Respiratory Medicine, Nottingham University Hospitals NHS Trust, Nottingham, UK.
Public Health, Erasmus MC, University Medical Center Rotterdam, Rotterdam, The Netherlands.
Thorax. 2022 Sep;77(9):882-890. doi: 10.1136/thoraxjnl-2021-217142. Epub 2021 Oct 29.
Lung cancer screening is effective if offered to people at increased risk of the disease. Currently, direct contact with potential participants is required for evaluating risk. A way to reduce the number of ineligible people contacted might be to apply risk-prediction models directly to digital primary care data, but model performance in this setting is unknown.
The Clinical Practice Research Datalink, a computerised, longitudinal primary care database, was used to evaluate the Liverpool Lung Project V.2 (LLP) and Prostate Lung Colorectal and Ovarian (modified 2012) (PLCO) models. Lung cancer occurrence over 5-6 years was measured in ever-smokers aged 50-80 years and compared with 5-year (LLP) and 6-year (PLCO) predicted risk.
Over 5 and 6 years, 7123 and 7876 lung cancers occurred, respectively, from a cohort of 842 109 ever-smokers. After recalibration, LLP produced a c-statistic of 0.700 (0.694-0.710), but mean predicted risk was over-estimated (predicted: 4.61%, actual: 0.9%). PLCO showed similar performance (c-statistic: 0.679 (0.673-0.685), predicted risk: 3.76%. Applying risk-thresholds of 1% (LLP) and 0.15% (PLCO), would avoid contacting 42.7% and 27.4% of ever-smokers who did not develop lung cancer for screening eligibility assessment, at the cost of missing 15.6% and 11.4% of lung cancers.
Risk-prediction models showed only moderate discrimination when applied to routinely collected primary care data, which may be explained by quality and completeness of data. However, they may substantially reduce the number of people for initial evaluation of screening eligibility, at the cost of missing some lung cancers. Further work is needed to establish whether newer models have improved performance in primary care data.
如果向疾病风险增加的人群提供肺癌筛查，则该筛查是有效的。目前，需要与潜在参与者直接接触以评估风险。减少接触不合格人数的一种方法可能是直接将风险预测模型应用于数字初级保健数据，但这种情况下的模型性能尚不清楚。
使用计算机化的纵向初级保健数据库临床实践研究数据链接（Clinical Practice Research Datalink）评估利物浦肺癌项目 V.2（LLP）和前列腺肺结直肠癌和卵巢（改良 2012 年）（PLCO）模型。在 50-80 岁的曾吸烟者中测量了 5-6 年内的肺癌发生情况，并将其与 5 年（LLP）和 6 年（PLCO）的预测风险进行了比较。
在 5 年和 6 年内，分别从 842109 名曾吸烟者队列中发生了 7123 例和 7876 例肺癌。经过重新校准，LLP 的 C 统计量为 0.700（0.694-0.710），但平均预测风险过高（预测：4.61%，实际：0.9%）。PLCO 表现出类似的性能（C 统计量：0.679（0.673-0.685），预测风险：3.76%）。应用 LLP 的 1%和 PLCO 的 0.15%风险阈值，可以避免对没有进行肺癌筛查资格评估的曾吸烟者中的 42.7%和 27.4%进行初始评估，但会遗漏 15.6%和 11.4%的肺癌。
当将风险预测模型应用于常规收集的初级保健数据时，其仅具有中等的区分度，这可能是由于数据的质量和完整性所致。但是，它们可以大大减少对筛查资格进行初始评估的人数，而代价是遗漏一些肺癌。需要进一步研究以确定新模型在初级保健数据中的性能是否有所提高。
