Neuroimaging reports in a general hospital: results from a quality-improvement program

Neurologists working in district general hospitals (DGHs) in the UK frequently rely on neuroimaging reports from general radiologists. Neuroradiologists and general radiologists may disagree in the interpretation of magnetic resonance imaging and computerized tomographs of brain and spine. We sought to analyse disagreements between reports from general radiologists and neuroradiologists in selected patients and to examine the impact of a four-point quality-improvement program in a DGH neuroimaging reporting service.

A single neurologist selected patients for reporting by neuroradiologists following a recommendation from general radiologists, or because of a concern by the neurologist. Differences between general radiologists and neuroradiologists in pre-planned primary and secondary findings and advice for further investigations were compared.

Primary finding disagreements occurred in 41 of 307 patients (13.4%) and secondary finding disagreements were identified in 62 patients (20.2%). There was no evidence from either result of improvement compared to an earlier study, p=0.45 and p=0.52 respectively. Neuroradiologists suggested further investigations in 42 patients (13.7%). Recurrent areas of disagreement included distinguishing perivascular spaces from ischemia, while recurrent missed lesions included subdural hemorrhage and cortical dysplasia.

Despite implementation of a quality-improvement program neuroradiologists frequently identified major discrepancies and recommended additional investigations in this DGH neuroimaging service. Future research should identify interventions which are more effective in improving neuroimaging reports in DGHs.