Use of Electronic Health Record Data to Evaluate the Impact of Race on 30-Day Mortality in Patients Admitted to the Intensive Care Unit

Mundkur, Mallika; Callaghan, Fiona; Abhyankar, Swapna; McDonald, Clement J.

doi:10.1007/s40615-016-0256-6

Cited by 4 publications

(10 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Multiparameter Intelligent Monitoring in Intensive Care (MIMIC) is a public dataset that includes de-identified information about patient demographics and ICU stays for patients [ 41 ]. In [ 21 ], the authors used MIMIC-II to identify the impact of Race on 30-day mortality of the patients. A synthetic version of the analyzed dataset was created based on MIMIC-III using HealthGAN and the results were duplicated [ 16 ].…”

Section: Resultsmentioning

confidence: 99%

“…MIMIC-III is an extension to this dataset which includes ICU stay details of over 40,000 patients from the year 2001 to 2012. In [ 21 ], the authors used MIMIC-II to create a dataset for understanding the impact of race on 30-day mortality. Andrew et.…”

Section: Methodsmentioning

confidence: 99%

“…The age in the dataset is split into bins: 15-24, 25-34, 35-44, 45-54, 55-64, 65-74 and 75+. We average the sleep time for the various age groups with the results for youngsters (15)(16)(17)(18)(19)(20)(21)(22)(23)(24) and elderly (75+) shown in Figure 1.…”

Section: Case Study 2: Average Sleep Time Of Americans Based On Atus Datasetmentioning

confidence: 99%

“…Case Study 1 examines the fairness of synthetic version of an extract from the popular MIMIC-III dataset designed to duplicate a prior research study [ 16 ]. The research study explored the impact of race on 30-day mortality based on odds-ratio of various comorbidities and demographic information [ 21 ]. We adapt a metric (log disparity) and a statistical test developed to evaluate the representativeness of the subgroups in randomized clinical trials to find inequities in the synthetic data as compared to the original data.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

The Problem of Fairness in Synthetic Healthcare Data

Bhanot

Miao

Erickson

et al. 2021

Entropy

View full text Add to dashboard Cite

Access to healthcare data such as electronic health records (EHR) is often restricted by laws established to protect patient privacy. These restrictions hinder the reproducibility of existing results based on private healthcare data and also limit new research. Synthetically-generated healthcare data solve this problem by preserving privacy and enabling researchers and policymakers to drive decisions and methods based on realistic data. Healthcare data can include information about multiple in- and out- patient visits of patients, making it a time-series dataset which is often influenced by protected attributes like age, gender, race etc. The COVID-19 pandemic has exacerbated health inequities, with certain subgroups experiencing poorer outcomes and less access to healthcare. To combat these inequities, synthetic data must “fairly” represent diverse minority subgroups such that the conclusions drawn on synthetic data are correct and the results can be generalized to real data. In this article, we develop two fairness metrics for synthetic data, and analyze all subgroups defined by protected attributes to analyze the bias in three published synthetic research datasets. These covariate-level disparity metrics revealed that synthetic data may not be representative at the univariate and multivariate subgroup-levels and thus, fairness should be addressed when developing data generation methods. We discuss the need for measuring fairness in synthetic healthcare data to enable the development of robust machine learning models to create more equitable synthetic healthcare datasets.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Case Study 2: Average Sleep Time Of Americans Based On Atus Datasetmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

The Problem of Fairness in Synthetic Healthcare Data

Bhanot

Miao

Erickson

et al. 2021

Entropy

View full text Add to dashboard Cite

show abstract

“…The first paper used MIMIC data to try to evaluate the impact of race on 30-day mortality [26]. The authors took different demographic variables and comorbidities and found the odds ratios for each of them to find out whether race specifically impacted the 30-day mortality of the patient.…”

Section: Impact Of Race On Icu Mortality Studymentioning

confidence: 99%

Generation and evaluation of privacy preserving synthetic health data

et al. 2020

View full text Add to dashboard Cite

We develop metrics for measuring the quality of synthetic health data for both education and research. We use novel and existing metrics to capture a synthetic dataset's resemblance, privacy, utility and footprint. Using these metrics, we develop an end-to-end workflow based on our generative adversarial network (GAN) method, HealthGAN, that creates privacy preserving synthetic health data. Our workflow meets privacy specifications of our data partner: (1) the HealthGAN is trained inside a secure environment; (2) the HealthGAN model is used outside of the secure environment by external users to generate synthetic data. This second step facilitates data handling for external users by avoiding de-identification, which may require special user training, be costly, or cause loss of data fidelity. This workflow is compared against five other baseline methods. While maintaining resemblance and utility comparable to other methods, HealthGAN provides the best privacy and footprint. We present two case studies in which our methodology was put to work in the classroom and research settings. We evaluate utility in the classroom through a data analysis challenge given to students and in research by replicating three different medical papers with synthetic data. Data, code, and the challenge that we organized for educational purposes are available.

show abstract