ABSTRACT
Background. We characterize the evolution of symptoms in those with selfreported SARS-CoV-2 infections and the likelihood of seeking treatment or medical care during different waves of the pandemic. Methods. The NC-CCRP is a longitudinal observational study of 37,820 participants who completed a daily symptom log from April 2020 through February 2022, during which there were 5,167 self-reported COVID-19 infections. Three variant periods were defined as pre-delta, delta, and omicron, based on the predominant variant in North Carolina. Quasi-Poisson and logistic regression models adjusted for demographics and vaccination were used to assess COVID-19 symptoms and their duration and seeking treatment or hospitalization Results. Cough was the most reported symptom in all waves and increased from 77% pre-delta to 85% during omicron (p=0.001). Sore throat was more common during self-reported infections during omicron (71%), compared with 62% during delta and 54% pre-delta (p< 0.001). The largest change in proportion reporting a symptom was loss of taste or smell which decreased from 55% during pre-delta to 17% during omicron (p< 0.001). Compared with the pre-delta period, delta (incidence risk reduction, IRR 0.86;95% CI 0.79-0.93) and omicron (IRR 0.67;95% CI 0.61-0.73) were associated with lower symptom duration. Participants infected during the delta wave were more likely to seek treatment compared with either pre-delta (odds ratio, OR 1.32 95% CI 1.06-1.64) or omicron (OR 1.42;95% CI 1.21-1.67). Omicron period infections were associated with a lower likelihood of self-reported hospitalization compared with pre-delta (OR 0.26;95% CI 0.10-0.59) or delta (OR 0.26;95% CI 0.11-0.60). Vaccination was associated with a reduced likelihood of hospitalization (OR 0.35;95% CI 0.18-0.70). Proportion and Duration of Symptoms by Variant Wave;Unadjusted by Vaccination Status. Conclusion. Our study indicates evolution in symptom presentation and duration by variant period. The omicron wave was associated with shorter duration and lower severity of illness. Longitudinal tracking of symptomology and severity of a novel pathogen provide insights into the evolution of the pathogen in the community and is vital for public health and clinical response.
ABSTRACT
Background. We characterize the incidence and risk factors of SARS-CoV-2 breakthrough infections in the NC-CCRP. Cumulative Incidence of Breakthrough infections after Self-reported Symptomatic SARS-CoV-2 Test Cumulative incidence curves (1 minus the unadjusted Kaplan-Meier risk), number at risk at each time point for the first self-reported symptomatic positive SARS-CoV-2 test, starting from full vaccination among participants who reported full vaccination. Methods. The NC-CCRP is an observational cohort study assessing COVID-19 symptoms, test results, vaccination status, and risk behavior via daily email or text surveys. Cox models were used to estimate hazard rates. Fixed covariates were age at enrollment, race/ethnicity, sex, county of residence classification, vaccine product, and healthcare worker status. Time varying covariates were vaccination rate in county of residence, mask usage in the week prior, the Delta time frame, the Omicron time frame, and receipt of a vaccine booster. Results. Among 15,808 eligible adult participants, 638 (4.0%) reported a positive SARS-CoV-2 test after vaccination from 01/15/2021 to 01/03/2022. The breakthrough rate increased with time from vaccination (Figure), with a cumulative incidence of 6.95% over 45 weeks of follow-up. Factors associated with a lower risk of breakthrough infection (p< 0.05) included older age (HR 0.7 for participants 45-64 years and 0.41 for those > 65 years compared to those 18-44 years), prior SARS-CoV-2 infection (HR 0.58), higher rates of mask use (HR 0.66), and receipt of a booster vaccination (HR 0.33). Higher rates of breakthrough infection were reported by participants vaccinated with BNT162b2 (HR 1.35) or Ad26.COV2.S (1.74) compared to mRNA-1273, those residing in suburban (HR 1.33) or rural (1.24) counties compared to urban counties, and during circulation of the Delta (3.54) and Omicron (16.68) variants compared to earlier time periods. There was no association of breakthrough infection with sex, race/ethnicity, healthcare worker status, or vaccination rate in the county of residence. Conclusion. In this real-world analysis, risk of breakthrough infections increased with time since vaccination, with some variability among the specific vaccine products. Risk increased dramatically during the Omicron surge. Higher rates among younger individuals may reflect more frequent, or higher risk exposures, including those related to childcare. Significantly lower rates of breakthrough associated with mask wearing and receipt of a booster highlight specific measures that individuals can take to minimize the risk for COVID-19.
ABSTRACT
Background. There remain important gaps in knowledge concerning the effects of SARS-CoV-2 infection or vaccination on the human blood proteome. Methods. The CCRP is a longitudinal surveillance study with information on SARS-CoV-2 infections, vaccinations and associated humoral immune responses in over 37,000 individuals. We selected a sample of blood spots cards (n=510) from serum antibody studies obtained between October 2020 and April 2021 for mass spectrometry proteomics analysis covering 540 unique plasma proteins. We analyzed the quantified protein differences based on dried blood samples obtained before and after infection or vaccination among previously non-infected individuals (immune naive) after adjustment for batch effects, age, sex, or prior diagnosis of cancer, cardiovascular or autoimmune disease, or diabetes. The majority of infected individuals were minimally symptomatic. Differentially expressed proteins were considered significant with an FDR adjusted p-value of < 0.05 and log2 fold change (L2FC) >0.2. Results. We found 11 and 12 proteins differentially expressed proteins in the naive/infected and naive/vaccinated people respectively, of which 10 were shared. Hepatocyte growth factor receptor (HGF) was upregulated (L2FC 0.24;p < 0.001) only in those who were infected while fibrillarin (L2FC -0.24;p< 0.001) and lambdacrystallin homolog (L2FC -0.29, p < 0.001) were downregulated only in the vaccinated samples (Fig 1). The remaining DE protein were associated with a wide array of functions including metabolic, cytostructural, extracellular matrix and DNA regulatory processes. Conclusion. We found changes in the proteome both from infection and vaccination. HGF, elevated in the infected, has been associated with endothelial inflammation, upregulation of pro-inflammatory cytokines to reduce lung fibrosis and is known to promote tissue repair. Fibrillarin, downregulated in the vaccinated, has been associated with higher rates of bacterial and viral clearance, inflammation reduction, and increased cell survival. These findings suggest detectable complex inflammation from mild to moderate infections. Further investigation is required to understand the mechanism of action and clinical implication of these findings.
ABSTRACT
Background. The COVID-19 Community Research Partnership (CCRP) is a large multicenter healthcare system-based study of the COVID-19 pandemic, including factors impacting risk of infection and hospitalization. The CCRP includes a subset of immunocompromised (IC) participants with varying vaccination status over time. Methods. We conducted an observational cohort study of 2,515 IC and 41,941 non-IC CCRP participants who contributed electronic health record data and daily electronic surveys to self-report COVID-19 symptoms, test results, and vaccinations from April 2020 to March 2022. The IC population included those with stem cell transplant, HIV, cancer, autoimmune disease, or solid organ transplant. The latter 3 must have also had an active systemic therapy to meet the IC condition (e.g. chemotherapy, immune modulator, steroid). Logistic regression was used to investigate risk of COVID-19 and hospitalization among IC participants and according to vaccine status within viral variant time periods (pre-delta, delta, omicron). Results. IC conditions included cancer (51%), autoimmune (41%), solid organ/ stem cell transplant (9%), and HIV (7%). The IC group was older and had more comorbidities. 95% of vaccine recipients received an mRNA vaccine. More vaccine breakthrough infections occurred in the IC group than non-IC group (36.1% vs 29.5%, p< 0.001). IC participants were less likely to remain COVID-19 free over time if vaccinated but not boosted (Fig 1A). However, after receiving a booster there was no difference in COVID-19 cases between the groups (Fig 1B). IC participants were more likely to be hospitalized with COVID-19 (OR 2.85;95% CI 1.69-4.76), but vaccination reduced risk for hospitalization (OR 0.26;95% CI 0.08-0.8). Receipt of a booster dose reduced risk of COVID-19 among IC participants during the delta wave (IRR 0.52;95% CI 0.28-0.94) but not during omicron. However, during omicron risk of hospitalization in the IC group was reduced by a booster dose (OR 0.13;95% CI 0.02-0.72). Conclusion. IC individuals were at increased risk for COVID-19 hospitalizations and breakthrough infections. After receiving a booster, IC participants were conferred the same level of protection from infection as their non-IC counterparts, highlighting the importance of boosters for these individuals. (Figure Presented).
ABSTRACT
Background. Well-regulated clinical trials have shown authorized COVID-19 vaccines to be immunogenic and highly efficacious. Information about antibody responses after vaccination in real-world settings is needed. Methods. We evaluated seroconversion rates in adults reporting ≥ 1 dose of an authorized COVID-19 vaccine in a U.S. multistate longitudinal cohort study, the COVID-19 Community Research Partnership. Participants were recruited through 12 participating healthcare systems and community outreach. Participants had periodic home-based serologic testing using either a SARSCoV-2 nucleocapsid and spike IgM/IgG lateral flow assay (63% of participants) or a SARS-CoV-2 spike IgG enzyme-linked immunosorbent assay (37% of participants). The timing and number of tests before and after vaccination varied based on participant time in study. Participants were included if they were seronegative on the last test before and had >1 test result after vaccination (some had previously been seropositive, but seroreverted). A weighted Cox regression model with right censoring was used to obtain adjusted hazard ratios for sex, age, race/ethnicity, and prior seropositivity. Time-to-event (seroconversion) was defined as time to first positive test > 4 days after vaccination;participants were censored at the date of their last available test result. Results. 13,459 participants were included and 11,722 seroconverted (Table). Median time in study was 272 days (range 31-395). Median follow-up time from vaccine to last available test was 56 days (range 1-147). Participants had a median of 3 tests (range 1-12) before and 2 tests (range 1-8) after vaccination. Based on the Kaplan-Meier method, median time to seroconversion after first COVID-19 vaccination was 35 days (interquartile range: 25-45). Likelihood of seroconversion decreased with older age (Table). Female participants, non-Hispanic Black participants, and participants who were previously seropositive were more likely to seroconvert (Table). Conclusion. All subgroups had high rates of seroconversion, with some small differences in likelihood of seroconversion between subgroups. These data demonstrate the excellent immunogenicity of COVID-19 vaccines in real-world settings in the US.
ABSTRACT
Background. Public health officials are concerned that adults may refuse to be vaccinated with an approved COVID-19 vaccine thereby limiting the community health benefit. Here, we studied the self-reported intent to be vaccinated of persons in North Carolina (NC) and then measured whether they did or did not get vaccinated. Methods. The Community COVID-19 Research Partnership (CCRP) is a large prospective study exploring COVID-19 epidemiology and sequelae in participants of several mid-Atlantic and Southern States. All participants complete an online daily survey where they are asked questions about COVID-like symptoms, infections, and their vaccination status. In addition to the daily survey, in December 2020, we implemented a short online cross-sectional survey questioning NC participants on whether they intended to be vaccinated. After completing the cross-sectional survey, we used daily survey data through 15 May 2021 to see if participants reported receiving vaccine. Unvaccinated participants who did not complete the daily survey 30 days or more prior to 15 May 2021 were excluded. Results. 18,874 participants completed the cross-sectional survey and reported vaccination status. Of these participants, 90% were white, 68% were female, 26% were healthcare workers, and 2% self-reported COVID-19 diagnosis The median age was 54 years (IQR: 41 - 65). 79%, 13%, 9%, and 2% answered yes, unsure, no, and prefer not to answer, respectively, about intention to be vaccinated (Table). 99% of the participants who intended to receive the COVID-19 vaccine reporting being vaccinated. Those who were unsure or intended not to get vaccinated had vaccination rates of 80% and 53%, respectively. 78% of the participants who preferred not to answer were vaccinated. Conclusion. More than three-quarters of NC participants intended to get vaccinated and by mid-May 2021, the vast majority had received at least one dose. Similarly, those who were unsure or preferred not to say were mostly vaccinated. Even among those who reported they would not get vaccine in January, more than half had received vaccine by May. The nature of our sample makes it difficult to generalize results to the population of NC;nevertheless, further investigation as to the causes of the shift in attitudes is warranted.