Propagated by experts, veterans of industry, and healthcare players spanning every facet of the healthcare ecosystem is the myth that healthcare interoperablity is a fantansy.
Is the United States healthcare system so complex that healthcare interoperability is a fantasy that will never be achieved?
Submitted to the ONC Blockchain Challenge, a paper that I co-authored with Dr. William Dailey, a practicing physician and chief of medical information at Golden Valley Memorial Healthcare, proposes solutions to the ongoing concerns regarding healthcare interoperability in the United States. Our paper, which is the second one I submitted to the challenge, is titled “Micro-Identities Improve Healthcare Interoperability With Blockchain: Deterministic Methods for Connecting Patient Data to Uniform Patient Identifiers.”
The new definition of healthcare interoperability
Interoperability is an old concept dating back to the eighth century BC. Society has been struggling for centuries with the idea of combining individual parts or components to create a whole unit. However, interoperability with healthcare only came about in published works 23 years ago, in 1993.
More recently, in 2013, HIMSS defined of healthcare interoperability as “the ability of different information technology systems and software applications to communicate, exchange data and use the information that has been exchanged.” This definition is alluring but inaccurate and incomplete. I’d like to offer this updated definition:
“Healthcare interoperability is the ability for multiple healthcare ecosystems to work in harmony without unreasonable efforts by the ecosystems’ producers and consumers.”
Healthcare has a track record of forgetting the patient.
FHIR gets us close
The topic of healthcare interoperability affects every patient treated in the United States. The ONC Blockchain Challenge gets a nation thinking about solutions, not expounding on the problems. Our paper on micro-identities dives into how this new healthcare ecosystem will interoperate. Global experts in bioinformatics believe a solution exists and that we as a society just can’t stop searching for it. Gil Alterovitz, a Harvard Medical School professor in the school’s Division of Medical Sciences, a core faculty member at the Computational Health Informatics Program at Boston Children’s Hospital and co-chair of HL7’s Clinical Genomics Work Group, said that he found the paper insightful and a necessary step exploring options toward a national healthcare interoperability solution. Alterovitz stated,
“I found the temporal geospatial use for identification to be an interesting little nugget. I don’t think I have seen that before explicitly called out as a feature — where the patient would use the previous location in helping make an ID. On another hand, it’s practicality is a question mark — time will tell. Also, it is one of the first pieces out there linking blockchain to implementation via FHIR.” – Gil Alterovitz, Ph.D., Harvard Medical School professor in the Division of Medical Sciences
Our paper describes HL7 FHIR’s flexible framework with the ability to support mobility and mobile health, social media, personal health records, public health, payment systems and clinical research. Leveraging this framework for integration between providers is straightforward. The HL7 FHIR standard is faster to learn, faster to develop and faster to implement. The framework also is free. It does not involve proprietary software or adapters.
Our paper highlights that while the specification is comprehensive, it does have gaps evident during healthcare implementations.
“The HL7 FHIR specification will soon be complete and implemented by major electronic medical record (EMR) vendors supported by the Argonaut Project. At that time, each healthcare entity will have RESTful interface(s) to the information they choose to make available to trusted partners, accessible by way of their HL7 FHIR server(s); however, specific gaps exist in locating the appropriate server, obtaining entity trust and requesting collocated patient information at the next point of care (POC). Patient demographics such as date of birth, gender, country, postal code, ethnicity and blood type may be slightly different due to multiple factors and ensuring acquisition of the correct data. Identifying the patient correctly is essential to care.
Blockchain technology has the potential to address these technology gaps. Novel methods for identity verification and sharing healthcare “event” transactions between entities, will advance healthcare stakeholders towards complete solutions.”
HL7 FHIR provides a needed standard for the clinical sharing of patients’ medical information. However, the challenge of matching patient identities between facilities remains unsolved.
Identity matching
Throughout the paper, Dailey and I explore matching patient identities at Golden Valley Memorial Healthcare, in a condition referred to as high sensitivity and low specificity. Sensitivity refers to a test’s ability to designate an individual with the disease as positive. A highly sensitive test means that there are few false negative results, and thus fewer cases of the disease are missed. The specificity of a test is its ability to designate an individual who does not have a disease as negative. The process and result are documented in detail within the paper.
This experiment inspired a data test at a small hospital system in Missouri, using real patient data. Golden Valley Memorial Healthcare is a small, 50-bed, rural healthcare facility. Interestingly, it has two separate, yet highly patient-concordant EMR systems (inpatient and ambulatory), making it ripe for analysis. Each EMR system contained approximately 70,000 to 90,000 records.
How would this patient matching between facilities work? Allow me to explain.
The result was that John, a sample patient, was verified. The only data viewable by the receptionist is an option to select the date of the visit. Once that is selected, a list of matching locations appeared in a drop-down box. The receptionist performed a verification, allowing all available data to be immediately pulled from all encounters at that facility, as a cascade of FHIR queries for the patient, John.
What’s accomplished by this method? This process unlocks healthcare interoperability for the alignment of patient identity, confidentiality, integrity and accessibility, resulting in better patient outcomes.
Exploring trust verse truth for healthcare
Today, we’ll explore one of the three research propositions examined in the paper, trust vs. truth.
“At some point, virtually every health system will be compromised. Healthcare leaders have a duty to verify the integrity of their healthcare systems independently. Today, this is done by adding new security components into the environment, e.g. virus protection software, hard or soft firewalls, virtual private networks, etc. The fundamental assumption in the decision to patch security gaps with hardware or software assumes that components will not be compromised. It is troubling that when transmitting data, it’s not possible to determine if new or old components have been compromised. Now, with blockchain, healthcare system administrators can prove the healthcare data has not been compromised. This is accomplished by establishing data authenticity with the chain of custody utilizing blockchain technologies.
The Keyless Signature Infrastructure (KSI) is designed to provide scalable digital signature based authentication for electronic data, machines, and humans. Every health care data transfer can be captured and timestamped, creating proof of authenticity and restoring truth into our healthcare system. This paradigm shift offers data integrity and visibility, previously unheard of – moving healthcare towards transparent truth, not trust. KSIs can resolve the lack of consistent methods for conducting patient matching, and decrease the occurrence of out of date and incorrect patient matching errors. Leveraging KSI can prevent man-in-the-middle (MiM) attacks. MiM attacks are a process where a user gains unauthorized access to communications between two parties who believe they are directly communicating with each other. When applied to healthcare, MiM attacks can alter valid matches, resulting in unauthorized users consuming the data for unknown and potential nefarious purposes, manufacturing “no matches found” despite the availability of valid matches. KSI helps to ensure data integrity and authenticity, protecting the patient.”
Healthcare can utilize blockchain technology to provide a nationally shared healthcare resource. As our paper, “Micro-Identities Improve Healthcare Interoperability With Blockchain: Deterministic Methods for Connecting Patient Data to Uniform Patient Identifiers,” suggests, this resource will enable patient identity matching, identity linking, redundant connectivity, location and the retrieval of granular patient data to and from any EMR.
Using blockchain technologies in the healthcare setting will represent a significant accelerator for healthcare interoperability.
For information on the first paper I submitted to the ONC Blockchain Challenge, please read my Aug. 13 post: “Healthcare Interoperability Research Propositions of the ONC Blockchain Challenge.“