Beusterien Kathleen M, Dziekan Kristina, Flood Emuella, Harding Gale, Jordan Jamie C
MEDTAP International, Bethesda, MD 20814, USA.
Value Health. 2005 Jul-Aug;8(4):453-61. doi: 10.1111/j.1524-4733.2005.00036.x.
Choosing among HIV medications involve making trade-offs among various efficacy, convenience, resistance, and side-effect attributes. This study tested the feasibility of using adaptive conjoint analysis (ACA) to assess preferences (utilities) for HIV medication attributes.
HIV individuals were recruited through newspaper advertisements. Participants completed a computerized ACA survey that assessed 12 attributes, including side effects, regimen convenience, resistance, and efficacy. Literature on third-agent HIV drugs was used to identify percentage risk and severity level descriptions for each attribute. Based on the ACA-derived utilities, we assessed the relative importance of the attributes by averaging individually calculated importance and estimated the percentages that would prefer selected HIV medications over others. To check validity of the ACA utilities, the survey also had respondents choose among medications with different attribute profiles.
The 35 respondents were primarily African Americans (94%) and unemployed (54%). Of these, 28 (80%) provided consistent responses and were analyzed. Of the 12 medication attributes evaluated, the risk of developing resistance, regimen convenience, and the risk of sleep disturbance had the greatest impact on preferences; each accounting for more than 8.5% of the variation in preferences. These were followed by risk of drug failure (8.2%), cholesterol elevation (7.1%), diarrhea (7.1%) and nausea (6.9%). The ACA utilities accurately predicted patients' actual medication choices 75% of the time.
Adaptive conjoint analysis was successful in predicting HIV treatment preferences under different medication scenarios. Resistance, regimen convenience, and sleep disturbance would likely make the most difference in the perceived value of a third-agent HIV medication.
在抗逆转录病毒药物中进行选择需要在多种疗效、便利性、耐药性和副作用特征之间进行权衡。本研究测试了使用自适应联合分析（ACA）评估抗逆转录病毒药物特征偏好（效用）的可行性。
通过报纸广告招募感染HIV的个体。参与者完成了一项计算机化的ACA调查，该调查评估了12个特征，包括副作用、治疗方案便利性、耐药性和疗效。利用关于第三代抗逆转录病毒药物的文献确定每个特征的风险百分比和严重程度描述。基于ACA得出的效用，我们通过平均个体计算的重要性来评估特征的相对重要性，并估计相比于其他药物更倾向于选择特定抗逆转录病毒药物的百分比。为检验ACA效用的有效性，该调查还让受访者在具有不同特征概况的药物中进行选择。
35名受访者主要是非洲裔美国人（94%）且失业（54%）。其中，28人（80%）提供了一致的回答并纳入分析。在评估的12个药物特征中，产生耐药性的风险、治疗方案便利性和睡眠障碍风险对偏好影响最大；每个特征占偏好差异的比例均超过8.5%。其次是治疗失败风险（8.2%）、胆固醇升高（7.1%）、腹泻（7.1%）和恶心（6.9%）。ACA效用在75%的时间里准确预测了患者的实际用药选择。
自适应联合分析成功预测了不同用药场景下的HIV治疗偏好。耐药性、治疗方案便利性和睡眠障碍可能对第三代抗逆转录病毒药物的感知价值影响最大。
