Risk-based prediction for optimal timing of booster vaccination for COVID-19 to prevent severe disease (preprint)

While waning protection from vaccination and natural infection against SARS-CoV-2 infection is well-documented, recent analyses have also found waning of protection against severe COVID-19. This highlights a broader need to understand the optimal timing of COVID-19 booster vaccines specific to an individual to mitigate the risk of severe COVID-19, while accounting for waning of protection and differential risk by age group and immune status. Here we show that more frequent COVID-19 booster vaccination (every 6-12 months) in older age groups and the immunocompromised population would effectively mitigate the burden of severe COVID-19, while frequent boosters in the younger population may only provide modest benefit. Analyzing United States COVID-19 surveillance and seroprevalence data in a microsimulation model, we estimated that in persons 75+ years, annual and semiannual bivalent boosters would reduce annual absolute risk of severe COVID-19 by 311 (277-369) and 578 (494-671) cases, respectively, compared to a one-time bivalent booster dose. In contrast, for persons 18-49 years, the model estimated that annual and semiannual bivalent boosters would reduce annual absolute risk of severe COVID-19 by 20 (13-26) and 37 (24-50) cases per 100,000 persons, respectively, compared to a one-time bivalent booster dose. Persons with prior infection had a much lower benefit of more frequent boosting, while immunocompromised persons had larger benefit. This study underscores the benefit of customizing timing of COVID-19 booster vaccines based on individual risk.

Infectious viral shedding of SARS-CoV-2 Delta following vaccination: a longitudinal cohort study (preprint)

The impact of vaccination on SARS-CoV-2 infectiousness is not well understood. We compared longitudinal viral shedding dynamics in unvaccinated and fully vaccinated adults. SARS-CoV-2-infected adults were enrolled within 5 days of symptom onset and nasal specimens were self-collected daily for two weeks and intermittently for an additional two weeks. SARS-CoV-2 RNA load and infectious virus were analyzed relative to symptom onset stratified by vaccination status. We tested 1080 nasal specimens from 52 unvaccinated adults enrolled in the pre-Delta period and 32 fully vaccinated adults with predominantly Delta infections. While we observed no differences by vaccination status in maximum RNA levels, maximum infectious titers and the median duration of viral RNA shedding, the rate of decay from the maximum RNA load was faster among vaccinated; maximum infectious titers and maximum RNA levels were highly correlated. Furthermore, amongst participants with infectious virus, median duration of infectious virus detection was reduced from 7.5 days (IQR: 6.0-9.0) in unvaccinated participants to 6 days (IQR: 5.0-8.0) in those vaccinated (P=0.02). Accordingly, the odds of shedding infectious virus from days 6 to 12 post-onset were lower among vaccinated participants than unvaccinated participants (OR 0.42 95% CI 0.19-0.89). These results indicate that vaccination had reduced the probability of shedding infectious virus after 5 days from symptom onset.

Estimation of COVID-19 cases prevented by vaccination in California (preprint)

Importance: Despite widespread vaccination against COVID-19 in the United States, there are limited empirical data quantifying the public health impact in the population. Objective: To estimate the number of cases of COVID-19 averted due to COVID-19 vaccination Design, Setting, and Participants: The California Department of Public Health (CDPH) provided person-level data on COVID-19 cases and COVID-19 vaccine administration. To estimate the number of COVID-19 cases that would have occurred in the vaccine era in absence of vaccination, we applied a statistical model that estimated the relationship of COVID-19 cases in the pre-vaccine era between the unvaccinated age group (<12 years) and vaccine-eligible groups ([≥]12 years) to COVID-19 case data after the start of vaccination. The primary study outcome was the difference between predicted number of COVID-19 cases in absence of vaccination and observed COVID-19 cases with vaccination. As a sensitivity analysis, we developed a second independent model that estimated the number of vaccine-averted COVID-19 cases by applying published data on vaccine effectiveness to data on COVID-19 vaccine administration and estimated risk of COVID-19 over time. Intervention: COVID-19 vaccination Main Outcomes and Measures: COVID-19 cases Results: There were 4,585,248 confirmed COVID-19 cases in California from January 1, 2020 to October 16, 2021, during which 27,164,680 vaccine-eligible individuals [≥]12 years were reported to have received at least 1 dose of a COVID-19 vaccine in the vaccine era (79.5% of the eligible population). We estimated that 1,523,500 [95% prediction interval (976,800-2,230,800)] COVID-19 cases were averted and there was a 34% [95% prediction interval (25-43)] reduction in cases due to vaccination in the primary model. Approximately 66% of total cases averted occurred after the delta variant became the dominant strain of SARS-CoV-2 circulating in California. Our alternative model identified comparable findings. Conclusions and Relevance: This study provides robust evidence on the public health impact of COVID-19 vaccination in the United States and further supports the urgency for continued vaccination.

COVIDNearTerm: A Simple Method to Forecast COVID-19 Hospitalizations (preprint)

1COVID-19 has caused tremendous death and suffering since it first emerged in 2019. In response, models were developed to help predict the course of various disease metrics, and these models have been relied upon to help guide public health policy. Here we present a method called COVIDNearTerm to "forecast" hospitalizations in the short term, two to four weeks from the time of prediction. COVIDNearTerm is based on an autoregressive model and utilizes a parametric bootstrap approach to make predictions. We evaluated COVIDNearTerm on San Francisco Bay Area hospitalizations and compared it to models from the California COVID Assessment Tool (CalCAT). We found that that COVIDNearTerm pre-dictions were more accurate than the CalCAT ensemble predictions for all comparisons and any CalCAT component for a majority of comparisons. For instance, at the county level our 14-day hospitalization median absolute percentage errors ranged from 16% to 36%. For those same comparisons the CalCAT ensemble errors were between 30% and 59%. COVIDNearT-erm is also easier to use than some other methods. It requires only previous hospitalization data and there is an open source R package that implements the algorithm.

Race-ethnicity and Perceptional Determinants of COVID-19 Vaccination Intentions: A Cross-sectional Study Among Health Workers and the General Population in the San Francisco Bay Area (preprint)

Importance: Surveys in the US have found that Black and Latinx individuals have more reservations than their white counterparts about COVID-19 vaccination. However, little is known about the degree to which racial-ethnic differences in COVID-19 vaccination intentions are explained by differences in beliefs or perceptions about COVID-19 vaccines. Objective: To compare intention to receive COVID-19 vaccination by race-ethnicity, to identify perceptional factors that may mediate the association between race-ethnicity and intention to receive the vaccine, and to identify the demographic and perceptional factors most strongly predictive of intention to receive a vaccine. Design: Cross-sectional survey conducted from November, 2020 to January, 2021, nested within two longitudinal cohort studies of prevalence and incidence of SARS CoV-2 among the general population and healthcare workers. Setting: Six San Francisco Bay Area counties. Study Cohort: 3,161 participants in the Track COVID cohort (a population-based sample of adults) and 1,803 participants in the CHART Study cohort (a cohort of employees at three large medical centers). Results: Rates of high vaccine willingness were significantly lower among Black (45.3%), Latinx (62.5%), Asian (65%), multi-racial (67.2%), and other race (61.0%) respondents than among white respondents (77.6%). Black, Latinx, and Asian respondents were significantly more likely than white respondents to endorse reasons to not get vaccinated, especially lack of trust. Participants' motivations and concerns about COVID-19 vaccination only partially explained racial-ethnic differences in vaccination willingness. Being a health worker in the CHART cohort and concern about a rushed government vaccine approval process were the two most important factors predicting vaccination intention. Conclusions and Relevance: Special efforts are required to reach historically marginalized racial-ethnic communities to support informed decision-making about COVID-19 vaccination. These campaigns must acknowledge the history of racism in biomedical research and health care delivery that has degraded the trustworthiness of health and medical science institutions among non-white population and may continue to undermine confidence in COVID-19 vaccines.

Comparison of COVID-19 vaccine prioritization strategies in the United States (preprint)

A critical question in the COVID-19 pandemic is how to optimally allocate the first available vaccinations to maximize health impact. We used a static simulation model with detailed demographic and risk factor stratification to compare the impact of different vaccine prioritization strategies in the United States on key health outcomes, using California as a case example. We calibrated the model to demographic and location data on 28,175 COVID-19 deaths in California up to December 30, 2020, and incorporated variation in risk by occupation and comorbidity status using published estimates. We predicted the proportion of COVID-19 clinical cases, deaths and disability-adjusted life years (DALYs) averted over 6 months relative to a scenario of no vaccination for five vaccination strategies that prioritized vaccination by a single risk factor: random allocation; targeting special populations (e.g. incarcerated individuals); targeting older individuals; targeting essential workers; and targeting individuals with comorbidities. Targeting older individuals averted the highest proportion of DALYs (40% for 5 million individuals vaccinated) and deaths (65%) but the lowest proportion of cases (12%). Targeting essential workers averted the lowest proportion of DALYs (25%) and deaths (33%). Allocating vaccinations simultaneously by age and location or by age, sex, race/ethnicity, location, occupation, and comorbidity status averted a significantly higher proportion of DALYs (48% and 56%) than any strategy prioritizing by a single risk factor. Our results corroborate findings of other studies that age targeting is the best single-risk-factor prioritization strategy for averting DALYs, and suggest that targeting by multiple risk factors would provide additional benefit.

Routine asymptomatic testing strategies for airline travel during the COVID-19 pandemic: a simulation analysis (preprint)

BackgroundAirline travel has been significantly reduced during the COVID-19 pandemic due to concern for individual risk of SARS-CoV-2 infection and population-level transmission risk from importation. Routine viral testing strategies for COVID-19 may facilitate safe airline travel through reduction of individual and/or population-level risk, although the effectiveness and optimal design of these "test-and-travel" strategies remain unclear. MethodsWe developed a microsimulation of SARS-CoV-2 transmission in a cohort of airline travelers to evaluate the effectiveness of various testing strategies to reduce individual risk of infection and population-level risk of transmission. We evaluated five testing strategies in asymptomatic passengers: i) anterior nasal polymerase chain reaction (PCR) within 3 days of departure; ii) PCR within 3 days of departure and PCR 5 days after arrival; iii) rapid antigen test on the day of travel (assuming 90% of the sensitivity of PCR during active infection); iv) rapid antigen test on the day of travel and PCR 5 days after arrival; and v) PCR within 3 days of arrival alone. The travel period was defined as three days prior to the day of travel and two weeks following the day of travel, and we assumed passengers followed guidance on mask wearing during this period. The primary study outcome was cumulative number of infectious days in the cohort over the travel period (population-level transmission risk); the secondary outcome was the proportion of infectious persons detected on the day of travel (individual-level risk of infection). Sensitivity analyses were conducted. FindingsAssuming a community SARS-CoV-2 incidence of 50 daily infections, we estimated that in a cohort of 100,000 airline travelers followed over the travel period, there would be a total of 2,796 (95% UI: 2,031, 4,336) infectious days with 229 (95% UI: 170, 336) actively infectious passengers on the day of travel. The pre-travel PCR test (within 3 days prior to departure) reduced the number of infectious days by 35% (95% UI: 27, 42) and identified 88% (95% UI: 76, 94) of the actively infectious travelers on the day of flight; the addition of PCR 5 days after arrival reduced the number of infectious days by 79% (95% UI: 71, 84). The rapid antigen test on the day of travel reduced the number of infectious days by 32% (95% UI: 25, 39) and identified 87% (95% UI: 81, 92) of the actively infectious travelers; the addition of PCR 5 days after arrival reduced the number of infectious days by 70% (95% UI: 65, 75). The post-travel PCR test alone (within 3 days of landing) reduced the number of infectious days by 42% (95% UI: 31, 51). The ratio of true positives to false positives varied with the incidence of infection. The overall study conclusions were robust in sensitivity analysis. InterpretationRoutine asymptomatic testing for COVID-19 prior to travel can be an effective strategy to reduce individual risk of COVID-19 infection during travel, although post-travel testing with abbreviated quarantine is likely needed to reduce population-level transmission due to importation of infection when traveling from a high to low incidence setting.

Estimation of effects of contact tracing and mask adoption on COVID-19 transmission in San Francisco: a modeling study (preprint)

The current COVID-19 pandemic has spurred concern about what interventions may be effective at reducing transmission. The city and county of San Francisco imposed a shelter-in-place order in March 2020, followed by use of a contact tracing program and a policy requiring use of cloth face masks. We used statistical estimation and simulation to estimate the effectiveness of these interventions in San Francisco. We estimated that self-isolation and other practices beginning at the time of San Francisco's shelter-in-place order reduced the effective reproduction number of COVID-19 by 35.4% (95% CI, -20.1%--81.4%). We estimated the effect of contact tracing on the effective reproduction number to be a reduction of approximately 44% times the fraction of cases that are detected, which may be modest if the detection rate is low. We estimated the impact of cloth mask adoption on reproduction number to be approximately 8.6%, and note that the benefit of mask adoption may be substantially greater for essential workers and other vulnerable populations, residents return to circulating outside the home more often. We estimated the effect of those interventions on incidence by simulating counterfactual scenarios in which contact tracing was not adopted, cloth masks were not adopted, and neither contact tracing nor cloth masks was adopted, and found increases in case counts that were modest, but relatively larger than the effects on reproduction numbers. These estimates and model results suggest that testing coverage and timing of testing and contact tracing may be important, and that modest effects on reproduction numbers can nonetheless cause substantial effects on case counts over time.