How better software choices could cut US health care costs

(Press-News.org) Denied insurance claims are among the biggest challenges facing the U.S. health care system – driving up costs and leaving providers with billions in unpaid bills.

 

Hilal Atasoy, an associate professor at the Rutgers Business School, thinks she’s uncovered an elegant if simple solution: better software management. Her paper on the topic appears in the September issue of the journal MIS Quarterly.

 

Medical claims contain a range of information, including patient demographics, medications, medical histories and procedures performed. Claims can be denied when they include erroneous information or uncovered services, among other errors.

 

To strengthen the quality of health data that doctors collect, in 2009, U.S. lawmakers adopted the Health Information Technology for Economic and Clinical Health Act, or HITECH, which incentivized the use of electronic health record (EHR) systems. The goal – lauded at the time – was to make technology the “foundation for health care reform.”

 

It was only partially successful. While the move incentivized adoption, the resulting mix of systems from different vendors created interoperability challenges that can contribute to claim denials, Atasoy said. When software used by one clinic or hospital system differs from another, patient data can get jumbled or go missing – making denials more likely.

 

Between 2000 and 2020, some $745 billion worth of care was provided without compensation, and claim denials were a significant contributor. Atasoy wanted to understand the potential role that EHR technologies played in these health payer rejections. 

 

To quantify the connection between unpaid bills and gaps in software interoperability, Atasoy and colleagues from Temple University and the University of Wisconsin-Madison analyzed the health records of more than 19 million anonymized patient visits to 48 Maryland hospitals. The dataset included information on whether initial claims were approved or rejected.

 

By comparing this data with information on EHR adoption and software sourcing, the researchers determined that software decisions play an outsized role in whether claims are initially approved or denied. The dataset did not capture appeals or eventual resubmissions.

 

For instance, they found that when hospitals relied on multiple vendors rather than a single vendor, claim denials rose. By contrast, hospitals that standardized software from a single vendor saw denials decrease to 1.03% from 1.18% — a seemingly small shift that translates into significant savings across millions of claims.

 

The researchers also investigated whether physicians’ familiarity with EHR applications from different vendors could explain claim denials. They found that when physicians practiced across hospitals with similar EHR systems, claim denials decreased, since familiarity improved data accuracy and compliance.

 

“Our findings suggest that [EHR] alignment improves clinical data collection and adherence to payer requirements, ultimately reducing claim denials,” they wrote.

 

Reducing the cost of health care in the U.S. will require a basket of solutions; technology is only part of the answer. And yet, as the researchers found, EHR coordination has the potential to reduce unnecessary spending, and in turn, bring costs back down to earth.

 

“If there is any inefficiency in this process, that can lead to surprise bills for clinics and health care organizations, which in turn shifts the cost burden to patients and eventually, to taxpayers,” said Atasoy. “Reducing denials is among the best ways to help contain skyrocketing health care costs.”

 

The findings suggest that the federal government should work to strengthen EHR certification and minimize variation among vendors, Atasoy added.

 

“Placing greater emphasis on the usability and standardization of user interfaces and layouts will improve the accuracy of information flow across disparate applications, suggesting a potential to reduce the extent of errors in claim processing,” the researchers wrote.

END