Impact of rapid influenza diagnostic test on physician estimation of viral infection probability in paediatric emergency department during epidemic period

•Performance of the RIDT after clinical examination significantly improved the accuracy of physicians’ estimates of influenza during the peak of a flu epidemic.•Performing RIDT reinforces the clinician’s clinical diagnosis.•The RIDT result positive enabled a significant decrease in orders for chest X-rays and laboratory tests.

BackgroundThe clinical diagnosis of influenza is difficult in the younger children.ObjectivesEvaluate the impact of rapid influenza diagnostic test (RIDT) on clinicians’ estimation of the clinical probability of influenza in children.Study designThis prospective study included children aged from 1 month to 5 years who were admitted in a university paediatric emergency department during an influenza epidemic period and presented with fever without source. The RIDT Quickvue® was performed on nasopharyngeal aspiration and results were confirmed with immunofluorescence and/or PCR. The clinical probability of influenza and serious bacterial infection (SBI) was evaluated for each child before and after the physician(s) was informed of the RIDT results.Results170 children were included from January 15th through March 18th, 2013. After the only clinical examination, the overall clinical probability of influenza was 66.0% [CI 95%: 63.04–68.4], and was significantly increased at 92.4% [CI 95%: 89.5–95.3] in case of positive RIDT and significantly decreased at 30.8% [CI 95%: 29.0–32.5] in case of negative RIDT without knowing the results of laboratory tests. Whereas the initial clinical probability of influenza were appropriate regarding the prevalence (66.0% vs. 57.0%), the probability of SBI was overestimated (30.2% vs. 8.8%). The RIDT result positive enabled a significant decrease in orders for chest X-rays (64,4% vs. 45.8%, p < 0,05) and laboratory tests (71,1% vs. 41.1%, p < 0,05).ConclusionsThe RIDT seems to be a useful diagnostic tool for ED clinicians in epidemic conditions. Improving clinician estimation of flu probability would reduce orders for imaging and testing.