SARS-CoV-2 reinfection rate before and after VOC Omicron emergence: a retrospective study in Brazil

Abstract

SARS-COV-2 reinfection has been reported worldwide, although its rate remains unclear. VOC Omicron’s emergence and its sub-variants led to an unprecedented number of COVID-19 cases in several countries, raising concerns regarding reinfection rates. 324,979 RT-qPCR-confirmed positive cases (72.57% from Minas Gerais State) diagnosed between April 1, 2020, and August 31, 2022, at the Hermes Pardini, Grupo Fleury (Brazil) were used to estimate the reinfection rate. Instances of reinfection were characterized by two positive tests occurring with a minimum interval of 60 days. We identified 11,669 cases of reinfection. The states of Minas Gerais, São Paulo, Rio de Janeiro and Goiás represented almost 41% of the reinfections. Up until epidemiological week 46 of 2020, only 14 cases of reinfection were recorded. The majority of reinfections, totalling 6,316 cases, were detected during the circulation period of the Omicron and its sublineages BA.1 and BA.2. Another 4,273 reinfections occurred during the circulation period of sublineages BA.4 and BA.5, revealing two distinct groups of observations. The first group comprised cases of reinfection with a shorter time interval (two infections within a period of up to 200 days), while the second group was associated with a longer time interval (two infections within a period of more than 500 days). The reinfection rate during this period was nearly 8%, which is six times higher than the rate observed at the beginning of the study. In conclusion, our study underscores the dynamic nature of SARS-CoV-2 reinfections and their correlation with emerging variants such as Omicron.

Introduction

The possibility of SARS-CoV-2 reinfections has been under consideration since the inception of the COVID-19 pandemic. Initially regarded as sporadically reported events, it wasn’t until December 2021 that they gained increased attention in Brazil [1, 2]. Although an unprecedented effort and quick advance were observed in several COVID-19 research fields, many reinfection aspects remain not completely understood.

One main issue is how to identify reinfection. Viral genome sequencing of longitudinal samples is the gold standard, making it possible to exclude infection persistence or confirm short-term reinfection, which are other understudied topics [3, 4]. Recently, one of the largest studies to date using genomic data was published [5]. It assessed around 45,000 reinfections in the UK and identified viral strain, age, vaccination status, and time since the last infection as factors influencing the risk of reinfection. However, because of the expense and intricacy associated with longitudinal follow-ups, studies using this approach tended to have relatively small to modest sample sizes, particularly in Brazil [6].

Another method to analyze reinfections involves utilizing real-world evidence from testing laboratories and establishing a time threshold to define reinfection events. A meta-analysis, which employed a 90-day threshold, discovered that protection against the BA.1 variant was significantly lower [7]. This analysis pooled data from 65 studies conducted across 19 countries, none from Brazil. Additionally, the study was unable to estimate the effect of the VOC Gamma, a previously circulating variant that contributed significantly to COVID-19 deaths in Brazil [8].

Understanding and monitoring reinfection rates are essential for shaping public health strategies, guiding vaccination efforts, and formulating effective, long-term control measures to address the enduring COVID-19 global impact. In this study, our objective was to estimate the reinfection rate based on real-world evidence from Brazil up to the emergence of the VOC Omicron. Similar to trends observed globally, the introduction of the VOC Omicron resulted in the highest number of cases reported in Brazil to date. It is noteworthy that, despite the higher vaccination coverage compared to previous VOCs circulating in the country, the surge in cases occurred.

Materials and methods

The data for this study were sourced from Hermes Pardini, Grupo Fleury, a private laboratory with a presence in all Brazilian states. We utilized data from samples that tested positive for SARS-CoV-2 through RT-qPCR from April 1, 2020 (Epidemiological Week - EW − 14/20) to August 31, 2022 (EW 35/22). Throughout this timeframe, were conducted over 1 million tests. Reinfection cases were identified when an individual (using the same identification number) tested positive for COVID-19 two or more times, with at least a sixty-day gap between the first infection and subsequent positive tests. This definition was chosen to exclude prolonged COVID-19 infections and aligns with the criteria set by the European Centre for Disease Prevention and Control [9, 10]. Data related to the COVID-19 epidemiological scenario in the country was retrieved from the public platform https://covid.saude.gov.br/.

Results

According to the Brazilian Ministry of Health, there were 19,113,368 registered COVID-19 cases from the onset of the pandemic (EW14–2020) until November 2022 (EW35–2022). During the same period, we accessed 1,201,947 COVID-19 tests, with 324,984 tests yielding positive results (positivity rate = 27.04%). These positive tests were reported across 672 municipalities, representing 12% of all cities in Brazil. The state of Minas Gerais (MG) had the highest number of samples during our study, with 235,839 positive tests. Notably, the majority of samples originated from the Southeast region, totalling 310,625 (95.58%) positive tests spread across 441 cities.

We can categorize the epidemic scenario in Brazil into distinct phases (Fig. 1). The first phase is associated with the early stages of the pandemic and the circulation of the B.1.1.28 and B.1.1.33 strains. The second phase coincides with the emergence of the VOC Gamma and Delta. During this period, COVID-19 vaccination efforts commenced and, by the end of EW51–2021, had reached 65% of the population, with at least one vaccine dose applied. The third phase corresponds to the emergence of the VOC Omicron, marked by the highest number of diagnosed positive cases in the country. At this point, 77% of the population had received two vaccine doses. The final phase is linked to the emergence of sub-variants of Omicron, such as BA.4 and BA.5 (Fig. 1A). During this period, according to the data obtained by the Brazilian Ministry of Health, 78% of the population had received at least two vaccine doses. Based on the trends observed, our data aligns with the information provided by the Brazilian Health Minister. Consequently, we proceeded with the analysis using data from Hermes Pardini, Grupo Fleury to estimate the reinfection rate in Brazil.

Fig. 1.

Overview of the SARS-CoV-2 scenario in Brazil across epidemiological weeks. (A) Number of SARS-CoV-2 confirmed infections according to the Brazilian Ministry of Health (blue line) and diagnosed by the Hermes Pardini, Grupo Fleury (red line) in all Brazilian territory. (B) Absolute (blue line) and relative (red line) frequencies of the first infections based on samples diagnosed by the Hermes Pardini, Grupo Fleury. (C) Absolute (blue line) and relative (red line) of reinfection cases identified in the samples diagnosed by the Hermes Pardini, Grupo Fleury. (D) The period between infections, categorized into four pivotal moments of the COVID-19 pandemic in Brazil: B.1.28 and B.1.33 lineages circulation (blue violin); VOCs Gamma and Delta (red violin); the emergence of VOC Omicron and sublineages BA.1 and BA.2 (green violin); and BA.4 and BA.5 circulation (light blue violin). An increased time between reinfections was observed, although a bimodal distribution was found in the most recent period, indicating both shorter (less than 200 days) and longer reinfections (greater than 500 days)

In total, we identified 11,669 reinfection cases, distributed across 96 municipalities (1.72% of all Brazilian cities). The initial two periods (EW 10–2020 to 51–2021) were characterized by a surge in first infections (see Fig. 1B). Remarkably, during this phase, the occurrence of reinfections in the studied population was minimal. However, in the subsequent two phases (EW 52/2021 to 20/2022 and EW 21–35/2022), there is a noticeable uptick in reinfection frequency, reaching almost an 8% reinfection rate in the population (Fig. 1C).

Notably, following the VOC Omicron onset (EW 52–2021), the number of reinfections surged almost sixfold compared to the preceding period (EWs 47/20–51/21). The final period reveals two distinct groups. The first group (lower portion of Fig. 1D), pertains to reinfection cases occurring within a shorter time interval (two infections in a period of up to 200 days), while the second group (upper portion of the chart), is associated with a longer time interval (more than 500 days) (Fig. 1D).

We explored whether the observed trends were reproducible in a state level using four states with the most samples in our dataset (at least 10,000 diagnosed cases). The states of Minas Gerais (MG), São Paulo (SP), Rio de Janeiro (RJ) in the southeast region, and Goiás (GO) in the midwest region represented almost 41% of the reinfections. Across all four states, a notable increase in diagnosed cases occurred after EW 01/2022, coinciding with the introduction of the VOC Omicron (see Fig. 2A). Furthermore, during the same period, an increased number of reinfections was evident in all four states, as illustrated in Fig. 2B. These reinfections also exhibited a bimodal distribution, similar to the pattern observed in Fig. 1D.

Fig. 2.

Distribution of positive tests and reinfection cases in four Brazilian states. (A) Number of diagnosed cases for Goiás, Minas Gerais, Rio de Janeiro, and São Paulo states per epidemiological week. (B) Identified reinfection cases in each state. (C) Distribution of reinfection cases based on age and the time between infections for each state

The surge in COVID-19 cases among the Brazilian population, despite more than 70% having received two vaccine doses, may be attributed to two factors: the emergence of a new variant with a high transmissibility rate and a reduction in vaccine effectiveness due to new mutations in the viral genome. In both scenarios, the administration of booster doses is expected to diminish the likelihood of reinfection. Brazil initiated the administration of booster doses in September 2021, initially targeting individuals over 70 years of age. Consequently, it was anticipated that the number of cases and reinfections among the elderly would be lower compared to other age groups in the Brazilian population. To assess this assumption, we examined the incidence of reinfections in the four states with the highest number of samples, categorized by age. The age group between 25 and 60 years showed the highest concentration of reinfection cases (Fig. 2C). This result suggests that the booster dose has been effective in reducing reinfections among individuals aged over 70 years.

Discussion

Numerous global studies have highlighted cases and rates of reinfection [11–13]. A study conducted in the state of São Paulo found a 5% reinfection rate [14]. In our results, reinfection rates varied during the studied period, reaching up to 8% at the peak. A recent systematic review, which included 23,231 reinfections reported across 23 studies, found rates ranging from 0.1 to 6.8% [15]. Another systematic review, encompassing 64 studies, estimated a SARS-CoV-2 reinfection proportion of 1.16%. Of these reinfections, 8.12% resulted in hospitalization, 1.31% required admission to the intensive care unit, and 0.71% resulted in death [16]. Clinical prognosis data for SARS-CoV-2 reinfections remain scarce, with a majority of cases lacking detailed outcome information [17]. Likewise, we were unable to access SARS-CoV-2 reinfection clinical relevance in our dataset.

Several countries reported an increase in the number of cases and deaths following the introduction of Omicron and its sublineages [18–20]. With over 60 mutations, Omicron has the potential to evade the host’s immune system, increase transmissibility, and resist vaccines, convalescent serum, and monoclonal antibodies [18]. Reinfections have been reported as more prevalent during the Omicron variant period [15]. Our findings corroborate this trend, showing a significant surge in reinfections, particularly after the introduction of the Omicron. We observed a sixfold increase compared to the initial two waves associated with the emergence of the Gamma and Delta.

The global vaccination efforts initially resulted in a significant decrease in COVID-19 cases. Vaccines against COVID-19 are effective in inducing immunity, reducing the risk of severe disease and transmission [21]. However, the emergence of the Omicron variant has reversed this trend. Booster doses, when administered, enhance and prolong this protection, particularly against emerging variants such as Omicron. Our study revealed that the highest number of reinfections occurs in the adult age group (25–60 years). Notably, this age group in Brazil had not yet received full vaccination with booster doses, which could be a contributing factor to the increased frequency of reinfections.

In conclusion, our study underscores the dynamic nature of SARS-CoV-2 reinfections and their correlation with emerging variants such as Omicron. The findings emphasize the importance of ongoing surveillance and adaptive public health strategies, including widespread vaccination and timely administration of booster doses. Continued research and vigilance will be essential in navigating future waves and mitigating the impact of COVID-19.

Author contributions

Writing—original draft: P.L.C.F and R.P.S.; writing—review and editing: P.L.C.F., R.P.S., N.K.M., E.I., J.C., and A.L.B.; conceptualization: R.P.S, R.S.A and D.A.G.Z.; investigation: P.L.C.F., F.S.V.M., I.B-P., C.S.A.S., R.S.A., D.A.G.Z. and R.P.S.; methodology: P.L.C.F., F.S.V.M., I.B-P., C.S.A.S., R.S.A., D.A.G.Z. and R.P.S.; formal analysis: P.L.C.F., F.S.V.M. and R.P.S.; project administration: R.S.A. D.A.G.Z. and R.P.S; funding acquisition: R.S.A. and R.P.S.

Funding

We acknowledge support from the Rede Corona-ômica BR MCTI/FINEP affiliated with RedeVírus/MCTI (1227/21); Instituto Todos pela Saúde - ITpS (Chamada 01/2021 - C1294); CNPq (315592/2021-4, INCT-One CNPq 405786/2022-0); FINEP (0494/20 01.20.0026.00); CAPES (Finance Code 001), FAPEMIG (BPD-00820-22).

Data availability

Not applicable.

Declarations

Ethical approval

This study was approved by a Research Ethics Committee under protocol CAAE 33202820.7.1001.5348.

Competing interests

The authors declare no competing interests.