Observation of racial differences in the use of physical restraint: Response to Taylor et al. commentary "Premature conclusions".
Observation of racial differences in the use of physical restraint: Response to Taylor et al. commentary "Premature conclusions".
复制标题
观察身体束缚使用中的种族差异：对泰勒等人的回应
DOI：
10.1111/acem.14150
复制
发表时间：
2020
期刊：
影响因子：
--
通讯作者：
Bird,Suzanne
中科院分区：
文献类型：
--
作者：
Macias-Konstantopoulos,Wendy;Schnitzer,Kristina;Merideth,Flannery;Hayden,Douglas;Shtasel,Derri;Bird,Suzanne
作者感谢Cor-bin博士和他的同事以及我们的同事的评论，认为我们的发现是“过早的结论”。我们的同事主要关注的是在我们的分析中调整的混杂因素数量有限。事实上，虽然我们能够调整其中两个变量（物质使用和精神健康诊断），但我们在这项研究中没有可用的数据来分析感染，神经血管或代谢并发症或创伤的潜在影响。未来的分析应考虑检查这些变量;然而，应该注意的是，考虑的每个新变量或诊断需要足够大的“N”的身体限制患者来检测效应，这需要非常大的样本量。这项工作的复杂性是无法控制每个研究中的每个可能因素和每个统计模型，考虑到可用数据的限制，可达到的样本量和统计建模技术。对于分析决策，我们选择一次检查一个变量的混杂效应，以进行清晰的解释。用变量组合来解释非显著性发现并不简单，因为这可能是由于随着模型大小的增加而增加的多重共线性。也很难界定什么是对混杂因素的“充分调整”，因为理论上总是会有额外的变量或组合需要考虑。我们在本分析中的目标是不继续添加混杂因素以消除潜在影响。采用这种方法可能会对可能存在的临床护理差异视而不见。此外，我们提出了潜在的意义或解释的问题，如果我们发现种族的影响消失。如果一个外在变量或外在变量的组合似乎消除了种族的影响，这是否不仅暴露了种族群体之间的协变量不平衡，而这种不平衡本身可归因于社会中的系统性种族主义？对于我们的初步分析，我们希望提供一个全面的种族图片与多个种族群体的现有数据;然而，未来的分析当然可以而且应该继续与亚组，正如我们的同事提到的，允许检查其他变量单独和组合。我们同意并且已经注意到，该分析的局限性包括使用一个出院诊断，无法进一步梳理“不可用”种族类别，统计学显著性并不等同于临床显著性。然而，尽管统计和临床意义之间的差异，在这种情况下，我们确实觉得身体约束固有地携带临床轴承，因为这种做法会导致已知的身体，心理和情感痛苦的那些受。
The authors appreciate the comments of Dr. Cor-bin and his colleagues as well as those of our colleagues suggesting our findings are “premature conclusions.” A primary concern of our colleagues pertains to the limited number of confounders adjusted in our analyses. Indeed, while we were able to adjust for two of the variables noted (substance use and mental health diagnoses), we did not have the data available in this study to analyze potential effects of infection, neurovascular or metabolic complications, or trauma. Future analyses should consider examination of these variables; however, it should be noted that each new variable or diagnosis considered requires a large enough “N” of physically restrained patients to detect an effect, which necessitates very large sample sizes. Inherent in the complexity of this work is an inability to control for every possible factor in each study and with each statistical model, given limitations of available data, attainable sample size, and statistical modeling techniques. With respect to analytical decisions, we chose to examine confounding effects of one variable at a time to allow for clear interpretation. Interpretation of a nonsignificant finding with a combination of variables is not straightforward because this could be due to increasing multicollinearity with increasing model size. It is also difficult to delineate what constitutes “adequate adjustment” for confounders, because theoretically there will always be additional variables or combinations to consider. Our goal in this analysis was not to continue to add confounders in the hopes of removing a potential effect. To employ such an approach would potentially turn a blind eye to a difference in clinical care that may exist. Additionally, we raise the question of potential meaning or interpretation were we to find the effect of race to disappear. If an extrinsic variable or combination of extrinsic variables appears to remove the effect of race, does this not only expose covariate imbalance across racial groups plausibly itself attributable to systemic racism in society? For our initial analysis, we wanted to provide a comprehensive racial picture with available data on multiple racial groups; however, future analyses can and should certainly continue with subgroups, as our colleagues mention, to allow for examination of additional variables alone and in combination. We agree and have already noted that limitations of this analysis include use of one discharge diagnosis, inability to further tease apart the “unavailable” race category, and that statistical significance is not synonymous with clinical significance. However, despite the difference between statistical and clinical significance, in this case, we do feel physical restraint inherently carries clinical bearing as this practice causes known physical, psychological, and emotional distress to those subjected.