More complex EHRs can result in declines of quality measures

Electronic health records boosted to meet stage 1 meaningful use standards produced an increase in quality. But there were setbacks once systems were further upgraded and became more complex, according to a new study.

Dartmouth College researchers examined data from 3,921 nonfederal hospitals from 2006 to 2010 to see how EHR adoption affected quality in four clinical categories.

For the study, hospitals were categorized into five levels based on the clinical applications available on their EHR systems. Level zero described a system with primitive capabilities and level four, the highest, described systems that had capabilities of all the proceeding levels as well as computerized physician order entry. Level three had the necessary capabilities to meet stage 1 meaningful use requirements.

Data on hospitals’ adoption levels were merged with quality data reported to the Centers for Medicare & Medicaid Services’ Hospital Compare website associated with heart attack, heart failure, pneumonia and surgical care infection prevention.

Researchers found that adoption of a level three EHR system was associated with gains in quality for heart attack, heart failure and pneumonia care by about 0.35 to 0.49 percentage points. There was not a gain in surgical infection prevention. But researchers found that a transition from level three to level four systems resulted in declines of 0.9 to 1 percentage point in quality for heart attacks, heart failure and surgical infection prevention care. Pneumonia care did see a quality increase of 0.13 percentage points.

The reason for the disparity is most likely linked to the presence of CPOE, said study co-author Ajit Appari, PhD, a research fellow at the Center for Digital Strategies at the Tuck School of Business at Dartmouth College. The study, “Meaningful Use of Electronic Health Record System and Process Quality of Care: Evidence from a Panel Data Analysis of U.S. Acute-Care Hospitals,” appeared online July 20 in Health Services Research (link).

Appari said CPOE is not a fully matured technology, and it needs to work with an effective clinical decision support system to yield positive results. He said most of the CPOE systems in use today have basic decision support capabilities, such as drug interactions, but not clinical guidelines and pathways.

The only CPOE requirement under stage 1 meaningful use is that 30% of patients with at least one drug in their medication lists have at least one prescription ordered through a CPOE system. The stage 2 regulations, released Aug. 23, double that percentage to 60%, and add a CPOE requirement of 30% of laboratory and radiology orders.

The Dartmouth study was conducted before the stage 2 regulations were released. Researchers did not attempt to draw a connection between further regulations and quality.

The findings of the Dartmouth study contradict studies conducted by the Healthcare Information and Management Systems Society, including one published in April. That study, conducted on behalf of Thomson Reuters, now Truven Health Analytics, found that hospitals in advanced stages of EHR adoption were more likely to set national benchmarks for performance than their peer hospitals with less advanced EHR systems (link).

Another HIMSS study, published in June in partnership with Press Ganey, found a correlation between advanced EHR adoption and scores in the Hospital Value-Based Purchasing Program. That is a Centers for Medicare & Medicaid Services initiative that rewards acute-care hospitals with incentive payments based on how closely they follow best clinical practices and how well hospitals enhance patients’ experiences of care.

John Hoyt, executive vice president of organizational services for HIMSS, said that based on his society’s studies, he believes the quality declines the hospitals in the Dartmouth study experienced were temporary.

“When hospitals go through the turmoil of moving from a paper-based system to an electronic system, there’s a huge amount of process redesign,” Hoyt said. “There could be some dissatisfaction in the staff, and errors might increase until people are used to the new processes, and you could get some short-term, temporary reductions in quality measures.”

Appari said his research focused on the incremental moves from one stage to the next and did not factor in how quickly a hospital moved from each stage. But “EHR systems are quite complex, and it is certainly a challenging endeavor in adopting one for any organization, regardless of smaller or larger hospital,” he said.

Success also could depend on the processes that are in place before EHR implementation. The Dartmouth study found that the most significant quality improvements were found at hospitals with a lower quality baseline before implementation. The gains made by hospitals that started out with a higher baseline were not as significant. The authors believe the hospitals did not have as much room to grow and that the technology simply helped the organizations sustain the high quality of care.

Although the study focused on hospitals, “It’s quite plausible to see similar results in physician practices with respect to outpatient and ambulatory quality measures,” Appari said. But, he said, a decline in quality with more advanced systems isn’t a definite result. There are things practices and hospitals can do to make their implementations more successful.

“First of all, it is important to understand that EHR systems are ‘living’ systems requiring careful nourishment, continuous building of strengths, and consistently exploiting those strengths routinely at point of action,” Appari said. He said technology can be leveraged to improve quality by continuous use of the systems and constant updating to the data sets used for decision-making.

Hoyt said it’s important that physicians are the champions when it comes to technology implementation. He said too many physicians go through the process of selecting a system only to pass the tasks of implementing and customizing the system to an office manager.

“Lack of [physician champions] will cause maybe an error increase, or a lack of optimizing the benefits a system can bring,” he said.