Covid vaccine: How to ensure equitable distribution
One of my acquaintances, a professor at the University of Florida, was reasonably excited when she got the Covid vaccine shot offered by the university fairly quickly. Every country, however, needs its own plan for procurement, distribution and prioritisation for vaccination according to its socioeconomic and demographic structure. This is undoubtedly an unprecedented scale of vaccination in India — a much more extensive exercise than the polio vaccination to which we are accustomed for years. “It will be 2024 till everybody gets vaccinated,” experts believe. Even if the procedure is expedited in the presence of multiple vaccines, the vaccination procedure would take quite some time to conclude, for sure.
In India, at least the government has already outlined a suitable plan for the initial stages of vaccination — frontline healthcare professionals first, then the frontline workers and people aged above 50 and a special category of those below 50 with comorbidities and requiring specialised care. Certainly, we need equity, not equality, in vaccine distribution, and that is maintained in this plan with some sort of emphasis on the risk factor and also due recognition of the importance of frontline healthcare professionals and frontline workers in the society. However, how ordering for the healthy people below 50 will be done may not be fixed yet. Will they be vaccinated randomly, or according to their registration chronology, or is there any priority according to their vulnerability? It might not be a bad idea to set an appropriate objective function and carry out the vaccine allocation to them by optimising that objective function.
The Western world’s vaccination programme was started with the UK, and a 91-year-old woman was the first to get the vaccine. Interestingly, while most countries have planned to put their vulnerable older people first in line for vaccination, Indonesia planned to vaccinate its young working-age population before the elderly. They’ve their own argument for not bucking the trend, of course. One important point is that Indonesia has initial access only to a vaccine developed by China’s Sinovac Biotech, which does not have enough data yet of the vaccine’s efficacy on elderly people, as the ongoing clinical trials in the country involves people aged 18-59. Fair enough. To achieve “herd immunity”, Indonesia planned to immunise 67 per cent of its target 18-59 group, who are more socially mobile and economically active. Such types of targeted inoculation procedures are in contrast to the straight-forward one, ie vaccinating proportionately in different strata (constituted according to age groups and occupations). It’s difficult to predict the right
approach though, as nobody knows how the situation will evolve in the coming days. And, in a country like India, the older people, in general, may not always be too less exposed to the risk.
I was reading an interesting research article published in Nature Medicine in December (https://www.nature.com/articles/s41591-020-01191-8), written by scientists from Johns Hopkins University, University of Maryland and Policymap, Inc, Philadelphia, where, in the context of the US, an effective community-level risk-based analysis using some sophisticated statistical modelling was done to identify relatively small fractions of the population (for example 4.3 per cent) that might experience a disproportionately large number of deaths (48.7 per cent). In fact, these authors introduced a web-based Covid-19 mortality risk calculator for the US adult (aged 18 years and older) population and interactive maps for viewing community-level risks. They integrated information from pandemic forecasting models so that an individual’s absolute risk can be informed based not only on their underlying risk factors, but also on community-level risk due to underlying pandemic dynamics. A webbased risk calculator (https://covid19risktools.com/riskcalculator) is available that allows an individual to input information on risk factors and obtain estimates of individualised risk for Covid-19 mortality in numerical values.
I was playing with the model. For example, a 45-year-old, 5’8’’ tall, black or Africanamerican weighing 180lbs, who never smoked, living in Glen Saint Mary town in Florida, has an estimated 0.26 (95 per cent CI: 0.24–0.28) times the risk of dying from Covid-19 compared to the average risk for the US population. With other conditions kept fixed, the man will have estimated 1.4 times the risk compared to an average American if the age is 60, and the risk will be 7.1 times if the age is 75. If the 45-year-old person is a smoker and has chronic heart disease and (controlled) diabetes, the risk will be 0.69 times the average value, whereas it will be 9.8 times the average value if the age is also increased to 75.
Thus, the first 10 per cent of the (remaining) people to be inoculated can be those having top 10 per cent mortality risk (among the remaining people), according to this calculator; the next 10 per cent maybe the next 10 per cent in terms of mortality risk; and so on. Such a model, however, emphasises on the risk of mortality only. As in the Indonesian case, economic activities may also be of importance in the process of inoculation — specially in the quest to bring normalcy, be that a new normal. Thus, the objective function may consider mortality and economy together, and the risk calculator may also take the economic importance of an individual based on profession. The relative weightages of economic activities and mortality maybe a serious issue which the policy-makers should decide and the model can be rebuilt based on the country’s perspective. In fact, a separate model could be constructed for each country — using huge Covid-19-related data, census data and various other healthcare data. And that could certainly provide a real “Big Data moment” for Covid-19.