Data Sheet 1_Exploring public preferences and demand for ovarian cancer screening: a discrete choice experiment.pdf

IntroductionRoutine population-level screening may in the future reduce the high mortality rates associated with late-stage ovarian cancer diagnosis. However, the voluntary nature of screening means that understanding the public’s acceptability of the benefits, harms and likely uptake of any potential screening programme is crucial to implementation.
ObjectiveTo measure public preferences towards the benefits and harms of a potential screening programme and to predict uptake.
MethodsAn online Discrete Choice Experiment was completed by 250 women 40-80 years old in England and Wales. Subjects were asked 12 questions where they were asked to choose between two hypothetical screening tests described in terms of four attributes; ovarian cancer deaths, false-positive, false-negative and overdiagnosis rates, and no screening. Responses were analysed using mixed logit regression.
ResultsIn total, 250 women completed the survey. Ovarian cancer deaths (0.42, [95% CI: 0.40 – 0.44]) was the most important attribute overall, followed by the rate of false positive results (0.30, [95% CI: 0.30-0.30]). However, there were high levels of heterogeneity with individuals exhibiting low levels of worry about ovarian cancer (OR=1.76 [95% CI: 1.17–2.69]), low perceived risk of ovarian cancer (OR=1.44 [95% 1.03–2.03]) or risk-averse individuals (OR=1.46 [95% CI: 1.05–2.04]) significantly more likely to opt for the no screening alternative. Oppositely, individuals who regularly participate in cervical screening (OR=0.63 [0.47–0.90]) were less likely to opt for no screening. Overall, results indicated participants would be willing to accept 2.59 (95% CI: 1.82 – 3.36) false-negative results, 205 (95% CI: 161 – 248) false-positive results and 2.35 (95% CI: 1.76-2.94) per 10,000 people screened to avoid 1 ovarian cancer-related death. Uptake analysis confirmed a high willingness to undergo screening across varying levels of benefits and harms.
ConclusionsCurrently ovarian cancer screening is not recommended as available screening methods do not offer benefits in terms of mortality reduction. The results of this study demonstrate a high demand for ovarian cancer screening and a willingness to trade between the benefits and risks of a potential test. Results of this study provide a useful resource for assessing the acceptability of future screening modalities which may become available.

引言
未来常规人群层面筛查或可降低因晚期卵巢癌诊断导致的高死亡率。然而，筛查具有自愿属性，因此了解公众对潜在筛查方案的收益、危害及参与意愿的接受度，对方案落地至关重要。

研究目标
评估公众对潜在卵巢癌筛查方案收益与危害的偏好，并预测其参与意愿。

研究方法
英格兰与威尔士地区250名年龄在40至80岁之间的女性完成了一项线上离散选择实验（Discrete Choice Experiment）。受试者需回答12道问题，在基于四项属性（卵巢癌死亡数、假阳性率、假阴性率及过度诊断率）描述的两种假想筛查方案与无筛查选项之间进行选择。采用混合Logit回归（mixed logit regression）对回收的应答数据进行分析。

研究结果
共计250名女性完成了本次调研。整体而言，卵巢癌死亡数（0.42，[95%置信区间（CI）：0.40~0.44]）是最重要的属性，其次为假阳性率（0.30，[95%CI：0.30~0.30]）。但研究存在显著的异质性：对卵巢癌担忧程度较低（比值比（Odds Ratio，OR）=1.76，[95%CI：1.17~2.69]）、感知自身卵巢癌患病风险较低（OR=1.44，[95%CI：1.03~2.03]）或风险规避型的受试者，选择无筛查选项的概率显著更高。与之相反，定期参与宫颈癌筛查的受试者（OR=0.63，[95%CI：0.47~0.90]）选择无筛查的概率更低。整体结果显示，每避免1例卵巢癌相关死亡，受试者愿意接受每10000名筛查者中出现2.59例假阴性结果（95%CI：1.82~3.36）、205例假阳性结果（95%CI：161~248）以及2.35例过度诊断情况（95%CI：1.76~2.94）。参与意愿分析证实，无论筛查收益与危害水平如何变化，受试者均表现出较高的筛查参与意愿。

研究结论
目前因现有筛查手段无法实现死亡率降低的获益，故不推荐开展卵巢癌筛查。本研究结果显示，公众对卵巢癌筛查存在较高需求，且愿意在潜在筛查方案的收益与风险之间进行权衡。本研究结果可为评估未来可能投入使用的新型卵巢癌筛查手段的接受度提供重要参考依据。