By Dr. Elizabeth Marshall

Preserving patient privacy is of paramount importance to clinicians, which is one reason healthcare providers often shy away from opportunities to make health information available for secondary uses, such as the development of new medications.

While some clinicians may feel uncomfortable providing patient data to outside entities, even when anonymized, it is important to keep in mind that there are many benefits to reusing health data to derive insights that can drive healthcare improvements through, for example, operational efficiencies, predicting disease outcomes, or innovative treatments regimens.

Fortunately, advancements in governance and technology can now deliver unprecedented levels of privacy and security to preserve the anonymity of patients. In addition, most patients are willing to share their data for research, according to a study published in JAMA Network Open. Given evolving sentiments and technology enhancements, more providers are now realizing the benefits of sharing non-identifiable structured and unstructured health data from EHRs.

When organizations take measures to protect health data to preserve privacy, the sharing of health data to other entities can be a win for both providers and patients.

Advancing innovation with clinical data
Clinical data is often rich in details that help pharma companies, device manufacturers, and other life science companies advance the discovery and development of new life-saving therapies, identify potential cohorts for clinical trials, and test the efficacy of drugs.

Consider, for example, the tremendous push to bring COVID-19-targeted drugs and vaccines to market. As of the end of the third quarter of 2020 over 600 drugs and vaccines were in development and data from EHRs has undoubtedly helped researchers better understand the disease and its progression, including the virus’s impact on different populations and the effectiveness of various therapies. The rapid worldwide spread of COVID-19 and its potential for serious complications have underscored the need for high-quality clinical data that accelerates the development of safe and effective solutions.

To provide value on a large scale, clinical information must be in a structured format that supports data analytics and includes a full 360o-view of a patient’s health status. Unfortunately, as much as 80% of patient data within clinical records is hidden as unstructured text within doctor narratives, lab reports, discharge summaries. This makes it difficult to glean vital health information, such as potentially undiagnosed conditions, social determinants of health, or risk factors driving poorer patient outcomes. Using natural language processing (NLP) technology, however, organizations can transform data from an unstructured to structured format to facilitate additional analysis.

Once data is in a structured format, privacy and analytics tools can be applied to protect health information, such as removing names and phone numbers, and generalizing geographic data. This preserves the integrity and utility of the data, while also ensuring details cannot be tied back to an individual. Data can be easily aggregated from multiple patients or a population and safely and responsibly delivered to innovation partners.

A sample scenario
Consider the following scenario that highlights how health data can help advance innovation.

George visits his doctor complaining of shortness of breath, weakness and fatigue. After a medical exam, the doctor begins the work-up by ordering a chest x-ray, electrocardiogram, arterial blood gas, complete blood count and numerous other blood tests to assess metabolic status. Unfortunately, the physician ultimately diagnoses George with an aggressive but early stage small cell lung cancer.

The doctor, of course, is committed to providing the best treatment possible, but believes that George’s prognosis is poor. Using NLP and privacy and analytics tools, the physician - or hospital or health system - extracts key data from George’s records and compares health details against the selection criteria of known internal or external active clinical trials that might help George qualify for an experimental therapy to hopefully prolong his life. This way George can be asked if he is interested in specific trials that matched his health data.

The physician also shares general non-identifiable healthcare record details with life sciences partners. Though the general details may not provide direct benefit to George, the information could potentially improve outcomes for other “Georges” in the future. Sharing privacy-preserving health data outside of the organization can thus accelerate the efforts of researchers across healthcare and the life sciences working to develop life-preserving and life-saving therapies.

Clinicians should rightly be concerned with protecting the privacy of their patients. With the right technologies, however, organizations can safely and responsibly make health data available to innovation partners outside their organization to advance their work that and possibly lead to new, life-enhancing therapies for patients.

About the author: Dr. Elizabeth Marshall, MD, MBA, is the associate director of clinical analytics at Linguamatics, an IQVIA company.