Updated on May 15th 2020, written by Dr David Owens.
The impact of all infectious illness is ultimately determined by two independent factors:
- Exposure Risk
- Individual immune response
The goal of the public health controls is to minimise exposure risk. Ideally, an individual will never encounter a specific infection. If they do, hopefully it will be the smallest possible quantity (the lowest viral load).
What are antibodies?
Recovery from any viral illness will be a result of an internal battle between the virus and the immune response. An individual who has experienced an illness previously, or has been immunised against that virus, is likely to already have antibodies against that infection. Antibodies are the soldiers of immune defence system. If an individual is exposed to the same virus again their natural immunity is able to win this battle relatively easily. The infection never takes hold. If they are exposed to an infection to which they have no immunity, the strength of the immune system will be impacted by some general factors such as age, fitness, nutritional status, other health conditions or drugs that they may be taking. The impact of the internal battle in this situation will be a balance of the ability and strength of the immune system in fighting the infection and the individual characteristics of the specific virus. In an ideal situation, the infection is recognised as a threat and the body is able to produce an effective response. The first line of this response typically involves IGM antibodies. These are the forward response teams of our defence systems. We know that in COVID-19 they start to be produced within a few days. Later, as the battle is hopefully being won these antibodies are replaced by longer lasting IGG antibodies. In many infections, such as chickenpox these antibodies may last for life. In some situations, this immunity may fade over time. COVID-19 is a new illness. We know that many of the most severe complications of this disease are caused by an excessive immune reaction in the later stages of the illness. Understanding of the immune response to this virus is evolving. It is too early to know how effective the antibody response to the SARS-COV-2 virus will be. We also do not know how long these antibodies will last and how effective they will be. Time and further study is necessary to understand these questions.
Why are antibody tests so difficult or unreliable?
When new viruses appear it is now possible to map their genetic code in a remarkably short period of time. Only a few weeks in the case of COVID-19. Once we have the genetic code, it is relatively easy to build a probe which lights up when it recognises the same sequence of code. These are the PCR tests which have been conducted widely. A PCR test is a point in time test. If it is positive it indicates that an individual is infectious and is excreting the virus. This may be positive for a few weeks but in an individual tested before being exposed to the illness or the week after clearing the virus, the test will be negative. To know how many people in a population have experienced an infection we need antibody tests. These tests are more difficult to produce. We must identify a response within the body which is unique to a specific virus. An ideal test will recognise the natural response in every single person who has previously contracted the COVID-19 virus but will not be confused by immunity to any other similar viruses. This is a much more difficult process. The ideal test would be 100% sensitive and 100% specific and the ideal test does not yet exist although advances are being made internationally.
What is sensitivity and specificity?
The sensitivity of a test is a measure of how accurate the test is in measuring a true positive result. The higher this number the more likely that all the true positives have been identified. Specificity is a measure of how accurate the test is in measuring a true negative result. The higher this number the more likely that all the true negatives have been identified. Of course, things are never quite as simple in statistics and as the prevalence (number of cases in the population) falls. The predictive value of these tests also changes. The simple summary is that the tests being used in the early studies are not yet accurate enough to give a good picture and more detailed studies with better and more accurate testing is needed. We are currently involved in such a study with Hong Kong University. In this study all positive tests will be confirmed by more accurate neutralization testing.
What do the early studies show?
Early studies from China, Europe and the US show relatively low antibody levels in population terms. The early studies show a prevalence of 2-4% in China and 10-14% in Italy, Germany and the US. None of these studies are yet high quality, peer reviewed research papers. It remains possible that some individuals are not mounting an antibody response to the infection. It is also possible that the testing we have at the moment is not accurate enough. Currently we are not sure how effective the antibody response is, how long it will last for or how effective it will be in preventing future infections. All of these questions need to be explored with future research.
Will antibody tests help in diagnosing COVID-19?
In some illnesses we use antibody tests to assess acute infection. Typically, an IGM response indicates a recent infection whilst an IGG response indicates past infection or immunisation. It is too early to understand the role of antibody testing during acute infections in COVID-19. This may change in the future but PCR tests remain the gold standard for assessing acute infection. Currently a positive antibody test gives us information about the past. It tells us that somebody has been infected previously. There is a caveat that there may be some false positive tests and it is not possible to be sure when an infection occurred, but broadly a positive antibody test is confirmation of past infection.
What information will the antibody studies give us?
Antibody studies will be important in answering a number of questions. We can consider that these questions may be of relevance to the individual and to the population.
It seems reasonable that an individual who has a positive antibody test is likely to have some degree of immunity. How effective that immunity will be in preventing future infection and how long it will last for we cannot yet be certain. Looking at other similar infections it seems likely that immunity will last for 1 to 2 years and possibly longer.
Antibody studies are going to be most effective in giving information in population terms. There are really two main pieces of information that will come out of the population prevalence studies:
The first is the total number of positive antibody tests. This gives a broad measure of population immunity. I have explained previously that an epidemic caused by an infectious disease with R0 of 2 will begin to die when 50% of the population develops immunity, for an R0 of 3 this increases to 66%. The early antibody studies are suggesting levels of immunity some way below this threshold. This would suggest that, unless these studies are under reporting natural immunity, the concept of herd immunity is still some way off.
There remains a role for antibody studies in this situation. Changes in seroprevalence between communities and within the same community over time can be used to measure the effectiveness of public health interventions. Hong Kong University School of Public Health have developed a real time measure of effective R0 known as Rt1. A combination of big data and health data, including changing antibody prevalence such as we are researching in our combined research project will be used to assess public health interventions as the epidemic evolves in Hong Kong. At a local level antibody tests may be helpful in deciding the strategic deployment in higher risk situations of immune health care workers. In the future it may be a factor in normalising office working.
Infection fatality rate
The other information which we will get from population studies is the incidence of unrecognised infections. This will allow us to identify the size of the iceberg beneath the surface. The larger the ratio of mild cases to diagnosed cases the less severe this illness will be on a case-by-case basis. The early studies, as expected show a significant number of unrecognised cases for every diagnosed case. More and better studies will be needed to better help us understand the true severity of this illness on a case by case basis.
The management of epidemics of infectious disease is ultimately a balance between the impact of the disease and the impact in population terms of the economic, social and political impact of the public health measures. The fatality rate of a disease is frequently downgraded as epidemics progress and more data is acquired. This is happening with COVID-19. This is not to diminish the importance of COVID-19 as a significant public health threat. A disease with a low infection fatality rate can have a devastating impact if it spreads widely and because of the numbers involved it is inevitable that some young and previously fit people will die. The nature of this disease also has a significant impact on health systems. We know that public health controls are effective. We also know that poverty is the single greatest factor in population health2. In general this disease has the greatest impact on older males with co-existing illness. It is inevitable that if further antibody studies continue to support the suggestion of a lower infection fatality rate that this will become a component of the political narrative around the evolving debate between ongoing public health controls versus the social, political and economic impact of such controls. The challenge will be to reboot the economy whilst mitigating the impact of the disease below the threshold at which health systems are able to function.
2. Poverty and social determinants. (2020, April 22). Retrieved from http://www.euro.who.int/en/health-topics/environment-and-health/urban-health/activities/poverty-and-social-determinants