Models of COVID-19 vaccine prioritisation: a systematic literature search and narrative review

Abstract

Background: How best to prioritise COVID-19 vaccination within and between countries has been a public health and an ethical challenge for decision-makers globally. We systematically reviewed epidemiological and economic modelling evidence on population priority groups to minimise COVID-19 mortality, transmission and morbidity outcomes. Methods: We searched the National Institute of Health iSearch COVID-19 Portfolio (a database of peer-reviewed and pre-print articles), Econlit, the Centre for Economic Policy Research and the National Bureau of Economic Research for mathematical modelling studies evaluating the impact of prioritising COVID-19 vaccination to population target groups. We narratively synthesised the main study conclusions on prioritisation and the conditions under which the conclusions changed. Findings: The search identified 1820 studies. 36 studies met the inclusion criteria and were narratively synthesised. 83% of studies described outcomes in high-income countries. We found that for countries seeking to minimise deaths, prioritising vaccination of senior adults was the optimal strategy and for countries seeking to minimise cases the young were prioritised. There were several exceptions to the main conclusion, notably reductions in deaths could be increased, if groups at high risk of both transmission and death could be further identified. Findings were also sensitive to the level of vaccine coverage. Interpretation: The evidence supports WHO SAGE recommendations on COVID-19 vaccine prioritisation. There is however an evidence gap on optimal prioritisation for low- and middle- income countries, studies that included an economic evaluation, and studies that explore prioritisation strategies if the aim is to reduce overall health burden including morbidity.

COVID-19 Vaccination Strategies in Settings with Limited Rollout Capacity: A Mathematical Modelling Case Study in Sierra Leone

“…Our ndings are consistent with previous modelling studies which support prioritising vulnerable adults (by age or comorbidity) when vaccine rollout is slow, if vaccine effectiveness against acquisition is low, and/or if community transmission is high [27][28][29][30][31][32]. Our study is one of the rst single-country studies of a low-income country based on a recent review of COVID-19 vaccine prioritisation models [4].…”

“…As vaccine supplies in low-and middle-income countries (LMIC) begin to increase, vaccine coverage rates are becoming determined by the capacity of countries to rollout vaccine programmes [3]. COVID-19 vaccine rollouts have been modelled extensively in high-income countries (HIC), however there remains limited published literature in LMIC [4]. Vaccine prioritisation strategies from HIC cannot be generalised to LMIC due to differences in age demographics, contact patterns, and seroprevalence between these settings [5].…”

“…There have been two main approaches to vaccine prioritisation: directly vaccinating those at highest risk of developing severe outcomes, or vaccinating those who contribute most to transmission [4]. Modelling of earlier COVID-19 variants demonstrated that the rst approach was justi ed when supply was low (< 20% population coverage), but the second approach was more bene cial with higher supplies of COVID-19 vaccines [5].…”

Modeling infections and deaths averted due to COVID-19 vaccination strategies in Ghana

“…However, this was not evident in our study as the priority group remained the same for the delta variant with a higher reproduction number. As found in some of our scenario analyses, other studies concluded vaccine optimization also depended on the vaccine supply ( 37, 38 ).…”

Data-driven Targeting of COVID-19 Vaccination Programs: An Analysis of the Evidence on Impact, Implementation, Ethics and Equity

“…(See Appendix A for the full list of 35 of these 55 papers that aren’t cited in this section; see Saadi et al for a more in-depth narrative review of the modeling evidence.) 6 Most studies chose to define COVID-19 vulnerability based on age, which is one of the strongest and most well-known predictors of risk, however age is just one of several criteria that should be considered for targeting. For example, certain underlying medical conditions are associated with a higher risk of hospitalization and/or death from COVID-19.…”