How 5 key steps can jump-start patient matching accuracy
Absent consensus about how best to match patients to records, providers need to realize that using best practices is a key to interoperability.
One of the great consequences of COVID-19 is the rapid shift to digital enablement. Healthcare providers had no choice but to move to contactless admissions and new patient onboarding, try to support e-case reporting and look for ways to electronically share patient information – all in a virtual health environment.
These new digital requirements, combined with staffing challenges, require more efficient processes. Interoperability and the ability to share critical patient data cannot be overlooked as an essential way to optimize the front office. Thus, data sharing is elevated when organizations use patient matching best practices.
It may not be intuitive to a provider to think of patient matching as the first line of defense when it comes to interoperability. The reality is “according to the Office of the National Coordinator for Health IT (ONC), there’s a 50 percent to 60 percent match rate as data is shared across unaffiliated organizations. This less-than-optimal rate leads to duplicates, overlays and overlaps.” While patient matching is one of the most complicated parts of interoperability, and one of the most important to get right, but when it's done right, could help reduce costs and support better patient outcomes.
Consider this example. If a clinician needs records for James Smith, the patient matching algorithm must search multiple entries for this prevalent name by considering all demographic information input by staff. Because James Smith is a very common name, identifying the correct person can be incredibly difficult. Matching a patient not only depends on the information staff has entered but also the information other staff members at other facilities and systems may have typed in incorrectly. In this instance, a query for James Smith can match on ZIP Code alone.
The human element is what makes patient matching one of the most frustrating parts of interoperability. After electronic matching occurs, a human can make the wrong decision on which patient to select, leading to potentially adverse outcomes on the quality of care. In fact, a recent report from the eHealth Initiative Foundation and NextGate confirmed that 38 percent of U.S. healthcare providers reported an adverse event because of a matching patient issue.
There are many supposed cures to the common patient record mismatch on the market – each one seeming to be a one-stop solution to the untrained interoperability eye. However, implementation cannot happen without cooperation from clinical teams, frontline staff and patients. Here are a few of the most powerful patient matching solutions offered today and why they are not a one-stop-shop solution for your interoperability challenges.
Enterprise Master Patient Index (EMPI) helps hospitals and health systems eliminate duplicate patient records and inaccurate patient information across IT systems. EMPI is considered cross-platform, meaning it aggregates the patient ID from multiple systems and correlates it to a master patient index. However, this EMPI platform is only the start of patient identification accuracy. It can be challenging to implement, and the platform requires full participation with consistent refinement and uploading of patient data. If there is one thing we have seen these days, it's that getting patient consent to do just about anything – especially with their data – can be frustrating and challenging.
Record Location Service (RLS) is a tool that helps organizations find the patient across various EHR connections through a centralized database of known patient encounters across the country. In theory, RLS leverages medical record numbers and a customizable scoring algorithm to improve workflow across clinical teams. Unfortunately, this method is not widely implemented, and often newer patients suffer as a result.
The idea of a National Patient Identifier was introduced in 1996 when HIPAA called for a system that would assign each U.S. citizen a permanent and unique number to be used across the continuum of care. Because of privacy concerns, Congress did not allow the Department of Health and Human Services to implement national patient identifiers. While this is a solution that I support, the sheer scale of implementing standardized system ID numbers to all aspects of care would come with astronomical monetary, time and maintenance costs.
The reality is that while all of these solutions seem like magic bullets, they are not. Any would be a tremendous step towards interoperability, but providers need to pump the brakes before going down these paths.
5 patient matching best practices
Recognizing that patient matching will always be an imperfect science is critical. Any of the previously mentioned patient matching tactics can work, but not without the help of providers and clinical teams on the front lines. Before approaching any "big ideas" regarding patient matching in an overall interoperability strategy, these five patient matching best practices can help overall accuracy.
Train staff to be the frontline defenders of interoperability. Make sure staff understand why accuracy in patient information is so critically important. Equip them with knowledge, because algorithms only go so far. Help them understand that they are frontline defenders of sound interoperability – and that is important work. What feels manual and insignificant to them at the front desk is actually tied to automation and efficiencies for the entire office. In addition to collecting the specific patient information, staff need to:
- Watch for typos.
- Understand that patient demographics are fluid and can change every single appointment.
- Capture the legal name – John Thomas III goes by "Trey" in another system. “Is Trey your legal name?”
- Identify name changes – “Has a married person taken their partner's last name?”
Asking patients to verify data accuracy. Accurate patient identification is critical to providing the best patient care. However, patient matching implications can happen during the patient registration process. Patient identifiers, such as demographic data, Social Security numbers and patients' addresses, can be mistyped easily; this may cause discrepancies and duplicate records in the system. You can eliminate this challenge by asking one simple question every time the patient enters your door: is your patient information still current and correct? Just this small step can improve accuracy significantly. In addition, answering these questions can help you think about the impacts of patient matching to set up the best systems and processes.
Revisit and revise. Many approach patient matching implementations with a "set it and forget it" mentality. The No. 1 thing I hear when talking to providers is, "I put protocols in place so patient matching will automatically happen, right?" or "Doesn't my EHR already do patient matching for me?" The reality is patient matching – and interoperability at large – require consistent and ongoing adjustments. There is no solution that magically fixes interoperability – it is a surprisingly manual process. If insufficient data and processes go into it, poor results will ensue. Thus, it is essential to revisit algorithms to ensure they optimize patient matching automation.
Stop before you start from scratch. Many providers believe that they need to start from scratch, but before you pay a developer to build an algorithm, evaluate current technology. Many EHRs come with patient-matching capabilities "out of the box" that can be simply enabled and configured.
Implement a process for unmatched records. Again, many EHRs come with capabilities that already have an exceptions report that lists any electronic patient document that could not be matched automatically. Organizations should ensure that staff knows to run this report and that accountability is assigned to team members to work it on a daily basis.
There is no one single magic bullet to matching patients to their records – absent a magically overarching solution that does not exist at present, constant human attention should be the rule for all.
Matt Becker is the vice president of interoperability at Kno2, where his focus is on improving interoperability across the entire continuum of care.