Looking under the lamp-post: quantifying the performance of contact tracing in the United States during the SARS-CoV-2 pandemic (preprint)

Contact tracing forms a crucial part of the public-health toolbox in mitigating and understanding emergent pathogens and nascent disease outbreaks. Contact tracing in the United States was conducted during the pre-Omicron phase of the ongoing COVID-19 pandemic. This tracing relied on voluntary reporting and responses, often using rapid antigen tests (with a high false negative rate) due to lack of accessibility to PCR tests. These limitations, combined with SARS-CoV-2’s propensity for asymptomatic transmission, raise the question “how reliable was contact tracing for COVID-19 in the United States”? We answered this question using a Markov model to examine the efficiency with which transmission could be detected based on the design and response rates of contact tracing studies in the United States. Our results suggest that contact tracing protocols in the U.S. are unlikely to have identified more than 1.65% (95% uncertainty interval: 1.62%-1.68%) of transmission events with PCR testing and 0.88% (95% uncertainty interval 0.86%-0.89%) with rapid antigen testing. When considering an optimal scenario, based on compliance rates in East Asia with PCR testing, this increases to 62.7% (95% uncertainty interval: 62.6%-62.8%). These findings highlight the limitations in interpretability for studies of SARS-CoV-2 disease spread based on U.S. contact tracing and underscore the vulnerability of the population to future disease outbreaks, for SARS-CoV-2 and other pathogens.

Looking under the lamp-post: quantifying the performance of contact tracing in the United States during the SARS-CoV-2 pandemic (preprint)

Contact tracing forms a crucial part of the public-health toolbox in mitigating and understanding emergent pathogens and nascent disease outbreaks. Contact tracing in the United States was conducted during the pre-Omicron phase of the ongoing COVID-19 pandemic. This tracing relied on voluntary reporting and responses, often using rapid antigen tests (with a high false negative rate) due to lack of accessibility to PCR tests. These limitations, combined with SARS-CoV-2s propensity for asymptomatic transmission, raise the question how reliable was contact tracing for COVID-19 in the United States? We answered this question using a Markov model to examine the efficiency with which transmission could be detected based on the design and response rates of contact tracing studies in the United States. Our results suggest that contact tracing protocols in the U.S. are unlikely to have identified more than 1.65% (95% uncertainty interval: 1.62%-1.68%) of transmission events with PCR testing and 0.88% (95% uncertainty interval 0.86%-0.89%) with rapid antigen testing. When considering an optimal scenario, based on compliance rates in East Asia with PCR testing, this increases to 62.7% (95% uncertainty interval: 62.6%-62.8%). These findings highlight the limitations in interpretability for studies of SARS-CoV-2 disease spread based on U.S. contact tracing and underscore the vulnerability of the population to future disease outbreaks, for SARS-CoV-2 and other pathogens.

Variations in COVID-19 impacts by social vulnerability in Philadelphia, June 2020-December 2022 (preprint)

Introduction: The study objective was to elucidate the relationship between social vulnerability and COVID-19 impacts in Philadelphia between June 2020 and December 2022. Methods: Using publicly available COVID-19 case, test, hospitalization, and mortality data for Philadelphia (June 7, 2020-December 31, 2022) and area-level social vulnerability data, we compared the incidence, test positivity, hospitalization, and mortality rates in high and low vulnerability neighborhoods of Philadelphia, characterized as scoring above or below the national median score on the social vulnerability index. We used linear mixed effects models to test the association between social vulnerability and COVID-19 incidence, test positivity, hospitalization, and mortality rates, adjusting for time and age distribution. Results: 90.4% of Philadelphians (n = 1,430,153) live in neighborhoods classified as socially vulnerable, based on scoring above the national median score on the social vulnerability index. COVID-19 incidence, hospitalization, and mortality rates were significantly elevated in the more vulnerable communities, with p < 0.05, p < 0.005, and p < 0.001, respectively. The relative risks of COVID-19-related incidence, hospitalization, and death, comparing the more vulnerable neighborhoods to the less vulnerable neighborhoods, were 1.11 (95%CI: 1.10-1.12), 2.07 (95%CI: 1.93-2.20), and 2.06 (95%CI: 1.78-2.38), respectively. Thus, between June 7, 2020 and December 31, 2022, 32,573 COVID-19 cases, 9,409 hospitalizations, and 1,967 deaths would have been avoided in Philadelphias more vulnerable communities had they experienced the same rates of incidence, hospitalization, and death as the less vulnerable Philadelphia communities. Conclusions: These results highlight the disparate morbidity and mortality experienced by people living in more vulnerable neighborhoods in a large US city. Importantly, our findings illustrate the importance of designing public health policies and interventions with an equity-driven approach, with greater resources and more intensive prevention strategies applied in socially vulnerable communities.

Impact of Lifting School Masking Requirements on Incidence of COVID-19 among Staff and Students in Greater-Boston Area School Districts: A Difference-in-Differences Analysis (preprint)

Background. In February 2022, following the rescinding of a Massachusetts statewide school masking mandate, only two cities (Boston and neighboring Chelsea) out of 79 school districts in the greater-Boston area, maintained masking requirements in K-12 schools. This provided an opportunity to examine the impact of removing masking on COVID-19 case rates among students and staff in the public-school setting. Methods. We used difference-in-differences for staggered policy adoption to compare incidence of COVID-19 cases among students and staff in greater-Boston area school districts that lifted masking requirements to those that had not yet lifted masking requirements during the 2021-2022 school year. Results. Before the statewide school masking policy was lifted, there was no statistically significant difference in case rate trajectories between school districts. However, weekly and cumulative case rates were significantly higher in students and staff in school districts that removed masking requirements, compared to districts that had not yet lifted requirements. We estimate that lifting of school masking requirements was associated with an additional 44.9 (95% CI: 32.6, 57.1) COVID-19 cases per 1,000 students and staff over the 15 weeks since the lifting of the statewide school masking requirement, representing nearly 30% of all cases observed in schools during that time. School districts that sustained masking requirements for longer periods tended to have older school buildings in poorer condition, more crowded classrooms, higher proportion of low income and English learning students and students with disabilities, and a higher proportion of Black and Latinx students and staff. Conclusions. Masking is a relatively low-cost but effective intervention that can protect students and staff from substantial illness and loss of in-person days in school. Despite compelling evidence that masking significantly reduces the spread of SARS-CoV-2, political will and public adherence to masking has waned. Our study confirms that universal masking requirements can benefit all students and staff, and therefore represents an important strategy to mitigate the impacts of structural racism, ensure health equity, and to avoid potential deepening of educational inequities.

Systematic review of the association between ABO blood type and COVID-19 incidence and mortality (preprint)

A large proportion of COVID-19 research has been focused on identifying markers of high-risk individuals. However, this research often fails to consider basic epidemiologic concepts to prevent bias in the design, selection, and analysis of observational data. One suspected marker of risk that has been repeatedly assessed is ABO blood type. Given the ease of measuring this biomarker, it is an appealing target for identifying high-risk individuals. However, this same ease of measurement makes associational research on ABO blood type and COVID prone to a range of common epidemiologic errors. We conducted a systematic review of studies assessing correlations between ABO blood type and COVID incidence, hospitalization, and mortality to determine the quality of evidence these studies provide and whether the overall evidence suggests ABO blood type could provide a useful indicator of COVID risk. We conclude that most existing studies are low quality and suffer from major methodological flaws. The few higher-quality studies which do exist find no association between ABO blood type and COVID outcomes. We conclude that there is no evidence to support the use of ABO blood type as a marker for COVID risk or severity. Key PointsO_LIThere is no sufficient evidence to conclude a biological relationship between ABO blood types and COVID-19 infection or severity. C_LIO_LIBiases of existing research could be avoided by careful study design. C_LI