Incubation period

(Latest update 29 April 2021)

The incubation period of COVID-19 (time between exposure to the virus and the development of the symptoms (symptom onset)) is on average five to six days with a range from two to 14 days. Studies including modelling suggest that the large majority of symptomatic cases are expected to have an incubation period of between two and 12 days [1-4]. This is important for determining the timing of testing.

The body of evidence on the COVID-19 incubation period has not changed since the last update in June 2020. ECDC continues to monitor new evidence and update its interpretation would new evidence arise.

The exact duration of infectivity of COVID-19 patients is not yet known. Isolation of live virus from clinical samples is rarely performed. In studies of non-severe cases, the virus was successfully isolated up to 10 days from the onset of symptoms [5-10]. Limited data indicate that immunocompromised patients have a prolonged infectivity period, e.g. > 20 days after symptoms onset [11-13].

Viral shedding, viral load, and infectivity

(last update 29/04/2021)

Viral shedding

Viral shedding, defined as ‘a qualitative description of disease status determined by a positive SARS-CoV-2 RT-PCR test result’ [14], does not equate with infectivity. Viral shedding occurs in asymptomatic infections or before symptom onset and clinical diagnosis [15].

A study of individual patient data including 7 340 observation for viral shedding analysis (preprint) estimated a median duration of viral shedding in respiratory samples of 4.76 days (95% confidence interval [CI]: 3.44 – 5.11) and in gastrointestinal samples of 4.94 days (95% CI: 4.09 – 5.8) [14]. A systematic review and metanalysis estimated a viral shedding of 17.0 days (maximum shedding duration 83 days) in upper respiratory tract, 14.6 days (maximum 59 days) in lower respiratory tract, 17.2 days (maximum 35 days) in stool, and 16.6 days (maximum 60 days) in serum samples [16].

However, some COVID-19 patients have positive RT-PCR results long after (e.g. 60 to >100 days) initial diagnosis and clinical diagnosis [16-20]. Studies in hospitalised COVID-19 patients have found that the RT-PCR test for SARS-CoV-2 could remain positive in respiratory samples up to six weeks from illness onset [21,22]. Some evidence is emerging that these cases were not linked with secondary transmission [23,24]. Prolonged shedding of SARS-CoV-2 RNA has been shown even after seroconversion [23,25].

Several factors have been associated with prolonged viral shedding, such as symptomatic infection [26], severe disease [11,17,27], increased age (e.g. > 60 years) [26], delays in diagnosis, some chronic diseases (e.g. cancer), immunodeficiency disorders and corticosteroid treatment [11,26]. Available evidence indicate there is no association between gender and duration of viral shedding [11].

Viral load

Viral load, defined as a ‘quantitative viral titre (e.g. copy number)’ [14], is a useful marker for assessing viral kinetics, disease severity and prognosis.

A pooled analysis of individual patient data including 5 328 observations of viral load, estimated that SARS-CoV-2 viral load peaks seven days prior to symptom onset [14]. Similar results have been reported from a mathematical model, predicting that on average, viral load peaks one day before symptom onset [28]. SARS-CoV-2 viral load in respiratory samples peaks during the prodromal phase and then decreases constantly afterwards [14].

Patients with severe disease have significantly higher viral loads than patients with mild disease [14]. Older age has also been associated with higher viral loads [14,28,29]. However, it has also been shown that children have viral loads similar to that of adults [30]. Prolonged viral shedding with high viral load have been associated with poor outcomes in hospitalised patients [28].

Impact of prior infection and vaccination on viral shedding and load

At present, there is no evidence supporting the assumption that viral load or duration of viral shedding are reduced after re-infection with SARS-CoV-2 [31].

Results from ongoing trials and observational studies of vaccine effectiveness suggest a relevant effect on viral load and viral shedding. Lower viral load and shorter duration of viral shedding have been observed in a small group of laboratory-confirmed COVID-19 cases among people vaccinated with the ChAdOx nCoV-19 vaccine compared to laboratory-confirmed unvaccinated controls [32].


(last update 21/04/2021)

In September 2020, ECDC published a threat assessment brief in response to a small number of published case reports documenting suspected or possible reinfections in individuals that had recovered from a prior episode of SARS-CoV-2 infection [33]. This brief highlighted the challenges in determining whether such reports represent true reinfections, persistent viral shedding, or recurrence of positive (re-positive) polymerase chain reaction (PCR) diagnostic tests [34]. Additional lines of investigation to support a diagnosis of reinfection were also highlighted and included genetic sequencing to compare virus isolates from the initial and suspected reinfection episode.

A diagnosis of true reinfection with SARS-CoV-2 can only be established when viral clearance is complete for the first episode of infection, and sufficient time has elapsed to allow for immune responses to be mounted. Re-positive PCR tests have been widely reported in convalescent patients. However, in the absence of a documented symptom-free period and supportive diagnostic sequencing, it is difficult to exclude fluctuations in viral shedding or false negative results when viral loads are low [35-38].

Duration of immunity and reinfection risk in seroconverted individuals

Following infection with SARS-CoV-2 virus, it is the adaptive immune response that ideally delivers long-term protection. The adaptive immune response primarily comprises memory B cells that produce different classes of antibodies to neutralise the virus or virus-infected cells, and memory T cells that support antibody production and also have a direct role in killing virus-infected cells. While there is evidence of both memory B cell and T cell immune responses in individuals infected with SARS-CoV-2, clear correlates for protective immunity have yet to be defined [39-42]. In the absence of definitive correlates of protective immunity, the presence of neutralising antibodies against SARS-CoV-2 provides the best current indication for protection against reinfection for previously infected individuals. The S1 domain of the SARS-CoV-2 spike protein includes the receptor binding domain (RBD), and antibodies targeting this critically impair virus cell entry [43]. A number of studies have shown that the neutralisation ability of polyclonal serum correlates positively with anti-spike IgG or anti-RBD IgG [40].

A scoping review performed by the Irish Health Information and Quality Authority (HIQA) to evaluate the long-term duration of immune responses following SARS-CoV-2 infection identified five studies that investigated immune responses at ≥six months post-infection, including two studies at ≥eight months post-infection. In general, studies reported a waning of antibody responses in the late convalescent period (three to six months post-infection). However, T-cell and memory B-cell responses were still present, and in many cases increased, up to eight months post-infection in all study participants [44]. Taken together, results from cohort studies confirm the protective effect of previous SARS-CoV-2 infection ranges from 81% to 100% during a follow-up period of five to seven months [45-50], although longer follow-up is necessary to better define the duration of protection for longer periods of time. These studies were largely conducted prior to the emergence of variants of concern (VOCs) with immune escape potential, for which the World Health Organization established working definitions in February 2021 [51]. In this context, the importance of sequencing available isolates from all suspected reinfection cases is further highlighted in ECDC guidance on representative and targeted genomic SARS-CoV-2 monitoring, published in May 2021 [52].

Reinfection incidence and surveillance in the EU/EEA

While reinfection events appear to be rare (studies from Denmark, Czechia and UK suggest they amount to less than 1% of documented SARS-CoV-2 positive cases [49,53,54]), there is currently limited population-level data available that captures the burden of reinfection cases at national level and over time. A survey of EU/EEA countries conducted by ECDC in January 2021 revealed that amongst 17 responding countries, the majority of responding countries reported having a working case definition and a national reporting system to capture reinfection cases. These definitions, although similar, were not standardised [55]. In order to better ascertain the burden and impact of SARS-CoV-2 reinfection across the EU/EEA, particularly in the context of emerging variants with immune escape potential, ECDC has established a surveillance case definition for suspected reinfection, introducing new case-based and aggregate variables to improve systematic reporting via The European Surveillance System (TESSy) [56]. A suspected COVID-19 reinfection case is defined as:

Positive PCR or rapid antigen detection test (RADT) sample ≥60 days following:

  • Previous positive PCR;
  • Previous positive RADT;
  • Previous positive serology (anti-spike IgG Ab).

This case definition takes into account the time required to mount a neutralising antibody response and the variability of neutralising antibody dynamics following infection with SARS-CoV-2, the potential risk of early immune escape posed by emerging VOCs, as well as existing surveillance practices and reporting capabilities amongst EU/EEA countries. To collect data on suspected reinfection cases via TESSy, an update to the metadata was implemented on 12 March 2021; more information can be found in the latest reporting protocol [57]. Standardised surveillance reporting protocols for suspected reinfection cases within the EU/EEA will facilitate the assessment of:

  1. The total number and incidence of suspected reinfection cases;
  2. The risk of suspected reinfection by VOCs;
  3. The severity of suspected reinfection cases, as compared to first episodes of infection.

Depending on the quality of data submitted to TESSy on suspected reinfection cases, these outputs will be considered for inclusion in ECDC’s COVID-19 country overview reports [58].



Supporting document: List of references