Community hospitals also benefit from speech recognition

The benefits of speech recognition software (SRS) are not exclusively for the largest medical centers, according to a study that found a 24-fold improvement in median radiology report turnaround times after SRS was implemented at a smaller community hospital.

Published online Oct. 22 in the Journal of the American College of Radiology, the study also found that SRS can be added without negatively impacting radiologist productivity.

“Although the majority of prior publications have focused on the benefits of SRS in large academic practices, our findings suggest that similar results can be attained in the community hospital setting and that the effect may be disproportionally larger,” wrote Luciano M. Prevedello, MD, MPH, of Harvard Medical School in Boston, and colleagues.

The authors examined an SRS implementation at a 150-bed community hospital with no radiology training program that occurred between May and July 2011. Report turnaround times and normalized radiologist productivity were determined for five-month periods before and after the software’s implementation.

Median radiology report turnaround times dropped from 24 hours to one hour, reported the authors. Turnaround times in the 80^th percentile fell from 60 to 10 hours, and the 95^th percentile fell from 165 to 33 hours. The effect of SRS appears to be sustainable as turnaround times remained stable post-implementation.

Prevedello and colleagues credited several factors for the success of the implementation. A physician leader with experience in technology adoption championed SRS implementation and a detailed planning effort was used to create report templates and training documents.

“Improvements in radiology report [turnaround times] can be achieved without significant adverse impact on radiologist productivity when coupled with quality improvement techniques including committed leadership, IT planning, and support as well as access to performance metrics,” they wrote.