Increasing situational awareness through nowcasting of the reproduction number (preprint)

The time varying reproduction number R is a critical variable for situational awareness during infectious disease outbreaks, but delays between infection and reporting hinder its accurate estimation in real time. We propose a nowcasting method for improving the timeliness and accuracy of R estimates, based on comparisons of successive versions of surveillance databases. The method was validated against COVID-19 surveillance data collected in Italy over an 18-month period. Compared to traditional methods, the nowcasted reproduction number reduced the estimation delay from 13 to 8 days, while maintaining a better accuracy. Moreover, it allowed anticipating the detection of periods of epidemic growth by between 6 and 23 days. The method offers a simple and generally applicable tool to improve situational awareness during an epidemic outbreak, allowing for informed public health response planning.

Priority age targets for COVID-19 vaccination in Ethiopia under limited vaccine supply (preprint)

Background The worldwide inequitable access to vaccination claims for a re-assessment of policies that could minimize the COVID-19 burden in low-income countries. An illustrative example is what occurred in Ethiopia, where nine months after the launch of the national vaccination program in March 2021, only 3% of the population received two doses of COVID-19 vaccine. In the meantime, a new wave of cases caused by the emergence of Delta variant of SARS-CoV-2 was observed between July and November 2021. Methods We used a SARS-CoV-2 transmission model to estimate the level of immunity accrued before the launch of vaccination in the Southwest Shewa Zone (SWSZ) and to evaluate the impact of alternative age priority vaccination targets in a context of limited vaccine supply. The model was informed with available epidemiological evidence and detailed contact data collected across different socio-demographic settings. Results We found that, during the first year of the pandemic, 46.1-58.7% of SARS-CoV-2 infections and 24.9-48% of critical cases occurred in SWSZ were likely associated with infectors under 30 years of age. During the Delta wave, the contribution of this age group in causing critical cases was estimated to increase to 66.7-70.6%. However, our findings suggest that, when considering the vaccine product available at the time (ChAdOx1 nCoV-19; 65% efficacy against infection after 2 doses), prioritizing the elderly for vaccination remained the best strategy to minimize the disease burden caused by Delta, irrespectively to the number of available doses. Vaccination of all individuals aged 50 years or older would have averted 40 (95%CI: 18-60), 90 (95%CI: 61-111), and 62 (95%CI: 21-108) critical cases per 100,000 residents in urban, rural, and remote areas, respectively. Vaccination of all individuals aged 30 years or more would have averted an average of 86-152 critical cases per 100,000 individuals, depending on the setting considered. Conclusions Despite infections among children and young adults likely caused 70% of critical cases during the Delta wave in SWSZ, most vulnerable ages should remain a key priority target for vaccination against COVID-19.

Waning of SARS-CoV-2 vaccine-induced immunity: A systematic review and secondary data analysis (preprint)

The emergence of Omicron (B.1.1.529) variant of SARS-CoV-2 in late 2021 was followed by a marked increase of breakthrough infections. Estimates of vaccine effectiveness (VE) in the long term are key to assess potential resurgence of COVID-19 cases in the future. We conducted a systematic review of manuscripts published until June 21, 2022 to identify studies reporting the level of protection provided by COVID-19 vaccines against SARS-CoV-2 infection and symptomatic disease at different time points since vaccine administration. An exponential model was used to perform a secondary data analysis of the retrieved data to estimate the progressive waning of VE associated with different vaccine products, numbers of received doses, and SARS-CoV-2 variants. Our results show that VE of BNT162b2, mRNA-1273, ChAdOx1 nCoV-19 vaccines against any laboratory confirmed infection with Delta might have been lower than 70% at 9 months from second dose administration. We found a marked immune escape associated with Omicron infection and symptomatic disease, both after the administration of two and three doses. The half-life of protection against symptomatic infection provided by two doses was estimated in the range of 178-456 days for Delta, and between 66 and 73 days for Omicron. Booster doses were found to restore the VE to levels comparable to those acquired soon after administration of the second dose; however, a fast decline of booster VE against Omicron was observed, with less than 20% VE against infection and less than 25% VE against symptomatic disease at 9 months from the booster administration. This study provides a cohesive picture of the waning of vaccine protection; obtained estimates can inform the identification of appropriate targets and timing for future COVID-19 vaccination programs.

The decline of COVID-19 severity and lethality over two years of pandemic (preprint)

Undernotification of SARS-CoV-2 infections has been a major obstacle to the tracking of critical quantities such as infection attack rates and the probability of severe and lethal outcomes. We use a model of SARS-CoV-2 transmission and vaccination informed by epidemiological and genomic surveillance data to estimate the number of daily infections occurred in Italy in the first two years of pandemic. We estimate that the attack rate of ancestral lineages, Alpha, and Delta were in a similar range (10-17%, range of 95% CI: 7-23%), while that of Omicron until February 20, 2022, was remarkably higher (51%, 95%CI: 33-70%). The combined effect of vaccination, immunity from natural infection, change in variant features, and improved patient management massively reduced the probabilities of hospitalization, admission to intensive care, and death given infection, with 20 to 40-fold reductions during the period of dominance of Omicron compared to the initial acute phase.

The decline of COVID-19 severity and lethality over two years of pandemic (preprint)

Undernotification of SARS-CoV-2 infections has been a major obstacle to the tracking of critical quantities such as infection attack rates and the probability of severe and lethal outcomes. We use a model of SARS-CoV-2 transmission and vaccination informed by epidemiological and genomic surveillance data to estimate the number of daily infections occurred in Italy in the first two years of pandemic. We estimate that the attack rate of ancestral lineages, Alpha, and Delta were in a similar range (10–17%, range of 95% CI: 7–23%), while that of Omicron until February 20, 2022, was remarkably higher (51%, 95%CI: 33–70%). The combined effect of vaccination, immunity from natural infection, change in variant features, and improved patient management massively reduced the probabilities of hospitalization, admission to intensive care, and death given infection, with 20 to 40-fold reductions during the period of dominance of Omicron compared to the initial acute phase.

Estimation of the incubation period and generation time of SARS-CoV-2 Alpha and Delta variants from contact tracing data (preprint)

Background. During 2021, the COVID-19 pandemic was characterized by the emergence of lineages with increased fitness. For most of these variants, quantitative information is scarce on epidemiological quantities such as the incubation period and generation time, which are critical for both public health decisions and scientific research. Method. We analyzed a dataset collected during contact tracing activities in the province of Reggio Emilia, Italy, throughout 2021. We determined the distributions of the incubation period using information on negative PCR tests and the date of last exposure from 282 symptomatic cases. We estimated the distributions of the intrinsic generation time (the time between the infection dates of an infector and its secondary cases under a fully susceptible population) using a Bayesian inference approach applied to 4,435 SARS-CoV-2 cases clustered in 1,430 households where at least one secondary case was recorded. Results. We estimated a mean incubation period of 4.9 days (95% credible intervals, CrI, 4.4-5.4; 95 percentile of the mean distribution: 1-12) for Alpha and 4.5 days (95%CrI 4.0-5.0; 95 percentile: 1-10) for Delta. The intrinsic generation time was estimated to have a mean of 6.0 days (95% CrI 5.6-6.4; 95 percentile: 1-15) for Alpha and of 6.6 days (95%CrI 6.0-7.3; 95 percentile: 1-18) for Delta. The household serial interval was 2.6 days (95%CrI 2.4-2.7) for Alpha and 2.4 days (95%CrI 2.2-2.6) for Delta, and the estimated proportion of pre-symptomatic transmission was 54-55% for both variants. Conclusions. These results indicate limited differences in the incubation period and intrinsic generation time of SARS-CoV-2 variants Alpha and Delta compared to ancestral lineages.

Tracking the progressive spread of the SARS-CoV-2 Omicron variant in Italy, December 2021 - January 2022 (preprint)

The SARS-CoV-2 variant of concern Omicron was first detected in Italy in November 2021. Data from three genomic surveys conducted in Italy between December 2021 and January 2022 suggest that Omicron became dominant in less than one month (prevalence on January 3: 78.6%-83.8%) with a doubling time of 2.7-3.1 days. The mean net reproduction number rose from about 1.15 in absence of Omicron to a peak of 1.83 for symptomatic cases and 1.33 for hospitalized cases, while it remained stable for critical cases.

Investigating the relationship between interventions, contact patterns, and SARS-CoV-2 transmissibility (preprint)

ABSTRACT Background After a rapid upsurge of COVID-19 cases in Italy during the fall of 2020, the government introduced a three-tiered restriction system aimed at increasing physical distancing. The Ministry of Health, after periodic epidemiological risk assessments, assigned a tier to each of the 21 Italian regions and autonomous provinces (AP). It is still unclear to what extent these different measures altered mixing patterns and how quickly the population adapted their social interactions to continuous changes in restrictions. Methods and findings We conducted a survey between July 2020 and March 2021 to monitor changes in social contact patterns among individuals in the metropolitan city of Milan, Italy, which was hardly hit by the second wave of COVID-19 pandemic. The number of contacts during periods characterized by different levels of restrictions was analyzed through negative binomial regression models and age-specific contact matrices were estimated under the different tiers. Relying on the empirically estimated mixing patterns, we quantified relative changes in SARS-CoV-2 transmission potential associated with the different tiers. As tighter restrictions were implemented during the fall of 2020, a progressive reduction in the mean number of contacts recorded by study participants was observed: from 16.4% under mild restrictions (yellow tier), to 45.6% under strong restrictions (red tier). Higher restrictions levels were also found to increase the relative contribution of contacts occurring within the household. The SARS-CoV-2 reproduction number was estimated to decrease by 18.7% (95%CI: 4.6-30.8), 33.4% (95%CI: 22.7-43.2), and 50.2% (95%CI: 40.9-57.7) under the yellow, orange, and red tiers, respectively. Conclusions Our results give an important quantification of the expected contribution of different restriction levels in shaping social contacts and decreasing the transmission potential of SARS-CoV-2. These estimates can find an operational use in anticipating the effect that the implementation of these tiered restriction can have on SARS-CoV-2 reproduction number under an evolving epidemiological situation.

Model-based evaluation of alternative reactive class closure strategies against COVID-19 (preprint)

There are contrasting results concerning the effect of reactive school closure on SARS-CoV-2 transmission. To shed light on this controversy, here we develop a data-driven computational model of SARS-CoV-2 transmission to investigate mechanistically the effect on COVID-19 outbreaks of school closure strategies based on syndromic surveillance and antigen screening of students. We found that by reactively closing classes based on syndromic surveillance, SARS-CoV-2 infections are reduced by no more than 13.1% (95%CI: 8.6%-20.2 %), due to the low probability of timely symptomatic case identification among the young population. We thus investigated an alternative triggering mechanism based on repeated screening of students using antigen tests. Should population-level social distancing measures unrelated to schools enable maintaining the reproduction number (R) at 1.3 or lower, an antigen-based screening strategy is estimated to fully prevent COVID-19 outbreaks in the general population. Depending on the contribution of schools to transmission, this strategy can either prevent COVID-19 outbreaks for R up to 1.9 or to at least greatly reduce outbreak size in very conservative scenarios about school contribution to transmission. Moving forward, the adoption of antigen-based screenings in schools could be instrumental to limit COVID-19 burden while vaccines continue to roll out through 2021, especially in light of possible continued emergence of SARS-CoV-2 variants.

The risk of symptomatic reinfection during the second COVID-19 wave in individuals previously exposed to SARS-CoV-2 (preprint)

To what extent infection with SARS-CoV-2 protects against subsequent reinfection or symptomatic reinfection is still unclear. In this cohort study, we analyzed surveillance records of COVID-19 cases identified between June 2020 and January 2021 in five Italian municipalities, where 77.7% of the entire population was screened for IgG antibodies in May 2020. We compared the risk of observing symptomatic infections in two mutually exclusive groups defined by the initial serological response. We estimated that the cumulative incidence of identified symptomatic infections in the IgG negative and positive cohorts was 2.67% (95%CI: 2.12% - 3.37%) and 0.14% (95%CI: 0.04% - 0.58%), respectively. The adjusted odd ratio of developing symptomatic infection in individuals previously exposed to SARS-CoV-2 was estimated at 0.054 (95%CI: 0.009 - 0.169). Quantifying protective immunity against COVID-19 disease elicited by natural infection with SARS-CoV-2 is essential to inform strategies for controlling the pandemic in the forthcoming months.

Co-circulation of SARS-CoV-2 variants B.1.1.7 and P.1 (preprint)

SARS-CoV-2 variants of concern (B.1.1.7, P.1 and B.1.351) have emerged in different continents of the world. To date, little information is available on their ecological interactions. Based on two genomic surveillance surveys conducted on February 18 and March 18, 2021 across the whole Italian territory and covering over 3,000 clinical samples, we found significant co-circulation of B.1.1.7 and P.1. We showed that B.1.1.7 was already dominant on February 18 in a majority of regions/autonomous provinces (national prevalence 54%) and almost completely replaced historical lineages by March 18 (dominant in all regions/autonomous provinces, national prevalence 86%). At the same time, we found a substantial proportion of cases of the P.1 lineage on February 18, almost exclusively in Central Italy (with an overall prevalence in the macro-area of 18%), which remained at similar values on March 18, suggesting the inability by this lineage to outcompete B.1.1.7. Only 9 cases from variant B.1.351 were identified in the two surveys. At the national level, we estimated a mean relative transmissibility of B.1.1.7 (compared to historical lineages) ranging between 1.55 and 1.57 (with confidence intervals between 1.45 and 1.66). The relative transmissibility of P.1 estimated at the national level varied according to the assumed degree of cross-protection granted by infection with other lineages and ranged from 1.12 (95%CI 1.03-1.23) in the case of complete immune evasion by P.1 to 1.39 (95%CI 1.26-1.56) in the case of complete cross-protection. These observations may have important consequences on the assessment of future pandemic scenarios.

Return to normal: COVID-19 vaccination under mitigation measures (preprint)

Being unable to suppress SARS-CoV-2 transmission, the majority of countries worldwide have resorted to a mitigation approach towards COVID-19, allowing some degree of viral circulation in the population. Here, we investigate the expected outcomes of the interplay between vaccination rollout and adaptive mitigation measures constantly altering the epidemic trajectory and keeping the reproduction number around the unit. Using a novel mathematical modeling framework, we estimate that, for vaccination capacities of at least 4 daily doses administered per 1,000 inhabitants, a complete release of mitigation measures can be expected within 7 to 13 months since the start of vaccination, with a two-year cumulative incidence of deaths between 0.18 and 0.46 per 1,000 population. A heavier burden of deaths and a delayed <> is expected for lower vaccine capacities, if viral transmissibility exceeds by >60% the one estimated at the beginning of the pandemic, or if vaccine protection is short-lived. Failure to prioritize the elderly or a premature release of mitigation measures after vaccination of the most fragile will conspicuously increase the expected mortality. Finally, strategies oriented to prioritize the suppression of SARS-CoV-2 by maintaining strict restrictions will take a similar time as a mitigation approach, possibly resulting in acceptability issues. Persisting unknowns about the evolving epidemiology of SARS-CoV-2 variants and on the effectiveness of available and upcoming vaccines may warrant a future reassessment of these conclusions.

Can a COVID-19 vaccination program guarantee the return to a pre-pandemic lifestyle? (preprint)

COVID-19 vaccination has been initiated in several countries to control SARS-CoV-2 transmission. Whether and when non-pharmaceutical interventions (NPIs) can be lifted as vaccination builds up remains key questions. To address them, we built a data-driven SARS-CoV-2 transmission model for China. We estimated that, to prevent local outbreaks to escalate to major widespread epidemics, stringent NPIs need to remain in place at least one year after the start of vaccination. Should NPIs be capable to keep the reproduction number (Rt) around 1.3, vaccination could reduce up to 99% of COVID-19 burden and bring Rt below the epidemic threshold in 9 months. Maintaining strict NPIs throughout 2021 is of paramount importance to reduce COVID-19 burden while vaccines are distributed, especially in large populations with little natural immunity.

Can a COVID-19 vaccination program guarantee the return to a pre-pandemic lifestyle? (preprint)

COVID-19 vaccination programs have been initiated in several countries to control SARS-CoV-2 transmission and return to a pre-pandemic lifestyle. However, understanding when non-pharmaceutical interventions (NPIs) can be lifted as vaccination builds up and how to update priority groups for vaccination in real-time remain key questions for policy makers. To address these questions, we built a data-driven model of SARS-CoV-2 transmission for China. We estimated that, to prevent local outbreaks to escalate to major widespread epidemics, stringent NPIs need to remain in place at least one year after the start of vaccination. Should NPIs be capable to keep the reproduction number (Rt) around 1.3, a vaccination program could reduce up to 99% of COVID-19 burden and bring Rt below the epidemic threshold in about 9 months. Maintaining strict NPIs throughout 2021 is of paramount importance to reduce COVID-19 burden while vaccines are distributed to the population, especially in large populations with little natural immunity.

Effectiveness of regional restrictions in reducing SARS-CoV-2 transmission during the second wave of COVID-19, Italy. (preprint)

To counter the second COVID-19 wave, the Italian government has adopted a scheme of three sets of restrictions (coded as yellow, orange, and red) imposed on a regional basis. We estimate that milder restrictions in regions at lower risk (yellow) resulted in a transmissibility reduction of about 18%, leading to a reproduction number Rt of about 0.99. Stricter measures (orange and red) led to reductions of 34% and 45% and Rt values of about 0.89 and 0.77 respectively.

Demography, social contact patterns and the COVID-19 burden in different settings of Ethiopia: a modeling study (preprint)

Background COVID-19 spread may have a dramatic impact in countries with vulnerable economies and limited availability of, and access to, healthcare resources and infrastructures. However, in sub-Saharan Africa a low prevalence and mortality have been observed so far. Methods We collected data on individual social contacts in Ethiopia across geographical contexts characterized by heterogeneous population density, work and travel opportunities, and access to primary care. We assessed how socio-demographic factors and observed mixing patterns can influence the COVID-19 disease burden, by simulating SARS-CoV-2 transmission in remote settlements, rural villages, and urban neighborhoods, under the current school closure mandate. Results From national surveillance data, we estimated a net reproduction number of 1.62 (95%CI 1.55-1.70). We found that, at the end of an epidemic mitigated by school closure alone, 10-15% of the overall population would have been symptomatic and 0.3-0.4% of the population would require mechanical ventilation and/or possibly result in a fatal outcome. Higher infection attack rates are expected in more urbanized areas, but the highest incidence of critical disease is expected in remote subsistence farming settlements. Conclusions The relatively low burden of COVID-19 in Ethiopia can be explained by the estimated mixing patterns, underlying demography and the enacted school closures. Socio-demographic factors can also determine marked heterogeneities across different geographical contexts within the same country. Our findings can contribute to understand why sub-Saharan Africa is experiencing a relatively lower attack rate of severe cases compared to high income countries.

Healthcare strain and intensive care during the COVID-19 outbreak in the Lombardy region: a retrospective observational study on 43,538 hospitalized patients (preprint)

Background. During the spring of 2020, the SARS-CoV-2 epidemic has caused significant resource strain in hospitals of Lombardy, Italy, with the demand for intensive care beds for COVID-19 patients exceeding the overall pre-crisis capacity. In this study, we evaluate the effect of healthcare strain on ICU admission and survival. Methods. We used data on 43,538 patients admitted to a hospital in the region between February 20 and July 12, 2020, of which 3,993 (9.2%) were admitted to an ICU. We applied logistic regression to model the probability of being admitted to an ICU and the probability of survival among ICU patients. Negative binomial regressions were used to model the time between hospital and ICU admission and the length of stay in ICU. Results. During the period of highest hospital strain (March 16 - April 22), individuals older than 70 years had a significantly lower probability of being admitted to an ICU and significantly longer times between hospital and ICU admission, indicating elective admission due to constrained resources. Healthcare strain did not have a clear effect on mortality, with the overall proportion of deaths declining from 52.1% (95%CI 49.8-54.5) for ICU patients admitted to the hospital before March 16, to 43.4% (95%CI 41.5-45.6) between March 16 and April 22, to 27.6% (95%CI 20.0-35.2) after April 22. Conclusions. These data demonstrate and quantify the adoption of elective admission to ICUs during the peak phase of the SARS-CoV-2 epidemic in Lombardy. However, we show that for patients admitted to ICUs, clinical outcomes progressively improved despite the saturation of healthcare resources.

Infection fatality ratio of SARS-CoV-2 in Italy (preprint)

We analyzed 5,484 close contacts of COVID-19 cases from Italy, all of them tested for SARS-CoV-2 infection. We found an infection fatality ratio of 2.2% (95%CI 1.69-2.81%) and identified male sex, age >70 years, cardiovascular comorbidities, and infection early in the epidemics as risk factors for death.