Department of Public Health, Erasmus University Medical Center, P.O. Box 2040, 3000 CA, Rotterdam, The Netherlands.
Department of Medical Decision Making, Leiden University Medical Center, P.O. Box 9600, 2300 RC, Leiden, The Netherlands.
Int J Cancer. 2018 Jun 1;142(11):2383-2393. doi: 10.1002/ijc.31265. Epub 2018 Feb 8.
Quality-adjusted life years are used in cost-effectiveness analyses (CEAs). To calculate QALYs, a "utility" (0-1) is used for each health state induced or prevented by the intervention. We aimed to estimate the impact of quality of life (QoL) assumptions (utilities and durations of health states) on CEAs of cervical cancer screening. To do so, 12 alternative sets of utility assumptions were retrieved from published cervical cancer screening CEAs. Two additional sets were based on empirical QoL data that were integrally obtained through two different measures (SF-6D and EQ-5D) from eight groups of women (total n = 3,087), from invitation for screening to diagnosis with cervical cancer. Per utility set we calculated the number of quality-adjusted days lost (QADL) for each relevant health state in cervical cancer screening, by multiplying the study-specific assumed disutilities (i.e., 1-utility) with study-specific durations of the loss in QoL, resulting in 14 "QADL-sets." With microsimulation model MISCAN we calculated cost-effectiveness of 342 alternative screening programs (varying in primary screening test [Human Papillomavirus (HPV) vs. cytology], starting ages, and screening interval) for each of the 14 QADL-sets. Utilities used in CEAs appeared to differ largely. We found that ten QADL-sets from the literature resulted in HPV and two in cytology as preferred primary test. The SF-6D empirical QADL-set resulted in cytology and the EQ-5D one in HPV as preferred primary test. In conclusion, assumed utilities and health state durations determine cost-effectiveness of cervical cancer screening. Also, the measure used to empirically assess utilities can be crucial for CEA conclusions.
质量调整生命年（QALYs）用于成本效益分析（CEA）。为了计算 QALYs，干预措施所导致或预防的每个健康状态都使用一个“效用值”（0-1）。我们旨在评估生活质量（QoL）假设（效用值和健康状态持续时间）对宫颈癌筛查CEA 的影响。为此，从已发表的宫颈癌筛查 CEA 中检索了 12 组替代效用假设。另外两组则基于通过两种不同措施（SF-6D 和 EQ-5D）从 8 组女性（总 n=3087）中整体获得的实证 QoL 数据，这些女性涵盖了从邀请参加筛查到确诊宫颈癌的全过程。对于每个效用集，我们通过将研究特异性假设的不效用值（即 1-效用值）与 QoL 损失的研究特异性持续时间相乘，计算出宫颈癌筛查中每个相关健康状态的质量调整天数损失（QADL），从而得出 14 个“QADL 集”。通过 MISCAN 微观模拟模型，我们为 14 个 QADL 集的每个集计算了 342 种不同筛查方案（在初始筛查测试[人乳头瘤病毒（HPV）与细胞学]、起始年龄和筛查间隔上有所不同）的成本效益。CEA 中使用的效用值似乎存在很大差异。我们发现，文献中的 10 个 QADL 集得出 HPV 为首选的初始筛查测试，2 个集得出细胞学为首选的初始筛查测试。SF-6D 实证 QADL 集得出细胞学为首选的初始筛查测试，EQ-5D 实证 QADL 集得出 HPV 为首选的初始筛查测试。总之，假设的效用值和健康状态持续时间决定了宫颈癌筛查的成本效益。此外，用于实证评估效用值的措施对于 CEA 结论可能至关重要。
