Published rates of error for interpretation of radiological studies range from 5-8%, with a major error rate of 1-3%. Uncommonly but regularly, these errors result in death or litigation. The key to avoiding this is to actively reduce these errors.
When targeted feedback and training is provided to radiologists, error rates have been shown to decrease, for example in one study from 19% to 4%.
However, current QA programs just choose studies randomly for double read. The future is in identifying studies at high risk of error for double-read, and then to provide individualized feedback.
The ScanGuardTM System was developed to make this a reality.
The ScanGuardTM system uses an proprietary adaptive computer algorithm to select scans for double-read. Factors considered include scan acuity, clinical question, scan type, the personal and historical profile of the interpreting radiologist, a historical database of common medical errors, and an assessment of the chance of misunderstanding or miscommuncation of the findings.
Flagged scans are then double-read by a GuardianAngelTM, who is a radiologist trained to be the safety net of each primary interpreting radiologist. The GuardianAngel also has access to a customized list of findings to verify or rule out, based on the assessment of the ScanGuard system. Any discrepancies are discussed with the interpreting radiologist and if necessary, communicated to the ordering physician.
When receiving reports from our teleradiology service, rest assured that these reports are protected by ScanGuardTM!