This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIRx Med, is properly cited. The complete bibliographic information, a link to the original publication on https://med.jmirx.org/, as well as this copyright and license information must be included.
During the COVID-19 crisis, protests against restrictions emerged and rule violations increased, provoking peaks in new positive cases, forcing authorities in France to impose fines to slow down the spread of the disease. Due to these challenges, subsequent implementations of preventive measures in response to COVID-19 recurrences or other pandemics could present difficulties for decision makers. A better understanding of the factors underlying the public acceptance of COVID-19 nonpharmaceutical preventive measures may therefore contribute greatly to the design of more effective public communication during future pandemics.
The aim of this study was to evaluate the acceptance of COVID-19 nonpharmaceutical prevention measures in France. The specific objectives were (1) to examine the public’s acceptance of COVID-19 nonpharmaceutical prevention measures and (2) to assess the association of the public’s acceptance of these prevention measures and their perception of COVID-19.
Data were collected from 2004 individuals through an online survey conducted 6-8 weeks after the first lockdown in France. For objective 1, participants were asked the extent to which they supported 8 COVID-19 nonpharmaceutical preventive measures using a 4-point Likert scale. For objective 2, COVID-19–related perceptions were assessed using a 5-point Likert scale from an adapted version of Witte’s Extended Parallel Process Model. Sociodemographic and environmental variables were also collected. The public’s acceptance factors were estimated using an unweighted least squares factorial analysis, and their associations with perceptions of COVID-19, expressed as rate ratios (RR) and 95% CIs, were estimated using generalized linear Poisson regression models. Statistical analyses were performed using the SPSS statistical package.
The acceptance rate reached 86.1% for individual protective measures, such as making masks mandatory in public open spaces, and 70.0% for collective restrictions, such as isolating the most vulnerable people (1604/2004, 80%) or forbidding public gatherings (n=1590, 79.3%). The least popular restrictions were closing all schools/universities and nonessential commerce such as bars and restaurants (n=1146, 57.2%). Acceptance of collective restrictions was positively associated with their perceived efficacy (RR 1.02, 95% CI 1.01-1.03), fear of COVID-19 (RR 1.04, 95% CI 1.03-1.05), and perceived severity of COVID-19 (RR 1.04, 95% CI 1.03-1.06), and negatively with age >60 years (RR 0.89, 95% CI 0.81-0.98). Acceptance of individual protective measures was associated with their perceived efficacy (RR 1.03, 95% CI 1.03-1.04), fear of COVID-19 (RR 1.02, 1.01-1.03), and perceived severity of COVID-19 (RR 1.03, 1.01-1.05).
Acceptance rates of COVID-19 nonpharmaceutical measures were rather high, but varied according to their perceived social cost, and were more related to collective than personal protection. Nonpharmaceutical measures that minimize social costs while controlling the spread of the disease are more likely to be accepted during pandemics.
The COVID-19 pandemic has affected many countries, with more than 10 million cases worldwide and more than 500,000 deaths as of July 1, 2020 [
The lockdown was lifted in France on May 11, 2020, after a dramatic decrease in the number of cases and deaths, but mobility restrictions had some major adverse consequences [
Beliefs and risk perceptions associated with the disease (perceived personal vulnerability and perceived severity of the disease) have a major influence on the acceptance and uptake of and adherence to required restrictions [
The objectives of this study were (1) to measure the public’s acceptance of COVID-19 nonpharmaceutical measures and (2) to assess the association of the public’s acceptance of these measures and their perception of COVID-19.
Data were collected from a 2-week cross-sectional survey administered 6-8 weeks after the first lockdown (June 25-July 5, 2020) among adults residing in France.
The respondents were recruited among Arcade Research panelists, who agreed to participate regularly in surveys of customer attitudes and experiences. The respondents to this survey were enrolled on the basis of a stratified sampling method to reflect the distribution of the French general population regarding sex, age, occupation, and region.
The research protocol was registered by the École des Hautes Études en Santé Publique (EHESP) School of Public Health Office for Personal Data Protections and approved by the Institutional Review Board of the Méditerranée Infection University Hospital Institute (reference number: 2020-022).
The dependent variable for the analyses was support of the following eight restrictive measures implemented (or likely to be implemented) by national governments to contain the COVID-19 outbreak: (1) make face masks mandatory in public closed spaces; (2) make face masks mandatory in public open spaces; (3) isolate vulnerable people (eg, older adults); (4) forbid public gatherings (eg, fairs, markets); (5) implement mobility restrictions for nonessential workers; (6) introduce a stay-at-home order for nonessential workers; (7) close all schools/universities; and (8) close nonessential commerce (eg, bars, restaurants). For each of them, the participants were asked to rate their acceptance on a Likert-type response scale, which ranged from 1 (“totally disagree”) to 4 (“totally agree”), and for which the meaning of each value was explicitly indicated [
To assess participants’ beliefs and expectations related to the COVID-19 epidemic, we used a range of constructs and variables from Witte’s EPPM. Items related to these constructs were adapted to the COVID-19 pandemic and translated into French. EPPM factors were estimated using an unweighted least squares factorial analysis, followed by a Promax rotation, and five factors were extracted accordingly [
Sociodemographic and environmental variables were also collected, such as age in years (divided into groups: 18-39 years, 40-59 years, and ≥60 years), gender (self-reported sex), occupational status (active, unemployed, or retired), persons in household (≥3, 2, or 1), living density (urban, more than 100,000 people; urban, 20,000-100,000 people; urban, 2000-20,000 people; rural), chronic disease (yes/no), and perceived health (very poor, poor, good, very good).
Categorical data were expressed as frequencies (n) and percentages (%), while numerical data were expressed as mean (SD), and compared with 1-way ANOVA. EPPM raw scale scores were transformed to a 0-100 scale: ([raw score – lowest possible raw score]/possible raw score range) × 100. Acceptance factors were estimated using an unweighted least squares factorial analysis, followed by a Promax rotation, a nonorthogonal (oblique) solution in which the factors are allowed to be correlated. This method provides accurate and conservative parameter estimates when using ordinal data [
Of the 2004 individuals who completed the survey (
More than 1 in 5 participants (404/2004, 20.5%) reported financial difficulties related to COVID-19, and 3 in 10 had a chronic disease (n=615, 30.7%). Nearly 9 in 10 respondents (n=1796, 89.6%) perceived their health state as “good” or “very good.”
Participants’ characteristics (N=2004).
Variables | Values | ||
|
|||
|
Male | 992 (49.5) | |
|
Female | 1012 (50.5) | |
|
|||
|
≥60 | 518 (25.8) | |
|
40-59 | 750 (37.1) | |
|
18-39 | 736 (36.7) | |
|
|||
|
Active | 1329 (66.3) | |
|
Retired | 427 (21.3) | |
|
Unemployed | 248 (12.4) | |
|
|||
|
≥3 | 825 (41.2) | |
|
2 | 723 (36.1) | |
|
1 | 456 (22.8) | |
|
|||
|
Urban, more than 100,000 people | 385 (19.2) | |
|
Urban, 20,000-100,000 people | 520 (25.9) | |
|
Urban, 2000-20,000 people | 627 (31.3) | |
|
Rural zone | 472 (23.6) | |
Chronic disease, n (%) | 615 (30.7) | ||
|
|||
|
Poor/very poor | 208 (10.4) | |
|
Good/very good | 1796 (89.6) | |
|
|||
|
Yes, related to COVID-19 | 404 (20.2) | |
|
Yes, unrelated to COVID-19 | 480 (24) | |
|
None | 1120 (55.9) | |
|
|||
|
Efficacy | 73.8 (17.4) | |
|
Fear control | 54.5 (26) | |
|
Severity | 73.5 (23.1) | |
|
Vulnerability | 42.7 (22.4) | |
|
Avoidance | 48.9 (22.9) |
aEPPM: Extended Parallel Process Model.
The majority of the study population approved of all 8 proposed measures (
Unweighted least squares exploratory factorial analysis, followed by a Promax rotation, was performed on the 8 items. Eigenvalues for the first 3 factors were 4.58, 1.05, and 0.63, respectively; this suggested a 2-factor solution explaining 62.5% of the common variance of the data. Factor 1 included 6 items related to collective restrictions and was interpreted as expressing acceptance of collective restrictions, whereas factor 2 included the 2 items related to mandatory mask wearing and was interpreted as expressing acceptance of individual protective measures. The factors showed satisfactory internal validity (Cronbach α was 0.88 for factor 1 and 0.87 for factor 2). The interscale correlation coefficient (
Regarding COVID-19 perceptions, as assessed by the EPPM, efficacy (mean 73.8, SD 17.4) and severity (mean 73.5.1, 23.1) had the highest scores on a 100-point response scale, followed by lack of fear control (mean 54.5, SD 26.0), cognitive avoidance (mean 48.8, SD 22.9), and perceived vulnerability (mean 42.8, SD 22.4). Differences between T-scores were significant, except for efficacy and severity.
Numbers, percentages, and factor loadings for the 2-factor solution of the acceptance of 8 nonpharmaceutical COVID-19 measures (N=2004).
Item | Totally agree/agree, n (%) | Totally disagree/disagree, n (%) | Factors | |
|
|
|
F1 | F2 |
Make mask mandatory in public closed spaces | 1783 (89) | 221 (11) | N/Aa | 0.95 |
Make mask mandatory in public open spaces | 1667 (83.2) | 337 (16.8) | N/A | 0.81 |
Isolate vulnerable people (eg, older adults) | 1604 (80) | 400 (20) | 0.56 | N/A |
Forbid mass gatherings (eg, fairs, markets) | 1590 (79.3) | 414 (20.7) | 0.59 | N/A |
Mobility restrictions for nonessential workers | 1482 (74) | 522 (26) | 0.74 | N/A |
Stay at home order for nonessential workers | 1314 (65.6) | 690 (34.4) | 0.85 | N/A |
Close all schools/universities | 1286 (64.2) | 718 (35.8) | 0.80 | N/A |
Close nonessential commerce (eg, bar, restaurant) | 1146 (57.2) | 858 (42.8) | 0.82 | N/A |
Eigenvalue | N/A | N/A | 4.58 | 1.05 |
Percentage of explained variance | N/A | N/A | 52.6 | 9.9 |
Cronbach α | N/A | N/A | 0.88 | 0.87 |
aN/A: not applicable.
Respondents (N=2004) agreeing with proposed collective COVID-19 nonpharmaceutical prevention measures.
Number of measures accepted | Respondents, n (%) |
0 | 122 (6.1) |
1 | 149 (7.4) |
2 | 186 (9.3) |
3 | 209 (10.4) |
4 | 239 (11.9) |
5 | 276 (13.8) |
6 | 823 (41.1) |
Estimate of acceptance of collective restrictions in univariate analysis (
Estimate of acceptance of individual protective measures in univariate analysis (
Rate ratios and 95% CIs of the acceptance of collective restrictions (N=2004), Poisson regression.a
Variables | Univariate, rate ratio (95% CI) | Multivariateb, rate ratio (95% CI) | |||
|
|||||
|
Female | 1.03 (0.98-1.07) | N/Ac | ||
|
Male | 1 | N/A | ||
|
|||||
|
≥60 |
|
|
||
|
40-59 | 0.97 (0.92-1.02) | 0.96 (0.91-1.01) | ||
|
18-39 | 1 | 1 | ||
|
|||||
|
Active | 1.01 (0.94-1.07) | 1.02 (0.96-1.09) | ||
|
Retired |
|
0.98 (0.88-1.09) | ||
|
Unemployed | 1 | 1 | ||
|
|||||
|
Urban, more than 100,000 | 1.00 (0.94-1.07) | N/A | ||
|
Urban, 20,000-100,000 | 1.04 (0.98-1.10) | N/A | ||
|
Urban, 2000-20,000 | 1.04 (0.98-1.10) | N/A | ||
|
Rural zone | 1 | N/A | ||
|
|||||
|
≥3 |
|
1.04 (0.99-1.11) | ||
|
2 | 1.03 (0.97-1.09) | 1.03 (0.97-1.09) | ||
|
1 | 1 | 1 | ||
Chronic disease | 1.00 (0.95-1.05) | N/A | |||
|
|||||
|
Poor/very poor | 0.96 (0.89-1.03) | N/A | ||
|
Good/very good | 1 | N/A | ||
|
|||||
|
Yes, related to covid | 1.07 (1.02-1.13) | N/A | ||
|
Yes, unrelated to covid | 1.01 (0.96-1.07) | N/A | ||
|
None | 1 | N/A | ||
|
|||||
|
Efficacy |
|
|
||
|
Lack of fear control |
|
|
||
|
Severity |
|
|
||
|
Vulnerability |
|
1.01 (0.99-1.02) | ||
|
Avoidance |
|
1.00 (0.99-1.02) |
aSignificant results (
bGoodness of fit for the multivariate model (value/df for the deviance)=1.08.
cN/A: not applicable.
dEPPM: Extended Parallel Process Model.
Rate ratios and 95% CIs of the acceptance of individual protective measures (N=2004), Poisson regression.a
Variables | Univariate, rate ratio (95% CI) | Multivariateb, rate ratio (95% CI) | |||
|
|||||
|
Female | 1.04 (0.97-1.11) | N/Ac | ||
|
Male | 1 | N/A | ||
|
|||||
|
≥60 | 1.08 (0.99-1.17) | N/A | ||
|
40-59 | 1.04 (0.96-1.12) | N/A | ||
|
18-39 | 1 | N/A | ||
|
|||||
|
Active | 1.02 (0.92-1.13) | N/A | ||
|
Retired | 1.09 (0.97-1.23) | N/A | ||
|
Unemployed | 1 | N/A | ||
|
|||||
|
Urban, more than 100,000 people | 0.95 (0.86-1.06) | N/A | ||
|
Urban, 20,000-100,000 people | 0.99 (0.90-1.09) | N/A | ||
|
Urban, 2000-20,000 people | 1.01 (0.92-1.10) | N/A | ||
|
Rural zone | 1 | N/A | ||
|
|||||
|
≥3 | 1.04 (0.95-1.14) | N/A | ||
|
2 | 1.04 (0.95-1.14) | N/A | ||
|
1 | 1 | N/A | ||
Chronic disease | 1.07 (0.99-1.15) | N/A | |||
|
|||||
|
Poor/very poor | 0.96 (0.86-1.07) | N/A | ||
|
Good/very good | 1 | N/A | ||
|
|||||
|
Yes, related to covid | 0.98 (0.90-1.07) | N/A | ||
|
Yes, unrelated to covid | 1.01 (0.963-1.09) | N/A | ||
|
None | 1 | N/A | ||
|
|||||
|
Efficacy |
|
|
||
|
Lack of fear control |
|
|
||
|
Severity |
|
|
||
|
Vulnerability |
|
1.00 (0.98-1.02) | ||
|
Avoidance | 1.00 (0.98-1.02) | N/A |
aSignificant results are marked in italics.
bGoodness of fit for the multivariate model (value/df for the deviance)=0.34.
cN/A: not applicable.
dEPPM: Extended Parallel Process Model.
Acceptance rates in our study population reached, on average, 86.1% for individual protective measures (such as mandatory face mask wearing), and 74.0% for collective restrictions, such as isolate vulnerable people (80%), forbid public gatherings (79.3%), and mobility restrictions for nonessential workers (74.0%). The least popular restrictions were closing of nonessential commerce such as bars and restaurants (57.2%). Acceptance of collective restrictions was positively associated with the level of efficacy, fear, and perceived severity, and negatively with age older than 60 years. Acceptance of individual protective measures was associated with level of efficacy, fear, and perceived severity.
Data were collected after the first lockdown in France, in a period when COVID-19 cases and deaths were minimal. Most restrictions implemented to help combat COVID-19 have been lifted; although strict hygiene and social distancing methods remained in place, life returned to some level of normality. However, global and local health authorities continued to use various media to inform the public about the epidemic and to promote a range of health protective behaviors to prevent infections [
Although individual protective measures were rather consensual in our study population, collective restrictions had more mixed acceptance rates—ranging from 80%-57%. One possible explanation is that these measures were assessed in light of their restrictive nature [
The relationship observed between vulnerability and acceptance of collective and individual protective measures became nonsignificant when entered together with efficacy, lack of fear control, and perceived severity in the multivariate models. This indicates that the acceptance of collective restrictions was more related to collective than personal protection, likely to protect others [
The results of this study must be viewed in light of its main limitations. First, the cross-sectional design does not allow causal inferences about relationships between variables to be determined. Furthermore, missing data precluded the investigation of EPPM appraisal in the total study sample, and some novel measures such as “location tracking” [
The aim of this study was to evaluate the acceptance of COVID-19 nonpharmaceutical measures and, more specifically, to measure the public’s acceptance of these measures and their association with COVID-19 perceptions. Our findings suggest that acceptance rates of COVID-19 nonpharmaceutical measures were rather high, but varied according to their perceived social costs, and seemed to be more related to collective than personal protection. Altogether, it appears that the nonpharmaceutical measures that minimize social costs while controlling the spread of the disease are more likely to be accepted and therefore more sustainable during pandemics.
Extended Parallel Process Model
rate ratio
The data underlying this article are available in Open Science Framework [
AC, DC, KGM, and JR contributed to the conception and design of the study and interpreted the data and drafted the final manuscript. KGM suggested the theoretical framework. AC performed the statistical analysis and wrote the first draft of the manuscript. All authors read and approved the manuscript.
None declared.