By Sandra Albert and Glenn Kharkongor
While there are many features of the spread of the virus that are common to different countries and regions, there are also differences that need consideration and calibration in the planning of the local or regional response.
The public hear mixed messages through the media from all over the world regarding testing, and some countries have tested far more persons than India. So we need to sort this through and determine what is appropriate for our country and region.
Why testing?
1.Most importantly, to determine if an individual has the disease, so that her/his/ze treatment and follow-up can be properly conducted.
2.It is also important to track and test those who have come in contact with the affected (test positive) individual so that not only spread can be contained (quarantine) but their health can be monitored.
3.Random testing of persons to ascertain if community spread has occurred, and if so, what is the prevalence (percentage of the population that has been infected, whether they have symptoms or not) of the disease. This helps health experts to recommend policies and schemes to the government.
There are scientific methods to calculate the number of individuals (sample size) required for random population-based testing. These calculations are similar to those used to determine TV ratings or exit poll projections.
The sample can be stratified to include variables like age and sex, to see if there are prevalence differences between men and women, different vulnerability in certain age groups etc.
A large country like India presents certain issues, like regional differences eg climate and culture, and also varying efficiencies of health systems from state to state. The logistics of delivery and implementation are more challenging. Finally, there is a limit to resources and this requires prudent deployment.
Types of tests
There are two types of tests. The Real Time-Polymerase Chain Reaction (RT-PCR) tests for the presence of the virus and is useful for early detection, even before onset of symptoms, and can detect those cases also who never become symptomatic. This test also helps to track contacts of these individuals.
The other kind of test finds antibodies, ie proteins which the body creates to fight the invasive organism. These antibodies may appear in the blood stream 2-3 weeks after the initial infection. This type of test is not useful for treating an individual, but are useful in field studies to determine the disease prevalence in the community. Thus long-term policies for preparedness and control can be designed.
ICMR guidelines
While government and private labs are being geared up for readiness, the Indian Council for Medical Research has posited a consistent set of criteria, appropriate to the stage of the pandemic.
At this point in time, only those who have symptoms and their contacts are being tested. ICMR has run some random field tests to see if the infection has penetrated the community. So far these have been negative, but ICMR may ramp up such testing in ‘hotspot’ areas.
If ICMR expands its guidelines, say, in the event of the pandemic moving to Stage 3, ie community transmission, community testing may be recommended. The purpose of this would be to know ‘Point Prevalence’, which as the term implies is the percentage of people in the community with a positive test at the time of the survey.
Point Prevalence may change. If the epidemic is on the upsurge, the point prevalence may increase in the next survey. If it declines, the epidemic is waning. But this is not so straightforward, because once a person becomes positive, the antibodies may stay on for a while, or even for life. But this information may be helpful in another way. If I already have antibodies, I may not need vaccination.
The only country that did widespread testing of the general population was South Korea. As of Mar 31, 2020, 410,564 had been tested with 9,786 positives. So the Point Prevalence on that day was 2.4%. South Korea’s infection rate has been dropping for the last two weeks, so 2.4% may indeed, be the reference point for prevalence for this virus.
The Meghalaya Scenario
So far, there have been no positive cases in the state ie all cases with symptoms and their contacts have tested negative. This kind of testing will continue, because it will help in the treatment of those individuals and follow-up on their contacts.
In case it is decided to determine the prevalence percentage in Meghalaya, we will need to calculate the number of persons who should be tested in the state, ie the sample size, which was defined earlier.
For this we use the reference point of 2.4% from South Korea. Apart from this, it is also necessary to decide the ‘precision’ of the sample size. For rare diseases, high precision is needed (so a larger sample size), whereas as for common diseases, precision can be lower (smaller sample size).
To make the final calculation of the sample size, let’s assume a reference prevalence of 3% (slightly higher than South Korea to be on the safe side). To this we will seek a precision of 99%. Using a formula common in epidemiology, the total number of tests on randomly selected individuals, that need to be done across the entire state of Meghalaya, is 1164. From the results of this survey of 1164 individuals, the prevalence rate can be determined. To be able to detect the prevalence separately for different age groups (e.g. children, adults, elderly), the numbers to be tested will have to be increased somewhat.
In summary, Meghalaya should continue the current level of testing and if there are any positive cases, test the contacts also. If a cluster of cases occurs, perhaps it will be traceable to a crowded event and then a further number around that cluster will also need to be tested. To cover these possible situations, the present testing guidelines are sufficient.