I’ve been asked several times what I think of the government’s vaccination strategy. This is not easy to answer, because the strategy is neither transparent nor explained well. Recent events have shown also that there is conflict within even MOH in implementing it. I know that many officials within MOH are trying to do their best, but it’s also evident that they face huge internal difficulties in crafting and implementing a sensible approach.
As I have said previously, Minister Dr Sudarshini Fernandopulle and Dr Amal Harsha De Silva should be commended for clearly saying that the aim is to cover most Sri Lankans and not to rely only on free COVAX donations. But I still have concerns about what the overall goals are, plus whether risks have been properly factored into the assessment of what is an optimal strategy. I also suspect that our typical bureaucratic processes may not be fast or adaptive enough to manage the uncertainties and rapid changes in information that we are seeing.
Our vaccination strategy needs to be clear about its goals. These need to be both: (1) minimizing deaths and serious illness, and (2) minimizing transmission, since our National COVID Strategy, whose goal is Elimination, has failed to achieve its objective. If we had been able to maintain elimination (which would have required an investment in testing which we never made), then we would have had the option of delaying vaccination a little bit.
There are and will remain huge shortages in vaccine supplies through 2021–2022. Our analysis at IHP is that under the most optimistic scenarios under which all manufacturers deliver on their promised production targets, there is barely enough vaccines to cover everyone in the world by early 2022. The optimal strategy needs to take into account the significant risks of non-delivery or delayed delivery by manufacturers. The best way to mitigate this risk is what other sensible countries are doing, including other developing countries, like Malaysia. That is to pay upfront in cash so that manufacturers prioritize Sri Lankan orders in the delivery queue, order from multiple manufacturers, and order over 100% of requirements to allow for likely delays and problems.
There is another problem that some vaccines might turn out not to be so effective against future variants that we may import in future. The strategy needs to take into account the probability that new variants will be more immune resistant. This requires two types of risk mitigation: (i) border restrictions to prevent import of new variants, and (ii) purchasing vaccines which are most likely to perform well against future variants. The government appears to have given up on or is not taking seriously preventing entry of new variants, including additional cases of B117 which is already in the community—this is actually counterproductive as it will in longterm increase the cost of the optimal vaccination strategy. The second issue implies obtaining a mix of Pfizer, Moderna, Sputnik and Novavax, since available data indicate that either they perform better against known variants than other vaccines, or are likely to perform better against future variants.
The strategy also needs to properly evaluate vaccine effectiveness in real time based on emerging evidence, and not rely on WHO views which are constrained by a number of factors, including the dependency of COVAX on a limited range of vaccines, or the NMRA regulatory process which take time. It’s worth recalling here that the countries who have are furthest ahead in vaccinating their populations, such as USA and UK, did not wait on obtaining regulatory approvals before their governments spent billions in hard cash to order vaccines. In a situation like this, purchasing has to balance the costs of delay versus the risks that vaccine candidates might not obtain approval or might not work as well as projected. I also note here that relying on the WHO EUL process automatically biases against non-Western vaccines, since the WHO process has an inbuilt bias (probably inadvertent) that imposes more steps in the approval process on non-Western producers.
For a number of reasons including those above, I personally think our best strategy right now would be to take the COVAX free 20% supply of Astra-Zeneca (but not assume quick delivery), and then pay cash to order a mix of Pfizer, Moderna, Sputnik and Novavax to cover 16 million Sri Lankans, i.e., everyone else. We should also place orders without waiting to obtain NMRA or WHO approvals, if the available evidence supports the effectiveness of a vaccine.
All of the vaccines I just listed have efficacies against preventing illness of more than 90%. I would also seriously consider paying extra if that would ensure faster delivery, since analyses, including by CDC Director Prof Walensky, suggests faster deployment is more cost-effective.
16 million paid orders implies covering kids for which most vaccines are not currently certified for, but given the difficulties achieving herd immunity. I would bet on some of the vaccines being approved for kids at some point during the year. It should be noted here that against B117, most vaccines (including Astra-Zeneca) are unable to achieve herd immunity even if all adults are vaccinated—only vaccines with >90% efficacy, like the ones I mentioned earlier, have a chance of doing this.
I would also aggressively follow up the Russian offer to support local production of Sputnik. Russia appears to have problems in scaling-up production, which is one of the reasons why German Chancellor Merkel has offered to help Russia produce the vaccine inside the EU. This would also support the government’s goal of making Sri Lanka a pharmaceutical manufacturing hub. However, this might require the government to follow the British approach of intervening to help local private firms start local vaccine manufacture, instead of just turning to SPC, which I suspect is not a good shot for long-term, competitive export manufacturing.
I would also ask Chinese firms whether we can fill and package some of their vaccines in country, as Indonesia is doing. This is because Chinese firms face a problem in cheaply manufacturing vaccines inside China as Chinese regulators do not permit the production of vaccines in multi-dose vials, although they appear to allow firms to do this abroad, which can significantly reduce the price.
All this obviously entails money the government doesn’t currently have. However, since the economic return from going for 100% vaccination coverage with the most effective vaccines will far outweigh the costs, we should borrow if needed. Options include ADB, AIIB and World Bank. Of these the World Bank is more restrictive, as its Board under US government influence has imposed rules to discourage the use of Russian and Chinese vaccines. However, a better strategy would be to go to Parliament and ask for a tax increase to allow us to purchase the best vaccines for the nation. I am not a politician, so will leave it up to politicians to decide on that. But from a technical perspective this is what I would recommend them to do.
In future, variants may emerge that are immune resistant. This is a global problem, and in the short term this means that international travel will remain restricted through 2022, and global economic recovery constrained. I agree with international experts who argue that the only solution to this problem is global eradication of the virus. I believe this is technically feasible, but the biggest barrier is geopolitics as it would require unprecedented global cooperation. But in the absence of such global cooperation, we should aim for the best possible vaccination coverage for Sri Lankans, and adopt other policies to maximize suppression of the virus within our borders. Of course, all this suggests a vaccination strategy very similar to that of developed countries. But I have to say whilst their approach can be legitimately criticized for being selfish and also hypocritical given they are in the habit of lecturing everyone else what to do, their approach is still the most rational for any country in the absence of adequate global cooperation.
*Updated 24/02/21: Rearranged some text that had got mixed up. plus add some additional thoughts on purchasing non-approved vaccines.