Scholarlink

Background <p>Understanding the causal relationships between clinical outcomes and environmental exposures is critical for advancing public health interventions and personalized medicine. These causal relationships can be applied to augment medical decision-making or suggest hypotheses for healthcare research. In this study, we applied a causal inference algorithm to an EHR dataset on patients with asthma or related common respiratory conditions (<i>N</i> = 14,937).</p> Methods <p>The EHR data were accessed via an open service named the Integrated Clinical and Environmental Service (ICEES). A multivariate feature table was extracted that included integrated data on features representing demographic factors, clinical measures, and environmental exposures; namely, sex, race, obesity, prednisone use, airborne particulate matter exposure, major roadway/highway exposure, residential density, and annual number of emergency department (ED) or inpatient hospital visits for respiratory issues, which we used as a proxy for asthma attacks. We estimated underlying causal relationships from the data by applying a Principal Component algorithm to identify significant causal relationships between the extracted features and asthma attacks. We also performed simulated interventions on the inferred causal network to detect the causal effects, in terms of shifts in the probability distribution for annual ED or inpatient hospital visits for respiratory issues.</p> Results <p>We found that obesity and prednisone were causally related to annual ED or inpatient visits in our causal inference model, and sex and race were indirectly related to annual ED or inpatient visits via a causal relationship to obesity. We further found that interventions in which all patients are simulated as obese or using prednisone (but not female) caused a shift to the right in the probability distribution of annual ED or inpatient visits for respiratory issues, thus supporting the results of our causal analysis, which demonstrated direct effects of obesity and prednisone (but not sex) on asthma attacks.</p> Conclusions <p>We successfully applied a causal model to the open ICEES service and identified direct causal relationships between prednisone and obesity on the frequency of asthma attacks, with indirect effects of sex and race by way of obesity. Our simulated interventions provided further support for our causal analysis by demonstrating a shift to the right in the probability of asthma attacks with interventions that assume all patients are using prednisone or obese.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Causal analysis for multivariate integrated clinical and environmental exposures data

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年1月15日

查看全文
  • Meghamala Sinha,
  • Perry Haaland,
  • Ashok Krishnamurthy,
  • Bo Lan,
  • Stephen A. Ramsey,
  • Patrick L. Schmitt,
  • Priya Sharma,
  • Hao Xu,
  • Karamarie Fecho

摘要

Background

Understanding the causal relationships between clinical outcomes and environmental exposures is critical for advancing public health interventions and personalized medicine. These causal relationships can be applied to augment medical decision-making or suggest hypotheses for healthcare research. In this study, we applied a causal inference algorithm to an EHR dataset on patients with asthma or related common respiratory conditions (N = 14,937).

Methods

The EHR data were accessed via an open service named the Integrated Clinical and Environmental Service (ICEES). A multivariate feature table was extracted that included integrated data on features representing demographic factors, clinical measures, and environmental exposures; namely, sex, race, obesity, prednisone use, airborne particulate matter exposure, major roadway/highway exposure, residential density, and annual number of emergency department (ED) or inpatient hospital visits for respiratory issues, which we used as a proxy for asthma attacks. We estimated underlying causal relationships from the data by applying a Principal Component algorithm to identify significant causal relationships between the extracted features and asthma attacks. We also performed simulated interventions on the inferred causal network to detect the causal effects, in terms of shifts in the probability distribution for annual ED or inpatient hospital visits for respiratory issues.

Results

We found that obesity and prednisone were causally related to annual ED or inpatient visits in our causal inference model, and sex and race were indirectly related to annual ED or inpatient visits via a causal relationship to obesity. We further found that interventions in which all patients are simulated as obese or using prednisone (but not female) caused a shift to the right in the probability distribution of annual ED or inpatient visits for respiratory issues, thus supporting the results of our causal analysis, which demonstrated direct effects of obesity and prednisone (but not sex) on asthma attacks.

Conclusions

We successfully applied a causal model to the open ICEES service and identified direct causal relationships between prednisone and obesity on the frequency of asthma attacks, with indirect effects of sex and race by way of obesity. Our simulated interventions provided further support for our causal analysis by demonstrating a shift to the right in the probability of asthma attacks with interventions that assume all patients are using prednisone or obese.