"In the months of March and April, people rarely came. People didn't even come to the hospital anymore, they feared even coughing: 'I'd better not cough, people will say that I have COVID.' So, at the beginning, they were afraid of the stigma." - health worker

Individuals are likely to perceive an incurable, deadly, and highly contagious disease such as COVID-19 as highly threatening, resulting in fear and consequently stigma, especially at the beginning of an outbreak. Health-related stigma can interfere with public health responses by creating barriers to protective behaviours as well as by undermining testing, care-seeking, or treatment. From November through December 2020, Johns Hopkins Center for Communication Programs (CCP), with funding from the United States Agency for International Development (USAID), led qualitative research to explore stigma related to COVID-19 in Abidjan, the largest city and epicentre of the COVID-19 pandemic in Côte d'Ivoire. This paper shares the results of this research, which illustrate how the response to the COVID-19 pandemic evolved over time and how stigma interacted with perceived threat (fear) and efficacy.

The study draws on the Extended Parallel Process Model (EPPM), which is a framework that describes how fear and perceived efficacy interact to shape behaviour. In short, if an individual evaluates an external threat as serious and as posing a risk to them personally, they will likely feel fear. When perceived threat is high, fear is also high, and individuals proceed to evaluate their efficacy to face the threat. "Starkly missing from the EPPM is the role of stigma in shaping responses to fear appeals, yet stigma and fear often go together." Prior research indicates that risk communicators crafting fear appeals need to carefully consider how stigma is likely to interact with fear and perceived efficacy. Furthermore, responses to fear appeals change based on the phase of the pandemic; as the immediate threat wanes, so does the ability to maintain protective behaviours in the longer term.

CCP conducted the study 9 months into the pandemic, 3 months after the end of the first wave and before the second wave, from January to March 2021. After obtaining written informed consent, trained Ivorian qualitative researchers conducted 27 in-depth interviews with people who had recovered from COVID-19, individuals who had lost a family member to COVID-19, and health workers to explore individual experiences with stigma related to COVID-19. The researchers also conducted focus group discussions (FGDs) with members of the general population who knew - and those who did not know - someone who previously had COVID-19 and members who did not know someone who had been infected with COVID-19 to explore social norms and community perceptions related to COVID-19.

Based on the data from these interviews and FGDs, the researchers present a simplified version of the EPPM. In quadrant I, when threat and efficacy are high, the response is to control the danger by adopting behaviours that reduce the threat. Quadrant I represents the optimal combination of threat and efficacy for an effective response to a pandemic like COVID-19. In quadrant II, threat is high, but efficacy is low. Quadrant III represents low threat and high efficacy. In quadrant IV, both threat and efficacy are low.

The data suggest that the first 10 months of the pandemic can be roughly divided into 3 phases, corresponding to movement through 3 quadrants of the simplified EPPM diagram. Each of these phases is described in detail in the article, outlining the levels of fear, efficacy, stigma, and response. In brief, in Phase 1, from March through May 2020, the Ivorian participants largely situated themselves and others in quadrant I: high threat, high efficacy, and appropriate danger control response. In Phase 2, from June to August 2020, participants described moving into quadrant II: a space of high threat, but low efficacy, and adopting a fear control response. In Phase 3, from September to December 2020, by the time CCP collected data, many participants had moved into quadrant IV: a low-threat, low-efficacy space with dropped prevention behaviours. For example, at this point, the COVID-19 vaccine was viewed as unnecessary by many and dangerous by some. Careless words by French researchers who, in April 2020, suggested testing vaccines in Africa reverberated among participants, as did rumours about the nefarious motives behind the vaccine, including suggestions that it would infect Africans with COVID-19, decimate the population, and sterilise Africans. (Post-study, vaccines against COVID-19 became available in Côte d'Ivoire; since then, risk communication efforts have focused on increasing vaccine uptake.)

Participants described that, initially, mass media highlighting the caseload and death counts abroad contributed to high perceived disease threat but that, over time, the perceived threat of COVID-19 decreased, given that cases and deaths did not skyrocket in Côte d'Ivoire as they did elsewhere. The initial shutdown imposed by the government appears to have heightened perceived disease threat; however, as the government relaxed enforcement of barrier measures, perceptions of disease threat diminished along with trust in the government, giving way to narratives denying or minimising the threat of COVID-19 for mainstream Ivorians.

The EPPM suggests that a high (disease) threat and high efficacy lead to an appropriate danger control response, which was observed in these data. However, the EPPM does not consider how the threat of stigma interacts with disease threat and efficacy. The study sheds light on this issue, finding that stigma may be linked to increased fear and decreased efficacy. Both among participants who had previously had COVID-19 and those who had not described how anticipated stigma, or the fear of being stigmatised, prevented them from seeking testing/treatment and adopting prevention measures. Participants who had recovered from COVID-19 were ostracised at home and at work (enacted stigma). Participants who had recovered from COVID-19, health workers, and people who had lost a family member to COVID-19 were also marginalised by the prevailing narrative that COVID-19 did not exist in Côte d'Ivoire.

These findings may have implications for risk communicators and for future research. Namely:

• Risk communicators should anticipate how perceived threat and efficacy may change over time by considering the characteristics of communication around the disease, the disease trajectory, and prevention behaviours. For example, how sustainable are the behaviours in both the short and long term? What is the economic impact of the behaviours? Are the behaviours culturally and socially acceptable? Risk communicators should also anticipate stigma, strive to minimise stigma in the response and prevention efforts, and identify strategies to proactively diffuse stigma.
• Future research should explore whether there is a predictable pattern of movement through the EPPM during a pandemic. In addition, it should explore, through quantitative methods, exactly how threat, stigma, and efficacy interact at different points in time throughout the response to a pandemic.