IT investments pay dividends by identifying errors

April 28, 2006

Of the $120 billion spent on healthcare in the U.S. annually, about 10% to 20% is wasted on inappropriate treatment. What if some of those funds were invested in enterprise information systems? Imagine the possibility of routinely using such systems to identify trends in medical imaging procedures and highlight potential errors, inaccuracies, and waste.

Of the $120 billion spent on healthcare in the U.S. annually, about 10% to 20% is wasted on inappropriate treatment. What if some of those funds were invested in enterprise information systems? Imagine the possibility of routinely using such systems to identify trends in medical imaging procedures and highlight potential errors, inaccuracies, and waste.

Such enterprise-wide systems could also identify a better test based on proven outcomes or alert physicians that an exam order duplicates a recently reported procedure, said Dr. Ramin Khorasani, vice chair of radiology at Brigham and Women's Hospital in Boston, during a presentation at the Society for Imaging Informatics in Medicine meeting on Thursday.

The chance to develop such IT safeguards could come in 2007, when Medicare will add a 2% reimbursement incentive to healthcare facilities that support treatment with better outcomes. The funds should encourage imaging departments to identify misused, overused, and underused procedures, Khorasani said.

"You need to find out what goes wrong in your department in order to take steps to correct it," he said.

Brigham and Women's recently completed an 18-month study evaluating diagnostic procedure volumes and comparing them with positive results, follow-up recommendations by radiologists, and lost revenue tied to billing.

"When we started the project, we discovered that the data had never been previously collected for this type of analysis," he said.

Khorasani shared data from an evaluation of 1000 CT head examinations. The frequency of ordering by individual emergency room physicians could identify potential overuse and underuse, both of which can affect patient treatment and safety.

"We discovered that one physician ordered a CT head exam for one out of every 10 patients who admitted to the ER. Further data mining determined that a portion of these patients had no symptoms that could possibly merit a head CT exam," he said.

The percentage of positive outcomes was compared with individual categories of referring physicians. For example, exams ordered by neurosurgeons produced 20% higher outcomes than general practice physicians. But that does not necessarily mean that general practice physicians are ordering the procedures in error, he said. It's logical that patients of neurosurgeons will demonstrate a high percentage of positive outcomes from head imaging procedures.

The evaluation also tracked radiologists' recommendations for additional procedures. A surprising percentage of referring physicians did not order additional studies, Khorasani said. In other findings, Khorasani learned that some radiology residents do not follow ACR protocols. And about 1% of all procedures misidentify the ordering physician. Overall, the study determined that for every CPT code billed, $10 is wasted in some manner.

On a positive note, ER physicians who learned that a head CT had recently been performed cancelled a new head CT order, relying on the report results of the original exam.

Khorasini encouraged attendees to utilize their enterprise information systems to perform similar studies, take the results to their chief operating officers, and ask that 50% of wasted costs be invested in information technology.

"You can use information available in enterprise systems to increase quality, improve patient safety and become a more efficient department," he said.