How much COVID testing do we need: (2) Technical Estimates

Sri Lanka needs to increase testing of COVID from ~250 to at least 2,000 RT-PCR tests a day. There is a large backlog in testing of contacts of known COVID cases that needs to be rapidly dealt with. We also need to put in place testing capacity that will allow the country and economy to get back to some level of normality during the next 12 months. 

We report target levels for daily testing, as well as how much capacity we need overall. Our analysis does not estimate requirements for antibody (IgG/IgM) testing, as this will be additional and is not appropriate as the first line of testing in this war against COVID-19. The most urgent priority for the country is to expand RT-PCR testing. Our initial estimates and analysis are given in our report (click to download) and the key numbers are summarized below.

The minimum numbers are our estimate of the cheapest testing strategy that might work. However, there is a significant risk that this will not work, since testing rates will remain one tenth of those in places like Hong Kong. We recommend that the country aims for the ideal numbers provided in each table. These will give us a much better chance of success.

Testing requirements

Minimum Ideal
Average daily tests 2,055 6,000
National capacity (tests/day) 4,000 9,000
High throughput machines 6 13

Note: Capacity set at 200% and 150% of average testing rates since rapid testing times require decentralized network with most machines not working at full capacity. Also allows for equipment downtime and spare surge capacity..

Testing has to do much more than simply diagnose and treat COVID cases. The strategic purpose of testing in this pandemic is: (i) to detect potential outbreaks of infection in the community, (ii) to slow the transmission of the virus between people by reducing the time that infected people continue to infect others; and (iii) to minimize the import of the virus through our borders.

If we can do this well, then the objective is not to “flatten the curve” as many countries are trying desperately to do. We have a better option. That is to crush the virus completely within Sri Lanka and keep it crushed. That will allow normal life to restart. We can do this because so far we do not have much local transmission of the virus.

There are only a few places which are aiming to do this and can do it. They include Singapore, Hong Kong, Taiwan, China and possibly Korea. Their strategy involves extensive testing of international arrivals and contacts of known cases, and surveillance testing of the community to pick up new cases that are not connected with foreign travel. We can copy them. We have the doctors and the laboratory experts. We also have a health system and a military that have shown before that they can do the impossible if the country’s leaders give them clear direction.

It will be expensive – we estimate that this strategy that does the minimum to copy the key elements of Hong Kong and Singapore’s approach will cost Rs 4–12 billion (US$ 23–65 million) over one year. 

Cost

Minimum Ideal
Operational cost/month Rs 321 million Rs 936 million
Equipment purchases Rs 492 million Rs 1,066 million
Total rupee cost over 1 year Rs 4,340 million Rs 12,302 million
Total US dollar cost over 1 year USD 23 million USD 65 million

Note: Equipment cost ignores any existing machines. Purchase cost is on basis of Roche machines (31 March 2020). Operational cost only considers Roche reagent costs which may be on high side, but excludes costs of sample collection, transport, time of laboratory staff and test reporting. These additional costs may increase cost estimates by 30%..

This seems unaffordable, but it isn’t. The cost of this can easily be financed by the new World Bank COVID funding Sri Lanka has obtained (US$ 128 million). This strategy also has the best chance of putting the country back to work and allowing us to re-open our airport. Crucially, restarting the economy will increase the taxes that the government can collect. We estimate that the resulting gains in tax collection will at the minimum be 20–50 times more than the cost of the testing. 

Affordability

Minimum Ideal
% of World Bank COVID loan ($128m) 18% 51%
% of gain in tax revenues 2% 5%

Notes: Baseline tax revenues for 2020 based on IMF projections (Nov 2019). We conservatively assume that COVID will result in tax collection falling to 60% of this, but increasing to 70% if the testing programme is adopted.

Key elements of the proposed strategy are provided below, with further details in the technical report.

GOALS OF NATIONAL COVID STRATEGY

  1. Ensure local transmission of COVID-19 remains at near-zero levels with no community transmission.
  2. Avoid national curfews or lockdowns after April 2020.
  3. Normal functioning of schools and businesses after May 2020 regardless of global situation.
  4. Re-opening of airport after May 2020 with adequate safeguards.

KEY ASSUMPTIONS

  1. Current measures (airport closure, lockdown, social distancing) achieve zero transmission by 30th April, indicated by zero new cases in previous 14d.
  2. Airport will increase arrivals consistent with the strategic goals, with an initial target of 30,000 arrivals/month, rising to 60,000 arrivals/month which would allow essential business travel, return of Sri Lankans living abroad, and some minimal tourism.
  3. PCR testing will be complemented by IgG/IgM testing as appropriate.
  4. Testing programme to last 12 months to April 2021, the minimum time required for vaccine to be available globally.

NATIONAL COVID CONTROL INTERVENTIONS

We need to do all of the following measures in combination:

  1. Border controls to minimize imported cases at level that the other measures can safely handle without risking sustained community transmission.
  2. Stepped-up tracing and testing of all direct and indirect contacts to reduce virus transmission.
  3. Effective quarantine and isolation to handle all detected cases and exposed contacts.
  4. PCR test-based surveillance of population to provide early warning of potential outbreaks.
  5. Measures to increase social distancing on permanent basis till pandemic is over.
  6. Targeted school closures, internal cordon sanitaires and lock-downs kept in reserve to handle threatening outbreaks.

COVID TESTING PROGRAMME

*ALL” means both public and private sectors.

  1. PCR testing of all arrivals plus 14d quarantine. Arrivals to be restricted to capacity of system to test everyone. Restrictions incrementally relaxed to lift mandatory quarantine on “safe” arrivals and shifting to PCR testing of only “high risk” arrivals. Risk assessment of arrivals to be based on a combination of originating country, health screening results and other information.
  2. PCR testing of all close contacts of new cases (including asymptomatic), and IgG/IgM testing of all non-close/secondary contacts.
  3. PCR testing of ALL ICU admissions
  4. PCR testing of ALL pneumonia/influenza admissions
  5. PCR testing of ALL hospital deaths suspected to have infectious origin.
  6. EITHER PCR testing of saliva of ALL outpatients with both fever and respiratory symptoms, OR of a systematic sample, e.g.,designated surveillance sites or random sample (>20%) of relevant cases.
  7. Doctors allowed to order PCR testing on basis of clinical suspicion on any case that does not fit the case criteria.
  8. PCR testing/antibody testing of pneumonia clusters in community.
  9. Two negative PCR tests on all recovered COVID patients before discharge.

4 thoughts on “How much COVID testing do we need: (2) Technical Estimates

  1. Testing capacity in Germany will be increased by up to factor 10 to up to 400,000 a day (!) by doing pooled testing.

    E.g mix 16 samples and if negative – all are negative, otherwise binary search for the positive(s). Could of course be used worldwide.

    • It has potential. However, even now Germany is doing around 50,000 tests each day compared with just 250 in Sri Lanka. Adjusting for the difference in population, that is 50 times more than here. Germany has still to implement the pooling method — which likely will not work for all types of testing, but that would increase their testing rate to 500 times the rate in Sri Lanka. Sri Lanka’s problem right now is that we first need to increase the number of tests done, and then see if pooling can increase it further. Pooling will not be enough to address the gap.

Leave a Reply to Dr.Preethi Wijegoonewardene Cancel reply