Privacy of Clinical Research Subjects: An Integrative Literature Review with Best Practices

With changes in clinical research practice, the importance of a study-subject’s privacy and the confidentiality of their personal data is growing. However, the body of research is fragmented, and a synthesis of work in this area is lacking. Accordingly, an integrative review was performed, guided by Whittemore and Knafl’s work. Data from PubMed, Scopus, and CINAHL searches from January 2012 to February 2017 were analyzed via the constant comparison method. From 16 empirical and theoretical studies, six topical aspects were identified: the evolving nature of health data in clinical research, sharing of health data, the challenges of anonymizing data, collaboration among stakeholders, the complexity of regulation, and ethics-related tension between social benefits and privacy. Study subjects’ privacy is an increasingly important ethics principle for clinical research, and privacy protection is rendered even more challenging by changing research practice.

The article concludes with suggested best practices based upon the findings.  

Best Practices

• Encourage collaboration - Collaboration among stakeholders is one of the prerequisites for protecting privacy in clinical research.
• Transparency - A well applied, transparent process with properly surveilled procedures ensures compliance with legal guidelines.   
• Reduce tension - The tension between study-subject privacy protection and benefits to society must be eased. This requires, public trust and more flexible privacy rules. But most important is a guarantee of adequate research-ethics training for researchers and clinical-study staff, with solid recognition of the moral rationale behind the regulations.

The full article can be downloaded below.  

How AI is Transforming Clinical Trial Recruitment: Including Best Practices

The medical world is shifting underneath our feet.

To keep up with the rising demands of empowered patients, physicians and pharma businesses regularly test innovative treatments and medicines during rigorous clinical trials.

But one misguided move can trigger a domino effect, such as when the wrong patients are selected for a clinical trial.

Today’s infographic comes to us from Publicis Health, and it highlights why the current model of clinical trial recruitment urgently needs to change.  The article also included some important insights for a patient-based approach, which can yield advantages and added value to recruitment, engagement, and data collection.  

Best Practices

• Omni-channel targeting - Actively reaching out to patients, wherever they are.
• Predictive analytics - Continually refining media channels and messaging to further patient interest
• Ongoing communications - Nurturing relationships with patients, starting with the initial outreach.

The full infographic and article can be viewed at this link.  

How a Pharma Company Applied Machine Learning to Patient Data: Best Practices

The growing availability of real-world data has generated tremendous excitement in health care. By some estimates, health data volumes are increasing by 48% annually, and the last decade has seen a boom in the collection and aggregation of this information. Among these data, electronic health records (EHRs) offer one of the biggest opportunities to produce novel insights and disrupt the current understanding of patient care.

But analyzing the EHR data requires tools that can process vast amounts of data in short order. Enter artificial intelligence and, more specifically, machine learning, which is already disrupting fields such as drug discovery and medical imaging but only just beginning to scratch the surface of the possible in health care.

Let’s look at the case of a pharmaceutical company we worked with. It applied machine learning to EHR and other data to study the characteristics or triggers that presage the need for patients with a type of non-Hodgkin’s lymphoma to transition to a later line of therapy. The company wanted to better understand the clinical progression of the disease and what treatment best suits patients at each stage of it. The company’s story highlights three guiding principles other pharma companies can use to successfully deploy advanced analytics in their own organizations.

Best Practices

• Preliminaries - Generating meaningful hypotheses (and organizational buy-in) requires engaging the right stakeholders
• Richness - The best data set might be a combination of data sets
• Test-and-learn - Feedback loops (many times over) are the key to great results

The fully detailed Harvard Business Review article can be viewed at this link.  

Social Determinants of Health (SDOH) are the conditions in which people are born, work, live, and age. The healthcare industry increasingly recognizes that improvements in health and health equity will only be possible after addressing SDOH, including socioeconomic status, education, neighborhood and physical environment, social support networks, and access to healthcare. Currently, payers use ZIP code characteristics to determine investments at the neighborhood level. The approach does not consider customized resource allocation at the individual level unless the member/patient has had multiple, high-cost interactions with the healthcare system (i.e., “hotspotting”).

UnitedHealthcare (UHC) is piloting a more targeted approach to addressing SDOH and will be tracking the results of the pilot work closely.

The steps of the pilot are:

1. Identify “at-risk” members/patients using specific ICD-10 codes, CPT codes, and LOINC codes on claims
2. Have care managers perform direct outreach to conduct assessments to evaluate specific needs
3. Generate/update care plans that treats the entire patient, including social, medical, and behavioral services to address member-specific needs
4. Connect members/patients to other payers, such as Medicare and Medicaid, if appropriate
5. Enroll members/patients into relevant UHC programs, such as literacy programs or programs around self-care
6. Arrange follow up
7. Go back to step 3

These Considerations for Success were discussed as part of eHealth Initiative’s September 2018 Value & Reimbursement Workgroup meeting as presented by Anupam Goel, Chief Health Information Officer, Clinical Services, UnitedHealthcare.

List of links to research and articles related to September 6, 2018 Artificial Intelligence and Machine Learning Executive Roundtable.

Book by Josh Sullivan: The Mathematical Corporation: Where Machine Intelligence and Human Ingenuity Achieve the Impossible

https://www.amazon.com/kindle-dbs/entity/author/B06WRP9HPH?_encoding=UTF8&node=283155&offset=0&pageSize=12&sort=author-pages-popularity-rank&page=1#formatSelectorHeader

More about Adam and Eve

https://www.cam.ac.uk/research/news/artificially-intelligent-robot-scientist-eve-could-boost-search-for-new-drugs

Book by Pedro Domingos: The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake our World

https://www.amazon.com/Master-Algorithm-Ultimate-Learning-Machine-ebook/dp/B012271YB2

The mPower Study, Parkinson disease mobile data collected using ResearchKit

https://www.nature.com/articles/sdata201611  

Diagnosing and treating depression with AI and Machine Learning

https://www.techemergence.com/diagnosing-and-treating-depression-with-ai-ml/

How AI could increase the risk of nuclear war

https://www.rand.org/blog/articles/2018/04/how-artificial-intelligence-could-increase-the-risk.html

AI will give us better french fries

https://www.bloomberg.com/news/articles/2018-04-26/ai-comes-to-the-rescue-of-stressed-potatoes-and-bad-french-fries

Understanding & Predicting Length of Stay (LOS) using Machine Learning

https://dexur.com/a/ml-research-los/6/

Length of Hospital Stay Prediction at the Admission Stage for Cardiology Patients Using Artificial Neural Network

https://www.hindawi.com/journals/jhe/2016/7035463/

Predicting Length of Stay among Patients Discharged from the Emergency Department—Using an Accelerated Failure Time Model

https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0165756

ML #6 – Healthcare.ai for Predicting Extended Length of Stay

https://www.youtube.com/watch?v=DSMDdCUq47o

Griffin Hoopes – the developer that did the work Matt Keating was referring to.

https://www.linkedin.com/in/griffinhoopes/

Artificial Intelligence in Cardiology

http://www.onlinejacc.org/content/71/23/2668 from the ACC

On the Prospects for a (Deep) Learning Health Care System

https://jamanetwork.com/journals/jama/fullarticle/2701667 - JAMA​

Deep Learning—A Technology With the Potential to Transform Health Care

https://jamanetwork.com/journals/jama/fullarticle/2701666  - JAMA​

Artificial intelligence in healthcare: past, present and future

https://svn.bmj.com/content/early/2017/09/11/svn-2017-000101

AMA Passes First Policy Recommendations on Augmented Intelligence

https://www.ama-assn.org/ama-passes-first-policy-recommendations-augmented-intelligence

Artificial Intelligence and the Augmentation of Health Care Decision-Making

https://catalyst.nejm.org/ai-technologies-augmentation-healthcare-decisions/ 

HHS Data Sharing Report:

https://www.hhs.gov/idealab/data-insights/

Cerner Smart Viewer

            https://www.youtube.com/watch?v=_WLQuQYPz6o&feature=youtu.be

Securing legacy medical devices is daunting – but not optional: Best Practices

Between high-profile hacks and hospitals’ growing dependency on connected medical devices, cybersecurity is as relevant to healthcare as it’s ever been. And while providers and device manufacturers are rightfully making protection a priority for newly developed or purchased medical devices, older legacy devices pose the greatest security risk to healthcare organizations.  MobiHealthNews looks at expert opinions on this topic.  

Best Practices

• Start at the beginning - Take the time to take inventory of relevant devices.  Identifying problems are crucial to being able to solve them.  
• Prioritize threats - Assess the risk for attack for each medical device.
• Consider consequences - What are the consequences if an attack succeeds?  What are the backup plans if a device fails?  
• Build a baseline - What does normal traffic look like for these devices and networks?  
• Plug holes - Be proactive and preventative when it comes to cybersecurity, while also being realistic about what can be achieved.  

The full article can be viewed at this link.  

As healthcare costs in the U.S. continue to inch towards 20% of gross domestic product, unique and inventive ways of bending the cost curve have begun to take shape.  The financial implications of value-based care delivery have increased, while fee-for-service models continue to face consumer and regulatory scrutiny. Measures of quality – such as readmission rates, patient-reported experiences of care (CAHPS), and spend-per-beneficiary – are now inseparable from reimbursement, workflow, and organizational viability.  Risks are now incurred inside the acute care environment, as well as before and after care is received.  Increasingly, a patient and caregiver’s knowledge of preventative activities, current condition(s), treatment plans, and self-care becomes paramount to achieving ideal outcomes outside the clinical environment.  Consequently, improving health literacy is increasingly recognized as a key cost and efficiency driver ripe with technologic and workflow advancement opportunities.

How big is the effect of health literacy on growing costs?  A recent study conducted by Accenture stated that amongst health plans, low health literacy incurred an estimated $4.8 billion annually in administrative costs.   Another study hosted by the National Institutes of Health (NIH) estimated that the total cost of low health literacy in the U.S. healthcare system for both providers and health plans totals more than 200 billion dollars per year.  Although population health efforts have provided a path for beginning to address the health literacy gaps, recognition of the scope of the health literacy problem is not readily understood by most provider and health plan organizations.  Increased focus surrounding how patients and caregivers interact and communicate with clinicians is vital to tackling the health literacy problem.

One way that many large health systems, integrated delivery networks, and care management organizations are improving patient health literacy is through more robust patient and caregiver education efforts.  Patient and caregiver education improves understanding around the importance of adherence to treatment plans and the appropriate utilization of care.  Through more formalized approaches to education and collaboration amongst patients, caregivers, and care teams, new benchmarks and standards are established and scaled across organizations.  From these efforts, patient outcomes improve, unnecessary costs can be decreased, and clinical quality heightened.  Fortunately, a formalized approach to tackling health literacy through education is becoming easier with the progression of technology that can deliver education to varying channels, care settings, and workflows.

Providers and health plans are encouraged to think about the following in addressing gaps in health literacy:

Care Team Considerations

• Assign care management resources to each patient
• Engage in motivational interviewing at the point of care and through care coordination efforts
• Provide patient and caregiver with appropriate education consistently and frequently to reinforce key concepts and behaviors
• Ensure that patients and caregivers fully understand the treatment plan so that they are engaged and invested in the course of action
• Involve the patient’s family to understand the patient’s information needs

Technology Considerations

• Consistently track goals & biological data
• Enroll patients into personalized longitudinal care plans with frequent education touchpoints
• Automate content during the visit, after the visit, or during follow-ups
• Track engagement metrics and patient’s response to information
• Provide different types of content for the patients, leveraging devices they use
• Deliver social media pushes, campaigns, and targeted materials to patients and their families

At-Home Considerations

• Send communications and education on a regular basis to patients from their healthcare organization and associated value-based program
• Use at-home devices such as blood glucose meters, pedometers, or other technology that can be integrated into the EMR
• Perform follow-up with a care manager diligently alongside treatment plan

These considerations for success were discussed as part of a presentation by eHI members Jake Blanchard and Josh Schlaich of Healthwise during eHealth Initiative’s September 2018 Workflow for Provider and Patient Engagement Workgroup meeting.

Navigating Your Path to Consumer-Driven Health Plans

As employers move toward offering a consumer-driven health plan (CDHP) as part of their employee benefits, many have questions such as:

• How do employees respond to consumerism?
• Are my employees ready to assume greater control and responsibility?
• What can my company do to make the transition easier?
• Will the benefits of making the transition be worth the expense and the upheaval?

This research from Humana answers those questions and more. Over a 15-month period, researchers studied individuals representing different age groups, family structures, employment positions, and income levels. Participants remained anonymous. The findings are reliable, but due to the nature of qualitative research, they cannot be considered representative of the general population. By studying the same people over time, an evolution can be seen in employees’ attitudes, beliefs, and behavior, and employers can make note of potential obstacles to avoid.

The study findings are summarized below.

Best Practices

• Employees appreciate choice - Your company’s implementation of a consumer-driven health plan can be smoother by designing offerings with choice in mind—and by making sure each offering is distinct. You can further build employees’ confidence in their enrollment choices by helping them predict potential costs and by providing tools that can help them forecast future healthcare needs.
• Employees want to “try it on” - The ability to predict healthcare needs and costs leads to confident decision-making in choosing a plan.  Permitting employees to make use of company computers, hosting training classes, or setting up kiosks can help overcome lack of computer familiarity. Likewise, employee expectations can be made more manageable by facilitating ongoing communication or by providing employees with a “coach” that can guide them in predicting healthcare needs and selecting an appropriate plan.
• Getting results - Changes in attitude and behavior are driven by experience and the desire to contain out-of-pocket costs.  One way to enhance an employee’s sense of control over their own healthcare is to communicate year-round about the benefits that the employee has selected and how he or she has used the plan. Such communications could include an individualized statement of benefits used or year-to-date cost savings
• The need for ongoing support - A system of supportive communication can help build familiarity and boost employee confidence.
• Money matters - Consider exploring special incentives that encourage employees in the responsible use of their plan benefits.
• Planning pays off - Encourage your employees to plan ahead for emergencies and provide the tools to make planning easy.
• Transparency is necessary - When your employees understand the value of transparency in healthcare, they are likely to have much greater confidence in their ability to be smart consumers. You help facilitate this understanding by making sure employees are aware of the tools available to them and by providing training to ensure that they are able to use these tools easily and effectively.
• Keep it simple - Employees’ top priorities are ease of use and claims follow-up.  There is a desire to see their spending information presented in a familiar, user-friendly frame of reference, such as a credit card statement.  You may also wish to explore options that can help simplify claims and reimbursement.
• Consumerism meets individual needs - There may be a need to acknowledge the various temperaments and styles in which people organize information. You may need to deliver information in a variety of ways. Consider surveying your employees about their preferred methods of receiving information.

The full Humana Research Summary can be downloaded below.  

Patient matching compares data sources from different health IT systems, doctors’ offices, and hospitals to decipher if they belong to the same patient. The process recovers data from health records and identifies commonalities based on personal traits, such as demographics, to find a “match.”  Patient matching improves patient safety and clinical care by establishing continuity of care among patients with multiple records at one provider or across various providers. The patient matching process allows organizations to develop a complete record of the patient’s health history and medical care.

Our August Technology & Analytics Workgroup Meeting included executives from 4Medica and the Nebraska Health Information Initiative (NeHII) who suggested the following considerations for success when developing a patient matching solution:

• Is your enterprise-wide duplication rate below two percent?
• Does your current Master Patient Index solve all of your identity matching and resolution problems – across all data sources and locations within your organization?
• Do you feel that you are spending too many resources (money and human capital) to continuously clean up identities in your data sources?
• Do you currently use unlimited historical and clinical data enrichment available to help with more precise matching?
• Is your organization at risk by having duplicate identities?

Patient matching has raised fears about false matching, privacy, and consent. According to a survey from Black Book Research, around 33% of all denied claims are linked with inaccurate patient identification, which costs the average hospital $1.5 million and the U.S. healthcare system $6 billion annually. Implementing the appropriate patient matching steps will help your organization better serve patients, while saving money.

To build a 360-degree view of a patient:

• Share data transparently
• Promote informed patient-centric care coordination
• Segregate and identify clinical information from each data source to ensure most up to date information for a patient in real-time.

Press release on the patient matching collaboration between NeHII and 4Medica.

Accurate patient matching has simultaneously become 10x more challenging and 10x more important than it was even just a few years ago. But conventional patient matching technologies have reached their breaking points, and duplicate rates are skyrocketing as a result—as are the costs associated with having duplicate records and incomplete patient records.Verato Auto-Steward is a cloud-based plugin for EHR and EMPI technologies that uses a powerful new paradigm in patient matching called Referential Matching to automatically resolve 50-75% of the toughest matches your EHR or EMPI cannot resolve.