Covid-19 Reinfections

Insight Health Solutions

By Craig Getz.

How often have you heard someone state that they have previously been infected with Covid-19 and that they no longer need to concern themselves with the virus? Many people believe that their previous infection is a guarantee of long-lasting immunity. This article explores whether this belief is well-founded based on observable data. Simply put, are previously infected people immune to reinfection or serious reinfections?

Insight consults to a broad spectrum of medical schemes, who collect extensive data on the number of diagnosed Covid-19 infections and the hospital admissions and deaths resulting from these infections. By virtue of this data, Insight is well-placed to measure how often South Africans are being reinfected with Covid-19 and how severe those reinfections are.

We have defined reinfection as an instance where a person tests positive for Covid-19 at least 45 days after their last positive test. This is broadly consistent with the definition used by the Centre for Disease Control and Infection (CDC) in the United States[1].

We find that 1.3% of people who have tested positive for Covid-19 have since been reinfected. This equates to one in every 74 people diagnosed. Some people, though, have recently tested positive and have not had an opportunity to be reinfected.

We can more reliably measure the risk of reinfection by considering only those infected at least 12 months ago. 2.3% Of which have been reinfected. This translates to one in every 44 people diagnosed. Reinfections are observed shortly after the initial infection. 0.3% Of infections result in reinfection within three months, 0.9% in six months and 1.8% within 12 months.

risk of reinfection

By extrapolating the 1.3% reinfection rate to the more than 2.7 million diagnosed cases in South Africa[2], we can infer that there have been approximately 36,000 reinfections. This is likely a severe underestimate as not every infection and reinfection are diagnosed.

We have contemplated approaches to better estimate the true number of reinfections whilst acknowledging the uncertainties associated with such estimates.

  • Approximately 10% of medical scheme beneficiaries tested positive for Covid-19. This is as opposed to the 4.5% of all South Africans who tested positive. The disparity is believed to be primarily attributable to variations in testing capacity and access. To the extent that the 10% infection rate is a more realistic representation of experience and that the 1.3% reinfection rate can reasonably be applied to that, we can deduce that there have been more than six million infections and more than 80,000 reinfections.
  • It has been widely reported that between 70% and 80% of South Africans have been infected with Covid-19. This equates to between 42 and 48 million people. It is unknown whether the 1.3% reinfection rate can be extrapolated to this cohort with any degree of certainty. Assuming it can be and assuming that at least 70% of South Africans have been infected, we would expect there to be over 500,000 reinfections. This, however, is mere conjecture.

To answer the first question, are people who have previously been infected with Covid-19 immune to reinfection? No, a notable and increasing proportion of people are reinfected after their initial infection.

We can make inferences about the severity of reinfections based on the extent to which these reinfections result in hospital admissions and deaths. Severe reinfections necessitate hospitalisations, and extremely severe infections result in death. 31.6% Of diagnosed reinfections are associated with hospitalisations, which equates to one hospitalisation for every three diagnosed reinfections. 3.4% Of diagnosed reinfections are associated with death[3]. This equates to one death for every 30 reinfections.

22.0% Of the initial infections (amongst those who were later reinfected) resulted in hospitalisations. This is materially lower than the 31.6% admission rate associated with reinfections.  This suggests that diagnosed reinfections are not, on average, less severe than the initially diagnosed infections.

The 31.6% case admission rate for reinfections is also higher than the case admission of 23.6%[4] across medical scheme beneficiaries. This suggests that diagnosed reinfections are not less severe than the average diagnosed infection.

3.4% Of the diagnosed reinfection resulted in death[5].

diagnosed reinfection

To answer the second question, are people who have previously been infected with Covid-19 likely to experience milder reinfections when compared to their initial infection? Unlikely, diagnosed reinfections have been associated with higher hospitalisation rates.

None of this suggests that people who have previously been infected are not at reduced risk for future infections compared to people who have not previously been infected. For example, people who were previously infected are only 25% as likely to be infected in the 90 days after their recovery as a person who hasn’t previously been infected. This equates to a 75% risk reduction.

These findings then suggest that a notable number of people who were previously infected with Covid-19 have since been diagnosed with reinfections. Those reinfections have, on average, been more severe than the initial infections. However, the chance of a diagnosed infection (within three months of a previously diagnosed infection) is 75% lower than for people who have previously been infected.

In turn, previous infections are not a guarantee of long-lasting immunity and people who were previously infected should still endeavour to protect themselves from Covid-19.

[1] There is a possibility that a subsequent positive test is indicative of a prolonged prior infection rather than a reinfection

[2] As of 25 August 2021,

[3] This is the case fatality rate. This should not be confused with the infection fatality rate, which refers to the proportion of infections that result in death. The infection fatality rate is not considered, given that the number of infections cannot be determined with certainty.

[4] The case admission rate is adjusted to account for the risk profile of persons reinfected. Age, gender and chronicity are adjusted for.

[5] None of the initial infections resulted in death. Persons who are later reinfected could not have succumbed to their initial infection.

For more information and/or to schedule a discussion, contact: 

Craig Getz



Get an email whenever we publish a new thought piece


By signing up you consent to our terms and conditions

More from Insight Health Solutions