- Managing patient identities is one of the hidden problems of health information exchange (HIE) and electronic health record (EHR) technology. While every patient needs a single, unique identifier tied to his medical records and billing information, a simple typo or a misspoken birthday can leave a patient with duplicate records, potentially compromising his care. Many hospitals use their own master patient indexes (MPIs) to check if patients have an existing record, but the emergence of HIE means new challenges in ensuring accuracy across multiple providers for millions of new patients.
Beth Just, MBA, RHIA, FHIMA, and president and CEO of Just Associates, Inc., tells For the Record that an MPI involves “all of the patient’s identifying information, such as names, birth dates, Social Security numbers, addresses, and phone numbers, [being] stored in a central data table.” But unique identifiers must be the same across the entire HIE for the network to function, which can be accomplished with the help of a system-wide enterprise MPI (EMPI). An EMPI provides its own identifier that spans the entire network, requiring accurate and complete data across every department or healthcare provider contributing to a patient’s care.
Part of the core effectiveness of an HIE system is its ability to match patients across multiple departments of a single hospital, or different providers all together. Accuracy and completeness, however, are two of the biggest problems in any large data set, and especially for patients, who may marry, have children, use different names over time, change addresses, and forget key information.
They are also a problem for the staff performing the data entry, often under conditions that require them to be as speedy as possible: mishearing the spelling of a surname or transposing a letter in a street address can accidentally create an entirely new record and identification number for a patient who may have visited the facility before, and whose record may contain critical information for his wellbeing.
Quality control during the admissions process is the first step towards ensuring accuracy, and training staff in standard practices can help prevent mistakes. But without consulting the patient, it’s nearly impossible to tell if the original data is correct or if a mistake was made previously and the new input is the proper information. Implementing safeguards in patient record software to prevent these small mistakes may seem like an easy task, but human error will always find a way to defy technology. HIE vendors are trying to work around these problems with a variety of innovative ideas.
While some companies are exploring biometric data as a solution, using face recognition and scanning patients for unique features such as iris pattern and fingerprints, others are turning to cloud-based solutions that match patients and providers across any system. Probabilistic algorithms are another avenue being explored. These formulas assign mathematical weights to different pieces of information in a patient’s record to predict the likelihood that the new information matches an existing file.
These systems are typically more expensive that simpler deterministic methods that simply match input to existing data, but they may be worth the price. “If you use less sophisticated algorithms in your record matching, you’re going to have potentially more false-positives,” Just says, describing the situation where records from multiple individuals are merged into a single HIE record. “You want to avoid that at all costs.”
Patient records are continually developing collections of changeable data. The goal of HIE and EHR is to manage this data cleanly, effectively, and with the maximum benefit to the patient, but the margin for error only grows wider as more and more providers try to collaborate and share information. HIE systems must be capable of preventing mistakes caused by duplication or accidental deletion of records, but healthcare providers themselves are ultimately responsible for being certain that the patient in front of them matches the information on their computer screen.