The project was performed within the William Beaumont Health System's department of radiology in part to satisfy the American Board of Radiology maintenance of certification (MOC) requirements. A team consisting of presenter Dr. Benjamin Lack, James Nugent, and Dr. Kurt Tech sought to analyze the quality of radiology reports at the institution. Written radiology reports are a vital component of every radiology exam, and improving these reports could be a way to improve communication and patient care, according to Tech.
"Written report quality was also chosen because there was a concern that changing practices related to generating reports could have an adverse affect on the quality of reports," Tech said.
These changing practices included the removal of the transcriptionist from the report-generating process, thus taking out the services of an independent proofreading editor. This step in the process had previously served to decrease report errors, he said.
In addition, pressure to deliver rapid turnaround time requires the radiologists to immediately proofread and sign their reports, he said.
"Proofreading immediately after dictating is inherently flawed, as one's brain tends to fill in or replace erroneous computer transcription segments or phrases with what they intended to say and not what is transcribed," Tech told AuntMinnie.com.
While turnaround time is a legitimate quality metric for both medical and service quality, excessively fast turnaround time is thought to be inversely related to the quality of the written report, regardless of the quality of the speech recognition software, according to Tech.
The researchers randomly selected nearly 1,000 radiology reports created at their institution in 2011, and they compared radiologist scores on report content, clarity, and typographical errors with scores provided by internal medicine physicians. They found that radiologists scored the reports better on content and clarity than their internal medicine colleagues, suggesting room for improvement in written communication, he said.
After a plan was implemented to improve the quality of reports, there was a trend toward improving report scores. However, this difference was not statistically significant.
"Improvement of written radiology reports has the potential to benefit patient care through improved communication, but it poses a number of challenges and there are barriers to improvement," he said.
In other findings, both radiologists and internists preferred the report impression at the end of the report, in line with classical organizational principles, Tech noted. Also, one-fourth of the respondents read the report findings section less than half of the time, suggesting a high likelihood that ordering clinicians may not receive information that's not included in the impression.
