡风险降低了51%（调整后的优势比[aOR]=0.49，95%置信区间：0.44-0.55）。敏感性分析并未显著改变估计的关联。对自我选择偏倚的校正降低了效应大小，但估计值仍具有显著性。
我们的结果表明，在机会性筛查率较高的情况下，参与乳腺癌筛查项目可大幅降低乳腺癌特异性死亡率。对于政策而言，这些结果应与筛查的潜在危害（包括过度诊断和过度治疗）相权衡。
