Department of Epidemiology and Biostatistics, University of Maryland, College Park, MD; Center for Drug Evaluation and Research, U.S. Food and Drug Administration, College Park, MD.
Center for Drug Evaluation and Research, U.S. Food and Drug Administration, College Park, MD.
Am Heart J. 2021 Nov;241:108-119. doi: 10.1016/j.ahj.2021.03.016. Epub 2021 May 10.
An endpoint that has received some attention in recent cardiovascular trials is 'days alive and out of hospital' (DAOH). Percent DAOH is a natural extension of DAOH that adjusts for differences in length of follow-up. This endpoint measure incorporates mortality and morbidity together in a way that has the potential to give more insight regarding treatment effects compared to conventional time-to-event endpoints. Other advantages of this measure include the relative ease of collection and interpretation. However, research on how to analyze this measure is still limited.
We propose using the one-inflated beta model to analyze percent DAOH. This model is appropriate for highly left-skewed data with a large proportion of boundary values. Data from the Prospective Comparison of ARNI [Angiotensin Receptor-Neprilysin Inhibitor] with ACEI [Angiotensin-Converting-Enzyme Inhibitor] to Determine Impact on Global Mortality and Morbidity in Heart Failure Trial (PARADIGM-HF) and Candesartan in Heart Failure Assessment of Reduction in Mortality and morbidity (CHARM) trials are used to illustrate this method.
Statistically significant differences in percent DAOH were observed for PARADIGM-HF and CHARM in favor of treatment. In PARADIGM-HF, treatment with sacubitril plus valsartan increased DAOH on average by 11 days (95% CI: 1.4-20.9 days) and increased percent DAOH by 1.64% at a fixed follow-up length of 1,000 days (95% CI: 0.61%- 2.67%). For the CHARM overall program, the candesartan group has 1.79% more DAOH (95% CI: 0.91%- 2.68%).
DAOH, and especially percent DAOH, can enhance our understanding of treatment effects in future cardiovascular trials, and the one-inflated beta model is an appropriate choice for its analysis.
在最近的心血管试验中，一个受到关注的终点是“存活且出院天数”（DAOH）。百分比 DAOH 是对 DAOH 的自然扩展，它可以根据随访时间的不同进行调整。与传统的时间事件终点相比，这种终点测量方法将死亡率和发病率结合在一起，具有潜在的更深入了解治疗效果的能力。该测量方法的其他优点包括相对容易收集和解释。然而，关于如何分析这种测量方法的研究仍然有限。
我们建议使用一单位膨胀贝塔模型来分析百分比 DAOH。该模型适用于高度左偏态数据，且具有大量边界值。使用 Prospective Comparison of ARNI [血管紧张素受体-脑啡肽酶抑制剂]与 ACEI [血管紧张素转换酶抑制剂]以确定对心力衰竭试验（PARADIGM-HF）和坎地沙坦心力衰竭评估降低死亡率和发病率（CHARM）试验中全球死亡率和发病率影响的研究数据来举例说明这种方法。
PARADIGM-HF 和 CHARM 试验中，治疗组的百分比 DAOH 存在统计学显著差异，对治疗有利。在 PARADIGM-HF 中，与缬沙坦联合沙库巴曲相比，使用沙库巴曲缬沙坦可使 DAOH 平均增加 11 天（95%置信区间：1.4-20.9 天），在固定随访时间为 1000 天时，使百分比 DAOH 增加 1.64%（95%置信区间：0.61%-2.67%）。对于 CHARM 整体计划，坎地沙坦组的 DAOH 多 1.79%（95%置信区间：0.91%-2.68%）。
DAOH，特别是百分比 DAOH，可以增强我们对未来心血管试验中治疗效果的理解，一单位膨胀贝塔模型是其分析的合适选择。
