Five Trends in Patient Matching for 2020

Five Trends in Patient Matching for 2020

By Lisa A. Eramo, MA

Consider this scenario: A 50-year-old female presents to the emergency room with shortness of breath, chest and shoulder pain, and nausea. A registrar quickly collects the patient’s name, date of birth, and insurance information before a nurse begins an assessment to rule out a heart attack.

Meanwhile, the registrar performs a quick search and doesn’t find an exact match. They create a new medical record number under which all of the patient’s care is documented and billed without realizing a medical record number already exists. The result? A duplicate record. The number of times this scenario occurs at most hospitals? Far too often.
Patient matching challenges aren’t new, says Letha Stewart, RHIA, director of customer relations at QuadraMed Corporation. What’s novel is the increased focus on these challenges and the downstream effects on outcomes and cost—particularly as more providers consolidate, participate in health information exchange (HIE), and join accountable care organizations (ACOs). Stewart identified five trends that will affect patient matching in 2020 and beyond.

1. Growing desire to learn more about patients across care settings, systems. Prior to ACOs and HIE, providers were primarily concerned with patient matching within their organization and mostly for the purpose of preventing medical errors, says Stewart. However, with value-based payment models, providers are starting to consider the bigger picture—that is, how patient matching (or lack thereof) across inpatient and outpatient settings contributes to costs. For example, a primary care physician can educate high-cost patients about the importance of medication adherence to minimize unnecessary trips to the emergency room. This intervention is nearly impossible without the ability to match patients correctly and analyze cost trends, she adds.

Where health systems go wrong is assuming that if they migrate all providers to the same electronic health record (EHR), they’ve solved the problem of patient matching. In reality, the problem is only solved at a local level, says Stewart. If a hospital or health system is part of an ACO, it also needs to share information with other entities that are part of that same ACO, many of which may not share the same EHR, she adds.

According the Office of the National Coordinator (ONC) for Health IT, 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, says Stewart.

The value proposition for matching patients within an ACO is relatively clear: All providers benefit financially when patients achieve positive outcomes at the lowest cost. Being able to match and track those patients is critical. However, one hurdle is that entities within the same ACO may also be vying for the same patients. This means these entities are often only willing to share the minimum amount of information necessary to satisfy the conditions of the ACO, says Stewart.

For example, a hospital may not want to share demographic information for its entire patient population with a competitor, such as a third-party lab vendor. If the lab has access to that information, it could theoretically market its services to patients who haven’t ever used the lab, she adds.

2. Framing patient matching in the context of patient satisfaction. Patients don’t want multiple bills from multiple providers for a single episode of care, says Stewart. When records are accurately linked, organizations can send one consolidated bill for hospital and physician services. This may improve patient satisfaction rates and the likelihood that patients will pay their bills, she adds. In addition, accurate matching prevents patients from receiving duplicate services that also drive up costs, such as a lab or other test linked to a duplicate record, says Stewart. When a provider can’t find the results because they’re in the duplicate record, they may ask the patient to repeat the lab or test.

3. Use of referential data. With referential matching, organizations use information from multiple sources, including but not limited to credit bureaus and the US Postal Services, to accurately identify patients.

“Referential data is an important tool in patient matching because it’s current and updated automatically with no specific action on the part of the patient,” says Stewart.

As patients move, for example, referential databases capture new addresses from utility and mortgage companies as well as the US Postal Service. As patients get married, referential databases capture name changes from DMV records. Healthcare organizations aren’t privy to these changes unless and until patients return for care, and even then, there’s no guarantee that patients will provide updated information, says Stewart. The more referential data sources, the better. For example, hunting and fishing licenses can help organizations match minors who may not have a credit history or a wealth of other public records, she adds.

One caveat is that patients themselves don’t validate information flowing into the referential database, making it easy for registrars to draw inaccurate conclusions. For example, a referential database may list a patient as living at a specific address when in reality they’ve lived at a nursing home for several years. Registrars may assume that it’s two different individuals when in fact, it’s the same individual, she adds.

“As an industry, we all need to agree on what constitutes a match when we’re looking at referential data,” says Stewart. “How much data is necessary, and what specific data is necessary? There’s no standard currently. All vendors that use referential data have their own proprietary algorithms for matching. Standards will help the industry gain the most value from referential data while also minimizing the risk of inaccurate matches.”

4. Tailoring matching algorithms to specific patient populations. Although referential data is beneficial to patient matching, one limitation is that it tends to focus on national data rather than regional nuances, says Stewart.

For example, the percentage of Hispanic individuals is higher in Texas than it is nationally. This means certain names may appear more frequently in Texas, reducing providers’ ability to rely on those names as unique identifiers. Increasingly, providers will use statistical frequency analyses of specific populations within a geographic area or even within a single health system to weight certain identifiers appropriately.

5. Greater reliance on cell phone numbers. Organizations are starting to rely more frequently on cell phone numbers as unique identifiers because patients tend to keep the same number for years, says Stewart. In addition, children and teenagers increasingly use cell phones. Unlike the Social Security number (SSN), patients willingly provide their cell phone number so they can receive appointment reminders and other alerts.

Although healthcare organizations often collect email addresses to sign patients up for portals, this information isn’t necessarily helpful in terms of patient matching, says Stewart. Registrars can easily make mistakes when typing email addresses into the EHR. In addition, patients often have more than one email address, and they may not provide the same address at each encounter, she says.

The same is true for social determinants of health (SDOH) data. Organizations increasingly collect this data, but it isn’t necessarily helpful for patient matching because it can change very quickly. For example, a patient may be homeless at one encounter but then find housing a week later.

An Ongoing Debate

Stewart says the national patient identifier (NPI) will continue to receive attention in the next couple of years. “I think it’s a good idea, but I don’t think it’s going to be an easy transition,” she adds. She provides the following challenges that could result in patient matching errors even when using an NPI:

  • Patient is unable to provide their NPI (e.g., they’re unconscious or forgot the number)
  • Registrars mistype the NPI during a current or previous encounter
  • Patient doesn’t yet have an NPI because they were born shortly after the government issued numbers
  • Patient isn’t a US citizen

In addition, there are privacy and security concerns. “Hackers can do anything,” says Stewart. “Once they have your NPI, they could access all of your clinical and financial information across systems if they wanted to do so.”


Lisa Eramo ( is a freelance writer and editor in Cranston, RI, who specializes in healthcare regulatory topics, health information management, and medical coding.

Leave a comment

Send a Comment

Your email address will not be published. Required fields are marked *