RadNet, which operates more than 230 outpatient imaging centers in eight states and performs over 5 million imaging exams each year, needed to improve medical results turnaround time and radiologists' work environment, while at the same time reducing costs.
Normally, when an x-ray, ultrasound, CT scan, MRI, PET/CT scan, or mammogram is completed, a radiologist interprets the image and dictates his or her findings into a digital recording system. Transcriptionists then transcribe these recordings, and the radiologist who dictated the original report electronically approves the report. Once the radiologist approves the original report, it is then sent (via fax, electronic messaging protocols, or snail mail) to the physician who requested the diagnostic imaging exam in the first place.
This protocol posed two problems. RadNet was spending over $10 million each year on transcription costs, and there was a typical lapse of 24 to 72 hours from when the radiologist dictated the report to when he or she signed the report.
To address these issues, the company decided to develop its own system using a small vendor's speech-to-text recognition engine. That enabled RadNet to provide a highly customizable workflow for each of its 500+ radiologists, who could customize the workflow and workspace to their liking.
RadNet anticipates potential annual savings of $7 million to $8 million. And because the company controls the software, it can respond to radiologists' suggestions in mere weeks, rather than having to depend on a ponderous software-enhancement process from a commercial vendor. Equally important, because the radiologist generates the report and signs off on it right away, RadNet can eliminate the typical long delays.