Investment in research and programs to discover and apply the principles that underpin sustained behavior change is needed to address the continuing threat from COVID-19 and future pandemics and will require collaboration among behavioral, social, biomedical, public-health and clinical scientists.
Human behavior was instrumental in causing COVID-19, and changing it has been vital to tackling this pandemic. The countries that have done best in mitigating COVID-19’s harms to health and to their economies have rapidly and successfully persuaded their populations to enact large-scale behavior change. Some of these interventions have been highly effective, others have been less so, and some have produced substantial social and financial harm. In particular, national ‘lockdowns’ have been effective in keeping people from interacting, to reduce the spread of disease1, but they have been very damaging to people’s lives and to national economies. Therefore, lockdowns should ideally be used only to bring transmission levels low enough to be controllable by other policies. These policies include adequate ‘find, test, trace, isolate and support’ systems2; border controls and quarantine to prevent reseeding of infections; the creation of safe working, domestic and transport spaces; and the promotion of personal protective behaviors such as the use of face coverings3.
Capability, opportunity and motivation
Large-scale, sustained behavior change is needed to reduce the risk of, and to prepare for, future pandemics4. The COVID-19 pandemic has shown that populations will adopt at least some of the required behaviors under certain conditions5. However, adoption has been variable across countries, over time and across social groups5.
Achieving sustained behavior change requires a sound understanding by policymakers and intervention designers of what underpins the behaviors concerned. For example, what does it take in all cultures to ensure that, where appropriate, people keep safe physical distances from each other, wear face coverings masks and disinfect their hands? What is needed to ensure that adequate ventilation is provided in enclosed spaces, and that people in high-risk settings use personal protective equipment effectively?
We provide here important behavioral targets for the prevention and mitigation of pandemics such as COVID-19 (Table 1). The behaviors vary in what is needed to enact them and the contextual influences on them.
A simple, comprehensive way of understanding behavior is through the ‘capability–opportunity–motivation–behavior’ (COM-B) model6,7 (Fig. 1). This model takes as its premise that for any behavior to occur, people must have the following: the relevant physical and psychological capabilities; the opportunity; and greater motivation to perform that behavior than anything else they could be doing at the time. Capability and opportunity feed into motivation, so that having greater capability and opportunity can increase motivation (Fig. 1). Behavior then feeds back, and so changing behavior can, in turn, influence capability, opportunity and/or motivation. Thus, behavior is part of an evolving and interacting system. Creating sustained behavior change requires changes to capability, opportunity and motivation that are mutually reinforcing.
Physical capabilities include strength, stamina, and physical and psychomotor skills, whereas psychological capabilities include knowledge, resilience, and mental skills. Motivation comprises both reflective processes, such as self-conscious decision-making, and automatic processes that involve emotions and habits. Opportunity involves having the resources, time and physical space to enact a behavior. It also involves having social structures, norms and cues that make the behavior possible or promote it. The World Health Organization has used the COM-B model to make recommendations for improving adherence to rules and guidance on behaviors that are protective against COVID-198.
Requirements for behavior change
The development of effective interventions requires an understanding of what changes in capability, motivation or opportunity, or their combination, will drive behavior change. Getting the balance wrong can be disastrous. For example, in 2020, the UK government largely failed in its attempts to get people with COVID-19 to self-isolate when needed, in part because it did not heed this lesson. Survey data showed that fewer than 50% of people with COVID-19 symptoms reported staying at home for the required isolation period9. Failure to self-isolate appeared to be driven largely by financial need or domestic or caring responsibilities—i.e., a lack of opportunity9. People on low incomes were just as motivated to self-isolate when needed as were those with more financial resources, but did so less10. Those who found it easiest to isolate were those who could work from large homes and had the funds to order in meals and other services. Adequate financial and practical support should have been a key part of the solution, and it has been a cornerstone of effective pandemic control in many countries, with up to 100% of lost wages being reimbursed, and with the provision of free alternative accommodation and daily visits. In the USA, introduction in some jurisdictions of emergency statutory entitlement to sick leave for people with symptoms or who had to quarantine was found to be effective in reducing cases11.
The UK government appeared to have erroneously assumed that the main reason people did not isolate was a lack of motivation, whereas it was instead a lack of opportunity. It threatened a fine of £10,000 for failure to comply with isolation mandates. An evaluation of free, mass testing in the city of Liverpool found that the take up of diagnostic tests among disadvantaged groups was as low as 4% in some areas12, and a £36 billion ‘test, trace and isolate system’ was estimated to have reached as little as 3% of contacts of people who had tested positive for COVID-1913. It seems that the threat of a large monetary fine did have a motivational influence, but it was opposite to that intended, with the best way of avoiding a fine being to avoid a test.
The behavior change wheel
The UK case study described above illustrates the fact that for mass behavioral changes to achieve their intended outcome, systematic approaches need to move from an understanding of behavior in its context to identifying interventions and policies that can change the behavior. One such framework is the behavior change wheel, developed through the integration of 19 frameworks identified in a literature review7,14,15 (Fig. 2). This wheel can be used to link the COM-B model with nine possible types of intervention and seven policy options for implementation. Intervention designers can use the wheel to identify the intervention types most likely to achieve the desired change and can then choose the most appropriate mixture of implementation options, given whatever constraints may be operating.
Use of the behavior change wheel framework can be illustrated by the challenge of getting healthcare staff to adhere to infection prevention and control guidelines. A published study identified a range of interventions that could be effective16, but it was important to go further and consider the capability, opportunity and motivation issues that needed to be addressed in a given context. First, in terms of capability, there was the nature of the guidelines, how they were communicated and what training was needed for this. Second, in the context of opportunity, it was essential to consider the physical spaces the staff worked in and the support they received from managers16. Third, for motivation, it was important that staff had trust in their personal protective equipment and that it was comfortable to use as required. Finally, to fully address the capability, opportunity and motivational issues, it was essential to include all staff, including support staff, in the process of intervention development and implementation15. The findings of this study further illustrated the need to treat all the actors and behaviors in a given scenario as part of an interconnected system, and pointed to workplace culture as an important influence, specifically managerial and peer support, social-norms solidarity and a focus on safety rather than complacency16.
Leadership, trust and solidarity
When it comes to motivation to adhere to stringent pandemic control regulations, variation in practices across different countries has highlighted the importance of good leadership fostering public trust in government and experts, and a sense of solidarity17. Trust derives from people seeing authorities as part of and serving the community18, and this in turn derives from these authorities treating the public as a respected partner rather than as a problem19.
The importance of fostering trust and solidarity has informed much of the advice provided by behavioral scientists participating in the UK government’s Scientific Advisory Group for Emergencies. This group has urged the government to understand the diversity of communities that make up the UK population, and to engage with, consult, listen to and learn from them. In addition, the government was urged to support and enable the population to adhere; to avoid blame and punishment; to recognize, reward and celebrate achievements, often in challenging circumstances; and to communicate clearly, consistently and concisely, in line with evidence. Finally, the government was advised to be transparent, honest and open in written and verbal communication; to engender trust; and to avoid socially divisive language, policies and behavior.
Sustaining behavior change involves taking these principles and applying them to create a mutually reinforcing system in which capability, opportunity and motivation support each other (Table 2). For example, sustained improvements in knowledge will need education about viral transmission and its prevention to be built into educational programs at all levels. Improved opportunity will require structural changes to inhabited spaces to make it easier to achieve social distancing when required. Sustained changes to motivation will involve the development of social identities that place greater value on behaviors that mitigate infection risk and the development of habits and routines that embed these behaviors into everyday life.
The principles noted above are generic. Detailed scientific investigation is needed to develop and successfully apply them to create interventions suited to different scenarios. Such interventions have been called ‘behavioral, environmental, social and systems interventions’ (BESSIs)4. Despite the importance of these kinds of intervention, research investment in all BESSIs collectively has been estimated to be less than 4% of the $3.3 billion in global spending on COVID-19 research20. In a comparison of BESSI versus pharmacological trials, the contrast is stark: as of December 2020, there were 1,725 registered pharmacological trials and 213 conducted pharmacological trials, but only 11 registered BESSI trials and 3 conducted BESSI trials21.
An important lesson from the COVID-19 pandemic is that no person or community or country can protect themselves or itself on their or its own: humans are all interconnected, and solutions must be global. Building pandemic-resilient societies for the future must include changing population-wide behaviors, as well as reducing inequalities. Embedding behaviors into societies requires changes to physical and social environments, such as fostering safe workplaces and norms around hygiene behaviors. These need to be underpinned by community activation, organization regulation and national legislation.
Societies that are already unequal have been made more so by the COVID-19 pandemic, with the virus spreading most aggressively in overcrowded, multigenerational housing and poorly regulated workplaces. Furthermore, those with pre-existing health problems and those from specific ethnic backgrounds, both linked to poverty, were hit harder by the virus. Structural racism in many countries meant that those who were already more vulnerable were made doubly so due to low incomes and poor working conditions. On top of these increased inequalities came government policies that exacerbated inequalities further in many countries. In the UK, workers on regular salaries were more financially protected than those working in the precarious economy, and mortgage-holders were better protected than renters.
Testing for the virus, as well as medicines and vaccines, will continue to be vital to the strategy for tackling COVID-19 and future pandemics. But a substantial rebalancing of funding on research and programs needed to embed behavioral changes across society and address inequality will also be critical in the long run.
Flaxman, S. et al. Nature 584, 257–261 (2020).
Rajan, S., Cylus, J. & McKee, M. Eurohealth 26, 34–39 (2020).
West, R., Michie, S., Rubin, G. J. & Amlôt, R. Nature Human Behaviour 4, 451–459 (2020).
Michie, S. & West, R. Br. Med. J. 370, m2982 (2020).
Becher, M., Stegmueller, D., Brouard, S. & Kerrouche, E. Preprint at https://doi.org/10.2139/ssrn.3652543 (2020).
West, R. & Michie, S. Qeios https://doi.org/10.32388/WW04E6.2 (2020).
Michie, S., van Stralen, M. M. & West, R. Implement. Sci. 6, 42 (2011).
World Health Organization. Regional Office for Europe. https://apps.who.int/iris/handle/10665/337574 (November 2020).
Smith, L. E. et al. Brit. Med. J. 372, n608 (2021).
Atchison, C. J. et al. BMJ Open 11, e043577 (2021).
Pichler, S., Wen, K. & Ziebarth, N. R. Health Aff. 39, 2197–2204 (2020).
University of Liverpool. https://www.liverpool.ac.uk/research/news/articles/covid-19-liverpool-community-testing-pilot-interim-findings-published/ (23 December 2020)
Wood, P. CityAM https://www.cityam.com/dido-harding-slammed-for-three-per-cent-test-and-trace-success-rate/ (10 November 2020).
West, R., Michie, S., Chadwick, P., Atkins, L. & Lorencatto, F. Public Health England https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/933328/UFG_National_Guide_v04.00__1___1_.pdf (2020).
Michie, S., Atkins, L. & West, R. The Behaviour Change Wheel: A Guide To Designing Interventions (Silverback Publishing, 2014).
Houghton, C. et al. Cochrane Database Syst. Rev. https://doi.org/10.1002/14651858.CD013582 (2020).
Wright, L., Steptoe, A. & Fancourt, D. Preprint at medRxiv https://doi.org/10.1101/2020.10.19.20215376 (2020).
Brewer, M.B. in Rationality and Social Responsibility: Essays in Honor of Robyn Mason Dawes 215–232 (Psychology Press, 2008).
Tyler, T. R. & Blader, S. L. Pers. Soc. Psychol. Rev. 7, 349–361 (2003).
Head, M. G. et al. Lancet Glob. Health 8, e1295–e1304 (2020).
BESSI. https://www.bessi-collab.net (accessed 23 April 2021).
S.M. and R.W. are unpaid directors of the Unlocking Behavior Change community interest company, and are participants in the Independent Scientific Pandemic Insights Group on Behaviours of the UK Scientific Advisory Group for Emergencies, and in the Independent Scientific Advisory Group for Emergencies.
About this article
Cite this article
Michie, S., West, R. Sustained behavior change is key to preventing and tackling future pandemics. Nat Med 27, 749–752 (2021). https://doi.org/10.1038/s41591-021-01345-2
This article is cited by
Objective risk assessment vs standard care for acute coronary syndromes—The Australian GRACE Risk tool Implementation Study (AGRIS): a process evaluation
BMC Health Services Research (2022)
Scientific Reports (2022)
Development and cross-national investigation of a model explaining participation in WHO-recommended and placebo behaviours to prevent COVID-19 infection
Scientific Reports (2022)
Longitudinal behavioral changes and factors related to reinforced risk aversion behavior among patients with chronic kidney disease during the COVID-19 pandemic
Scientific Reports (2022)