School of Government, Wellington School of Business and Government, Victoria University of Wellington, Wellington, New Zealand.
Department of Physical Fitness and Health, School of Sport Science, Beijing Sport University, Beijing, China.
Appl Health Econ Health Policy. 2024 Jul;22(4):555-568. doi: 10.1007/s40258-024-00881-5. Epub 2024 Apr 19.
There are limited studies comparing the health utility values of EQ-5D-5L and SF-6Dv2 within the same patient cohorts. The widespread transmission and recurring infections associated with Omicron variants amid the COVID-19 pandemic have resulted in substantial health detriments and increased utilisation of health care resources. This highlights the crucial need to assess the loss in quality-adjusted life years (QALYs). Therefore, this study aims to compare the ceiling and floor effects, agreement, correlation and responsiveness between EQ-5D-5L and SF-6Dv2 based on COVID-19 patients during the Omicron outbreak in China.
We recruited 694 COVID-19 patients across mainland China to participant in an online questionnaire survey from January to February 2023. The questionnaire encompassed queries concerning the sociodemographic and health details of the participants, who were requested to recollect their health status during and after experiencing COVID-19 using the EQ-5D-5L and SF-6Dv2 questionnaires. Epanechnikov kernel density plots were used to visualise the ceiling and floor effects for both instruments. Agreement was assessed by Bland-Altman graph and intraclass correlation coefficient (ICC). Correlation was evaluated using linear regression, Pearson's correlation and Spearman's correlation. The standardised response mean (SRM) and relative efficiency (RE) were used to examine the responsiveness of EQ-5D-5L and SF-6Dv2 at detecting the health improvement after COVID-19 infection and the difference in dichotomous health indicators.
In total, 648 valid responses from patients aged 35.6 ± 15.0 years were involved in analysis. The EQ-5D-5L utility indices were 0.58 ± 0.33 and 0.92 ± 0.14 during and after COVID-19 infection, respectively, which were significantly higher than indices of the SF-6Dv2 utility (0.43 ± 0.31 and 0.81 ± 0.19, p < 0.001). A ceiling effect of EQ-5D-5L larger than that of SF-6Dv2 was observed during COVID-19 infection (49.5% vs 21.6%). Intraclass correlation coefficients between EQ-5D-5L and SF-6Dv2 during and after COVID-19 infection were 0.69 and 0.55, respectively. The utility indices of EQ-5D-5L and SF-6Dv2 were highly correlated, with Pearson's correlation coefficients of 0.76 and 0.70 (p < 0.001) during and after COVID-19 infection, respectively. The spearman's correlations were moderate to high between dimensions of EQ-5D-5L and SF-6Dv2 (p < 0.01). Both EQ-5D-5L and SF-6Dv2 were responsive to detect health improvement after COVID-19 and the differences in dichotomous health indicators.
The utility indices generated by EQ-5D-5L and SF-6Dv2 in COVID-19 patients demonstrate strong correlation and responsiveness. However, the agreement between these two instruments does not reach a satisfactory level. Consequently, these two measures cannot be used interchangeably. In situations where apprehensions about ceiling effects affecting outcome measurement arise, it is advisable to consider SF-6Dv2 as a preferable outcome measure for studies on patients with COVID-19.
在同一患者队列中，比较 EQ-5D-5L 和 SF-6Dv2 的健康效用值的研究有限。在 COVID-19 大流行期间，奥密克戎变异株广泛传播并反复感染，导致健康严重受损，医疗资源利用增加。这凸显了评估质量调整生命年（QALYs）损失的迫切需要。因此，本研究旨在比较基于中国奥密克戎疫情的 COVID-19 患者的 EQ-5D-5L 和 SF-6Dv2 的天花板和地板效应、一致性、相关性和反应性。
我们在内地招募了 694 名 COVID-19 患者，于 2023 年 1 月至 2 月期间参与了一项在线问卷调查。问卷涵盖了参与者的社会人口学和健康详细信息，要求他们回忆在 COVID-19 期间和之后的健康状况，使用 EQ-5D-5L 和 SF-6Dv2 问卷。Epanechnikov 核密度图用于直观地显示两种仪器的天花板和地板效应。通过 Bland-Altman 图和组内相关系数（ICC）评估一致性。使用线性回归、Pearson 相关和 Spearman 相关评估相关性。使用标准化反应均值（SRM）和相对效率（RE）来检查 EQ-5D-5L 和 SF-6Dv2 在检测 COVID-19 感染后健康改善和二分健康指标差异方面的反应性。
共有 35.6 ± 15.0 岁的 648 名有效患者参与了分析。EQ-5D-5L 效用指数在 COVID-19 期间和之后分别为 0.58 ± 0.33 和 0.92 ± 0.14，明显高于 SF-6Dv2 效用指数（0.43 ± 0.31 和 0.81 ± 0.19，p<0.001）。在 COVID-19 感染期间，观察到 EQ-5D-5L 的天花板效应大于 SF-6Dv2（49.5%比 21.6%）。COVID-19 期间和之后，EQ-5D-5L 和 SF-6Dv2 之间的 ICC 分别为 0.69 和 0.55。EQ-5D-5L 和 SF-6Dv2 的效用指数高度相关，COVID-19 期间的 Pearson 相关系数为 0.76，COVID-19 之后的 Pearson 相关系数为 0.70（均 p<0.001）。EQ-5D-5L 和 SF-6Dv2 之间的 Spearman 相关性为中度至高度（p<0.01）。COVID-19 后，EQ-5D-5L 和 SF-6Dv2 均能灵敏地检测到健康改善和二分健康指标的差异。
EQ-5D-5L 和 SF-6Dv2 在 COVID-19 患者中生成的效用指数具有很强的相关性和反应性。然而，这两种仪器之间的一致性未达到令人满意的水平。因此，这两种措施不能互换使用。在对影响结果测量的天花板效应感到担忧的情况下，建议将 SF-6Dv2 作为 COVID-19 患者研究的首选结果测量。
Zotero 插件