MRI dedicated to the emergency department for diplopia or dizziness: a cost-effectiveness analysis.

OBJECTIVES: The purpose of this study was to compare the costs and organizational benefits of diagnostic workup without and with MRI dedicated to the ED. METHODS: We conducted a prospective observational uncontrolled before-after study in one ED of a university hospital in France from July 1, 2018, and January 3, 2020. We included all consecutive patients presenting with dizziness or diplopia. The main outcomes were the clinical decision time of ED physicians and the total costs for each strategy. Outcomes were compared using propensity score with inverse probability weighting in the 2 arms and an incremental cost-effectiveness ratio (ICER) was calculated. RESULTS: Among the 199 patients during the "before" period (average age: 60.4 years ± 17.6): 112 men (57%), and 181 during the "after" period (average age, 54.8 years ± 18.5): 107 men (59%), the average costs were €2701 (95% CI 1918; 3704) and €2389 (95% CI: €1627; 3280) per patient, respectively. The average time to clinical decision was 9.8 h (95% CI: 8.9 10.7) in the group "before" and 7.7 h (95% CI: 7.1; 8.4) in the group "after" (ICER: €151 saved for a reduction of 1 h in clinical decision time). The probabilistic sensitivity analysis estimated a 71% chance that the MRI dedicated to ED was dominant (less costly and more effective). CONCLUSION: Easy access to MRI in the ED for posterior circulation stroke-like symptoms must be considered a relevant approach to help physicians for an appropriate and rapid diagnostic with reduction of costs. TRIAL REGISTRATION: NCT03660852 KEY POINTS: • A dedicated MRI in the ED for diplopia or dizziness may be considered an efficient strategy improving diagnostic performance, reducing physicians' decision time, and decreasing hospital costs. • This strategy supports clinical decision-making with early treatment and management of patients with posterior circulation-like symptoms in the ED. • There is 71% chance that the MRI dedicated to ED was dominant (less costly and more effective) compared with a strategy without dedicated MRI.

Solid pseudopapillary tumour should be part of differential diagnosis of focal pancreatic lesions with increased 18 F-FDOPA uptake.

OBJECTIVE: To assess the specificity of increased 18 F-dihydroxyphenylalanine (18 F-FDOPA) uptake in patients who underwent PET/CT for suspicion of isolated pancreatic neuroendocrine tumour (pNET). False-positive results mimicking a pNET have been investigated. MATERIAL AND METHODS: Carbidopa-assisted 18 F-FDOPA PET/CT scans performed in patients with suspicion of localized pNET were retrieved. Only patients with a definitive diagnosis were retrospectively included. When available, the histopathological result after pancreatic surgery was the gold standard. In other cases, the diagnosis was based on endoscopic ultrasonography (EUS)/cytology and/or on concordant imaging results of at least two of the following: contrast-enhanced computed tomography (CE-CT), magnetic resonance imaging (MRI) and somatostatin receptor scintigraphy (SRS). RESULTS: Forty-four among 731 patients were selected. Among these, 36 patients (82%) were surgically treated, revealing pNET (n = 28), solid pseudopapillary tumour (SPT) (n = 4), adenocarcinoma (n = 2), serous cystadenomas (n = 1) and solitary fibrous tumour (n = 1) cases. An additional three cases of pNET were diagnosed by EUS/cytology. In the remaining five patients, a consensus was reached on follow-up imaging results: pNET (n = 1), serous cystadenoma (n = 2) and undetermined/no pNET (n = 2). Both specificity and negative predictive value of 18 F-FDOPA PET/CT for localized pNET were 67%. Surprisingly, all four false-positive results were SPTs showing intense 18 F-FDOPA uptake and negative SRS. There was no significant difference in 18 F-FDOPA uptake intensity between PET-positive pNETs and SPTs. CONCLUSION: 18 F-FDOPA PET/CT is not specific for pNET in patients with localized pancreatic lesions. SPT could mimic pNET and should be part of differential diagnosis in such a clinical situation. If these results are confirmed in a broader population, the imaging pattern 18 F-FDOPA PET-positive/SRS-negative lesions might be considered as the imaging phenotype of SPT.