Self-reported symptoms in French primary care SARS-CoV-2 patients: association with gender and age group.

Introduction: The early identification of COVID-19 patients is of outmost importance in the current pandemic. As with other pathogens, presenting symptoms of SARS-CoV-2 may vary, depending on sociodemographic factors. We aimed to describe the clinical characteristics of COVID-19 patients by age/gender and to assess whether the diagnostic performance of these symptoms varied according to these variables.

Methods: We analysed data from a cross-sectional study involving primary care patients undergoing RT-PCR testing in Lyon, France. Among patients who tested positive, we examined whether there was an association between age/gender and various symptoms. In addition, we calculated the diagnostic performance of the most specific symptoms (smell/taste disorder).

Results: Among 1543 consecutive patients, 253 tested positive (16%). There were significant age/gender-related differences in symptoms. In middle-aged women, the diagnostic performance of smell/taste disorders were AUC = 0.65 [95%CI 0.59-0.71] and PPV = 72% [95%CI 53-87%], that is higher than in the entire sample (smell/taste disorders: AUC = 0.59 [95%CI 0.57-0.62] and PPV = 57% [95%CI 47-67%]. In contrast, the negative predictive values of smell/taste disorders were similar in both groups (85% [95%CI 81-89%] for middle-age women and 86% [95%CI 85-88%] for the entire sample).

Conclusion: We found significant age/gender-related differences in the clinical characteristics of COVID-19 patients. Screening strategies based on smell/taste disorders performed better in middle-aged women, but could not ensure a diagnosis of COVID-19 in any subgroup of patients. Future diagnostic strategies should use age/gender differentiated approaches.