Mass testing options spark COVID debate
JUDGING from the confusing public debate on the projected “mass (or massive) testing” to determine the extent of coronavirus disease (COVID-19) infection, some points still need clarification.
For insight, we chose an article by Dr. Edsel Maurice T. Salvaña on what testing is and why it is done. He is an award-winning infectious diseases specialist and molecular biologist at the University of the Philippines and the Philippine General Hospital. He and other Balik-Scientists are advising the government on the COVID-19 menace.
We print below excerpts from his article on Facebook (edited to fit our space):
1. There are two types of testing: Clinical testing and Epidemiologic testing.
Clinical testing determines if someone has the disease so he can be treated properly. For COVID-19, the best test if someone is sick and spreading virus is the RT-PCR (Reverse Transcriptase-Polymerase Chain Reaction). This is best done with the onset of symptoms since RT-PCR detects the genetic material of the virus. If the virus is present in high enough quantities, the test will be positive. If RT-PCR is done too early or too late, it will not detect the virus and will be falsely negative.
Doctors look at a patient and label him a “probable case” even if the test is negative if there is exposure and the clinical picture (fever, cough, pneumonia) is consistent with the disease. That’s why we test people with symptoms and also healthcare workers who are highly exposed. We act on the results of this individual test – isolate, trace and treat.
Epidemiologic testing gives a picture of the spread of a disease in the community. It may not change what we do with the individual patient, but it is a good guide to policy. This can be done with a reliable antibody test. Testing everyone is not needed, as a properly taken representative sample can give a good idea of what is going on in the community.
The best antibody test for epidemiologic testing is a laboratory-based ELISA (Enzyme-linked Immunosorbent Assay) because these are much more sensitive and specific than rapid antibody tests.
Sampling methodologies are important in ensuring that the sample is truly representative of the community spread of the virus. If only people with exposure (in contrast with clinical testing) are tested, we get a falsely high result.
2. What is mass testing, and is it feasible? If you define mass testing as population-based testing, meaning testing everyone regardless of symptoms and exposure to pick out those infected, then it is not feasible and not useful.
Testing someone who is not exposed and has no risk is a waste of money and may come back as a false positive. Even if you test everyone and they are negative, it does not mean they will no longer be exposed to the virus the next day and become carriers and cases afterwards.
Even RT-PCR will fail to detect virus in the first three days of the incubation period and a lot of asymptomatic (someone without symptoms) carriers will be missed.
Antibody testing for asymptomatic carriers is even worse than RT-PCR – they will not be positive because they have not produced enough antibodies to make the test positive. Antibody tests become positive about seven days after the symptoms show up and are best done after 14 days of symptoms because antibody levels are low until someone is on the way to recovery.
Testing everyone to determine the percentage of the population exposed is epidemiologic testing and can be done with a representative sample. A proper sample for community-based testing of the entire Philippines has been calculated to be as small 4,000 to 5,000 people.
If by mass testing is meant testing everyone who is at risk, then the government is already trying to do it and is paying for it. This is called enhanced targeted (massive) testing.
This is the same approach of South Korea and Germany. We are now testing all patients with symptoms, which is a liberal approach compared with most countries. In fact some local governments have been testing asymptomatics with exposure using RT-PCR.
No country has used rapid antibody tests to determine clinical disease in patients. Only RT-PCR is reliable enough to do this. The “mass testing” approach of South Korea and Singapore used RT-PCRs, and only in those at risk.
The World Health Organization recommends against using rapid antibody tests for clinical decision-making. Such tests can be used for research and epidemiologic testing, but not to decide who is sick or not. This may change as we get better tests, but for now it is too risky as basis for policy.
3. Are we doing enough testing? Isn’t more better?
The WHO has pegged adequate testing to at least 10 tests for every one test that comes back positive. Testing capacity has improved with increasing healthcare facilities. We are doing about 16.3 tests for every one positive that comes back. That is better than the US’s 7.8 and nearly the same as Germany’s 18.6.
As a third-world country, we need to conserve our resources. Too much testing will divert limited money that we can use for our social amelioration programs and economic stimulus packages.
There is a global shortage of materials, and if we use up too early what we have, we won’t have more materials to keep tracking the pandemic. This isn’t going to end anytime soon. We need to be efficient with testing and not overdo it to the detriment of resources and future capacity.
4. What is the best test to clear people for work?
The best test to clear people for work is to check them for symptoms. Symptomatic COVID-19 is responsible for >85 percent of transmission. The Department of Health and the Philippine Society for Microbiology and Infectious Diseases have released the guideline — no need for laboratory testing to clear people for work.
As for asymptomatic carriers, there is no good test to completely eliminate asymptomatic transmission. This is why we need to “COVID-proof” the workplace with daily symptom checks, wearing of masks, physical distancing, handwashing and disinfecting. Neither RT-PCR nor antibody testing will pick up all our asymptomatics.