Data Management and Analysis Centre, Discipline of Public Health, University of Adelaide, Australia.
Pharmacoepidemiol Drug Saf. 2010 Jul;19(7):699-707. doi: 10.1002/pds.1942.
Observational studies have investigated the comparative safety of antipsychotics with varying results. Instrumental variable analysis has been suggested as a possible alternative to conventional analyses when there is concern about the effect of unmeasured confounding in observational studies. Using the example of the risk of death with typical compared to atypical antipsychotics, we aimed to explore the performance of two different instruments. We used the doctor prescribing preference instrument, which has been used in previous studies, to investigate further the assumptions of this instrument in the Australian population. We also propose an alternative instrument, nursing home facility preference.
With the Australian Department of Veterans' Affairs administrative claims database, we used an instrumental variable analysis to compare the risk of death after 12 months between the two antipsychotic classes.
Using the doctor prescribing preference instrument we estimated that typical antipsychotics were associated with an extra 24 (95% Confidence Interval (CI) 18-30) deaths per 100 patients per year compared to atypical antipsychotics, and an extra 10 (95% CI 7-14) deaths per 100 patients per year among nursing home residents. Facility prescribing preference was a stronger instrument (OR=19.2 95% CI 17.1-21.6) and provided a better balance of covariates than doctor prescribing preference.
Our study has shown that valid instruments in one population may not be directly applicable to other health care settings and testing of assumptions is crucial when performing IV analyses. Facility prescribing preference appears to be a potentially valid instrument for further work in this area.
观察性研究已经调查了具有不同结果的抗精神病药物的相对安全性。当观察性研究中存在对未测量混杂因素的影响的担忧时，工具变量分析已被提议作为常规分析的一种替代方法。使用典型与非典型抗精神病药物死亡风险的例子，我们旨在探索两种不同工具的性能。我们使用了以前研究中使用的医生处方偏好工具，以进一步研究该工具在澳大利亚人群中的假设。我们还提出了一种替代工具，即养老院设施偏好。
我们使用澳大利亚退伍事务部的行政索赔数据库，使用工具变量分析比较了两种抗精神病药物类别在 12 个月后死亡的风险。
使用医生处方偏好工具，我们估计与非典型抗精神病药物相比，典型抗精神病药物每年每 100 名患者多导致 24 例（95%置信区间（CI）18-30）死亡，而在养老院居民中每年每 100 名患者多导致 10 例（95%CI 7-14）死亡。设施处方偏好是一种更强的工具（OR=19.2 95%CI 17.1-21.6），并且在进行 IV 分析时提供了更好的协变量平衡。
我们的研究表明，一种人群中的有效工具可能不适用于其他医疗保健环境，并且在进行 IV 分析时，对假设的测试至关重要。设施处方偏好似乎是该领域进一步工作的潜在有效工具。
