Article Text

Download PDFPDF

Allocation of COVID-19 vaccination: when public prioritisation preferences differ from official regulations
Free
  1. Philipp Sprengholz1,
  2. Lars Korn1,2,
  3. Sarah Eitze1,2,
  4. Cornelia Betsch1,2
  1. 1Media and Communication Science, University of Erfurt, Erfurt, Germany
  2. 2Center for Empirical Research in Economics and Behavioral Sciences, University of Erfurt, Erfurt, Germany
  1. Correspondence to Philipp Sprengholz, Media and Communication Science, University of Erfurt, 99089 Erfurt, Germany; philipp.sprengholz{at}uni-erfurt.de

Abstract

As vaccines against COVID-19 are scarce, many countries have developed vaccination prioritisation strategies focusing on ethical and epidemiological considerations. However, public acceptance of such strategies should be monitored to ensure successful implementation. In an experiment with N=1379 German participants, we investigated whether the public’s vaccination allocation preferences matched the prioritisation strategy approved by the German government. Results revealed different allocations. While the government had top-prioritised vulnerable people (being of high age or accommodated in nursing homes for the elderly), participants preferred exclusive allocation of the first available vaccines to medical staff and personnel caring for the elderly. Interestingly, allocation preferences did not change when participants were told how many individuals were included in each group. As differences between allocation policies and public preferences can affect trust in the government and threaten the social contract between generations, we discuss possible strategies to align vaccination prioritisations.

  • COVID-19
  • resource allocation
  • public health ethics

Data availability statement

Materials, data, and data analysis scripts are available at https://doi.org/10.17605/OSF.IO/CKHBA.

This article is made freely available for use in accordance with BMJ’s website terms and conditions for the duration of the covid-19 pandemic or until otherwise determined by BMJ. You may use, download and print the article for any lawful, non-commercial purpose (including text and data mining) provided that all copyright notices and trade marks are retained.

https://bmj.com/coronavirus/usage

Statistics from Altmetric.com

The approval of the first vaccines against COVID-19 in late 2020 marked an important milestone in the fight against a pandemic that had, at that point, taken nearly 2 million lives.1 However, because of limited production capacities, it will take months or even years before everyone can be vaccinated. This elicited discussions on how the low number of vaccines should be allocated,2–4 leading to country-specific prioritisation strategies.5 For instance, in Germany, the Standing Committee on Vaccination, the German Ethics Council and the German National Academy of Sciences Leopoldina considered four allocation objectives: (1) the distribution should reduce severe and deadly infections, (2) protect people with high exposure due to their work, (3) prevent transmission and (4) sustain public life.6 Based on these objectives, a prioritisation strategy was developed assigning groups of people to six different priority levels (table 1). Individuals assigned to level 1 would be offered vaccination before people at level 2, and so on. No prioritisation within the levels was made.

Table 1

German prioritisation strategy16

While development of the German strategy had been informed by ethical and epidemiological considerations, public acceptance of the prioritisation strategy was not taken into account. As public support is vital for the successful implementation of resource allocations,7 we examined the public support for the four allocation objectives and conducted an experiment to investigate whether public prioritisation preferences match the official strategy. We further explored whether the communication of group sizes affected prioritisation preferences. As table 1 reveals, over 37 million people will be prioritised and receive two doses of the vaccines before adults under 60 years with no specific risk factors become eligible for vaccination. Thus, understanding the group sizes may lead to a more precise estimate of personal vaccination delays and, consequently, to a change in prioritisation preferences. Because we began our research in the preliminary stages of vaccinations, the experiment focused only on those groups assigned to priority levels 1 and 2.

Figure 1

Public vaccination prioritisation. The Plackett-Luce model estimates the prioritisation for each group. Log-transformation was applied to compare the priority of each group with a baseline (residents of nursing homes for the elderly). Positive values indicate higher priorities compared with the baseline, while negative values indicate lower prioritisation. Error bars displaying quasi-standard errors indicated three priority levels: participants top-prioritised staff in medical facilities and caregivers for the elderly (green area; 1), followed by medium priorities for elderly residents and other staff in nursing homes as well as caregivers for people with dementia or mental disabilities (grey area; 2). Individuals aged 75 years and older as well as people with dementia and mental disabilities were assigned lower priorities compared with the baseline (red area; 3). Communicating group sizes did not affect prioritisation results.

Method

The experiment was conducted as part of the cross-sectional COVID-19 Snapshot Monitoring study series.8 From 8 December to 10 December 2020, N=1379 individuals from a German non-probabilistic sample, quota-representative for age×gender and federal state, were randomly assigned to a two-factorial (communication of group sizes vs no communication) between-subjects design. Since data were collected from an online panel, only individuals below 75 years were recruited. Participants were 18–74 years old (M=45.39, SD=15.68), with 693 men and 686 women.

Before the experiment, participants were asked how important they found each of the four allocation objectives. Answers were rated on scales ranging from 1 (not important at all) to 7 (very important). Afterwards, participants were randomly assigned to one of the two conditions. In both conditions, they were told that vaccines against COVID-19 were scarce and that some groups should be prioritised for vaccination. The groups assigned to priority levels 1 and 2 (except for people with Down syndrome as they were added to the prioritisation strategy after the study took place) were presented in random order and participants were asked to rank them by priority. While participants in the no communication condition received no further information, individuals in the group size communication condition were given any available information about the size of the groups. To put the numbers into perspective, they were also told that the vaccine supply should be sufficient to vaccinate 5.5 million people in the first quarter of 2021.1

Results

An analysis of variance comparing average importance ratings for the four allocation objectives revealed significant differences, F(3, 5512)=13.48, p<0.001. While importance was generally high for all four objectives, post-hoc Tukey tests showed that the reduction of severe and deadly infections (M=5.89, SD=1.57) and the protection of people with high exposure due to their work (M=5.90, SD=1.56) were rated as slightly more important than preventing transmission (M=5.70, SD=1.61) and sustaining public life (M=5.57, SD=1.66).1

A Plackett-Luce model9 was employed to estimate the average vaccination priority of each person group from participants’ rankings.1 As residents of nursing homes for the elderly had the highest mortality rates,10 11 we used this group as a reference. Figure 1 shows which groups were prioritised equal (grey area; 2), higher (green area; 1) or lower (red area; 3) than the reference group. Staff in medical facilities and caregivers in outpatient care and nursing homes for the elderly were prioritised over vulnerable groups (such as residents of nursing homes for the elderly and people aged 75 years and older) and other workers in elderly nursing homes. People with dementia or mental disabilities were assigned the lowest vaccination priority among the groups included in the experiment. Prioritisation results did not differ between communication conditions; there was no change when participants were told how large the groups were. Further exploratory analyses revealed that prioritisation results did not differ between participants below and above 60 years. 1

Discussion

While scholars have emphasised the importance of public acceptance towards vaccination prioritisation strategies,2 3 little research on public preferences has been conducted. In Germany, the data show that there is strong support for the official allocation objectives. However, when asked to prioritise certain groups for vaccination, public preferences seemed to differ from the official vaccination strategy. Whereas both the government and survey respondents top-prioritised medical staff with very high exposure risk and contact to vulnerable groups as well as caregivers in outpatient care and nursing homes for the elderly, results indicated that participants do not agree with assigning people 80 years and older or residents of elderly nursing homes an equal priority. In the public opinion, age-related vulnerability alone does not seem to be sufficient for prioritisation. Since all study participants were younger than 75 years, prioritisation preferences of populations above that age could differ. Within the sample at hand, however, prioritisation preferences did not vary with age, and previous research indicates that age alone has limited influence on older peoples’ views of resource allocation.12

With the exception of non-caregivers in nursing homes for the elderly, those caring for the vulnerable were assigned higher vaccination priorities than the vulnerable themselves. Similar prioritisation preferences have been found in the USA.13 This may reflect the fact that vaccination of both groups suffices different objectives. While vaccinating the vulnerable primarily reduces morbidity and mortality, the immunisation of medical personnel not only protects those persons from infection, but also sustains the medical infrastructures necessary to control the pandemic. Furthermore, as recent research suggests that vaccination limits transmission of the coronavirus,14 vaccinating medical personal with many contacts is likely to protect much more people than vaccinating the vulnerable. Participants may also have assumed that vaccinating a 50-year-old doctor saves more life years than vaccinating someone 80 years or older. However, this notion is not necessarily supported by epidemiological models and ethical assessments.2 15 It should be noted that participants were only asked to prioritise groups assigned to the official priority levels 1 and 2. When groups from lower levels, such as individuals with pre-existing conditions or teachers, are included in the ranking task as well, older and disabled people may be further moved down the prioritisation order.

Interestingly, the communication of group sizes did not affect prioritisation. As ethical allocation decisions should not be based on group sizes (even if only parts of a large and more vulnerable or exposed group can be vaccinated, smaller and less vulnerable or exposed groups that could be vaccinated entirely should not be preferred), this result reflects the moral integrity of public allocation preferences. However, not finding a difference between the two experimental conditions may be due to comparable group size knowledge. Most participants should have known that the groups comprising the elderly are much larger than those including medical and care personnel (table 1). This knowledge may have affected the allocation decisions of all participants and could explain the lower prioritisation of larger, older groups compared with the official vaccination strategy.

As previously noted, decision makers should pay attention when differences between public vaccination allocation preferences and prioritisation policies emerge. If public concerns about vaccination allocation are not addressed properly, trust in the government and science can wane, ultimately threatening the social contract between generations. To align public preferences and the official allocation strategy, two approaches can be employed. First, policies can be revised; just as new epidemiological insights are incorporated into the prioritisation strategy (eg, people with Down syndrome were added to priority level 2 as new findings showed that the condition is associated with an elevated risk of mortality), public preferences could be considered as well. Prioritisation and actual implementation of vaccination should also align with formal policies to increase public trust. Second, if epidemiological insights and ethical considerations speak against alterations of the allocation strategy, they should be communicated to the public rapidly, clearly and in an easy-to-understand way. When people are given the reasons behind a specific prioritisation plan, public support for the allocation strategy is likely to increase. In conclusion, monitoring public policy preferences as presented in this article enables decision makers to design accepted, thus efficient, policies.

Data availability statement

Materials, data, and data analysis scripts are available at https://doi.org/10.17605/OSF.IO/CKHBA.

Ethics statements

Ethics approval

The research obtained ethical clearance from the University of Erfurt’s Institutional Review Board (#20200302/20200501) and all participants provided informed consent prior to data collection.

Acknowledgments

The study was conducted as part of Germany’s COVID-19 Snapshot Monitoring (COSMO), a joint project of the University of Erfurt (Cornelia Betsch (principal investigator), Lars Korn, Philipp Sprengholz, Philipp Schmid, Lisa Felgendreff, Sarah Eitze), the Robert Koch Institute (RKI; Lothar H. Wieler, Patrick Schmich), the Federal Centre for Health Education (BZgA; Heidrun Thaiss, Freia De Bock), the Leibniz Institute of Psychology (ZPID; Michael Bosnjak), the Science Media Center (SMC; Volker Stollorz), the Bernhard Nocht Institute for Tropical Medicine (BNITM; Michael Ramharter) and the Yale Institute for Global Health (Saad Omer).

References

Footnotes

  • Contributors PS, LK, SE and CB designed the research. PS, LK and SE performed the research. PS planned and performed data analysis. PS wrote the initial draft, which was revised and approved by all authors.

  • Funding This work was supported by German Research Foundation (BE3970/12-1), Federal Centre for Health Education, Robert Koch Institute, Leibniz Institute of Psychology, Klaus Tschira Foundation, Thüringer Ministerium für Wirtschaft, Wissenschaft und digitale Gesellschaft, Thüringer Staatskanzlei and University of Erfurt (no award/grant numbers).

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Materials, data, and data analysis scripts are available at https://doi.org/10.17605/OSF.IO/CKHBA.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.