Why patient identification remains a daunting challenge for healthcare

As healthcare M&A activity continues and the creation of clinically integrated networks and other collaborative care arrangements play a greater role in the business strategy toward value-based care and population health, effectively and accurately identifying patients across the care continuum is essential.

Poor patient identification leads to diagnosis errors, redundant medical tests, skewed reporting and analytics, and billing inaccuracies. As data sharing matures, an enterprise view of patient information is needed for informed clinical-decision making, effective episodic care, and a seamless patient experience during every encounter in the network. The challenge for healthcare organizations is how to efficiently integrate and resolve patient identities from various aligned medical groups, outpatient clinics, post-acute facilities, affiliated subsidiaries, and acquired hospitals when establishing a larger enterprise.

Disparate data spread throughout multiple and diverse organizations in the continuum contributes to a proliferation of duplicate and incomplete patient records. The fragmented nature of IT systems means that individuals receiving care and services from more than a single provider in the network often have medical records in several locations. Large-scale consolidations exacerbate the issue, as acquired EHR systems and legacy applications often reside in silos.

Integration remains a core problem

In an industry daunted by software applications and IT systems that fail to communicate or share information effectively, the real challenge of patient identity management ultimately revolves around integration and lack of utilizing information exchange standards. More specifically, it revolves around the demographics and associated identifiers dispersed across numerous IT systems.

Because IT systems often lack an ability to effectively communicate with each other, and because they store their data using a fragmented architecture, an excessive proliferation of identifiers occurs. The result is unreliable demographic information, triggering further harm in data synchronization and integrity.

Clearly, keeping these identifiers and demographics as localized silos of data is an undesirable model for healthcare that will never function properly. While secondary information such as clinical data should remain local, the core identity of a patient and basic demographics including name, gender, date of birth, address, and contact information shouldn’t be in the control of any single system. To overcome this challenge, organizations should invest in enterprise, cloud-based tools like an Enterprise Master Patient Index (EMPI), that externalize this information from these insulated applications to maintain accuracy and consistency across all connected systems within the delivery network.

An EMPI provides faster, more accurate identification that can be extended to multiple scenarios within the healthcare environment. Managing individuals across diverse systems and geographic locations, as well as linking patients to their providers and incorporating social determinants outside of the healthcare domain, addresses interoperability in a scalable and manageable architecture and ensures the delivery of a full and complete medical record.

Beyond removing duplicates and automating record cleanup, an EMPI enables access to patient data in a single location. An enterprise unique identifier (EUID) generated by the EMPI serves as a link to an individual’s record in any system, streamlining clinical and administrative workflows for patient data access points like medication history, lab results, and visit summaries. As a centralized index of patient demographics, the EMPI also allows an individual’s data to be reconciled in real time, so the individual is identified the moment he or she enters the system, thus avoiding the creation of new duplicates.

Innovations in cloud computing are helping organizations build a secure and scalable EMPI platform that is continuously maintained, updated and accurate. This approach allows upfront costs to be minimized by streamlining the deployment process, eliminating the need to maintain and monitor the solution with internal staff, and deferring the costs of the system to a point in time where it is already providing tangible business value. Cloud’s ability to ingest mass volumes of data—residing in various, siloed systems—and present it across the enterprise with a high level of transparency, allows providers to access accurate patient information for more meaningful decision-making.

The move toward a cloud-based service also opens the door for sophisticated machine learning algorithms which can learn from the manual interventions typically made by data stewards. The ability to deduce decision patterns, such as when to designate two records unique or designate them as a match, will allow the EMPI to eventually perform nearly all its actions autonomously with minimal user intervention. With a more flexible, scalable solution to meet their data needs, an even broader adoption of an EMPI as a Service can be expected in the industry.

Hope for universal identifier solution

Focus on patient identification and matching has accelerated in the past few years, with several industry stakeholders working to tackle the issue. However, many of these initiatives have demonstrated that there is no single patient-matching solution that will ensure 100% accuracy.

While a national patient identifier has been touted as the silver bullet to overcoming this problem, an EMPI as a central piece of the infrastructure would still be required to ensure that every system is using that identifier, and that the central hub of identifiers remains consistent and duplicate free. That is not to say that pursuing a universal patient identifier isn’t a laudable goal. Rather, that we should treat a unique identifier as just another strong indicator of identity that, in conjunction with other data points, can assist with matching.

The optimal solution to accurately identify a patient at any time, in any care setting, will always be a combination of technology, processes, and people. This includes the right technology for effectively linking and de-duplicating patient data, streamlined processes for identifying individuals at registration, and having the right people in place for records administration.