Predicting COVID-19 Incidences from Patients’ Viral Load using Deep-Learning

The transmission of the contagious COVID-19 is known to be highly dependent on individual viral dynamics. Since the cycle threshold (Ct) is the only semi-quantitative viral measurement that could reflect infectivity, we utilized Ct values to forecast COVID-19 incidences. Our COVID-19 cohort (n=9531), retrieved from a single representative cross-sectional virology test center in Lebanon, revealed that low daily mean Ct values are followed by an increase in the number of national positive COVID-19 cases. A subset of the data was used to develop a deep neural network model, tune its hyperparameters, and optimize the weights for minimal mean square error of prediction. The final model’s accuracy is reported by comparing its predictions with an unseen dataset. Our model was the first to capture the interaction of the previously reported Ct values with the upcoming number of COVID-19 cases and any temporal effects that arise from population dynamics. Our model was deployed as a publicly available and easy-to-use estimator to facilitate prospective validation. Our model has potential application in predicting COVID-19 incidences in other countries and in assessing post-vaccination policies. Aside from emphasizing patient responsibility in adopting early testing practices, this study proposed and validated viral load measurement as a rigid input that can enhance outcomes and precision of viral disease predicting models.

Cytomegalovirus Viral Load in Transplanted Patients Using the NeuMoDx™ (Qiagen) Automated System: A 1-Month Experience Feedback

Cytomegalovirus (CMV) reactivations represent a significant morbidity and mortality problem in transplant patients. Reliable and rapid measurement of CMV viral load is a key issue for optimal patient management. We report here the evaluation of NeuMoDx™ (Qiagen) in a routine hospital setting (University Hospitals of Marseille, France) in comparison with our classical reference technique R-GENE. During one month, 719 CMV viral loads from 507 patients were measured in parallel in both techniques. Using the ROC (receiver operating characteristic) curve and our biological experience we suggest that values <52 IU/mL (geometric mean) correspond to negative samples, values >140 IU/mL (Fowlkes–Mallows index) correspond to quantifiable positive results and values ranging from 52 to 140 IU/mL represent non-quantifiable positive results. Follow-up of 15 transplant patients who developed CMV reactivation during the study showed that NeuMoDx™ provided higher viral load measurement during the first two weeks of follow-up for three patients. These important intra-individual variations resulted in a significant median increase considering the whole data set (6.7 points of difference expressed as a percentage of the initial viral load). However, no difference between the two techniques was noticeable after two weeks of treatment. Subsequent to this first study we conclude that NeuMoDx™, used with optimized logistics and an adapted threshold, allows a rapid CMV viral load measurement and that its use does not lead to any difference in patient management compared to the reference technique R-GENE®.

Incidence and Predictors of Loss-to-follow-up Among Pregnant and Breastfeeding Women on Option B+ PMTCT Program in Northwest Ethiopia: a Retrospective Follow-up Study.

Background: Although Ethiopia is implementing an option B+ program for the last seven years, lost to follow-up among HIV positive women remains the major problem for ART treatment. The study was conducted on lost to follow-up among human immunodeficiency virus infected pregnant and breastfeeding women after Option B+ modality since there was a scarcity of literatures in Northwest Ethiopia. The result is also expected to strengthen the effort to eliminate mother-to-child transmission of HIV to 5%. Methods: A retrospective follow-up study was conducted among 403 pregnant and breastfeeding women between June 2013 and December 2019 at the University of Gondar comprehensive specialized Hospital. Cox proportional hazards regression model was used for identifying predictors of lost-to-follow-up. Results were reported as hazard ratios with 95% confidence interval at a significance level of p=0.05. Results: The overall incidence rate of loss to follow-up was 9.4 per 1000 person-months of observation (95% confidence interval (CI): 7. 4‒11. 9). According to the multivariable Cox regression, rural residency (adjusted hazard ratio(AHR): 2.30; 95% CI:1.08‒4.88), being Muslim religion follower (AHR: 2.44; 95% CI:1.23‒4.81), having no baseline viral load measurement (AHR: 4.21; 95% CI:2.23–7.96, being on ART before enrolment (AHR: 0.30; 95% CI:0.15–0.62), having drug side effects (AHR: 1.82; 95% CI:1.01–3.33), same day ART initiation (AHR: 3.23; 95% CI:1.53–6.84) and having sub-optimal adherence level (AHR: 3.96; 95% CI:2.18–7.19) were significant predictors of lost to follow up.Conclusion: The incidence of lost to follow-up is lower as compared to evidence from most African countries but slightly higher than the WHO target. It is better to strength and expand viral load measurement for all women and giving attention to those women who are residing in a rural area and having fair/poor adherence level.