We aimed to elucidate the accuracy and optimal cut-off point of the self-diagnosis of influenza and the associated clinical symptoms of children by their guardians, compared with those of the rapid influenza diagnostic test (RIDT).

Seasonal influenza is a common outpatient problem during the winter season. A paediatric influenza epidemic has socio-economic impacts like temporary school closure, school event cancellations, and unscheduled work absences among parents. Hence, early identification and assessment of influenza to prevent its spread is important from a societal perspective.

We performed a cross-sectional observational study in a rural clinic in Japan every winter season from December 2013 to March 2016. We retrospectively extracted information from the medical records and pre-examination checklists of 24 patients aged <12 years (mean age, 5.4 years; men, 54.2%). The data extracted from the medical records and pre-examination checklist included the baseline characteristics (age, sex and past medical history of influenza), clinical signs and symptoms, diagnosis by guardians (%) and RIDT results.

The optimal cut-off point of the self-diagnosis of influenza by guardians was 80%, with a sensitivity and specificity of 63.6% (95% confidence interval: 30.8–89.1) and 92.3% (64.0–99.8). At a 50% cut-off point, the sensitivity and specificity were 90.9% (58.7–99.8) and 53.8% (25.1−80.8). The accuracy of feeling severely sick, as estimated by the guardians showed a sensitivity and specificity of 90.9% (58.7–99.8) and 69.2% (38.6–90.9). Our study indicates that the diagnosis of seasonal influenza by guardians to their children would be useful in the establishment of both confirmatory diagnoses when it has high probability above the optimal cut-off point (80%), and exclusion diagnosis when it has low probability (50%). Not feeling severely sick, estimated by the guardians might be a useful indicator for the exclusion of paediatric influenza.