A database de-identification framework to enable direct queries on medical data for secondary use

Clinical records anonymisation and text extraction (CRATE): an open-source software system

Background

Electronic medical records contain information of value for research, but contain identifiable and often highly sensitive confidential information. Patient-identifiable information cannot in general be shared outside clinical care teams without explicit consent, but anonymisation/de-identification allows research uses of clinical data without explicit consent.

Results

This article presents CRATE (Clinical Records Anonymisation and Text Extraction), an open-source software system with separable functions: (1) it anonymises or de-identifies arbitrary relational databases, with sensitivity and precision similar to previous comparable systems; (2) it uses public secure cryptographic methods to map patient identifiers to research identifiers (pseudonyms); (3) it connects relational databases to external tools for natural language processing; (4) it provides a web front end for research and administrative functions; and (5) it supports a specific model through which patients may consent to be contacted about research.

Conclusions

Creation and management of a research database from sensitive clinical records with secure pseudonym generation, full-text indexing, and a consent-to-contact process is possible and practical using entirely free and open-source software.

iGAS: A framework for using electronic intraoperative medical records for genomic discovery

OBJECTIVE

Design and implement a HIPAA and Integrating the Healthcare Enterprise (IHE) profile compliant automated pipeline, the integrated Genomics Anesthesia System (iGAS), linking genomic data from the Mount Sinai Health System (MSHS) BioMe biobank to electronic anesthesia records, including physiological data collected during the perioperative period. The resulting repository of multi-dimensional data can be used for precision medicine analysis of physiological readouts, acute medical conditions, and adverse events that can occur during surgery.

MATERIALS AND METHODS

A structured pipeline was developed atop our existing anesthesia data warehouse using open-source tools. The pipeline is automated using scheduled tasks. The pipeline runs weekly, and finds and identifies all new and existing anesthetic records for BioMe participants.

RESULTS

The pipeline went live in June 2015 with 49.2% (n=15,673) of BioMe participants linked to 40,947 anesthetics. The pipeline runs weekly in minimal time. After eighteen months, an additional 3671 participants were enrolled in BioMe and the number of matched anesthetic records grew 21% to 49,545. Overall percentage of BioMe patients with anesthetics remained similar at 51.1% (n=18,128). Seven patients opted out during this time. The median number of anesthetics per participant was 2 (range 1-144). Collectively, there were over 35 million physiologic data points and 480,000 medication administrations linked to genomic data. To date, two projects are using the pipeline at MSHS.

CONCLUSION

Automated integration of biobank and anesthetic data sources is feasible and practical. This integration enables large-scale genomic analyses that might inform variable physiological response to anesthetic and surgical stress, and examine genetic factors underlying adverse outcomes during and after surgery.

A risk-based framework for biomedical data sharing

The problem of biomedical data sharing is a form of gambling; on one hand it incurs the risk of privacy violations and on the other it stands to profit from knowledge discovery. In general, the risk of granting data access to a user depends heavily upon the data requested, the purpose for the access, the user requesting the data (user motives) and the security of the user's environment. While traditional manual biomedical data sharing processes (based on institutional review boards) are lengthy and demanding, the automated ones (known as honest broker systems) disregard the individualities of different requests and offer "one-size-fits-all" solutions to all data requestors. In this manuscript, we propose a conceptual risk-aware data sharing system; the system brings the concept of risk, from all contextual information surrounding a data request, into the data disclosure decision module. The decision module, in turn, imposes mitigation measures to counter the calculated risk.

Modular design, application architecture, and usage of a self-service model for enterprise data delivery: the Duke Enterprise Data Unified Content Explorer (DEDUCE)

PURPOSE

Data generated in the care of patients are widely used to support clinical research and quality improvement, which has hastened the development of self-service query tools. User interface design for such tools, execution of query activity, and underlying application architecture have not been widely reported, and existing tools reflect a wide heterogeneity of methods and technical frameworks. We describe the design, application architecture, and use of a self-service model for enterprise data delivery within Duke Medicine.

METHODS

Our query platform, the Duke Enterprise Data Unified Content Explorer (DEDUCE), supports enhanced data exploration, cohort identification, and data extraction from our enterprise data warehouse (EDW) using a series of modular environments that interact with a central keystone module, Cohort Manager (CM). A data-driven application architecture is implemented through three components: an application data dictionary, the concept of "smart dimensions", and dynamically-generated user interfaces.

RESULTS

DEDUCE CM allows flexible hierarchies of EDW queries within a grid-like workspace. A cohort "join" functionality allows switching between filters based on criteria occurring within or across patient encounters. To date, 674 users have been trained and activated in DEDUCE, and logon activity shows a steady increase, with variability between months. A comparison of filter conditions and export criteria shows that these activities have different patterns of usage across subject areas.

CONCLUSIONS

Organizations with sophisticated EDWs may find that users benefit from development of advanced query functionality, complimentary to the user interfaces and infrastructure used in other well-published models. Driven by its EDW context, the DEDUCE application architecture was also designed to be responsive to source data and to allow modification through alterations in metadata rather than programming, allowing an agile response to source system changes.

Use of Electronic Health-Related Datasets in Nursing and Health-Related Research

Datasets of gigabyte size are common in medical sciences. There is increasing consensus that significant untapped knowledge lies hidden in these large datasets. This review article aims to discuss Electronic Health-Related Datasets (EHRDs) in terms of types, features, advantages, limitations, and possible use in nursing and health-related research. Major scientific databases, MEDLINE, ScienceDirect, and Scopus, were searched for studies or review articles regarding using EHRDs in research. A total number of 442 articles were located. After application of study inclusion criteria, 113 articles were included in the final review. EHRDs were categorized into Electronic Administrative Health-Related Datasets and Electronic Clinical Health-Related Datasets. Subcategories of each major category were identified. EHRDs are invaluable assets for nursing the health-related research. Advanced research skills such as using analytical softwares, advanced statistical procedures, dealing with missing data and missing variables will maximize the efficient utilization of EHRDs in research.

Protecting privacy in a clinical data warehouse

Peking University has several prestigious teaching hospitals in China. To make secondary use of massive medical data for research purposes, construction of a clinical data warehouse is imperative in Peking University. However, a big concern for clinical data warehouse construction is how to protect patient privacy. In this project, we propose to use a combination of symmetric block ciphers, asymmetric ciphers, and cryptographic hashing algorithms to protect patient privacy information. The novelty of our privacy protection approach lies in message-level data encryption, the key caching system, and the cryptographic key management system. The proposed privacy protection approach is scalable to clinical data warehouse construction with any size of medical data. With the composite privacy protection approach, the clinical data warehouse can be secure enough to keep the confidential data from leaking to the outside world.

Using patient lists to add value to integrated data repositories

Patient lists are project-specific sets of patients that can be queried in integrated data repositories (IDR's). By allowing a set of patients to be an addition to the qualifying conditions of a query, returned results will refer to, and only to, that set of patients. We report a variety of use cases for such lists, including: restricting retrospective chart review to a defined set of patients; following a set of patients for practice management purposes; distributing "honest-brokered" (deidentified) data; adding phenotypes to biosamples; and enhancing the content of study or registry data. Among the capabilities needed to implement patient lists in an IDR are: capture of patient identifiers from a query and feedback of these into the IDR; the existence of a permanent internal identifier in the IDR that is mappable to external identifiers; the ability to add queryable attributes to the IDR; the ability to merge data from multiple queries; and suitable control over user access and de-identification of results. We implemented patient lists in a custom IDR of our own design. We reviewed capabilities of other published IDRs for focusing on sets of patients. The widely used i2b2 IDR platform has various ways to address patient sets, and it could be modified to add the low-overhead version of patient lists that we describe.

Knowledge management and informatics considerations for comparative effectiveness research: a case-driven exploration

BACKGROUND

As clinical data are increasingly collected and stored electronically, their potential use for comparative effectiveness research (CER) grows. Despite this promise, challenges face those wishing to leverage such data. In this paper we aim to enumerate some of the knowledge management and informatics issues common to such data reuse.

DESIGN

After reviewing the current state of knowledge regarding biomedical informatics challenges and best practices related to CER, we then present 2 research projects at our institution. We analyze these and highlight several common themes and challenges related to the conduct of CER studies. Finally, we represent these emergent themes.

RESULTS

The informatics challenges commonly encountered by those conducting CER studies include issues related to data information and knowledge management (eg, data reuse, data preparation) as well as those related to people and organizational issues (eg, sociotechnical factors and organizational factors). Examples of these are described in further detail and a formal framework for describing these findings is presented.

CONCLUSIONS

Significant challenges face researchers attempting to use often diverse and heterogeneous datasets for CER. These challenges must be understood in order to be dealt with successfully and can often be overcome with the appropriate use of informatics best practices. Many research and policy questions remain to be answered in order to realize the full potential of the increasingly electronic clinical data available for such research.

Inferring community structure in healthcare forums. An empirical study

BACKGROUND

Detecting community structures in complex networks is a problem interesting to several domains. In healthcare, discovering communities may enhance the quality of web offerings for people with chronic diseases. Understanding the social dynamics and community attachments is key to predicting and influencing interaction and information flow to the right patients.

OBJECTIVES

The goal of the study is to empirically assess the extent to which we can infer meaningful community structures from implicit networks of peer interaction in online healthcare forums.

METHODS

We used datasets from five online diabetes forums to design networks based on peer-interactions. A quality function based on user interaction similarity was used to assess the quality of the discovered communities to complement existing homophily measures.

RESULTS

Results show that we can infer meaningful communities by observing forum interactions. Closely similar users tended to co-appear in the top communities, suggesting the discovered communities are intuitive. The number of years since diagnosis was a significant factor for cohesiveness in some diabetes communities.

CONCLUSION

Network analysis is a tool that can be useful in studying implicit networks that form in healthcare forums. Current analysis informs further work on predicting and influencing interaction, information flow and user interests that could be useful for personalizing medical social media.