Patient Matching - Fixing an Identity Problem

Howdy! We are back with another topic around Patient Matching, which is the ability to identify and link records for the same patient within and across different health systems and organizations. It is the practice of connecting disparate patient records across different medical providers or facilities. The outcome of a successful patient matching is that when a patient is visiting two different doctors or hospitals, it yields the same medical record.

Patient matching and record linkage support interoperability by determining whether the patient records held within a single facility or in different healthcare organizations correctly refer to a specific individual.

According to HealthIT.gov, Patient matching is defined as the identification and linking of one patient's data within and across health systems to obtain a comprehensive view of that patient's health care record. At a minimum, this is accomplished by linking multiple demographic data fields such as name, birth date, phone number, and address. 

A robust patient matching strategy is fundamental to successful interoperability and the health information technology infrastructure. Mismatched or incomplete patient data can lead to decreased clinician trust in the information they are seeing, inadvertent privacy issues, operational inefficiencies, and even resulting in fatal patient health outcomes.

There is a lack of a standardized way to link disparate patient records to create a complete longitudinal record of care due to which each patient interaction with a hospital, clinic, pharmacy, laboratory, commercial or government health insurer, generates patient demographic data that is recorded, managed, and exchanged differently which creates the downstream challenge of efficiently and securely identifying and linking patient from multiple health systems.

Patient matching sounds straightforward. However, in reality, it is one of the most complicated parts of interoperability, and one of the most important to get right. Implementation of Patient Matching is based on ensuring that the patient’s primary data- the basic demographic information will ideally connect the right patient to the correct records across any care setting, provided the data is accurate.

Common Issues Encountered in Patient Matching

Implications of Patient Mismatching

Patient mismatch errors are often perpetuated or compounded when an organization migrates its data to new systems or merges with another organization, silo’ed electronic health record systems, and transitions from paper records of past and patients moving around between various and unrelated medical groups and hospitals.

Patient matching errors occur frequently in the healthcare system. A recent report from eHealth Initiative Foundation and NextGate confirmed that 38% of US healthcare providers reported an adverse event due to an issue in patient matching.

Inaccurate patient matching can lead to fragmented or duplicate patient records, which in turn lead to delayed, inappropriate, or unnecessary care, inefficiencies in care coordination and healthcare billing.

The implications of a patient mismatch can be categorized into two broad categories:

·        Mismatched data resulting in patient harm

·        Mismatched data resulting in financial strain on health systems

Mismatched data resulting in patient harm

A study revealed that roughly one in five patients have incomplete health records which can translate into Providers having an imperfect view of a patient’s medical history, patient records may be delayed, and unnecessary testing or improper treatment may be ordered, there could be misunderstandings about medications, miscommunication of follow-up instructions, or delays in care leading to potentially fatal consequences. Therefore, accurate patient identification is paramount to making sure that patients receive appropriate and safe care. 

Mismatched data resulting in financial strain on health systems

Mismatched patient data has also resulted in a significant cost burden on the health system.

According to a Black Book survey, duplicate patient records have cost hospitals an average of $1,950 per patient. It has been found that roughly 33 percent of denied claims are due to mismatched or incorrect patient information. Denied claims cost hospitals an average of $1.5 million and $6 billion annually for the healthcare system.

Approaches taken to improve Patient Matching

Unique Identifier system

Creation of a unique identifier system that can unambiguously identify patients and correctly link the patients to their respective records is a seemingly simple solution that can improve patient matching to a great extent.

Master Patient Index (MPI) and Enterprise Master Patient Index (EMPI)

Master Patient Index or MPI aims to identify individual patients by storing and analyzing demographic information and assigning a unique identifier to that person. However, study has shown that the MPI within the EHR is often too limited in its ability to compare records coming from different sources because of which the EHR inadvertently creates duplicate records and splits up a patient’s information making it harder for a care provider to track and find the right information.

The concept of Enterprise Master Patient Index (EMPI) provides the capability to move beyond a single EHR. Unlike MPI, EMPI is designed to be a centralized, cross-platform solution designed to link and reconcile records in real-time from diverse systems and settings of care, including HIEs, ACOs, radiology groups, outpatient clinics, physician practices, labs, and rehabilitation facilities to name a few. An EMPI leverages both probabilistic and deterministic matching algorithms to account for minor variations in patient data.

The EMPI technique can match records against a huge range of parameters including name, family name, address, sex, date of birth, post code and phone to name a few. An Enterprise Unique Identifier (EUID) is assigned to each patient record and this identifier together with its associated clinical data, forms a single, up to date patient data repository, done through a matching engine. The EMPI’s basic function is to manage and house the EUIDs. A data matching engine is the heart of EMPI software. This mechanism can identify matching patient records and link them together into a single record. The accuracy of the matches depends on the type of algorithm the system’s using.

Data standardization

An important technique widely used for patient matching includes the approach that relies on comparing demographic data—such as names and birthdates—that are stored in different records to determine whether those records refer to the same individual. However, variation in the types of demographic data collected by different organizations and discrepancies in data formatting among systems—such as abbreviating “Street” or “Avenue,” or including different fields for city, state, and ZIP code— can affect patient matching using this technique.

In the recent days, standardizing “address” has shown good result in improving patient matching. In organizations with an existing match rate of 85%, the standardization of “Address” alone has reduced the percentage of unlinked records by 20%.

Standardizing data elements such as “Address”, is an important step towards solving the Patient Matching problem. Recently ONC has started an initiative “Project US@”, to create health care specific standard for “Address”.

Biometric attributes

Biometrics is the analysis of individually distinguishable characteristics such as fingerprints, palm vein scanning, iris scanning, and facial recognition. Biometrics technology as an add-on to comparing patient records based on demographics has the potential to significantly reduce the percentage of patient mismatches. Biometric identification technologies are advantageous because such data is more difficult to “steal, exchange or forget”. Unlike unique numbers or ID cards, a patient can’t forget or lose track of his or her fingerprints, and a facial scan doesn’t rely on a patient being responsive when wheeled into the emergency department.

An example use case could be a change in patient address or a last name due to a divorce, which can potentially lead to a case of patient mismatching. Biometrics could provide an additional data point to help improve the accuracy of matches as they could create a unique combination along with patient demographics data. A potential shortcoming that needs to be addressed is when two separate health systems are using different biometric systems or scanners. If the two systems are not interoperable, then mismatches can occur. Also, another important aspect is how can one ensure guarantee that patient biometric data is protected and used according to the specific guidelines.

Therefore, Biometrics technique has privacy concerns that need to be addressed.

Referential databases of records

Matching and linking patient medical records within and across the healthcare ecosystem that are done using records from non-healthcare sources, such as drivers’ licenses, is typically called “Referential Matching.”

A growing number of organizations are implementing the Referential Matching software to increase odds of identifying patients correctly. In Referential matching software a third-party service provider adds an additional layer of demographic data (typically from outside of healthcare) including datasets from credit reporting and public utilities that are routinely updated and maintained to enhance patient matching.

Concerns have been raised that Referential Matching could lead to clinicians and payers having access to personal and financial information like credit information. Referential Matching also has limitations related to certain patient populations including children, homeless individuals, and undocumented immigrants because data sources used for referential matching do not contain or have limited information on these sections of the population.

Radio Frequency Identification

Emerging technologies, like radio frequency identification (RFID) are being analyzed to enhance patient identification. RFID technology uses wireless communication (radio waves) to identify patients. Unlike existing barcode technologies, RFID labels can hold more data than barcodes and be read automatically without user intervention. In addition, RFID technologies offer re-writeable functionality, allowing information to be modified. This technology also offers more advanced forms of data security like encryption, allowing for patients’ health data to be kept more secure. However, is not widespread due to its high cost and lack of standards or guidelines for implementation within healthcare. RFID technologies also raise privacy concerns over the wireless transmission of patients’ health information and collection of patients’ health data by third-party actors without patients’ approval or knowledge.

All the above techniques currently used in patient matching are not mutually exclusive but can work in tandem.

Accurate and unique identification of patients enhances data sharing and interoperability and is essential for patient care and safety. Patient identification techniques ranging from UPIs to algorithms to biometric identification have been implemented worldwide—each accompanied by their own set of opportunities and challenges with no single solution having a 100% match rate.

Patient matching issues have rapidly moved into the forefront with the current public health crisis. ONC’s newly released Cures Act final rule, aims to standardize certain data elements and API interfaces which has the potential to improve patient matching.

More research is however needed around strategies to ensure a robust patient matching process which in turn will lead to improved interoperability, quality of care and patient safety.

How are you handling patient identity problems in your app? If you have specific input and wish to contribute more on this topic, please feel free to reach us at healthviva@taliun.com.

Adios!

Team HealthViva