What healthcare can learn from credit card companies about catching medical errors
Healthcare has long struggled with detecting and addressing medical errors. It’s been more than 20 years since the groundbreaking U.S. Institute of Medicine’s piece To Err is Human, and as an industry, we’re far from achieving zero errors.
However, there are things that we can learn from a most unexpected place, the credit card in your pocket. Credit card companies have been making significant advancements in finding and managing anomalous activity, which the industry refers to as “fraud.” While the issue of detecting medical errors is quite different from detecting credit card fraud, both fields face similar challenges – the detection of behaviors that stray from the norm.
Credit card companies have risen to the challenge and are investing in what will be a $63B global fraud detection and prevention market. These new programs have led to recent and projected decreases in existing credit card fraud claims in the United States. As these efforts develop, what can healthcare leaders learn from the work being done by credit card companies?
Review Everything, Smartly
Fraud detection systems have been around since the 1990s but, in the past, heavily depended on static rules that only reviewed a small sample of credit card transactions. For a long time, the promise of new technology failed to deliver, like the healthcare industry’s experience with EHRs. The root of the problem for credit card companies was a high technology infrastructure barrier, coupled with the fact that automation and algorithms had yet to reach a level of maturity that could deliver scalable value. Early fraud warning systems were simplistic, relying on human-coded rules which failed to account for the myriad of situations in the real world. This combination of factors led to incorrect approvals on unlawful activity while blocking legitimate transactions, and costing, according to some estimates, $118B in lost revenue.
Rules-based checking led to inaccurate and costly consequences that no credit card company nor hospital wants. In today’s EHR age, a corollary to this is the constant warnings coming from EHR systems based on limited data streams. Unfortunately, this has led to the widespread problem of alert fatigue in many healthcare systems, wasting time and frustrating healthcare practitioners.
Today in the credit card industry, modern solutions are addressing both problems. Leading fraud detection systems can review every transaction and use state-of-the-art artificial intelligence models with higher sensitivity and specificity. This new paradigm of analysis detects fraud in a way that hand-written rules could never have and, equally important, does not trigger as many distracting false positives.
In healthcare, many clinicians and leaders know the pain of hearing and seeing a multitude of alarms both from the EHR and other medical devices, many of which are more distracting than helpful. America’s leading credit card companies use fraud detection systems to meaningfully reduce risk exposure, act more efficiently, and help customers transact in this dynamic world. Taking inspiration from the credit card industry, health systems can find value moving beyond the old models of reviewing a small sample of clinical actions and instead leverage new technologies, such as AI used in the quality workflow, to catch medical errors and improve patient safety.
Address the End-to-End Workflow
In complex systems such as hospitals or financial institutions, detection is the first step. The finding must be verified, and the appropriate parties must be notified for insights to be helpful. Here, credit card companies’ best practices can be a source of learning. When a credit card transaction is deemed at risk for being fraudulent and thus declined, there was, in the past, a long and arduous process to prove innocence or guilt, often driving good customers away. With their new fraud detection tools, credit card companies caused significant problems for case investigators later in the process. The key was to think about and address the continuum of the entire issue.
In hospital patient safety initiatives, such as peer review in radiology, identifying the medical error is simply the first step. In fact, another type of medical error that commonly occurs is errors in communication and coordination, where information critical to patient care is not seen or is not passed along. Hospitals can benefit from thinking about the handoff of information and assessing communication workflows so that dangerous medical errors do not go unresolved. Efficiently validating and then communicating the error to the appropriate parties is where value is realized in all industries.
More recently, credit card companies have implemented advanced case investigation management systems that make it easier for fraud agents to resolve fraud risks accurately and quickly. These tools integrate data across the relevant systems and help the user focus on the most critical cases first. The manual processes are automated, and the right people are updated. One case study found a 50% increase in productivity, and another saw a reduction in fraud by 25%.
In healthcare, AI tools can empower staff to take a similar approach to reduce medical errors once they are detected. The key lesson here is investing in an integrated infrastructure through which medical errors can be communicated is as important as finding the medical errors themselves. Today’s credit card industry has evolved to use advanced tools integrated with common-sense workflows to find and address anomalies. The healthcare industry can learn much from these lessons by implementing AI and communication tools that ensure that the right team members are empowered with the right insights to act on and prevent future medical errors.
Health systems can leverage these cross-industry learnings by asking a few questions:
- What percent of your clinical decisions get a second look?
- What is the administrative burden to “right the wrong” and improve patient safety, and how might you reduce that burden with better communication processes and tools?
- Where are the weak points in your end-to-end workflow for communicating errors and improving patient care?
How is your healthcare system addressing medical errors?
Drop us a note and share your learnings.