Protected Health Information Research Articles

The Disclosure of Personally Identifiable Information in Studies of Neighborhood Contexts and Patient Outcomes.

Clinical epidemiology and patient-oriented health care research that incorporates neighborhood-level data is becoming increasingly common. A key step in conducting this research is converting patient address data to longitude and latitude data, a process known as geocoding. Several commonly used approaches to geocoding (eg, ggmap or the tidygeocoder R package) send patient addresses over the internet to web-based third-party geocoding services. Here, we describe how these approaches to geocoding disclose patients’ personally identifiable information (PII) and how the subsequent publication of the research findings discloses the same patients’ protected health information (PHI). We explain how these disclosures can occur and recommend strategies to maintain patient privacy when studying neighborhood effects on patient outcomes.

Ask The Expert Geriatric Psychiatry: VA program to support clinicians

This symposium addresses a critical need in geriatric psychiatry staffing shortages. The Department of Veterans Affairs is increasingly catering to aging veterans with complex medical and emotional problems. Unfortunately, there is a critical shortage of geriatric psychiatry expertise. The COVID-19 pandemic has further amplified the need for geriatric psychiatry services particularly in rural areas. In an effort to extend the valuable and too-often-scarce resource of geriatric psychiatry expertise, the National Telemental Health Center launched a new consultation service, Ask the Expert program in Geriatric Psychiatry, earlier this year. A team of VA Geriatric Psychiatrists were made available to answer general questions that VA clinicians may have about psychiatric care of older Veterans with potentially complex and interacting psychiatric, medical, neurocognitive, and behavioral care needs. The program aimed to provide a response within 7 days. Providers from the field were encouraged not share any Protected Health Information (PHI) in your questions. The service was not intended to make case-specific clinical recommendations, review patient health records, or respond to psychiatric emergencies.Speaker Prasad Padala will discuss the conceptualization, process of development and the National support needed for actualization of the program. He will also provide an overview of common themes in the questions received so far.Panelists: Dr. Ruchi Aggarwal will review the infrastructure, coverage and day to day operation of the program. She will further discuss the performance of the team.Dr. Donovan Maust will provide illustrative examples of polypharmacy, de-prescription and management of complex medical comorbidities.Dr. Gauri Khatkhate will provide illustrative example for management of neuropsychiatric symptoms of dementia.Dr. Susan Schultz will serve as the discussant and will highlight the need for workforce development in geriatric psychiatry. She will also discuss AAGPs efforts in possible targeted certification (e.g., Geriatric Psychopharmacology) and models by other organizations.The symposium will include audience discussion that may include stakeholder input from clinicians, educators, association leaders and staff.

The role of collectivism and moderating effect of IT proficiency on intention to disclose protected health information.

This paper aims to identify and understand factors affecting insiders’ intention to disclose patients’ medical information and to investigate how these factors affect the intention to disclose. Based on the literature review on deterrence theory and health information security awareness (HISA), we identify relevant factors and develop a research model explaining insiders’ intention to disclose patients’ health information. We collect data (N = 105) through scenario-based experiments. Results show that two personal factors, collectivism, and IT proficiency, play a significant role in the model. While collectivism affects two components (health information security regulation awareness and punishment severity awareness) of HISA which influences intention to disclose, IT proficiency moderates the relationship between HISA and intention to disclose. In addition, HISA negatively affects reporting assessment and intention to disclose. This paper aims to fill a research gap in understanding factors affecting insiders' intentions to disclose protected health information. We identify and investigate factors (e.g., collectivism, HISA, reporting assessment, and IT proficiency) that may affect insiders' disclosing intentions. We find that collectivism affects two components of HISA which influence reporting assessment and disclosing intention. We also discover that IT proficiency moderates the relationship between HISA and intention to disclose. Our findings suggest that we need to carefully consider personal factors such as collectivistic nature and IT proficiency in managing insiders' security breaches.

Strategies to Address the Lack of Labeled Data for Supervised Machine Learning Training With Electronic Health Records: Case Study for the Extraction of Symptoms From Clinical Notes.

BackgroundAutomated extraction of symptoms from clinical notes is a challenging task owing to the multidimensional nature of symptom description. The availability of labeled training data is extremely limited owing to the nature of the data containing protected health information. Natural language processing and machine learning to process clinical text for such a task have great potential. However, supervised machine learning requires a great amount of labeled data to train a model, which is at the origin of the main bottleneck in model development.ObjectiveThe aim of this study is to address the lack of labeled data by proposing 2 alternatives to manual labeling for the generation of training labels for supervised machine learning with English clinical text. We aim to demonstrate that using lower-quality labels for training leads to good classification results.MethodsWe addressed the lack of labels with 2 strategies. The first approach took advantage of the structured part of electronic health records and used diagnosis codes (International Classification of Disease–10th revision) to derive training labels. The second approach used weak supervision and data programming principles to derive training labels. We propose to apply the developed framework to the extraction of symptom information from outpatient visit progress notes of patients with cardiovascular diseases.ResultsWe used >500,000 notes for training our classification model with International Classification of Disease–10th revision codes as labels and >800,000 notes for training using labels derived from weak supervision. We show that the dependence between prevalence and recall becomes flat provided a sufficiently large training set is used (>500,000 documents). We further demonstrate that using weak labels for training rather than the electronic health record codes derived from the patient encounter leads to an overall improved recall score (10% improvement, on average). Finally, the external validation of our models shows excellent predictive performance and transferability, with an overall increase of 20% in the recall score.ConclusionsThis work demonstrates the power of using a weak labeling pipeline to annotate and extract symptom mentions in clinical text, with the prospects to facilitate symptom information integration for a downstream clinical task such as clinical decision support.

Privacy and safety: Issues of dual compliance in high-secure and other forensic psychiatric hospitals

Health care organizations are obligated to provide safe and effective treatment to their patients and also protect the safety of their workers. This paper analyzes the tensions arising from legislative regimes that, respectively, protect privacy and workplace safety, using a large, tertiary high-secure forensic psychiatric hospital in Ontario, Canada, as an example. In Ontario, the Personal Health Information Protection Act (PHIPA) prohibits personal health information (PHI) from being disclosed to individuals who fall outside the “circle of care,” including nonclinical employees who have direct involvement with patients and may be at risk of violence. PHIPA permits the disclosure of information where there is a risk of violence, but the statute's scheme for privacy protection was not designed to address, and may not be compatible with, the operations and requirements of high-secure forensic and other psychiatric hospitals. At the same time, the Occupational Health and Safety Act (OHSA) creates a regulatory framework that sets health and safety standards, including an employer's duty to disclose the risk of violence. OHSA prosecutions and proceedings demonstrate how these duties have been enforced against psychiatric hospitals. We examine this regulatory backdrop, explaining that PHIPA provides little guidance to psychiatric hospitals, where the risk of violence is elevated. We also discuss issues of dual compliance that arise from a hospital's legal obligations under PHIPA and OHSA. Finally, we turn to the ongoing clinical and operational challenges, suggesting strategies for increasing staff safety. These include strengthening the therapeutic alliance and providing patients with the option of consenting to disclosure of PHI to those outside the circle of care.

Privacy as the Price of Drug Access

In response to the recent increase in FDA-approved specialty drugs and escalating specialty drug prices, drug companies now offer patient support programs (“PSPs”) for eligible patients prescribed a particular pharmaceutical drug. Such programs encompass both financial assistance for the purchase of a specialty drug and behavioral services, including nursing support and injection training, intended to improve drug adherence. Although ostensibly gratuitous, these programs have a steep and underappreciated cost: disclosure of protected health information. In effect, patient support programs compel patients to trade protected health information for drug access. This Article provides the first in- depth examination of the legal and ethical concerns associated with patient support programs. Enrollment in a drug company’s patient support program furnishes the company with linked patient- and prescriber- identifying information for each enrollee, data which may enable drug companies to target marketing to patients and healthcare providers with an otherwise unattainable degree of precision. Moreover, once a drug company acquires an enrollee’s protected health information pursuant to a valid Health Insurance Portability and Accountability Act (HIPAA) authorization, a drug company faces few limits on downstream uses of those data. This Article illuminates a possible role for patient support program-mediated data collection in two unlawful drug company practices: (1) kickback schemes in coordination with foundations that cover pharmaceutical drug copays, and (2) “product hopping” to a new brand-name drug formulation after patent expiration of an older formulation. The current regime for health data privacy in the United States lacks adequate safeguards to prevent drug companies from exploiting patient support program-derived data to the detriment of patients. The Article ends by proposing practical modifications to the HIPAA Privacy Rule to modernize HIPAA’s protections vis-à-vis health data transferred from covered entities to noncovered entities such as drug companies.

DISCLOSURE OF MEDICAL RECORD INFORMATION TO GUARANTEE LEGAL ASPECTS OF CONFIDENTIALITY OF MEDICAL RECORDS IN TIRTAYASA PUSKESMAS

Background Medical Record according to Permenkes No.269/Menkes/Per/III/2008, Medical Record is a file that contains notes and documents about the patient's identity, examination results, treatment that has been given to the patient. Health facilities are responsible for protecting health information contained in medical records against possible loss, damage, falsification and unauthorized access. Maintaining information security, accuracy of information and ease of access to information are demands of health service organizations and health practitioners as well as authorized third parties. Meanwhile, parties who need information must always respect patient privacy. Overall, security, privacy, confidentiality and safety are devices that fortify information in medical records [1]. The purpose of this study was to determine the review of the release of medical record information in ensuring the legal aspects of the confidentiality of medical records at Kencana Hospital. Research method The research design used is descriptive research, which is a research method whose purpose is to present a complete picture of the social setting by describing a number of variables that are relevant to the problem and the unit under study among the phenomena being tested. This type of research is used to determine the process of providing medical record information and guaranteeing the legal aspects of medical record confidentiality at Kencana Hospital. The population in this study were all officers involved in the process of providing medical record information at Kencana Hospital which had 8 employees, which were divided into 3 medical record officers (1 head of medical records and 2 people served as distribution and retrieval officers), 2 people who handles related to insurance and 2 officers handles related to BPJS. In this study, the sample used was the party from the medical record unit, namely the medical record officer who has the authority in the process of providing patient medical information to a third party. The instruments used are Checklist Sheets and interview guidelines. In this study, the sampling used was purposive sampling technique because in sampling the authors considered certain things related to the research, The results of the research on Standard
 
 
 Operating Procedures related to providing patient medical information at the Kencana Hospital have an SOP. This SOP contains points for the process of releasing patient medical information, each point with a slightly different procedure depending on who the 3rd party is requesting. What is meant by 3rd parties here are patients, insurance companies, and legal institutions (courts). Each of these 3rd parties has made their respective procedures. The parties involved in the process of releasing patient medical information at Kencana Hospital are all medical record officers totaling 15 people (8 officers in the filing section), (2 people in the distribution of medical record files), and (5 people in the registration section). The information released is in the form of a photocopy of a medical resume sheet. At Kencana Hospital, the only party who can decide on the release of patient medical information is the medical record officer. Kencana Hospital does not yet have special security related to the process of releasing patient medical information, but if a patient is represented in the process of requesting release of information, the representative must have a power of attorney from the patient concerned. If the request for release of information is not represented, the patient will be given a release approval form from the medical record unit. Suggestion It is better to make Standard Operating Procedures related to the Release of Patient Medical Information referring to the Regulation of the Minister of Health of the Republic of Indonesia Number 36 of 2012 concerning Medical Secrets and Guidelines for the Implementation and Procedures of Indonesian Hospital Medical Records issued by the Director General of Yanmed.

The cell phone quandary

Cell phones are used by almost everyone and have become an integral part of our daily life. They are an almost universal instrument for gathering and transmitting information. The amount of bacteria crawling on a typical cell phone has been a point of contention, although studies show that they contain at least 10 times the amount of bacteria found on most toilet seats. As medical workers in hospital and clinic settings, we use our cell phones extensively, for paging, texting, calling, and recreational activities; thus, the risk of contamination by pathogens is a legitimate issue. Cell phones in the operating room may give patients the impression that they are not the main priority or the center of care, in addition to being a potential source of infection. Cell phones also have been found to be the number-one productivity killer in workplaces, amongst a long list of other distractions. All workers must be fully aware of their Health Insurance Portability and Accountability Act (HIPAA) compliance responsibilities and obligations, including the protection of health information, while using their cell phones at workplaces. On clinic or hospital grounds, reasonable diligence and strict adherence to cell phone policies may help us maintain greater safety, productivity, and professionalism, resulting in better service for our patients.

Investigation of the Utility of Features in a Clinical De-identification Model: A Demonstration Using EHR Pathology Reports for Advanced NSCLC Patients.

BackgroundElectronic health record (EHR) systems contain a large volume of texts, including visit notes, discharge summaries, and various reports. To protect the confidentiality of patients, these records often need to be fully de-identified before circulating for secondary use. Machine learning (ML) based named entity recognition (NER) model has emerged as a popular technique of automatic de-identification.ObjectiveThe performance of a machine learning model highly depends on the selection of appropriate features. The objective of this study was to investigate the usability of multiple features in building a conditional random field (CRF) based clinical de-identification NER model.MethodsUsing open-source natural language processing (NLP) toolkits, we annotated protected health information (PHI) in 1,500 pathology reports and built supervised NER models using multiple features and their combinations. We further investigated the dependency of a model's performance on the size of training data.ResultsAmong the 10 feature extractors explored in this study, n-gram, prefix–suffix, word embedding, and word shape performed the best. A model using combination of these four feature sets yielded precision, recall, and F1-score for each PHI as follows: NAME (0.80; 0.79; 0.80), LOCATION (0.85; 0.83; 0.84), DATE (0.86; 0.79; 0.82), HOSPITAL (0.96; 0.93; 0.95), ID (0.99; 0.82; 0.90), and INITIALS (0.97; 0.49; 0.65). We also found that the model's performance becomes saturated when the training data size is beyond 200.ConclusionManual de-identification of large-scale data is an impractical procedure since it is time-consuming and subject to human errors. Analysis of the NER model's performance in this study sheds light on a semi-automatic clinical de-identification pipeline for enterprise-wide data warehousing.

Synthetic Medical Images for Robust,Privacy-Preserving Training of Artificial Intelligence: Application to Retinopathy of Prematurity Diagnosis.

PurposeDeveloping robust artificial intelligence (AI) models for medical image analysis requires large quantities of diverse, well-chosen data that can prove challenging to collect because of privacy concerns, disease rarity, or diagnostic label quality. Collecting image-based datasets for retinopathy of prematurity (ROP), a potentially blinding disease, suffers from these challenges. Progressively growing generative adversarial networks (PGANs) may help, because they can synthesize highly realistic images that may increase both the size and diversity of medical datasets.DesignDiagnostic validation study of convolutional neural networks (CNNs) for plus disease detection, a component of severe ROP, using synthetic data.ParticipantsFive thousand eight hundred forty-two retinal fundus images (RFIs) collected from 963 preterm infants.MethodsRetinal vessel maps (RVMs) were segmented from RFIs. PGANs were trained to synthesize RVMs with normal, pre-plus, or plus disease vasculature. Convolutional neural networks were trained, using real or synthetic RVMs, to detect plus disease from 2 real RVM test datasets.Main Outcome MeasuresFeatures of real and synthetic RVMs were evaluated using uniform manifold approximation and projection (UMAP). Similarities were evaluated at the dataset and feature level using Fréchet inception distance and Euclidean distance, respectively. CNN performance was assessed via area under the receiver operating characteristic curve (AUC); AUCs were compared via bootstrapping and Delong’s test for correlated receiver operating characteristic curves. Confusion matrices were compared using McNemar’s chi-square test and Cohen’s κ value.ResultsThe CNN trained on synthetic RVMs showed a significantly higher AUC (0.971; P = 0.006 and P = 0.004) and classified plus disease more similarly to a set of 8 international experts (κ = 0.922) than the CNN trained on real RVMs (AUC = 0.934; κ = 0.701). Real and synthetic RVMs overlapped, by plus disease diagnosis, on the UMAP manifold, showing that synthetic images spanned the disease severity spectrum. Fréchet inception distance and Euclidean distances suggested that real and synthetic RVMs were more dissimilar to one another than real RVMs were to one another, further suggesting that synthetic RVMs were distinct from the training data with respect to privacy considerations.ConclusionsSynthetic datasets may be useful for training robust medical AI models. Furthermore, PGANs may be able to synthesize realistic data for use without protected health information concerns.

Visible Light Imaging: Clinical Aspects with an Emphasis on Medical Photography—a HIMSS-SIIM Enterprise Imaging Community Whitepaper

Photodocumentation is a subset of visible light imaging and is an important growing segment of enterprise imaging. Medical videography is another subset of visible light imaging that shares many of the challenges of photodocumentation. Medical photographs are used to document clinical conditions, support diagnosis, guide, and document procedures and to enable collaboration among colleagues. They also play a significant role in patient engagement and are a mechanism for patients to share information with their provider without the need for a clinical office visit. The content of medical photographs raises issues for acquisition, management, storage, and access. Medical photographs may contain protected health information, and these images benefit from the standardized, secure processes inherent in any enterprise imaging program. The ability to securely acquire images on mobile, and sometimes personally owned devices, is a necessity. In addition to containing protected health information, photograph content can be sensitive or gruesome or the images may be used for forensic purposes. These types of images require additional protections. Access to these images should be role-based and auditable. To properly identify photographs and to convey information about their acquisition parameters new metadata requirements and mechanisms for its association with the imaging files are evolving. Institutional policies need to be developed to define the organization’s requirements for medical photography, including consent processes. Existing policies such as those defining the designated record set and legal health record should address the management of medical photography.

Protected Health Information Recognition by Fine-Tuning a Pre-training Transformer Model.

ObjectivesDe-identifying protected health information (PHI) in medical documents is important, and a prerequisite to de-identification is the identification of PHI entity names in clinical documents. This study aimed to compare the performance of three pre-training models that have recently attracted significant attention and to determine which model is more suitable for PHI recognition.MethodsWe compared the PHI recognition performance of deep learning models using the i2b2 2014 dataset. We used the three pre-training models—namely, bidirectional encoder representations from transformers (BERT), robustly optimized BERT pre-training approach (RoBERTa), and XLNet (model built based on Transformer-XL)—to detect PHI. After the dataset was tokenized, it was processed using an inside-outside-beginning tagging scheme and WordPiece-tokenized to place it into these models. Further, the PHI recognition performance was investigated using BERT, RoBERTa, and XLNet.ResultsComparing the PHI recognition performance of the three models, it was confirmed that XLNet had a superior F1-score of 96.29%. In addition, when checking PHI entity performance evaluation, RoBERTa and XLNet showed a 30% improvement in performance compared to BERT.ConclusionsAmong the pre-training models used in this study, XLNet exhibited superior performance because word embedding was well constructed using the two-stream self-attention method. In addition, compared to BERT, RoBERTa and XLNet showed superior performance, indicating that they were more effective in grasping the context.

Regionalization with Self-Organizing Maps for Sharing Higher Resolution Protected Health Information

This article addresses the challenge of sharing finer scale protected health information (PHI) while maintaining patient privacy by using regionalization to create higher resolution Health Insurance Portability and Accountability Act (HIPAA)-compliant geographical aggregations. We compare four regionalization approaches in terms of their fitness for analysis and display: max-p-regions, regionalization with dynamically constrained agglomerative clustering and partitioning (REDCAP), and self-organizing map (SOM) variants of each. Each method is used to create a configuration of regions that aligns with census boundaries, optimizes intraunit homogeneity, and maximizes the number of spatial units while meeting the minimum population threshold required for sharing PHI under HIPAA guidelines. The relative utility of each configuration was assessed with measures of model fit, compactness, homogeneity, and resolution. Adding the SOM procedure to max-p-regions resulted in statistically significant improvements for nearly all assessment measures, whereas the addition of SOM to REDCAP primarily degraded these measures. These differences can be attributed to the different impacts of SOM on top-down and bottom-up regionalization procedures. Overall, we recommend REDCAP, which outperformed on most measures. The SOM variant of max-p-regions (MSOM) could also be recommended, because it provided the highest resolution while maintaining suitable performance on all other measures.

Artificial Intelligence: A Review of Progress and Prospects in Medicine and Healthcare

Andrew NG, a leading philosopher in the field of Artificial Intelligence (AI) once quoted “AI is the new electricity” which has the potential to transform and drive every industry. The most important driving factor for the AI transformation will be data. Clive Humby, a data science entrepreneur was once quoted saying “data is the new oil” and data analytics being the “combustion engine” will drive the AI led innovations. The rapid rise of Artificial Intelligence technologies in the past decade, has inspired industries to invest in every opportunity for integrating AI solutions to their products. Research, development, and innovation in the field of AI are shaping various industries like automobile, manufacturing, finance, retail, supply chain management, and education among others. The healthcare industry has also been adopting the ways of AI into various workflows within the domain. With the evolution in computing and processing powers coupled with hardware modernizations, the adoption of AI looks more feasible than ever. Research and Innovations are happening in almost every field of healthcare and hospital workflows with the target of making healthcare processes more efficient & accessible, increase the overall state of healthcare, reduce physician stress levels, and increase the patient satisfaction levels. The conventional ways in which healthcare and clinical workflows have been operating are now starting to see the change with the integration of many data driven AI solutions. The digital innovations are making life easy for healthcare professionals allowing them to spend more time listening to the problems of patients and consequently increasing the patient satisfaction levels. However, there are limitations and concerns on security of Protected Health Information which have to be addressed for a seamless amalgamation of AI systems into the healthcare domain. Many papers have been published which mostly talk about one particular field/problem in the healthcare domain. No publications have covered the opportunities provided by AI technologies to the entire healthcare domain. This review paper discusses in detail about the progress AI has been able to make in the healthcare domain holistically and what the future of AI looks like. The paper also discusses about the implementation opportunities various AI technologies like Machine Learning, Deep Learning, Reinforcement Learning, Natural Language Processing, Computer Vision provide in different fields of healthcare and clinical workflows and how Artificial Intelligence systems will boost the capabilities of healthcare professionals in restoring the human touch in patient-physician encounters. A physician’s intuition and judgement will always remain better suited since each case, each health condition, and each person is unique in its own way, but AI methods can help enhance the accuracy of diagnosis, assist physicians in making improved and precise clinical decisions.

Clinical Notes De-Identification: Scoping Recent Benchmarks for n2c2 Datasets.

Publicly shared repositories play an important role in advancing performance benchmarks for some of the most important tasks in natural language processing (NLP) and healthcare in general. This study reviews most recent benchmarks based on the 2014 n2c2 de-identification dataset. Pre-processing challenges were uncovered, and attention brought to the discrepancies in reported number of Protected Health Information (PHI) entities among the studies. Improved reporting is required for greater transparency and reproducibility.

Principle-Based Approach for the De-Identification of Code-Mixed Electronic Health Records

Code-mixing is a phenomenon where at least two languages are combined in a hybrid manner in the context of a single conversation. The use of mixed language is widespread in multilingual and multicultural countries and poses significant challenges for the development of automated language processing tools. In Taiwan’s electronic health record (EHR) systems, unstructured EHR texts are usually represented in a mixture of English and Chinese which increases the difficulty for de-identification and synthetization of protected health information (PHI). We explored this problem by applying several state-of-the-art pre-trained mono- and multilingual language models and propose to exploit the principle-based approach (PBA) for the tasks of PHI recognition and resynthesis on a code-mixed EHR corpus annotated with 6 main categories and 25 subcategories of PHIs. A hierarchical principle slot schema is defined in the PBA to encode knowledge of code-mixed PHIs and utilize slots to learn from the training set to assemble principles for recognizing PHI mentions and synthesizing surrogates simultaneously. In addition, a semantic disambiguation process is implemented to disambiguate ambiguous PHI categories in the de-identification process and to dynamically extend the knowledge encoded in PBA during the knowledge augmentation process. The experiment results demonstrate that the proposed method can achieve the best micro- and macro-F-scores in comparison to the other mono- and multilingual language models fine-tuned on our code-mixed corpus.

Use of Patients' Protected Health Information to Solicit Hospital Funds: How did This Practice Come About?

Modifications to Health Insurance Portability and Accountability Act (HIPAA) have allowed for the disclosure of patient protected health information (PHI) for the purpose of hospital fundraising. The public has recently raised ethical concerns regarding these practices. We examined the forces that brought about these HIPAA modifications. We first examined 304 comments submitted to the proposed rule for the HIPPA regulation modifications. We additionally queried the OpenSecrets repository for lobbying activity by these commenters. We found that 57 out of the 304 comments pertained specifically to fundraising practices. The majority of comments were from hospital developmental (fundraising) offices (51%, 29 of 57 comments), and the majority (96%, 24 of 25 hospital comments; 83%, 34 of 41 total comments discussing PHI disclosure) supported additional PHI disclosure. There was a paucity of comments from physician organizations (1 of 57) and patient advocates (2 of 57). The majority of lobbying dollars (95% of over $81 million) were from commenters who favored the modifications. The lack of physician and patient representation in the rule-making process likely contributed to the creation of regulations that elicit ethical concerns in physicians, and potential harm for patients.

Resource Depletion Risk for Medical Equipment: Embracing the Circular Economy.

Resource Depletion Risk for Medical Equipment: Embracing the Circular Economy.

The Protection Mechanism of Personal Health Information in the Digital Economy Environment.

With the application of big data technology in the medical field, personal health information is experiencing unprecedented challenges while showing great value. Opportunities and risks are like two sides of the same coin, and the rapid development of science and technology has not only brought valuable opportunities, but also made modern society seem to have developed into a risk society. Although countries around the world are committed to promoting the development and application of medical data and pay great attention to the protection of personal health information, the risk attributes of personal health information have rarely been mentioned before. This article first introduces the origin and development of medical big data, paving the way for the special risks faced by personal health information in this context. Then, through the introduction of concepts and characteristics, the content of personal health information is accurately defined. Next, according to a large number of personal health information security incidents, the new semantic scope of personal health information risk in the medical big data environment is analyzed, the necessity of personal health information risk regulation is emphasized, and the value conflicts that may occur in the process of personal health information risk regulation are clarified, laying the groundwork for the establishment of the risk decision-making procedure in the following article.

Classifying Cyber-Risky Clinical Notes by Employing Natural Language Processing.

Clinical notes, which can be embedded into electronic medical records, document patient care delivery and summarize interactions between healthcare providers and patients. These clinical notes directly inform patient care and can also indirectly inform research and quality/safety metrics, among other indirect metrics. Recently, some states within the United States of America require patients to have open access to their clinical notes to improve the exchange of patient information for patient care. Thus, developing methods to assess the cyber risks of clinical notes before sharing and exchanging data is critical. While existing natural language processing techniques are geared to de-identify clinical notes, to the best of our knowledge, few have focused on classifying sensitive-information risk, which is a fundamental step toward developing effective, widespread protection of patient health information. To bridge this gap, this research investigates methods for identifying security/privacy risks within clinical notes. The classification either can be used upstream to identify areas within notes that likely contain sensitive information or downstream to improve the identification of clinical notes that have not been entirely de-identified. We develop several models using unigram and word2vec features with different classifiers to categorize sentence risk. Experiments on i2b2 de-identification dataset show that the SVM classifier using word2vec features obtained a maximum F1-score of 0.792. Future research involves articulation and differentiation of risk in terms of different global regulatory requirements.