Feasibility of Reidentifying Individuals in Large National Physical Activity Data Sets From Which Protected Health Information Has Been Removed With Use of Machine Learning

Using large national physical activity data sets, we found that machine learning successfully reidentified the physical activity data of most children and adults when using 20-minute data with several pieces of demographic information. Partial aggregation of the data over time (eg, reidentifying daily-level physical activity data) did not significantly reduce the accuracy of the reidentification. These results suggest that current practices for deidentification of Physical Activity Monitor (PAM) data might be insufficient to ensure privacy and that there is a need for deidentification that aggregates the physical activity data of multiple individuals to ensure privacy for single individuals.

The full article can be downloaded below.  

Health Industry Cybersecurity Practices: Managing Threats and Protecting Patients

HHS convened the Task Group in May 2017 to plan, develop, and draft this guidance document. To ensure a successful outcome and a collaborative public–private development process, HHS engaged a diverse group of health care and cybersecurity experts from the public and private sectors. Participation was open and voluntary. HHS collaborated with the HPH Sector Government Coordinating Council, the HPH Sector Coordinating Council, the Department of Homeland Security (DHS), and the National Institute of Standards and Technology (NIST).ii ii Participants included subject matter experts with backgrounds and experience in the following roles: chief executive officer; chief information security officer (CISO) and/or IT security professional; chief information officer; chief risk officer or other risk manager; office of technology leader or hospital administrator; doctor, nurse, and other health care practitioners The Task Group’s approach to the guidance document:

1. Examines current cybersecurity threats affecting the HPH sector;
2. Identifies specific weaknesses that make organizations more vulnerable to the threats; and
3. Provides selected practices that cybersecurity experts rank as the most effective to mitigate the threats.

This document provides best practices regarding risks such as:

• E-mail phishing attacks
• Ransomware attacks
• Loss or theft of equipment or data
• Insider, accidental or intentional data loss
• Attacks against connected medical devices that may affect patient safety

The full Health and Human Services document can be downloaded below.  

5 blockchain developments in 2018

Blockchain moved from mostly hype in 2017 to early technological developments, proofs-of-concept and pilot projects during the last 12 months.  

Deloitte, in a late summer report, predicted a breakout moment for the distributed ledger technology is approaching as almost 75 percent of research respondents see a compelling business case for DLT.  

So it’s no surprise that major tech vendors are embracing Blockchain. Here are five such moves made in 2018.

1. Walmart filed a blockchain patent
2. Change Healthcare made enterprise blockchain technology available
3. Amazon made a blockchain move of its own
4. The CDC enlisted IBM and Intel in blockchain pilots to fight the opioid epidemic
5. Gartner cautioned against waiting too long to get started

The full Healthcare IT News article can be viewed at this link.  

The value of learning health systems in disease control and aging

We are living in an unprecedented revolutionary age of science and technology. Real‐time databases of disease‐specific registries are expected to dramatically and efficiently accelerate clinical research studies. The use of real‐world data to augment data from randomized clinical trials is gaining traction and support globally. The article entitled “The Global Academic Research Organization Network: Data Sharing to Cure Diseases and Enable Learning Health Systems” in this issue describes the activities of the Global ARO Network, including a workshop with participants from Asia, Europe, and the United States. This network represents the global expansion of the ARO Council and global disease‐specific consortia that collaborate on disease‐specific registries. Such networks enable research on a global scale to test drugs and medical devices from academia, ushering in an age where we can collaborate on research and obtain approval for new therapies simultaneously around the world. The formation of global networks for patients with rare diseases is an essential step toward overcoming such diseases, and we now have a more specific picture of the expanded role that these networks play in realizing global learning health systems.

Not only can learning health systems be beneficial in identifying the best treatments for individuals with specific diseases, but there is a role for functioning learning health systems to be more broadly applied to identifying ways to prevent diseases by leveraging and learning from the data from healthy individuals. In developed countries, aging populations pose an increasing social burden and a threat to the vitality of the society, particularly when many of the elderly are inflicted with chronic or debilitating diseases. The slogan, “society in which people in their 100s can remain active,” presages a society where no one is bedridden.* This idea may seem like an impossible dream, like eternal youth and immortality. However, there is an important role of learning health systems in resolving the age‐associated dilemma of extending life, along with quality of life, and controlling diseases that prevent most elderly individuals from being independent and active centenarians.

The full article can be downloaded below.  

How blockchain technology will reshape health care

Blockchain advocates say a breakthrough “killer app” is imminent that will change the business of healthcare as we know it. In the meantime, however, there are at least five practical uses for the technology that permits the distribution of digital information, but not the copying of that information.

These include:

• “Smart” contracts. Contracts automatically go into effect when certain previously agreed upon conditions are met.
• Supply chain processes. The new technology could make supply chains more efficient and transparent, improving the warehousing and delivery of medical goods and supplies.
• Physician credentialing. A company called ProCredEx recently launched Professional Credentials Exchange to do this with a private Medicare claims processor, a private provider of Medicaid managed care and Medicare Advantage plans, and the Michigan-based Spectrum Health System.
• Peer-to-peer data exchange. A system with features that can include preventing people from getting multiple opioid prescriptions and verifying clinical trial data. The hope is that this function can also be used to shrink processing time for prior-authorization requests down to less than five minutes.
• Proof of work. In medical liability cases where attorneys may claim that physician defendants have altered their records, clinical notes entered in blockchain time-stamped blocks create a tamper-proof ledger of what a physician did and when.

The full article from the American Medical Association can be found at this link.  

Use of Electronic Health Data in Clinical Development

In clinical research and development, the scientific possibilities for analyzing large volumes of data are still not used to the extent that it is possible in other sectors (e.g. finance, consumer behavior). Health data are often widely distributed and locked in individual databases, standards are highly inconsistent, and data privacy protection complicates data consolidation and data use. This results in complex clinical protocols with often unrealistic selection criteria, and trials are still too often assigned to inappropriate sites. Furthermore, patient recruitment continues to be one of the major problems in the execution of clinical trials. The use of electronic health data (real world data) allows alignment of protocols to actual medical conditions, formulation of realistic inclusion and exclusion criteria and testing their effects on recruitment using real data. In addition, trials can be assigned to sites that have a proven number of patients in their databases, and patients can be identified at the site. Various providers are players in the field of “big data” and it is not always easy to assess which system is best suited to meet the demands of clinical development. Therefore, a requirements specification is presented in the following.

The full article can be downloaded below.  

5 Ways Artificial Intelligence May Affect Health Care in the Near Future and What That Means for You

Technology is changing fast, and the world is changing with it. Concepts that were mere science fiction only a couple of decades ago -- like artificial intelligence (AI) -- are quickly becoming commonplace. Computers have become powerful enough to handle complex AI computations; machine learning algorithms are more accurate and faster than ever; and the cloud and the internet of things have made it possible for even small devices to access artificial intelligence's enormous capabilities.

That's why responsible use of AI solutions in health care could improve, and even save people's lives. On the other hand, health care is an area where recklessness can occur; that's why new developments are regulated and implemented slowly and cautiously.

Here are five ways that AI and machine learning will likely be affecting your health care in the very near future:

• Digital Consultations
• Radiology and Images
• Personalized Medicine: Faster, More Accurate Diagnoses
• Robot Surgeons
• Cybersecurity

The full Entrepreneur article can be viewed at this link.  

A SCOPING REVIEW EXPLORATION OF THE INTENDED AND UNINTENDED CONSEQUENCES OF EHEALTH ON OLDER PEOPLE: A HEALTH EQUITY IMPACT ASSESSMENT

eHealth is one perceived mechanism to extend the range and reach of limited health-care resources for older adults. A decade-scoping review (2007–2017) was conducted to systematically search and synthesize evidence to understand the intended and unintended consequences of eHealth initiatives, informed by a health equity impact assessment framework. Scoping review sources included international academic and grey literature on eHealth initiatives (e.g., eHealth records, telemedicine/telecare, and mobile eHealth application) focused on the varying needs of older adults (aged 60+), particularly individuals experiencing sociocultural and economic difficulties. Findings suggest that eHealth has several potential benefits for older adults, but also the possibility of further excluding already marginalized groups, thereby exacerbating existing health disparities. Ongoing evaluation of eHealth initiatives for older adults is necessary and requires attention to unique individual-level, socioeconomic, and cultural characteristics to heighten benefits and better capture both the intended and unintended outcomes of advanced eHealth systems.

The full article can be downloaded below.  

Computing Statistics from Private Data

In several domains, privacy presents a significant obstacle to scientific and analytic research, and limits the economic, social, health and scholastic benefits that could be derived from such research. These concerns stem from the need for privacy about personally identifiable information (PII), commercial intellectual property, and other types of information. For example, businesses, researchers, and policymakers may benefit by analyzing aggregate information about markets, but individual companies may not be willing to reveal information about risks, strategies, and weaknesses that could be exploited by competitors. Extracting valuable utility from the new “big data” economy demands new privacy technologies to overcome barriers that impede sensitive data from being aggregated and analyzed.

Secure multiparty computation (MPC) is a collection of cryptographic technologies that can be used to effectively cope with some of these obstacles, and provide a new means of allowing researchers to coordinate and analyze sensitive data collections, obviating the need for data owners to share the underlying data sets with other researchers or with each other. This paper outlines the findings that were made during interdisciplinary workshops that examined potential applications of MPC to data in the social and health sciences.

The primary goals of this work are to describe the computational needs of these disciplines and to develop a specific roadmap for selecting efficient algorithms and protocols that can be used as a starting point for interdisciplinary projects between cryptographers and data scientists.

The full article can be downloaded below.  

Split learning for health: Distributed deep learning without sharing raw patient data

Can health entities collaboratively train deep learning models without sharing sensitive raw data? This paper proposes several configurations of a distributed deep learning method called SplitNN to facilitate such collaborations. SplitNN does not share raw data or model details with collaborating institutions. The proposed configurations of splitNN cater to practical settings of i) entities holding different modalities of patient data, ii) centralized and local health entities collaborating on multiple tasks and iii) learning without sharing labels. We compare performance and resource efficiency trade-offs of splitNN and other distributed deep learning methods like federated learning, large batch synchronous stochastic gradient descent and show highly encouraging results for splitNN.

The full article can be downloaded below.