Furthering Medical Progress

It seems to me….

I think history would say that medical research has, throughout many changes of parties, remained as one of the shining lights of bipartisan agreement, that people are concerned about health for themselves, for their families, for their constituents.” ~ Francis Collins[1].

There are numerous available pathways for medical research capable of significantly advancing healthcare progress. One of the formidable challenges healthcare providers currently face is putting the latest available medical data to its greatest possible use[2]. Somewhere between the quest to unlock the mysteries of medicine and design better treatments, therapies, and procedures lies the real world of applying data and protecting patient privacy. Much is dependent upon medical data being available to researchers throughout the field but those goals, of protecting patients and the quest for answers, are frequently at odds.

The newest and best medicines are frequently initially offered to patients in clinical trials and patients in those trials often do better than patients on standard treatments[3]. Yet trials, particularly in cancer, but also for other diseases, frequently have many empty patient slots as the vast majority of potential participants are not offered an opportunity to enroll. Most patients never get into lifesaving drug trials due to barriers at community hospitals. Low trial enrollment, which effectively cuts patients off from lifesaving medicine, is a significant national health problem. Obstacles to trials need to be overcome, especially in community hospitals, by reducing the burden on local doctors and improving patient-trial matching technology.

The drugs available in clinical trials often represent the latest in research, and many turn out to be considerably more effective than standard treatments. Half of all drugs that make it into the last of three phases of drug trials, when most patients enter those trials, end up being approved by the FDA because of these improved results. About one third of patients in the U.S. meet the criteria for a trial with a new drug, but only about 4 percent end up in such tests, according to National Cancer Institute estimates, and some specialists say the real number is even lower.

The main reason for the massive shortfall is that in the nonacademic community hospitals where most patients are treated, doctors do not feel they have the time, the incentives, or the support to learn about available trials, to qualify and enroll patients, or to provide the extra follow-up care such trials often entail.

A National Academies of Sciences, Engineering, and Medicine study concluded that “community practitioners lack the needed infrastructure and support to actively participate in clinical trials”. A study in the clinical cancer journal CA called trial enrollment “embarrassingly low” and blamed it, in part, on “a lack of knowledge about available studies by community oncologists, a lack of time or interest, or a lack of resources to support the cost of performing clinical trials”. Because nationally about 85 percent of patients end up at community hospitals, most of the low participation in trials is attributable to the failure of those hospitals to enroll their patients.

The enrollment problem also handicaps research. Lack of patients forces many trials to stop before getting results, ending the progress of many promising treatments. Most trials are at least delayed by patient enrollment shortages. About one out of six of all trials never manage to recruit a single patient.

Clinical trials can be redesigned to reduce the burden on community hospital physicians, shifting more of the workload to the research centers that originate the trials. The biggest physician barrier was time constraints.

What seems poised to effect the most change is a combination of approaches: trial researchers who get out into communities and market their work to local doctors, trial designs that reduce physician workload, and tools that automate patient-trial matching and related tasks.

Similar to clinical trials, medical advances are increasingly dependent on the analysis of enormous datasets – as well as data that extends beyond any one agency or enterprise. As connected healthcare devices flourish, at-home and remote monitoring has increased and big data analytics advances at a staggering rate, the stakes and the ability to use, misuse, and abuse confidential data has risen significantly.

Extensive data, mostly filed away in the form of medical records, has accumulated as the result of years of doctor’s examinations and treatments. If that data was combined in a large corpus of numerous patients suffering similar ailments, common patterns applicable beyond specific instances would become apparent revealing how medical conditions are possibly related beyond numerous individual occurrences. Medical advances are increasingly becoming dependent upon related developments in non-medical fields; e.g., robotics for more precise control than possible by human surgeons, big-data analytics for determining ailment correlations, artificial intelligence (AI) for diagnostic assistance, wearable biometric fitness trackers, etc.

Electronic healthcare records, personal fitness devices, connected home monitoring systems, and a variety of other sensors, machines, and systems are pushing the boundaries of medicine in new directions. As a result, researchers, physicians, and other practitioners, using big data analytics and machine learning, can spot patterns, trends, and causalities that would otherwise escape human detection. This makes it possible to improve therapies, procedures, and drugs while improving diagnostics and care for individual patients.

As a part of the American Recovery and Reinvestment Act passed by Congress in 2009, all public and private healthcare providers and other eligible professionals were required to adopt and demonstrate “meaningful use” of electronic medical records (EMR) by 1 January 2014 in order to maintain their existing Medicaid and Medicare reimbursement levels. Currently, medical records must only be maintained for at least 11 years or until the patient reaches the age of majority plus six years, whichever is longer. This is much too short a requirement for considerable long-term research which requires full lifetime reports.

While the mandate prompted significant growth in health informatics, an interdisciplinary field of study that merges information technology and healthcare, essentially no standard formats for maintaining that data yet exist. Neither healthcare providers or researchers therefore normally have access to a patient’s complete and accurate medical records.

Greater portability of data, greater interoperability between systems, and a more coordinated approach to patient care is crucial. A national medical record database or depository able to store medical data in a common format available for research is urgently needed. Additionally, all health-related data, including medical, psychiatric, and genetic sequences, must be included.

Regulators and patient advocates have for years pushed for data-sharing standards within the medical sector to make it easier for records to flow between hospitals and doctors’ offices. The lack of interoperability has made it a challenge for consumers to access high-quality care and has led to unnecessary medical errors.

While a comprehensive standard for medical data does not currently exist, progress is being made. Fast Healthcare Interoperability Resources[4] (FHIR) is a standard describing data formats and elements and an application programming interface (API) to enable the electronic exchange of protected health information among healthcare professionals. It includes diagnostic data, clinical health data, and administrative information. The FHIR specification is expected to become the next-generation standard for inter-organizational health information exchange and healthcare companies can now build on the progress of FHIR connecting data from specialties that use their own unique data-management techniques.

Medical errors currently contribute to about 250,000 deaths a year in the U.S. alone according to a 2016 study conducted by Johns Hopkins Medicine[5]. Roughly one-third of medical malpractice claims can be directly linked to failures in communication, either between the doctor and their patient or between medical professionals. While numerous errors contribute to such problems, many are procedural. AI assisted error checking systems are being further developed to help reduce many of the more common mistakes. Without record standards, these applications must be specifically written for each set of data records.

AI-determined correlation would facilitate researchers in identifying likely trial subjects and cause/effect relationships. Genome-wide association studies, which look for links between particular genetic variants and incidence of disease, are the basis of much modern biomedical research but databases of genomic information pose privacy risks and many people are understandably reluctant to contribute their genomic data to biomedical research projects[6].

Patient privacy becomes a significant concern for any data storage facility. As more fitness devices, medical monitoring devices, and advanced analytics are developed and come into use, systems must address the often-competing interests of putting data to maximum use but also protecting it. Researchers at the Massachusetts Institute of Technology (MIT) and Stanford University recently unveiled a system for shielding the privacy of people who contribute their data to genomic studies based on a process called secret sharing which diffuses sensitive data across multiple servers, none of which can deduce the data by themselves.

Amazon, Apple, Google, and Microsoft have announced plans to increase their share of the healthcare market. Apple has expanded into virtual medical research using its iPhone and Apple Watch. Microsoft has introduced cloud-based tools to help health systems share medical data. Last year, Amazon joined JPMorgan Chase and Berkshire Hathaway in a venture to try to improve care and reduce costs for their employees in the U.S.

Google’s health efforts include a push to use AI to read electronic health records and then try to predict or more quickly identify medical conditions. They will partner with Ascension medical systems which operates 150 hospitals in 20 states and the District of Columbia to store all Ascension patient’s data on Google’s cloud computing platform. While it is legal for health systems to share patients’ medical information with business partners like electronic medical record companies, many patients do not trust Google which has paid multiple fines for violating privacy laws with their personal medical details.

Apple updated its iPhone Health app so users can see their available medical data from hospitals and clinics in addition to existing health app medical data from multiple providers whenever they choose. Johns Hopkins Medicine, Cedars-Sinai, Penn Medicine, and several other participating hospitals and clinics are some of the first to make this data available to their patients. The U.S. Department of Veterans Affairs (VA) launched Health Records to select patients this past summer.

Amazon is marketing its data storage experience as way to achieve true interoperability encompassing not just data interoperability but semantic and syntactic interoperability. Specific announcements of additional prospective users are so far limited.

Microsoft began marketing its HealthVault designed to be an online, encrypted vault, where U.S. users could store and manage their health records in 2007 but recently terminated the service. It now is promoting its Azure to the health and life sciences industry as a method of developing connected solutions that engage patients and empower care teams while saving costs by improving clinical and operational efficiency and has announced a number of users including Children’s Mercy Hospital, IRIS Healthcare, Roche Diagnostics, Merck KGaA, and others.

Electronic medical record (EMR) systems are becoming increasingly popular, as the healthcare industry is moving toward digitization. Government initiatives, such as encouraging physicians to adopt electronic health records, investing in training healthcare information technology workers, and establishing regional extension centers to provide technical and other advices are triggering the EMR market’s growth.

Moreover, factors, like the rising need for integrated healthcare systems, big data trends in the healthcare industry, and technological advancements in the field of data storage are driving the growth of the EMR market. On the other hand, factors, like data privacy concerns, high initial investment, shortage of properly trained staff, and inter-operability issues are the primary restraints in market growth.

It is estimated[7] that the electronic health record market will surpass $30 billion by 2026 representing an average annual growth rate of 5.5 percent from 2019 to 2026. If the true potential of these developments is to be realized, it is imperative that standards be developed insuring interoperability for data sharing and research. Rather than the domain of any single provider, these data must be available to researchers throughout the entire field.

That’s what I think, what about you?


[1] Francis Sellers Collins is an American geneticist who discovered genes causing genetic diseases and led the U.S. National Institutes of Health (NIH) public research consortium in the Human Genome Project (HGP).

[2] Greengard, Samuel. Finding a Healthier Approach to Managing Medical Data, Communications of the ACM, https://cacm.acm.org/magazines/2018/5/227203-finding-a-healthier-approach-to-managing-medical-data/fulltext, May 2018, pp31-33.

[3] Freedman, David H. Clinical Trials Have The Best Medicine But Do Not Enroll The Patients Who Need It, Scientific American, https://www.scientificamerican.com/article/clinical-trials-have-the-best-medicine-but-do-not-enroll-the-patients-who-need-it/, January 2019, pp60-65.

[4] Fast Healthcare Interoperability Resources (FHIR), eCQI Resource Center, https://ecqi.healthit.gov/fhir, 5 November 2019.

[5] Prater, Valerie S. Confidentiality, Privacy And Security Of Health Information: Balancing Interests, Johns Hopkins Medicine, News Release, https://healthinformatics.uic.edu/resources/articles/confidentiality-privacy-and-security-of-health-information-balancing-interests/, 3 May 2016.

[6] Hardesty, Larry. Protecting Confidentiality In Genomic Studies, MIT News, http://news.mit.edu/2018/protecting-confidentiality-genomic-studies-0507, 7 May 2018.

[7] Electronic Health Record Market Future Prospective And Rapid Technological Advancements, Computerworld, https://www.computerworld.com.au/mediareleases/36279/electronic-health-record-market-future/, 18 November 2019.

About lewbornmann

Lewis J. Bornmann has his doctorate in Computer Science. He became a volunteer for the American Red Cross following his retirement from teaching Computer Science, Mathematics, and Information Systems, at Mesa State College in Grand Junction, CO. He previously was on the staff at the University of Wisconsin-Madison campus, Stanford University, and several other universities. Dr. Bornmann has provided emergency assistance in areas devastated by hurricanes, floods, and wildfires. He has responded to emergencies on local Disaster Action Teams (DAT), assisted with Services to Armed Forces (SAF), and taught Disaster Services classes and Health & Safety classes. He and his wife, Barb, are certified operators of the American Red Cross Emergency Communications Response Vehicle (ECRV), a self-contained unit capable of providing satellite-based communications and technology-related assistance at disaster sites. He served on the governing board of a large international professional organization (ACM), was chair of a committee overseeing several hundred worldwide volunteer chapters, helped organize large international conferences, served on numerous technical committees, and presented technical papers at numerous symposiums and conferences. He has numerous Who’s Who citations for his technical and professional contributions and many years of management experience with major corporations including General Electric, Boeing, and as an independent contractor. He was a principal contributor on numerous large technology-related development projects, including having written the Systems Concepts for NASA’s largest supercomputing system at the Ames Research Center in Silicon Valley. With over 40 years of experience in scientific and commercial computer systems management and development, he worked on a wide variety of computer-related systems from small single embedded microprocessor based applications to some of the largest distributed heterogeneous supercomputing systems ever planned.
This entry was posted in AI, AI, Amazon, American Recovery and Reinvestment Act, Apple, Artificial Intelligence, Artificial Intelligence, Ascension, Berkshire Hathaway, Cancer, Cancer, Cedars-Sinai, Children's Mercy Hospital, Clinical Trials, Data Base, Disease, Doctor, Fast Healthcare Interoperability Resources, FDA, FHIR, Fitness Trackers, Food and Drug Administration, Genetics, Google, Google, Health, Healthcare, Hospital, Hospitals, IRIS Healthcare, Johns Hopkins Medicine, JPMorgan Chase, Malpractice, Massachusetts Institute of Technology, Medicaid, Medical Records, Medical Research, Medicare, Merck KGaA, Microsoft, Microsoft, MIT, National Cancer Institute, Penn Medicine, Physicians, Research, Roche Diagnostics, Stanford and tagged , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , . Bookmark the permalink.

1 Response to Furthering Medical Progress

  1. lewbornmann says:

    One of the response I received to this posting came from Dr. Richard Yoder. I appreciate the time and consideration he took in writing it. His background and experience in this area lends credibility and relevance to his comments. I therefore thought them worth sharing with anyone reading my posting.
    =============================================

    Lew, You pretty much have it right. In summary, it’s under development and starting to work, but not fully here yet.

    My own background is that I researched and co-authored several papers on computer processing of medical records back in the 1960’s. No CRT displays (output via printer), no direct keyboard input (used punched cards), fixed-length words and fields, etc. It was obvious to us at that time that what we could do was limited but that new technology, whatever it might be, would chip away at the problems and eventually make the idea viable. We knew that when it came of age it would be valuable, and for the reasons you describe. It took about forty years and many technological inventions and advances to begin to do so.

    I had the good fortune to be included in a “committee” which promoted the development and implementation of an electronic medical data system for Northern California. One result has been the affiliation of many providers with SacValley Medshare. So, today it’s starting (just starting) to work. We have private systems of record-keeping up and running and are working toward expanding them. Unfortunately, most of these systems are not mutually compatible, but we are working on methods of pooling and interchange of information. It will come.

    As you state, the current statutory retention period for patient records is too short. It was probably determined by the capabilities and capacities of electronic data gathering and storage at the time. We have new devices, compression and security techniques, so the retention period can be lengthened. We also have greatly increased knowledge and familiarity with electronic data gathering methods on the part of the public so more information can be input by patients and other non-computer-professionals.

    When the development of electronic record-keeping was mandated a large number of private firms stepped up to provide their customers with the service. Unfortunately, this resulted in the development of a myriad of incompatible forms and formats, as mentioned above. Standard formats and translation software are being developed and are making records from different sources compatible but this doesn’t happen overnight. There is usually little incentive for a company to develop compatibility with a rival company.

    In addition to format compatibility, issues which must be overcome in order to effect pooling and interchange of data include protection of patient privacy; data security from hackers or unauthorized access; criteria for access, use, release, local storage and transmission of data; etc. If we are to store “health-related data” as well as medical records, we are talking about a lot of data and a lot of information about an individual. It’s use could be damaging as well as beneficial.

    Some of this is now being done. The Department of Veterans Affairs (VA) now stores a lot of patient data. Because of the volume, it uses several geographic sites for such storage. Data can be accessed from anywhere but if it is stored elsewhere it may take a day or so to get it. Many health facilities provide patient access to their records through a “patient portal” over the internet but the facilities maintain their individuality rather than pooling the records. For example, through their individual “patient portals” I can download lab results from Lab Corp, emergency room visit reports from Mercy, visit and lab reports from the VA, etc. A common Continuity of Care Document (CCD) is also widely used. This is information useful for initial and return visits.

    In northern California we have developed SacValley Medshare which already stores a large number of records from multiple facilities. This will provide pooling of information but it hasn’t reached the stage of universality and maturity needed to make it really useful for epidemiologic work.
    Unfortunately, we are still suffering from a problem that I tried to address back in the 1960’s: getting computers to work for doctors instead of forcing doctors to work for computers. For example, the system designers quite naturally institute data input techniques which facilitate system operation but are foreign and often awkward for doctors to use, and the doctors dislike it. To improve the quality of data we need to design the systems so that doctors like, accept, and use them.

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.