Ricerc@Sapienza

Since the most frequent symptoms of novel coronavi- rus 2019 disease (COVID-19) are common in influenza A/B (FLU), predictive models to distinguish between COVID-19 and FLU using standardized non-specif- ic laboratory indicators are needed. The aim of our study was to evaluate whether a recently dynamic nomogram, established in the Chinese population and based on age, lymphocyte percentage and monocyte absolute count, might apply to a different context. We collected data from 299 patients (243 with COVID-19 and 56 with FLU) at Policlinico Umberto I, Sapienza University of Rome. The nomogram included age, lymphocyte percentage and monocyte absolute count to differentiate COVID-19 from FLU. Odds ratios (ORs) and 95% confidence intervals (95% CIs) were calculated for all associations. Multivariate logistic regression models were used to adjust for potential confounding. A p-value of less than 0.05 was consid- ered statistically significant. Patients with COVID-19 had higher age, lymphocyte percentage and monocyte
absolute count than patients with FLU. Although uni- variate analysis confirmed that age, lymphocyte per- centage and monocyte absolute count were associat- ed with COVID-19, only at multivariate analysis was monocyte count statistically significant as a predictive factor of COVID-19. Using receiver operating charac- teristic (ROC) curves, we found that a monocyte count >0.35x103/mL showed an AUC of 0.680 (sensitivity 0.992, specificity 0.368). A dynamic nomogram includ- ing age, lymphocyte percentage and monocyte abso- lute count cannot be applied to our context, probably due to differences in demographic characteristics be- tween Italian and Chinese populations. However, our data showed that monocyte absolute count is highly predictive of COVID-19, suggesting its potential role above all in settings where prompt PCR nasopharyn- geal testing is lacking.