A Methodology to Generate Virtual Patient Repositories
📝 Abstract
Electronic medical records (EMR) contain sensitive personal information. For example, they may include details about infectious diseases, such as human immunodeficiency virus (HIV), or they may contain information about a mental illness. They may also contain other sensitive information such as medical details related to fertility treatments. Because EMRs are subject to confidentiality requirements, accessing and analyzing EMR databases is a privilege given to only a small number of individuals. Individuals who work at institutions that do not have access to EMR systems have no opportunity to gain hands-on experience with this valuable resource. Simulated medical databases are currently available; however, they are difficult to configure and are limited in their resemblance to real clinical databases. Generating highly accessible repositories of virtual patient EMRs while relying only minimally on real patient data is expected to serve as a valuable resource to a broader audience of medical personnel, including those who reside in underdeveloped countries.
💡 Analysis
Electronic medical records (EMR) contain sensitive personal information. For example, they may include details about infectious diseases, such as human immunodeficiency virus (HIV), or they may contain information about a mental illness. They may also contain other sensitive information such as medical details related to fertility treatments. Because EMRs are subject to confidentiality requirements, accessing and analyzing EMR databases is a privilege given to only a small number of individuals. Individuals who work at institutions that do not have access to EMR systems have no opportunity to gain hands-on experience with this valuable resource. Simulated medical databases are currently available; however, they are difficult to configure and are limited in their resemblance to real clinical databases. Generating highly accessible repositories of virtual patient EMRs while relying only minimally on real patient data is expected to serve as a valuable resource to a broader audience of medical personnel, including those who reside in underdeveloped countries.
📄 Content
A Methodology to Generate Virtual Patient Repositories Uri Kartoun PhD, Massachusetts General Hospital / Harvard Medical School ABSTRACT Electronic medical records (EMR) contain sensitive personal information. For example, they may include details about infectious diseases, such as human immunodeficiency virus (HIV), or they may contain information about a mental illness. They may also contain other sensitive information such as medical details related to fertility treatments. Because EMRs are subject to confidentiality requirements, accessing and analyzing EMR databases is a privilege given to only a small number of individuals. Individuals who work at institutions that do not have access to EMR systems have no opportunity to gain hands-on experience with this valuable resource. Simulated medical databases are currently available; however, they are difficult to configure and are limited in their resemblance to real clinical databases. Generating highly accessible repositories of virtual patient EMRs while relying only minimally on real patient data is expected to serve as a valuable resource to a broader audience of medical personnel, including those who reside in underdeveloped countries. Keywords: simulation in healthcare, electronic medical records, electronic health records INTRODUCTION The importance of patient privacy has been thoroughly emphasized by governmental resources such as the HIPAA Privacy Rule and by academia [1–3]. Numerous strategies to maintain patients’ privacy have been developed [4–8]; however, the medical profession is not yet able to guarantee full protection of privacy while providing detailed information about each patient [9]. Cohorts assembled from electronic medical records (EMRs) represent a powerful resource to study disease complications at a population level. Recent studies have demonstrated the usefulness of EMR analysis for discovering or confirming outcome correlations, subcategories of disease, and adverse drug events [10–16]. Due to confidentiality restrictions, accessing and analyzing EMR databases is a privilege given to only a small number of individuals. Individuals who work in institutions that do not have access to EMR systems cannot experiment with such valuable resources. When professors wish to teach a biomedical informatics course focused on EMR technology, they cannot distribute real EMR data among their students. Simulated medical databases [e.g., 17, 18] and open EMR platforms [19] are currently available; however, no confidentiality-free massive scale longitudinal EMR databases have yet been algorithmically created. Virtual patient repositories that bear a high degree of resemblance to real patient databases while relying only minimally on real patient data are expected to serve as a valuable resource for medical professionals in training, and to accelerate health care research and development. The aim of this study is to develop a novel methodology for creating virtual patient repositories. I demonstrate that a method entailing minimal configuration requirements can generate nonconfidential artificial EMR databases that could be used to practice statistical and machine-learning algorithms. I further demonstrate the potential broad public interest in the availability of this technology. MATERIALS AND METHODS The process of generating a virtual patient repository was based on preconfiguration of population-level and patient-level characteristics. First, an object-oriented program acquired a population-level configuration to generate patient objects. Next, the program created a clinical profile for each patient, including admissions associated with chief complaints and laboratory measurements. Finally, the patient objects were stored in a database. Following this methodology, three databases of 100, 10,000, and 100,000 virtual patients were created. Population-level configuration Population-level configuration specifies the number of individual records that will be generated and defines preconfigured values for demographic characteristics. Demographic characteristics include gender, marital status, major language, ethnicity, date of birth, and income level. Configuring categorical variables defines several potential values for the variable and the percentage of the population with the value. For example, in a population of n = 100,000 individuals, a potential configuration for ethnicity would be 49% white, 23% Asian, 15% African American, and the remainder unknown. For continuous variables such as age, the population percentages for several ranges of date of birth are defined. For example, dates of birth in the range of 1940 to 1950 were randomly created for 15% of the population. The configuration tables are presented in Supplemental Table 1(a–b). Generating a virtual patient repository Having acquired the population-level configuration, the program generated n objects each representing one virtual pat
This content is AI-processed based on ArXiv data.