Health and social behaviour through pandemic phases in Switzerland: Regional time-trends of the COVID-19 Social Monitor panel study

André Moser; Viktor von Wyl; Marc Höglinger

doi:10.1371/journal.pone.0256253

. 2021 Aug 25;16(8):e0256253. doi: 10.1371/journal.pone.0256253

Health and social behaviour through pandemic phases in Switzerland: Regional time-trends of the COVID-19 Social Monitor panel study

André Moser ^1,², Viktor von Wyl ², Marc Höglinger ^3,^*

Editor: Barbara Guidi⁴

PMCID: PMC8386858 PMID: 34432842

Abstract

Background

Switzerland has a liberal implementation of Coronavirus mitigation measures compared to other European countries. Since March 2020, measures have been evolving and include a mixture of central and federalistic mitigation strategies across three culturally diverse language regions. The present study investigates a hypothesised heterogeneity in health, social behavior and adherence to mitigation measures across the language regions by studying pre-specified interaction effects. Our findings aim to support the communication of regionally targeted mitigation strategies and to provide evidence to address longterm population-health consequences of the pandemic by accounting for different pandemic contexts and cultural aspects.

Methods

We use data from from the COVID-19 Social Monitor, a longitudinal population-based online survey. We define five mitigation periods between March 2020 and May 2021. We use unadjusted and adjusted logistic regression models to investigate a hypothesized interaction effect between mitigation periods and language regions on selected study outcomes covering the domains of general health and quality of life, mental health, loneliness/isolation, physical activity, health care use and adherence to mitigation measures.

Results

We analyze 2,163 (64%) participants from the German/Romansh-speaking part of Switzerland, 713 (21%) from the French-speaking part and 505 (15%) from the Italian-speaking part. We found evidence for an interaction effect between mitigation periods and language regions for adherence to mitigation measures, but not for other study outcomes (social behavior, health). The presence of poor quality of life, lack of energy, no physical activity, health care use, and the adherence to mitigation measures changed similarly over mitigation periods in all language regions.

Discussion

As the pandemic unfolded in Switzerland, also health and social behavior changed between March 2020 to May 2021. Changes in adherence to mitigation measures differ between language regions and reflect the COVID-19 incidence patterns in the investigated mitigation periods, with higher adherence in regions with previously higher incidence. Targeted communcation of mitigation measures and policy making should include cultural, geographical and socioeconomic aspects to address yet unknown long-term population health consequences caused by the pandemic.

Introduction

Europe faced the second wave of the Coronavirus pandemic during the autumn and winter months of 2020. Switzerland is among the countries with the highest case numbers and deaths per capita (https://coronavirus.jhu.edu, accessed December 29, 2020). Non-pharmaceutical mitigation measures such as social distancing, testing, or restricting mobility can substantially reduce Coronavirus transmission [1, 2]. Switzerland has a liberal implementation of mitigation measures to slow down Coronavirus transmission compared to other European countries. These mitigation measures center around self-responsibility. Freedom of movement is not restricted, and shops, businesses, and schools have remained open, while restaurants were forced to close only shortly before the Christmas holidays [3]. The Oxford COVID-19 Government Response Tracker Stringency Index for November 2020 was 37.5 for Switzerland, 60.7 for Germany, 66.7 for Italy and 78.7 for France, with a higher index indicating a more stringent implementation of mitigation measures (https://covidtracker.bsg.ox.ac.uk/stringency-scatter, accessed December 29, 2020).

Switzerland’s mitigation strategy can be divided into different mitigation periods with stepwise increasing or decreasing stringency of measures and implementation at different political levels. The Swiss Federal Council coordinated nationwide and centralized mitigation measures (for example, the nationwide lockdown on March 16, 2020) until June 19, 2020, when the state of emergency as per the Swiss Epidemic Law ended. Thereafter, the 26 cantonal authorities were mainly responsible for a federalistic implementation of mitigation measures and remained in charge until January 17, 2021 (but with close federal and intercantonal coordination). While the epidemic situation worsened in the autumn months, cantons reacted differently and with varying strength of mitigation measures. This led to a patchwork of heterogeneous mitigation measures within small spatial proximity, with restaurants in one canton being closed and others remaining open.

Switzerland’s federalistic system overlaps with culturally diverse language regions. Citizens from the same language region often share common cultural traits, and health, risk, social and prevention behaviour differs between language regions in many regards [4–8]. For example, cultural differences in vaccination uptake were reported before the Coronavirus pandemic in Switzerland [8, 9]. The administered COVID-19 vaccination rate is highest in the Italian-speaking part of Switzerland and varies substantially between language regions (https://www.covid19.admin.ch/de/overview/, accessed, May 26, 2021). While regional and temporal variation in COVID-19 incidence patterns influence decision-making on the cantonal or nationwide implementation of mitigation measures, the socio-cultural context may play an important role in communication, the awareness of the pandemic situation and the individuals’ adherence to mitigation measures [10]. Because Switzerland is surrounded by European countries with varying mitigation strategies, the emerging pandemic challenged the different regions in Switzerland in several ways. Mitigation measures as well as other consequences of the pandemic (e.g. infection rates, widespread fears, emergency department crowding) likely impact a range of relevant public health and behavioral outcomes leading to different coping strategies over the phases of the pandemic and across language regions.

Objectives and research hypotheses

Positive and negative consequences of implemented mitigation measures and the pandemic on changes in relevant health and social behavior, health care use and the population’s adherence to mitigation measures during the Coronavirus pandemic have not been investigated in Switzerland so far. COVID-19 incidence patterns differ substantially between cantons and language regions, and heterogeneously-implemented mitigation measures may lead to different behaviors across language regions. We hypothesize that an interaction effect between pre-specified mitigation periods and language regions on behavioral changes exists. To investigate this research hypothesis, we use data from the COVID-19 Social Monitor, a population-based online survey which has longitudinally collected various aspects of social and health behavior since the beginning of the pandemic [11]. We analyze changes in these outcomes over the course of the pandemic. Our results provide first evidence about the extent of observable variations over time and about differences between the Swiss language regions. Our findings aim to support the communication of regionally targeted mitigation strategies and to provide evidence to address longterm population-health consequences of the pandemic by accounting for different pandemic contexts and cultural aspects.

Methods

Study population

Our study population covers a stratified random sample of a large cohort of the resident population in Switzerland with online access aged 18 years or older. Stratification was based on age, sex, and language region, i.e. the cohort is representative for Switzerland with respect to these three stratification criteria.

Data source

We use the COVID-19 Social Monitor survey waves 1 to 16 (March 2020 to May 2021). In brief, the COVID-19 Social Monitor is a population-based online survey which collects relevant aspects for a broad range of domains over multiple survey waves [11]. Study participants have been sampled from an online-panel whose members have been actively recruited using random probability sampling based on national landline telephone directories and random digit dialling of mobile phone numbers. An initial survey sample (survey wave 1) of 2,026 participants was interviewed from March 2020 onwards in a total of 11 survey waves. In December 2020, an additional sample of 1,355 individuals participated in the survey. These were–together with the initial sample—interviewed in four subsequent survey waves. Survey participants were randomly drawn from age, gender and language region strata. Table 1 shows a schematic overview of the survey waves from March 2020 to May 2021. We use data from the Federal Office of Public Health (https://www.covid19.admin.ch/de/overview/, accessed, May 18, 2021) for cantonal new COVID-19 cases from February 24, 2020 (the first COVID-19 case in Switzerland) to May 03, 2021 (the latest interview date of survey wave 16). Because new cases are only reported on a cantonal level and language regions do not follow cantonal borders, we assign the canton Ticino to the Italian-speaking part, the cantons Fribourg, Geneva, Jura, Neuchâtel, Vaud and Valais to the French-speaking part, and the remaining cantons to the German/Romansh-speaking part of Switzerland. In order to plot maps, we used free geodata from the Federal Office of Topography swisstopo.

Table 1. Schematic overview of COVID-19 Social Monitor survey waves.

Survey wave	1	2	3	4	5	6	7	8
Month	March 2020	April 2020	April 2020	April 2020	May 2020	May 2020	June 2020	July 2020
N of initial sample	2026	2026	2026	2026	2026	2026	2026	2026
N of additional sample	-	-	-	-	-	-	-	-
N of analysis sample	2026	2026	2026	2026	2026	2026	2026	2026
No. of participants	2026	1537	1540	1729	1673	1616	1522	1508
Non-participation (%)	0	24%	24%	15%	17%	20%	25%	26%
Survey wave	9	10	11	12	13	14	15	16
Month	August 2020	October 2020	November 2020	December 2020	January 2021	February 2021	April 2021	May 2021
N of initial sample	2026	2026	2026	2026	2026	2026	2026	2026
N of additional sample	-	-	-	1355	1355	1355	1355	1355
N of analysis sample	2026	2026	2026	3381	3381	3381	3381	3381
No. of participants	1532	1511	1492	2802	2564	2346	2219	2154
Non-participation (%)	24%	25%	26%	17%	24%	31%	34%	36%

Open in a new tab

Mitigation periods

Table 2 shows an overview of four a priori defined time periods from March 2020 to December 2020, based on the stringency of mitigation measures in Switzerland. The first period started from the date of the nationwide lockdown (March 16, 2020) and ended one day before the date of the nationwide reopening of stores and public schools (May 10, 2020). The second period lasted from May 11, 2020 to one day before the date of the mandatory nationwide implementation of face mask wearing in public transport (July 5, 2020). The third period started on July 6, 2020 and ended one day before the slowdown (i.e. mandatory nationwide wearing of face masks in public buildings, ban on spontaneous gatherings with more than 15 persons and recommended work from home) on October 18, 2020. The fourth period covers from October 19, 2020 to January 17, 2021 and was marked by the entry into a second pandemic wave with high case numbers. On January 18, 2021, the Swiss Federal Council announced a period with nationwide stringent mitigation measures with the closing of restaurants, mandatory homeoffice regulations and a ban on gatherings with more than 5 people in households (fifth period).

Table 2. Overview of mitigation periods and implemented mitigation measures.

Period	Coordination level	Mitigation measures*
(1) March 16, 2020 to May 10, 2020	Nationwide	Ban on gatherings >5 persons
	Nationwide	Public school closures
	Nationwide	Closure of stores and markets
	Nationwide	Closure of restaurants and bars
	Nationwide	Partial border closure
	Nationwide	Testing of symptomatic cases
	Nationwide	Hygiene rules, isolation, quarantine
(2) May 11, 2020 to July 5, 2020	Nationwide	Ban on gatherings >30 persons
	Nationwide	Partial school closure
	Nationwide	Partial border closures
	Nationwide	Testing of symptomatic cases
	Nationwide	Hygiene rules, isolation, quarantine
(3) July 6, 2020 to October 18, 2020	Nationwide	Face masks in public transport
	Nationwide	Ban on gatherings in public places >15 persons
	Nationwide	Allowance of mass gatherings >1000 persons
	Nationwide	Testing of symptomatic cases
	Nationwide	Hygiene rules, isolation, quarantine
	Cantonal	Face mask wearing in stores and public buildings
(4) October 19, 2020 to January 17, 2021	Nationwide	Face masks in public transport
	Nationwide	Face mask wearing in busy places and public buildings
	Nationwide	Recommendation for home office
	Nationwide	Testing of symptomatic cases
	Nationwide	Hygiene rules, isolation, quarantine
	Cantonal	Restrictions for restaurants and bars
(5) January 18, 2021 onwards	Nationwide	Face masks in public transport
	Nationwide	Face mask wearing in busy places and public buildings
	Nationwide	Ban on gatherings >5 persons
	Nationwide	Home office mandatory
	Nationwide	Testing of symptomatic cases
	Nationwide	Hygiene rules, isolation, quarantine
	Nationwide	Closing of restaurants and bars
	Cantonal	Vaccination

Open in a new tab

Study outcomes

We use the following study outcomes grouped in six domains of interest. Study outcomes were a priori selected to cover a broad domain of relevant health and behaviorial aspects and were (mostly) consistently included in all survey wave questionnaires. All study outcomes stem from single questions which allowed for categorical answers (e.g. on a Likert-scale). The source and original question used in the survey questionnaire is provided in S1 Table. Study outcomes were dichotomized to communicate results in terms of proportions and odds ratios.

General health and quality of life: Measured by 1) 1: very poor to poor self-assessed general health status vs 0: very good/good/fair 2) 1: very poor to poor self-assessed quality of life vs 0: very good/good/fair. Mental health: Measured by 1) 1: often to always in a depressive mood vs 0: never/seldom/sometimes 2) 1: often to always lacking energy vs 0: never/seldom/sometimes, 3) 1: fear of losing employment vs 0: no fear. Loneliness/Isolation: Measured by 1) 1: very often feelings of loneliness vs 0: never/seldom/sometimes/often 2) 1: often feelings of isolation vs 0: never/sometimes (only population 65 years or older). Physical activity: Measured by 1) 1: not being physically active vs 0: being physically active. Health care use: 1) 1: General health care use vs 0: no use 2) 1: General health care non-use vs 0: no non-use 3) 1: COVID-19 related health care use (contact of general practitioner or hospital because of COVID-19 symptoms) vs 0: no use. Adherence to mitigation measures: Measured by 1) 1: always adherence to physical distance when meeting persons vs 0: not always 2) 1: always the wearing of face masks vs 0: not always 3) 1: always avoidance of private appointments vs 0 not always 4) 1: always non-use of public transport vs 0: not always.

Variables of interest and confounding variables

Our main variables of interest are language region (German/Romansh, French, Italian) and the above-defined mitigation periods (March 16, 2020 to May 10, 2020; May 11, 2020 to July 5, 2020; July 6, 2020 to October 18, 2020; October 19, 2020 to January 17, 2021; January 18, 2021 onwards). We a priori define the following confounding variables: Age category (<45 years, 45 to <65 years, 65 years or older), gender (women, men), highest attained education (compulsory, secondary, tertiary), nationality (Swiss, non-Swiss), living with a partner (yes/no), living in urban area (yes/no). We selected these variables because we expect an association between the study outcomes and the main variable of interest.

Statistical methods

We describe the survey population by frequencies (n) and percentages (%). Incidence rates were calculated from Poisson rates with 95% confidence intervals (CIs). We calculate crude proportions of all study outcomes for each period and language region from logistic regression models. We test an interaction effect between language regions and mitigation periods for all study outcomes using a likelihood ratio test (LRT). We test for the null hypothesis of no time trend across mitigation periods by a univariable language region stratified hierarchical logistic regression model (accounting for repeated measurements within participants) using the mitigation periods as independent variable and reporting a two-sided p-value from a LRT [12]. To investigate whether period effects are confounded by other variables, we report odds ratios (OR) and 95% CIs from adjusted language region stratified logistic regression models (i.e. mitigation period as independent variable adjusted for confounding variables). We adjust for the variables age category, gender, highest attained education, nationality, living with a partner and living in an urban area. In hierarchical regression models we scale the calibration weights so that the new weights sum to the effective number of repeated measurements for each participant [13]. We set an alpha level of 5% as statistical significant. We replace missing values by their survey population median values for statistical modeling.

All models are survey-weighted regression models with calibration weights to account for sampling and nonresponse bias and to account for the fact that answers from the same individuals are correlated. Sampling weights make the survey population representative of the Swiss 2018 census population and nonresponse weights account for dropouts and nonresponse. We calculated the probability of being sampled from the census population using a logistic regression model with an age and gender interaction and language region as predictors to construct sampling weights. Non-response weights were constructed in a similar way using predictors age, gender, language region, living with a partner, working situation and highest attained education (see description in S1 Text). All analyses were performed in R version 4.0.2 [14]. For survey-weighted regression, we used the package svyglm version 4.0 [15].

Ethics statement

Ethical approval: The Cantonal Ethics Commission of Zurich concluded that the current study does not fall within the scope of the Human Research Act (BASEC-Nr. Req-2020-00323).

Informed consent: As per the decision of the Cantonal Ethics Commission of Zurich, explicit informed consent was not needed from participants for this particular study. However, participants gave their general permission to be part of research studies when accepting the invitation to the online panel from which we sampled our respondents. Participation in the study was voluntary and participants could withdraw from the study at all times.

Results

Study population

Fig 1 shows the survey sample distribution by language regions. 2,163 (64%) participants are from the German/Romansh-speaking part of Switzerland, 713 (21%) from the French-speaking part and 505 (15%) from the Italian-speaking part. Table 3 describes the COVID-19 Social Monitor survey population by language region. The total survey sample consists of 3,381 participants. 506 (15.0%) participants are older than 65 years. 48.6% of the survey population are women. Most of the survey participants live in an urban area (80.6%). The average daily COVID-19 incidence per 100,000 inhabitants for the period from February 24, 2020 to December 31, 2020 is 4.24, 95% CI (4.23–4.25), with a substantial variation between language regions. Missing values in baseline characteristics and study outcomes ranged from 0.01% (general health) to 0.4% (education), see S2 Table.

Table 3. Survey population baseline characteristics and COVID-19 incidence, by language region.

	Language region	German/Romansh (n = 2163)	French (n = 713)	Italian (n = 505)	Switzerland (N = 3381)
Characteristic		n (%)	n (%)	n (%)	n (%)
Age categories	0 to <45 years	1069 (49.4%)	367 (51.5%)	253 (50.1%)	1689 (50.0%)
	45 to <65 years	765 (35.4%)	239 (33.5%)	182 (36.0%)	1186 (35.1%)
	65 years or older	329 (15.2%)	107 (15.0%)	70 (13.9%)	506 (15.0%)
Gender	Men	1117 (51.6%)	359 (50.4%)	260 (51.5%)	1736 (51.3%)
	Women	1046 (48.4%)	354 (49.6%)	245 (48.5%)	1645 (48.6%)
Highest attained education	Compulsory	154 (7.1%)	55 (7.7%)	23 (4.5%)	232 (6.9%)
	Secondary	1028 (47.5%)	340 (47.7%)	255 (50.5%)	1623 (48%)
	Tertiary	971 (44.9%)	316 (44.3%)	224 (44.4%)	1511 (44.7%)
	missing	10 (0.5%)	2 (0.3%)	3 (0.6%)	15 (0.4%)
Citizenship	Non-Swiss	150 (7.0%)	91 (12.8%)	55 (10.9%)	296 (8.8%)
	Swiss	2010 (92.9%)	621 (87.1%)	449 (88.9%)	3080 (91.1%)
	missing	3 (0.1%)	1 (0.1%)	1 (0.2%)	5 (0.1%)
Working situation	Employed	1591 (73.6%)	485 (68%)	339 (67.1%)	2415 (71.4%)
	Unemployed	54 (2.5%)	19 (2.7%)	19 (3.8%)	92 (2.7%)
	Retired	328 (15.2%)	109 (15.3%)	74 (14.7%)	511 (15.1%)
	Other	190 (8.8%)	100 (14.0%)	73 (14.5%)	363 (10.7%)
Living with partner	Yes	1527 (70.6%)	478 (67.0%)	378 (74.8%)	2383 (70.5%)
Living in urban are	Yes	1679 (77.6%)	590 (82.8%)	455 (90.1%)	2724 (80.6%)
		Mean (95%CI)	Mean (95%CI)	Mean (95%CI)	Mean (95%CI)
COVID-19 incidence per day and 100,000 inhabitants from February 24, 2020 to May 3, 2021		4.25 (4.24–4.26)	6.72 (6.70–6.6.75)	5.90 (5.83–5.96)	4.96 (4.94–4.97)

Open in a new tab

COVID-19 incidence, by mitigation period and language region

Fig 2 shows new COVID-19 cases per day and 100,000 inhabitants, by mitigation period and language region. The Italian-speaking region had from February 24, 2020 (the first COVID-19 case in Switzerland) to May, 10 2020 an average daily incidence of 12.1, 95% CI (11.7, 12.5), per 100,000 inhabitants. The incidence in this region decreased to 0.47, 95% CI (0.38, 0.58), in the second period, increased to 67.9, 95% CI (67.0, 68.9), in the fourth period, and decreased again to 16.9, 95% CI (16.4, 17.3) in the fifth period. We found strong evidence for an interaction effect (p<0.001) between mitigation period and language region.

Study outcomes

Figs 3–5 show the crude proportions of the study outcomes weighted for sampling and nonresponse by mitigation period, language region and for the whole of Switzerland. For example, the proportion of individuals with no health care use in the German/Romansh-speaking part of Switzerland in the first mitigation period is 14.1%, 95% CI (12.7%-15.7%), and decreases to 1.3%, 95% CI (1.0%-1.8%), in the last mitigation period (see Fig 4 and S3 Table). Individuals from the Italian-speaking part of Switzerland show the highest percentage of adherence to mitigation measures (see Fig 5). We found evidence for a period effect in all language regions for the study outcomes poor quality of life (all p<0.04), depressive mood (p<0.01), lack of energy (all p<0.001), no physical activity (all p<0.005), health care use and non-use (all p<0.001) and for the adherence to mitigation measures (all p<0.001), see S4 Table.

Figs 6–8 show the adjusted ORs from language region stratified hierarchical logistic regression models weighted for sampling and nonresponse. We found evidence for an interaction effect between language region and mitigation period for the study outcome adherence to mitigation measures (all p<0.003, see S5 Table). The adjusted OR for not being physically active (compared to the mitigation period March 16, 2020 to May 10, 2020) in the French-speaking part of Switzerland is 1.51, 95% CI (1.16–1.96), in the period from October 19, 2020 to January 17, 2021 (see Fig 6 and S6 Table).

Discussion

Summary of main findings

The COVID-19 Social Monitor, a population-based longitudinal online survey, allows us to investigate the impact of mitigation measures on changes in health and social behavior, health care use and the adherence to mitigation measures in Switzerland during the Coronavirus pandemic from March 2020 to December 2020. We hypothesized an interaction effect between mitigation periods and culturally diverse language regions, because of regional and temporal variation in COVID-19 incidence patterns which led to a hetereogeneous implementation of mitigation measures in Switzerland. We found evidence for an interaction effect between language regions and mitigation periods for the study outcome adherence to mitigation measures, but not for the other investigated health and social related study outcomes. We observe changes in adherence to mitigations measures, with stronger adherence in regions with previously higher COVID-19 incidence. The presence of poor quality of life, depressive mood, lack of energy, no physical activity, general health care use and non-use and the adherence to mitigation measures changed over the analyzed mitigation periods in all language regions. We found no changes in the presence of feelings of loneliness or fear of losing employment over the investigated mitigation periods.

Regional differences in the course of the epidemic situation

The first COVID-19 case was reported on Febuary 24, 2020, in the canton of Ticino. The canton of Ticino borders the Italian region of Lombardy, which was a highly affected European region in the first wave of the Coronavirus pandemic [16]. The epidemic situation in the canton of Ticino quickly worsened (12.1 daily new cases per 100,000 inhabitants from February 24, 2020 to May 10, 2020) and the cantonal government rapidly implemented stringent mitigation measures to slow down transmission chains. However, new COVID-19 cases quickly appeared in the cantons of Geneva (13.5 new cases 100,000 inhabitants from February 24, 2020 to May 10, 2020), Vaud (9.0 new cases per 100,000 inhabitants from February 24, 2020 to May 10, 2020) and Basel-City (7.5 new cases per 100,000 inhabitants from February 24, 2020 to May 10, 2020) with densely populated areas in and around the larger cities of Geneva, Lausanne and Basel-City. With the nationwide lockdown on March 16, 2020, the incidence rate could quickly be slowed down and stabilized at lower levels. The epidemic situation during the summer months 2020 remained stable at lower incidence rates (0.4 cases per 100,000 inhabitants from May 11, 2020 to July 5, 2020 for the whole of Switzerland) and with a less pronounced variation between regions. In autumn 2020, during the second wave of the Coronavirus pandemic, the incidence rate quickly increases in the French-speaking region (78.7 cases per 100,000 inhabitants from October 19, 2020, onwards), but also in neighbouring regions to Germany (cantons of Basel) and Austria/Italy (cantons Grison, Ticino and Valais). With the worsening situation in the French-speaking region, the cantons of Fribourg, Neuchâtel, Valais and Vaud almost jointly implement more stringent mitigations measures. On October 19, 2020, the Swiss Federal Council announces the mandatory wearing of face masks in public places and buildings and bans gatherings of more than 15 persons. Yet, cantonal authorities react differently for the upcoming Winter months, for example, with varying restrictions for ski resorts. With a stringent implementation of mitigation measures (nationwide closing of restaurants and bars, ban on gatherings with more than 5 people and mandatory homeoffice regulation) from January 18, 2021 onwards, the number of new cases quickly decreased in this period. The Swiss Economic Institute from the Swiss Federal Institute of Technology Zurich estimated that the effective reproduction number decreased in a more pronounced way in cantons with more stringent mitigation measures (https://kof.ethz.ch/en/forecasts-and-indicators/indicators/kof-stringency-index.html, accessed May 20, 2021).

Changes in health and social behavior

In 2017, the proportion of individuals with a poor self-assessed health status in Switzerland was estimated at 3.5% (German-speaking region: 3.5%, French-speaking region: 3.5%, Italian-speaking region: 5.0%), for feelings of loneliness, at 1.7% (German-speaking region: 1.2%, French-speaking region: 3.0%, Italian-speaking region: 2.5%) and for being physically inactive at 8.2% (German-speaking region: 6.8%, French-speaking region: 10.9%, Italian-speaking region: 13.6%) (https://www.bfs.admin.ch/bfs/de/home/statistiken/gesundheit/erhebungen/sgb.assetdetail.6426300.html). Our survey population estimates for poor self-assessed health status over all investigated mitigation periods are lower than the estimates for 2017, whereas the estimates for being physically inactive are slightly higher than in 2017. The worsening trend in general health, quality of life and mental health problems from March 2020 to October 2020 onwards may be explained by the strengthening of mitigation measures in October 2020, compared to the very liberal situation during the summer months. This coincides with the increasing incidence rate of COVID-19 in Switzerland. Our findings show that the adherence to mitigation measures quickly changed in regions with higher COVID-19 incidence. The first survey of the COVID-19 Social Monitor by the middle of March 2020 was two weeks after the nationwide lockdown so that citizens were already familiar with the lockdown mitigation measures. The strenghtening mitigation measures from October 2020 may have a large impact on citizens’ physical and mental well-being and behaviors. Some of these observed changes may also be influenced by seasonality effects of depressive symptoms and mental health issues [17]. Nevertheless, survey results from Norway and Canada found that stringent mitigation measures are associated with severe mental health problems and with physical inactivity [18, 19]. A systematic review of 68 observational studies of the time period from December 2019 to July 2020 including 19 countries found increased psychological distress which is associated with age, gender, living in rural versus urban areas and socioeconomic position [20]. Population-based cohort studies in Switzerland found that disease outcomes and risk behavior are different across language regions [21–25]. In contrast to other international findings, we could not find a change in the presence of feelings of loneliness during the pandemic [18, 26, 27]. Our survey estimates for feelings of loneliness are similar to estimates from 2017, before the Coronavirus pandemic. Nevertheless, we found changes in the presence of feelings of isolation in the elderly population with a lower chance of feelings of isolation when the mitigation measures were less stringent during the summer months. Social isolation has been shown to be associated with poor health conditions and behavior in the Swiss population [28]. Longterm mental health effects caused by social isolation may be amplified during the Coronavirus pandemic and require further research.

Health care use during the Coronavirus pandemic

The Coronavirus pandemic has a huge impact on the health care system, including access, delivery and utilisation of health care [29–31]. For example, patients with chronic diseases, acute health events or emergencies may not seek health care during the pandemic with a potential negative impact on longterm health outcomes [32–36]. In general, patients are more likely to be fearful of seeking health care professional advice, non-elective treatments are postponed, and intensive care units in hospitals face an alarming situation with COVID-19 cases. Our results reveal an increased percentage of health care non-use in the first phase of the pandemic with a substantial decrease during the summer months, similar in all language regions. This change may be explained by the improved epidemic situation with less stringent mitigation measures also for health care providers and a seasonality effect during the summer months. Longterm patient outcomes–especially for vulnerable subpopulations and/or the chronically ill–because of a change in health care utilization during the pandemic are still unknown. Regional variation in delivery of health care and health care utilization may be associated with health care (non-)use and may ultimately affect patient outcomes. Switzerland has a substantial variation in health care utilization by region [37–43]. A cross-sectional survey from 2018 in Switzerland found that Italian-speaking individuals reported visiting a specialist more often than individuals living in the French- or German-speaking part of Switzerland [44]. Such regional variation in health care (non-)utilization may have an important impact on population health during the Coronavirus pandemic and needs further investigation, also considering the potential of new telemedicine approaches [45, 46].

Geographical and socioeconomic factors

Switzerland is a culturally diverse country, surrounded by the countries Austria, France, Germany and Italy. Geographical factors may partly explain some of the regional variation of Coronavirus transmission in Switzerland. Switzerland’s mountainous topography divides Southern and Northern Europe and is thus an important European travel link. Italy, Germany and France had a rapid growth in new COVID-19 cases during the first of wave of the pandemic and may reflect the observed incidence patterns in Switzerland [47]. A partial closure of borders to its neighbouring countries aimed to slow transmission rates in Switzerland during the first wave of the pandemic. An important driver for the pandemic is more densily populated areas like cities and surrounding areas leading to an urban-rural gradient [48]. In 2020, the percentage of individuals living in an urban area was estimated at 83% (https://www.bfs.admin.ch/bfs/de/home/statistiken/kataloge-datenbanken/grafiken.assetdetail.12767388.html, accessed January 10, 2020). The COVID-19 incidence patterns in Switzerland show an urban-rural gradient with more reported new cases in cities and urban regions than in rural regions even in pandemic phases with less stringent mitigation measures. Our survey findings show a different adherence to mitigation measures between urban and rural areas with a lower chance of adherence to physical distance, avoidance of private appointments and non-use of public transport for individuals living in urban areas. Such urban-rural behavioral differences may be related to socio-economic factors as socio-economic status substantially varies between regions, cities and even neigbourhoods in Switzerland [49]. Socio-economic factors may play an important role in the adherence to mitigation measures and to social and health behavior during the Coronavirus pandemic [50]. Socio-economic and regional differences in health and social behavior before the pandemic were reported for Switzerland [4, 51, 52]. Our results show that the highest attained education–a proxy for socio-economic position–was associated with changes in health and social behavior in our survey population. For example, our findings show that individuals with tertiary education had a lower chance of being socially isolated, of being physical inactive and of having depressive symtoms compared to individuals with only a compulsory education. Targeted communication strategies may mitigate health inequalities across cultural and socio-economic groups during the pandemic [53]. Nevertheless, further research is needed to investigate health inequalities across socio-economic groups during the Coronavirus pandemic in Switzerland.

Strengths and limitations

The COVID-19 Social Monitor has several strengths. The longitudinal and population-based survey design allows for a rigorous investigation of behavioral changes during the Coronavirus pandemic. Sampling and nonresponse weights make the survey sample representative of the Swiss 2018 census population older than 18 years. The use of established survey items—which are, for example, also used in the Swiss Health Survey—allows for a comparison of our findings to the year 2017, before the Coronavirus pandemic.

Our study has limitations. First, because our survey is online-based, we probably include more individuals with a greater affinity to online processes and better educated individuals in the survey, which leads to some sampling selection bias [54]. Yet, with the use of sampling weights, we minimize this kind of bias. Second, the regular assessment of individuals with the same survey items may lead to response bias, for example, the tendency to positive answers [55]. Third, we chose mitigation periods based on dates where nationwide mitigation measures were implemented. Federalistically implemented mitigation measures not only varied by canton but also in time. Thus, our chosen mitigation period may not mirror immediate changes in our investigated study outcomes and may not reflect the true underlying cantonal pattern of those measures. For example, many French-speaking cantons had already implemented a mandatory face mask wearing measure in public stores in early autumn (before October 19, 2020). Fourth, we used dichotomized study outcomes which allows us to present our results in terms of proportions. Thus we have a potential loss of information from categorical variables. Neverthless we think that the benefit from presenting percentages outweighs this limitation.

Conclusion

We conclude that the implemented mitigation measures from March 2020 to October 2020 had an impact on health and social behavior in Switzerland. The adherence to mitigation measures changed differently between language regions and reflected the COVID-19 incidence patterns in the investigated mitigation periods, with higher adherence in regions with previously higher incidence. Cultural, geographical and socio-economic aspects should be included in future communication strategies and policymaking to diminish potential (and as yet unknown) population health consequences in Switzerland caused by the pandemic. Our study informs the public and health authorities about the positive and negative impacts of implemented mitigation measures on changes in health and social behavior in Switzerland and adds important evidence for public health decision- and policy-making for the targeted implementation of mitigation measures.

Supporting information

S1 Table. The domain, the source and the original question from the survey questionnaire.

(DOCX)

Click here for additional data file.^{(14.2KB, docx)}

S2 Table. Description of missing values.

(DOCX)

Click here for additional data file.^{(15.3KB, docx)}

S3 Table. Underlying results for Figs 3–5.

(DOCX)

Click here for additional data file.^{(68.8KB, docx)}

S4 Table. p-values from likelihood ratio test for period effect, by language region.

(DOCX)

Click here for additional data file.^{(15.2KB, docx)}

S5 Table. p-values from likelihood ratio test for interaction effect between language regions and mitigation periods from unadjusted and adjusted models.

(DOCX)

Click here for additional data file.^{(16.3KB, docx)}

S6 Table. Results for from adjusted logistic regression models.

(DOCX)

Click here for additional data file.^{(49.6KB, docx)}

S1 Text. Statistical methods for calibration weights.

(PDF)

Click here for additional data file.^{(287.4KB, pdf)}

S2 Text. Reproducible analysis example.

(PDF)

Click here for additional data file.^{(175.9KB, pdf)}

S3 Text. Codebook.

(PDF)

Click here for additional data file.^{(36.5KB, pdf)}

S1 Data

(CSV)

Click here for additional data file.^{(1.6MB, csv)}

Acknowledgments

We thank Paul Kelly for proofreading the manuscript.

Data Availability

All relevant data are within the manuscript and its Supporting Information files.

Funding Statement

The Social Monitor study has received funding from the Swiss Federal Office of Public Health and from Health Promotion Switzerland. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

1.Salathé M, Althaus CL, Neher R, Stringhini S, Hodcroft E, Fellay J, et al. COVID-19 epidemic in Switzerland: on the importance of testing, contact tracing and isolation. Swiss Med Wkly. 2020. doi: 10.4414/smw.2020.20225 [DOI] [PubMed] [Google Scholar]
2.Flaxman S, Mishra S, Gandy A, Unwin HJT, Mellan TA, Coupland H, et al. Estimating the effects of non-pharmaceutical interventions on COVID-19 in Europe. Nature. 2020;584: 257–261. doi: 10.1038/s41586-020-2405-7 [DOI] [PubMed] [Google Scholar]
3.Haug N, Geyrhofer L, Londei A, Dervic E, Desvars-Larrive A, Loreto V, et al. Ranking the effectiveness of worldwide COVID-19 government interventions. Nat Hum Behav. 2020;4: 1303–1312. doi: 10.1038/s41562-020-01009-0 [DOI] [PubMed] [Google Scholar]
4.Abel T, Hofmann K, Schori D. Social and regional variations in health status and health behaviours among Swiss young adults. Swiss Med Wkly. 2013. doi: 10.4414/smw.2013.13901 [DOI] [PubMed] [Google Scholar]
5.Jürgens V, Ess S, Phuleria H, Früh M, Schwenkglenks M, Frick H, et al. Tobacco-related cancer mortality: projections for different geographical regions in Switzerland. Swiss Med Wkly. 2013. doi: 10.4414/smw.2013.13771 [DOI] [PubMed] [Google Scholar]
6.Chatelan A, Beer-Borst S, Randriamiharisoa A, Pasquier J, Blanco J, Siegenthaler S, et al. Major Differences in Diet across Three Linguistic Regions of Switzerland: Results from the First National Nutrition Survey menuCH. Nutrients. 2017;9: 1163. doi: 10.3390/nu9111163 [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Spoerri A, Egger M, von Elm E. Mortality from road traffic accidents in Switzerland: Longitudinal and spatial analyses. Accid Anal Prev. 2011;43: 40–48. doi: 10.1016/j.aap.2010.06.009 [DOI] [PubMed] [Google Scholar]
8.Zürcher K, Zwahlen M, Berlin C, Egger M, Fenner L. Trends in influenza vaccination uptake in Switzerland: Swiss Health Survey 2007 and 2012. Swiss Med Wkly. 2019. doi: 10.4414/smw.2019.14705 [DOI] [PubMed] [Google Scholar]
9.Riesen M, Konstantinoudis G, Lang P, Low N, Hatz C, Maeusezahl M, et al. Exploring variation in human papillomavirus vaccination uptake in Switzerland: a multilevel spatial analysis of a national vaccination coverage survey. BMJ Open. 2018;8: e021006. doi: 10.1136/bmjopen-2017-021006 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Airhihenbuwa CO, Iwelunmor J, Munodawafa D, Ford CL, Oni T, Agyemang C, et al. Culture Matters in Communicating the Global Response to COVID-19. Prev Chronic Dis. 2020;17: 200245. doi: 10.5888/pcd17.200245 [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Moser A, Carlander M, Wieser S, Hämmig O, Puhan MA, Höglinger M. The COVID-19 Social Monitor longitudinal online panel: Real-time monitoring of social and public health consequences of the COVID-19 emergency in Switzerland. Donnelly TT, editor. PLoS One. 2020;15: e0242129. doi: 10.1371/journal.pone.0242129 [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Molenberghs G, Verbeke G. Models for Discrete Longitudinal Data. New York: Springer-Verlag; 2005. doi: 10.1007/0-387-28980-1 [DOI] [Google Scholar]
13.Carle AC. Fitting multilevel models in complex survey data with design weights: Recommendations. BMC Med Res Methodol. 2009;9: 49. doi: 10.1186/1471-2288-9-49 [DOI] [PMC free article] [PubMed] [Google Scholar]
14.R Team Core. R: A language and environment for statistical computing, Vienna, Austria. R Foundation for Statistical Computing; 2020. Available: https://www.r-project.org/ [Google Scholar]
15.Lumley T. Analysis of Complex Survey Samples. J Stat Softw. 2004;9. doi: 10.18637/jss.v009.i08 [DOI] [Google Scholar]
16.Odone A, Delmonte D, Scognamiglio T, Signorelli C. COVID-19 deaths in Lombardy, Italy: data in context. Lancet Public Heal. 2020;5: e310. doi: 10.1016/S2468-2667(20)30099-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Wirz-Justice A, Graw P, Kräuchi K, Wacker HR. Seasonality in affective disorders in Switzerland. Acta Psychiatr Scand. 2003;108: 92–95. doi: 10.1034/j.1600-0447.2003.00127.x [DOI] [PubMed] [Google Scholar]
18.Hoffart A, Johnson SU, Ebrahimi O V. Loneliness and Social Distancing During the COVID-19 Pandemic: Risk Factors and Associations With Psychopathology. Front Psychiatry. 2020;11. doi: 10.3389/fpsyt.2020.589127 [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Lesser IA, Nienhuis CP. The Impact of COVID-19 on Physical Activity Behavior and Well-Being of Canadians. Int J Environ Res Public Health. 2020;17: 3899. doi: 10.3390/ijerph17113899 [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Wang Y, Kala MP, Jafar TH. Factors associated with psychological distress during the coronavirus disease 2019 (COVID-19) pandemic on the predominantly general population: A systematic review and meta-analysis. Murakami M, editor. PLoS One. 2020;15: e0244630. doi: 10.1371/journal.pone.0244630 [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Richard A, Faeh D, Rohrmann S, Braun J, Tarnutzer S, Bopp M. Italianity is associated with lower risk of prostate cancer mortality in Switzerland. Cancer Causes Control. 2014;25: 1523–1529. doi: 10.1007/s10552-014-0456-5 [DOI] [PubMed] [Google Scholar]
22.Schmidlin K, Spoerri A, Egger M, Zwahlen M, Stuck A, Clough-Gorr K. Cancer, a disease of aging (part 1)–trends in older adult cancer mortality in Switzerland 1991–2008. Swiss Med Wkly. 2012. doi: 10.4414/smw.2012.13637 [DOI] [PubMed] [Google Scholar]
23.Schmidlin K, Spoerri A, Egger M, Zwahlen M, Stuck A, Clough-Gorr K. Cancer, a disease of aging (part 2)–risk factors for older adult cancer mortality in Switzerland 1991–2008. Swiss Med Wkly. 2012. doi: 10.4414/smw.2012.13607 [DOI] [PubMed] [Google Scholar]
24.Spoerri A, Egger M, Von Elm E. Mortality from road traffic accidents in Switzerland: Longitudinal and spatial analyses. Accid Anal Prev. 2011;43: 40–48. doi: 10.1016/j.aap.2010.06.009 [DOI] [PubMed] [Google Scholar]
25.Spoerri A, Zwahlen M, Panczak R, Egger M, Huss A. Alcohol-selling outlets and mortality in Switzerland-the Swiss National Cohort. Addiction. 2013;108: 1603–1611. doi: 10.1111/add.12218 [DOI] [PubMed] [Google Scholar]
26.Bu F, Steptoe A, Fancourt D. Who is lonely in lockdown? Cross-cohort analyses of predictors of loneliness before and during the COVID-19 pandemic. Public Health. 2020;186: 31–34. doi: 10.1016/j.puhe.2020.06.036 [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Groarke JM, Berry E, Graham-Wisener L, McKenna-Plumley PE, McGlinchey E, Armour C. Loneliness in the UK during the COVID-19 pandemic: Cross-sectional results from the COVID-19 Psychological Wellbeing Study. Murakami M, editor. PLoS One. 2020;15: e0239698. doi: 10.1371/journal.pone.0239698 [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Hämmig O. Health risks associated with social isolation in general and in young, middle and old age. Ginsberg SD, editor. PLoS One. 2019;14: e0219663. doi: 10.1371/journal.pone.0219663 [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Hajek A, De Bock F, Wieler LH, Sprengholz P, Kretzler B, König H-H. Perceptions of Health Care Use in Germany during the COVID-19 Pandemic. Int J Environ Res Public Health. 2020;17: 9351. doi: 10.3390/ijerph17249351 [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Kutikov A, Weinberg DS, Edelman MJ, Horwitz EM, Uzzo RG, Fisher RI. A War on Two Fronts: Cancer Care in the Time of COVID-19. Ann Intern Med. 2020;172: 756–758. doi: 10.7326/M20-1133 [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Iyengar K, Mabrouk A, Jain VK, Venkatesan A, Vaishya R. Learning opportunities from COVID-19 and future effects on health care system. Diabetes Metab Syndr Clin Res Rev. 2020;14: 943–946. doi: 10.1016/j.dsx.2020.06.036 [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Reichert M, Sartelli M, Weigand MA, Doppstadt C, Hecker M, Reinisch-Liese A, et al. Impact of the SARS-CoV-2 pandemic on emergency surgery services—a multi-national survey among WSES members. World J Emerg Surg. 2020;15: 64. doi: 10.1186/s13017-020-00341-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Vogel AC, Schmidt H, Loud S, McBurney R, Mateen FJ. Impact of the COVID-19 pandemic on the health care of >1,000 People living with multiple sclerosis: A cross-sectional study. Mult Scler Relat Disord. 2020;46: 102512. doi: 10.1016/j.msard.2020.102512 [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Granata T, Bisulli F, Arzimanoglou A, Rocamora R. Did the COVID-19 pandemic silence the needs of people with epilepsy? Epileptic Disord. 2020;22: 439–442. doi: 10.1684/epd.2020.1175 [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Leocani L, Diserens K, Moccia M, Caltagirone C. Disability through COVID‐19 pandemic: neurorehabilitation cannot wait. Eur J Neurol. 2020;27. doi: 10.1111/ene.14320 [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Sabatello M, Burke TB, McDonald KE, Appelbaum PS. Disability, Ethics, and Health Care in the COVID-19 Pandemic. Am J Public Health. 2020;110: 1523–1527. doi: 10.2105/AJPH.2020.305837 [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Panczak R, Luta X, Maessen M, Stuck AE, Berlin C, Schmidlin K, et al. Regional Variation of Cost of Care in the Last 12 Months of Life in Switzerland. Med Care. 2017;55: 155–163. doi: 10.1097/MLR.0000000000000634 [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Scheuter C, Wertli MM, Haynes AG, Panczak R, Chiolero A, Perrier A, et al. Unwarranted regional variation in vertebroplasty and kyphoplasty in Switzerland: A population-based small area variation analysis. Kou YR, editor. PLoS One. 2018;13: e0208578. doi: 10.1371/journal.pone.0208578 [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Bähler C, Rapold R, Signorell A, Reich O, Panczak R, Blozik E. Regional differences in healthcare costs at the end of life: an observational study using Swiss insurance claims data. Int J Public Health. 2020;65: 969–979. doi: 10.1007/s00038-020-01428-w [DOI] [PubMed] [Google Scholar]
40.Wei W, Gruebner O, von Wyl V, Dressel H, Ulyte A, Brüngger B, et al. Exploring geographic variation of and influencing factors for utilization of four diabetes management measures in Swiss population using claims data. BMJ Open Diabetes Res Care. 2020;8: e001059. doi: 10.1136/bmjdrc-2019-001059 [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Stoller N, Wertli MM, Zaugg TM, Haynes AG, Chiolero A, Rodondi N, et al. Regional variation of hysterectomy for benign uterine diseases in Switzerland. Garzon S, editor. PLoS One. 2020;15: e0233082. doi: 10.1371/journal.pone.0233082 [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Wertli MM, Schlapbach JM, Haynes AG, Scheuter C, Jegerlehner SN, Panczak R, et al. Regional variation in hip and knee arthroplasty rates in Switzerland: A population-based small area analysis. Shah TI, editor. PLoS One. 2020;15: e0238287. doi: 10.1371/journal.pone.0238287 [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Ulyte A, Wei W, Dressel H, Gruebner O, von Wyl V, Bähler C, et al. Variation of colorectal, breast and prostate cancer screening activity in Switzerland: Influence of insurance, policy and guidelines. Van Hemelrijck M, editor. PLoS One. 2020;15: e0231409. doi: 10.1371/journal.pone.0231409 [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Giezendanner S, Bretschneider W, Fischer R, Diaz Hernandez L, Zeller A. The ecology of medical care in Switzerland: prevalence of illness in the community and healthcare utilisation in Switzerland. Swiss Med Wkly. 2020. doi: 10.4414/smw.2020.20221 [DOI] [PubMed] [Google Scholar]
45.Nittas V, von Wyl V. COVID-19 and telehealth: a window of opportunity and its challenges. Swiss Med Wkly. 2020. doi: 10.4414/smw.2020.20284 [DOI] [PubMed] [Google Scholar]
46.Vokinger KN, Nittas V, Witt CM, Fabrikant SI, von Wyl V. Digital health and the COVID-19 epidemic: an assessment framework for apps from an epidemiological and legal perspective. Swiss Med Wkly. 2020. doi: 10.4414/smw.2020.20282 [DOI] [PubMed] [Google Scholar]
47.Plümper T, Neumayer E. Lockdown policies and the dynamics of the first wave of the Sars-CoV-2 pandemic in Europe. J Eur Public Policy. 2020; 1–21. doi: 10.1080/13501763.2020.1847170 [DOI] [Google Scholar]
48.Wells K, Lurgi M, Collins B, Lucini B, Kao RR, Lloyd AL, et al. Disease control across urban–rural gradients. J R Soc Interface. 2020;17: 20200775. doi: 10.1098/rsif.2020.0775 [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Panczak R, Galobardes B, Voorpostel M, Spoerri A, Zwahlen M, Egger M, et al. A Swiss neighbourhood index of socioeconomic position: development and association with mortality. J Epidemiol Community Heal. 2012;66: 1129–1136. doi: 10.1136/jech-2011-200699 [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Anderson G, Frank JW, Naylor CD, Wodchis W, Feng P. Using socioeconomics to counter health disparities arising from the covid-19 pandemic. BMJ. 2020; m2149. doi: 10.1136/bmj.m2149 [DOI] [PubMed] [Google Scholar]
51.Hämmig O, Gutzwiller F, Kawachi I. The contribution of lifestyle and work factors to social inequalities in self-rated health among the employed population in Switzerland. Soc Sci Med. 2014;121: 74–84. doi: 10.1016/j.socscimed.2014.09.041 [DOI] [PubMed] [Google Scholar]
52.Moser A, Panczak R, Zwahlen M, Clough K, Spörri A, Stuck A, et al. What does your neighbourhood say about you? A study of life expectancy in 1.3 million Swiss neighbourhoods. J Epidemiol Community Health. 2014;68: 1125–1132. doi: 10.1136/jech-2014-204352 [DOI] [PubMed] [Google Scholar]
53.Rimmer A. Covid-19: Tackling health inequalities is more urgent than ever, says new alliance. BMJ. 2020; m4134. doi: 10.1136/bmj.m4134 [DOI] [PubMed] [Google Scholar]
54.Bethlehem J. Selection Bias in Web Surveys. Int Stat Rev. 2010;78: 161–188. doi: 10.1111/j.1751-5823.2010.00112.x [DOI] [Google Scholar]
55.Kreitchmann RS, Abad FJ, Ponsoda V, Nieto MD, Morillo D. Controlling for Response Biases in Self-Report Scales: Forced-Choice vs. Psychometric Modeling of Likert Items. Front Psychol. 2019;10. doi: 10.3389/fpsyg.2019.02309 [DOI] [PMC free article] [PubMed] [Google Scholar]

PLoS One. doi: 10.1371/journal.pone.0256253.r001

Decision Letter 0

Barbara Guidi

17 May 2021

PONE-D-21-02911

Health and social behaviour through pandemic phases in Switzerland: Longitudinal analysis of the COVID-19 Social Monitor panel study

PLOS ONE

Dear Dr. Höglinger,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

The paper needs a MAJOR REVISION. Please follow the suggestions given by reviewers in order to improve the quality of the manuscript.

Please submit your revised manuscript by Jul 01 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Barbara Guidi

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please provide additional details regarding participant consent. In the ethics statement in the Methods and online submission information, please ensure that you have specified what type you obtained (for instance, written or verbal, and if verbal, how it was documented and witnessed). If your study included minors, state whether you obtained consent from parents or guardians. If the need for consent was waived by the ethics committee, please include this information.

3. Please note that in order to use the direct billing option the corresponding author must be affiliated with the chosen institute. Please either amend your manuscript to change the affiliation or corresponding author, or email us at plosone@plos.org with a request to remove this option.

4. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

5. Please include a copy of Table 2 which you refer to in your text in line 130.

6.We note that Figure(s) 1 in your submission contain map images which may be copyrighted. All PLOS content is published under the Creative Commons Attribution License (CC BY 4.0), which means that the manuscript, images, and Supporting Information files will be freely available online, and any third party is permitted to access, download, copy, distribute, and use these materials in any way, even commercially, with proper attribution. For these reasons, we cannot publish previously copyrighted maps or satellite images created using proprietary data, such as Google software (Google Maps, Street View, and Earth). For more information, see our copyright guidelines: http://journals.plos.org/plosone/s/licenses-and-copyright.

We require you to either (1) present written permission from the copyright holder to publish these figures specifically under the CC BY 4.0 license, or (2) remove the figures from your submission:

a) You may seek permission from the original copyright holder of Figure(s) 1 to publish the content specifically under the CC BY 4.0 license.

We recommend that you contact the original copyright holder with the Content Permission Form (http://journals.plos.org/plosone/s/file?id=7c09/content-permission-form.pdf) and the following text:

“I request permission for the open-access journal PLOS ONE to publish XXX under the Creative Commons Attribution License (CCAL) CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). Please be aware that this license allows unrestricted use and distribution, even commercially, by third parties. Please reply and provide explicit written permission to publish XXX under a CC BY license and complete the attached form.”

Please upload the completed Content Permission Form or other proof of granted permissions as an "Other" file with your submission.

In the figure caption of the copyrighted figure, please include the following text: “Reprinted from [ref] under a CC BY license, with permission from [name of publisher], original copyright [original copyright year].”

b) If you are unable to obtain permission from the original copyright holder to publish these figures under the CC BY 4.0 license or if the copyright holder’s requirements are incompatible with the CC BY 4.0 license, please either i) remove the figure or ii) supply a replacement figure that complies with the CC BY 4.0 license. Please check copyright information on all replacement figures and update the figure caption with source information. If applicable, please specify in the figure caption text when a figure is similar but not identical to the original image and is therefore for illustrative purposes only.

The following resources for replacing copyrighted map figures may be helpful:

USGS National Map Viewer (public domain): http://viewer.nationalmap.gov/viewer/

The Gateway to Astronaut Photography of Earth (public domain): http://eol.jsc.nasa.gov/sseop/clickmap/

Maps at the CIA (public domain): https://www.cia.gov/library/publications/the-world-factbook/index.html and https://www.cia.gov/library/publications/cia-maps-publications/index.html

NASA Earth Observatory (public domain): http://earthobservatory.nasa.gov/

Landsat: http://landsat.visibleearth.nasa.gov/

USGS EROS (Earth Resources Observatory and Science (EROS) Center) (public domain): http://eros.usgs.gov/#

Natural Earth (public domain): http://www.naturalearthdata.com/

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Thank authors for investigating such an important subject; I have the following comments:

Line 130 states that: “Table 2 shows an overview of four a priori …”, however, table 2 was not included in the paper.

Line143-155 introduces the study outcomes, however, it is important to specify which category is considered as a baseline in each outcome and to report the number and percentages of cases in each category of the outcome.

In Lines 183-185, the authors indicated that they used sampling weights as well as nonresponse weights for the outcomes; although the readers are referred to the supplementary material for detail (I did not have access to supplementary materials), this information is important and should be included in the paper.

Line 181 states that the results are shown by Odds Ratio(OR) with 95%CI; it means that the significant level is set to 0.05, however, in line 262-263, the author reported poor quality of life with a p-value higher than 0.05 as significant.

Line 259-275 explains the interpretation of OR in the figure of 6-8, and it stated that all the variables in figure 6 are significant, however, in these figures most of the confidence intervals include OR=1 so their corresponding outcomes cannot be significant; for example, In line 263 it is stated that “no physical activity” is significant and in line 268 it's OR and 95%CI is reported, however, both ORs’ confidence intervals in line 268 include OR=1 which means that odds ratio for this variable is not significant. As it is demonstrated in figure 6, the variable “no physical activity” is significant only in the last period (0ctober 19, 2020 onwards) for the French and Italian region; similarly, other outcomes in figure 6 have significant value for only the last period.

The main results of the paper including the Odds Ratios and their confidence intervals are shown in figure 6,7, and 8, so the significant variables in the figure needs to be marked so the readers can recognize them.

The title of the paper indicated that longitudinal analysis is performed in the study, however, logistic regression is applied in the current study which is appropriate for cross-sectional study; longitudinal statistical models are the most powerful models for showing the pattern of change during the time, but they are not used in this study.

Reviewer #2: PLOS One review

Thank you for the opportunity to review this manuscript of original study from the Swiss COID19 Social Monitor Study.

My comments follow the order of sections in the manuscript.

Abstract

By reviewing the Abstract it is not really clear what the purpose of the study comprises. Yes, to test predetermined interaction effects in COVID19 mitigation efforts across the three language regions in Switzerland. But why? How do these tests enhance our knowledge in the field of COVID19 or other virus mitigation? Clarifying this would be helpful for readers as many people will only read the abstract.

Introduction

L74-85 and 92-94: This description suggests that mitigation efforts were variant across the 26 cantons more than the three regions that are used in the proposed interaction tests. How will this potentially complicate the storyline, results and interpretation? The statement that variation in mitigation efforts may lead to “different behaviors across language regions” does not appear well supported given this description.

L94-95: “We hypothesize that an interaction effect between pre-specified mitigation periods and language regions on behavioral changes exists”. This appears quite unclear. The authors should provide a cultural description and background information or data which backs the notion that such interaction effects may be warranted. Presently the reader is left in the cold concerning why and which interaction effects may be found between the language regions. Any prospective description around what to except should be outlined in the Introduction section.

Methods

The “Data Source”/Sample subsection is quite challenging to follow. Please state clearly how many individuals were in the original sampling frame, how many responded, and how many have responded in subsequent waves of data collection, including attrition rates.

The subsection defined as “Mitigation periods” appears to suggest that several mitigation periods were indeed applied and led at the federal level. This is in contrast to the Introduction which appears to suggest that most efforts were led at the local level. Please explain.

In line with the comments above about the lack of study question context within the Introduction section, the material presently in the subsection titled “Study outcomes” comes across as very unfocused and lacking in connection with the overall study objective. Specifically, why these measures and not some others are included has not been discussed or contextualized. This is sorely needed.

The description of the measures provided in the “Study outcomes” section is also unclear. Why are all questions dichotomized? Are responses within the same domain collapsed together to form a scale? For example, have the three mental health questions been summed to form a scale or are they modeled individually? Such a description is not provided.

Similar to the above with regards to the confounding variables. Why are those variables and not some others included in the analyses? At this point the reader has not been introduced to any reasoning or context which supports the selection of those variables over other ones.

Analyses

“Because of the low percentage of missing values, we replace missing value by its survey population median value for statistical modeling.” – please reveal the % missing for all waves used on the analyses so that the reader can assess the appropriateness of this method of dealing with missing values.

The analyses subsection is quite chaotic and hard to follow and would require a substantial rewrite for clarity. Presently it is more or less unclear what the analyses entail. For example: “We construct sampling weights to make the survey sample representative of the 2018 census population of Switzerland aged 15 years or older and construct nonresponse weights to account for dropouts and nonresponse (see Supplementary Material for details). We use sampling and nonresponse weights in the above-specified logistic regression models to account for sampling and nonresponse bias.” – by reviewing this statement it is really hard to assess what was actually done and how.

Please clarify if Table 3 includes information about the survey population at baseline only.

It also appears that the authors are conducting trend analyses and interaction effects on time-trends by region. The title of the study should therefore be renamed to include the term “time-trends” as oppose to “longitudinal”. For example, the subtitle cold be “Regional Time-Trends in the COVID19 Social Monitor Panel Study”.

Results

This is very long section that could be truncated.

Discussion

The discussion around variations in mitigation effort further supports the notion that the bulk of variation in observed variables stem from within language region rather than across them.

Reviewer #3: How were participants recruited and followed longitudinally?

Line 299: Is the study truly “population-based” (i.e., people with certain characteristics are more likely to self-select into the survey such as access to information about this study)?

Line 49: For the loneliness/isolation measure, was this only asked among people older than 65 years?

Line 110: It is not clear how many times each participant answered questionnaires multiple times at each wave (followed longitudinally) or if this is a repeated cross-sectional study design? What is the difference between participants from wave 11 and the ‘additionally sampled participants’ from wave 12?

Line 182: Could you quantify the level of missingness?

Was multiple testing addressed?

304: typo “varyiance”

Line 318: What is the purpose of this section “The course of the epidemic situation in Switzerland?” It does not refer to the data/results and seems out of place.

What are some of the limitations that come with having a dichotomous yes/no for all the outcome variables?

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2021 Aug 25;16(8):e0256253. doi: 10.1371/journal.pone.0256253.r002

Author response to Decision Letter 0

4 Jun 2021

Our responses to the reviewers and the editor might be more easily readable in the separate letter instead of in this form.

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

We checked the PLOS One style requirements and corrected formatting and changed the file naming accordingly. Also, we included the required references to the supplementary files at the end of the article.

We now offer more details regarding participant consent in the manuscript:

Zurich University of Applied Sciences, to which the corresponding author is affiliated, is a PLOS institutional partner with direct billing option. We have used this in the past; hence, that should work.

We included captions in the Supporting Information (Tables in S1-S5 Table) and changed in-text citations accordingly.

5. Please include a copy of Table 2 which you refer to in your text in line 130.

We apologize for the missing Table 2. This table is now included in the manuscript.

We used shapefiles from the Federal Office of Topography (swisstopo) for plotting the maps. The free geodata and geoservices of swisstopo may be used, distributed and made accessible. (https://www.swisstopo.admin.ch/en/home/meta/conditions/geodata/ogd.html).

5. Review Comments to the Author

Reviewer #1:

We thank reviewer #1 for her or his feedback on our manuscript. We hope to clarify in our answers below that we did not a priori define a “significant” type-I error cutoff to avoid a dichotomization of findings. We would also like to emphasize that we did not use the wording “significant” in the manuscript. Instead, we utilize descriptors for different levels of ‘strength of evidence’, which allow readers to judge by themselves how ‘strong’ this evidence is. Not only is this approach recommended by eminent statisticians and reporting guidelines in our field (e.g. STROBE), we also believe that this is an appropriate way to discuss our findings, because under our observational study design - which is likely influenced by sampling, non-response and selection bias – any effect estimate needs a careful interpretation, which should not be influenced by p-values. P-values were only reported in two cases (due to the lack of better alternatives): for interaction tests (S6 Table) and in an overall test for period effects (S5 Table).

Thank authors for investigating such an important subject; I have the following comments:

Line 130 states that: “Table 2 shows an overview of four a priori …”, however, table 2 was not included in the paper.

We apologize for the missing Table 2. This table is now included in the manuscript.

Line 143-155 introduces the study outcomes, however, it is important to specify which category is considered as a baseline in each outcome and to report the number and percentages of cases in each category of the outcome.

Thank you for this important comment. We added the corresponding information (1 vs 0) in the methods section. The percentages of the study outcomes are shown in Figures 3-5 and absolute numbers are provided in S3 Table.

We apologize that the reviewer did not have access to the supplementary material, which was uploaded during the submission process. We agree that it is a very important piece of information for the reader. However, the corresponding section in the supplementary material is likely too technical for non-statistical readers. We decided to include a more detailed description of our weighting strategy in the methods section and then still refer to the supplementary material section.

We thank the reviewer for this thoughtful comment. We would like to stress that we decided not to a priori define a pre-specified cut-off type-I error of say, 5% (we additionally note that the manuscript does not include the wording “significant”). We based this decision on the ASA Statement on p-Values (https://doi.org/10.1080/00031305.2016.1154108), the observational design of our study and all the required assumptions on our used statistical models. We do not discuss “significant” results but discuss whether evidence for an association exists. Thus, we leave it up to the reader whether a p-value of, say 0.1, is enough “evidence” (under the assumptions that the null hypothesis is true, the statistical models are true and the frequentist properties of the p-value). We agree with the reviewer that our reported list of “p-values” is arbitrary, but we wanted to be parsimonious in the results section (and not mention all estimates). We agree with the reviewer that the 95% width choice of the confidence interval is arbitrary. However, we had to define a range to show the uncertainty of our estimates. We think that a 95% CI is an appropriate choice for presenting uncertainty around point estimates. But we would like to emphasize that we did not intend this range to be used for dichotomizing results into “significant” or “not significant” results.

We thank the reviewer for this comment. We relate our answer to the previously- addressed revision point that we did not use a dichotomizing cut-off type-I error of 5%. We stress that we do not mention “significant” results on lines 259-275. We discuss whether evidence for an association exists and leave it up to the reader to decide how strong this evidence is. We agree that our chosen way of talking about evidence is one of many options, but it is a preferred and recommended way in epidemiological research to present results (see e.g. ISBN 978-0-865-42871-3). Of course, p-values contradict in general the likelihood principle so that the terminology of “evidence” requires a careful use. Given that the likelihood function covers all necessary information for parameter inference, other measures should be reported for completeness (such as likelihood ratios or relative likelihood ratios, see, for example, 10.1007/978-3-662-60792-3). However, we decided not to include additional measures to avoid confusion for the reader. We agree with the reviewer that the presented 95% CIs is arbitrary, but we emphasise again that we had to decide for a range of uncertainty for presenting results.

We thank the reviewer for this comment. We decided not to dichotomize results into “significant” or “not significant”. Given that our study design is a survey design (thus, of observational nature) marking “significant” results (as the reviewer suggests) is misleading to the reader. Observational designs require the careful inclusion of observed (and unobserved) confounders to make estimates as “unbiased” (in terms of removing spurious association from the design and sampling process) as possible. However, this is only possible to some degree, so that marking “significant” results is not recommended. We assume that potential readers of our manuscript have basic knowledge of epidemiological, biostatistical and causal approaches, and are therefore in a position to assess the strength of evidence of our findings (without dichotomizing by “significance”). Hence, we preferred not to implement the suggestion by the reviewer.

Thank you for bringing up this important point, which has made us realize that further explanations of our regression approach are warranted. Logistic regression models “model” the likelihood function of the statistical model - which is in our case a Bernoulli distributed random variable - which transforms – in case ofidentically and independently distributed random variables - into a binomial likelihood. This is independent, whether measurements from the same individuals are correlated or not. Thus, a logistic regression model can be used for nested (and unnested) designs. Our study design requires the inclusion of calibration weights (which accounts for the sampling design and non-response), but also that (repeated) measurements from the same individual are not independent. Thus, we used a survey-design approach (modeled with the R package svyglm). In brief, this model approach accounts for the sampling design and non-response, while accounting for the fact that measurements are nested within participants (this is specified in the analysis as svyglm(id~id, …), where id is the variable for identifying participants. We modeled the period effect as a fixed effect because those effects are common for all participants and in order to investigate our a priori specified interaction effect between language regions and mitigation periods. Thus, we perform a “longitudinal” analysis using logistic regression models. We agree with the reviewer that we were not precise in how we included the fact that measurements of the same individuals are correlated. We added a sentence to the methods section for clarification and added a reference for the used statistical package.

Reviewer #2: PLOS One review

We are grateful for the feedback and comments of reviewer #2, which greatly improved the readability of our manuscript. We agree that in specific sections we were too vague and ideas were not precisely formulated. We hope that this revised version better explains our aims and storyline of the manuscript.

Thank you for the opportunity to review this manuscript of original study from the Swiss COID19 Social Monitor Study.

My comments follow the order of sections in the manuscript.

Abstract

We thank the reviewer for this helpful comment. We added a sentence to the background that our findings aim to support targeted implementation of mitigation measures, while accounting for cultural aspects. We believe that these finding are relevant not only for Switzerland but for Europe in general.

Introduction

We thank the reviewer for this important comment. It is true that – due the federal system in Switzerland – the cantonal organization play an important role in the variation of mitigation measures. However, language regions play a very important role in health and social behavior in Switzerland. This is supported by the cited literature in the discussion with evidence from large nationwide cohort studies about health and social aspects. We do not think that the nested structure of cantons within language regions complicate the storyline and our findings. Cultural behavior and tradition play a very important role in Switzerland, such that the investigation of a ‘higher-level factor’ language region is not only well-established in the literature (see e.g. references 4, 7, 8, 20-23 in the revised manuscript), but also necessary. In the discussion we highlighted how the pandemic emerged geographically, which was likely influenced – among other factors - by neighboring countries. Our findings show that the Italian-speaking part of Switzerland still has a different behavior in relation to mitigation measures than the other parts of Switzerland. We think that mostly overlaps with the cultural aspects which ultimately influence individual behavior.

We thank the reviewer for this important comment. We changed the introduction accordingly. We now give an example (vaccination uptake before the pandemic) and mention the differences in administered COVID-19 vaccination rates across language regions. Further, we now explicitly mention the influence of Switzerland’s surrounding countries and the coping strategies which vary over the pandemic phases and language regions, leading to our hypothesized interaction effect. We hope this clarifies our interaction hypothesis for the reader.

Methods

We rewrote this section accordingly. Additionally, we changed Table 1, which now shows all used survey waves with the number of participants and attrition.

We thank the reviewer for this comment. We agree that Table 2 likely provides a picture that cantons only have a minor role in the implementation of mitigation measures. We want to clarify as follows: 1) Some of the mitigation measures include the (repetitive) ‘standard’ measures of hygiene rules, isolation, testing, face masks and others. Those measures have been implemented during the very early phase of the pandemic on a federal level. Thus, Table 2 has many repetitive entries on a “nationwide” level, which likely gives the impression that federal measures have higher priority. The Swiss Federal Council announced the state of emergency from March 16, 2020, to June 19, 2020. From then on, cantons were responsible for the implemented mitigation measures until January 18, 2021 (measures for restaurants, bars, face mask wearing). Those cantonal mitigation measures had a major impact on individual behavior and mobility and ultimately on the emerging of new waves. For example, during the autumn months of 2020, individuals traveled from a more stringent canton to a less stringent canton for, for example, shopping and leisure activities, which was likely a driver for increased new cases. Thus, the cantonal mitigation played a huge role in the development of the pandemic for a long period. However, we still think that Table 2 should include all measures as listed, because all play an important role in Switzerland’s mitigation strategy.

We agree that our reasoning on the selection of study outcomes is vague. We a priori selected study outcomes to cover a broad domain of relevant health and behavioral aspects, while accounting for the necessity that study outcomes need to be included in all survey wave questionnaires (this is not the case for all study outcomes in our study). We changed this section accordingly. Also, we now mention in the introduction (objectives subsection) that our aim is to analyze in a descriptive way variations in relevant health and behavioral outcomes over the course of the pandemic and corresponding differences across language regions.

We dichotomized study outcomes so that results can be communicated in terms of proportions and odds ratios. We are aware that we lose information through this process. However, we think the benefit for the communication of results in proportions outweighs the potential information loss due to the dichotomization. The same comment has been pointed out by reviewer #3.

We apologize for being imprecise in the formulations. We now explicitly mention that we expect an association of those variables with the study outcomes and language region. Those variables are known confounders and have been used in many epidemiological studies (see e.g. references 20-23, 47, 50 in revised manuscript).

Analyses

We thank the reviewer for this important comment. We included a Supplemental Table which presents the number of missing values for baseline characteristics and study outcomes. Because of the low frequency, we decided to show only the overall missing number and not per wave. We expect a very small bias from the median imputation.

Thank you for this suggestion to make it less ‘chaotic’. Nevertheless, we are a bit unclear as to what it makes so chaotic, because the section is structured in a rather common way: Used descriptive statistics, used modeling approaches (including used test statistics, effect measures and adjustment variables) and dealing with missing values. This is a common and suggested way to structure the statistical analysis section. We rewrote the calibration weights part with more focus on why we use those weights and set it apart in a new paragraph. We hope this clarifies the used statistical approaches and simplifies it for the reader . Some rewriting is connected to comments from reviewer #1 who required a better description of the calibration weights.

Please clarify if Table 3 includes information about the survey population at baseline only.

Thank you for this suggestion. We added that we only show baseline characteristics.

Thank you for this suggestion. We changed to this more informative subtitle.

Results

This is very long section that could be truncated.

We substantially shortened the results section. We decided to delete the supporting information about adjusted effects of education and urban/rural information, because this is only briefly discussed in the discussion but not as a main finding. We decided on the other hand to provide the estimates for Figures 3-6, and also of the statistical tests for interaction and period effects.

Discussion

The discussion around variations in mitigation effort further supports the notion that the bulk of variation in observed variables stem from within language region rather than across them.

Thank you for this comment. We agree that due to the federal structure in Switzerland some of the variation might be explained by cantons. However, language regions play an important overall role in social and health behavior, structural organizational and even political decisions. Several studies from Switzerland show evidence for strong cultural differences (e.g. in risk behavior or vaccination uptake, see references 7 and 8 in the manuscript). Working and leisure time mobility of Swiss residents is often within the same language region, which can be seen in defined labor force areas (https://www.bfs.admin.ch/bfs/de/home/grundlagen/raumgliederungen.assetdetail.8706492.html). This supports the hypothesis that cultural regions have a higher-level impact on every-day life activities of Swiss residents (and thus the mitigation or emerging of the pandemic), despite the federal system in Switzerland. The Swiss Economic Institute of the Swiss Federal Institute of Technology in Zürich summarized that the effective reproduction number decreased in a more pronounced way in cantons with stronger mitigation measures (https://kof.ethz.ch/en/forecasts-and-indicators/indicators/kof-stringency-index.html). Given that the strengthening of mitigation measures in the autumn months of 2020 were often joint efforts from cantons within the same language region, this also supports a cultural effect of mitigation measures. We conclude that we can only speculate as to whether differences in mitigation adherence across language regions might be in truth more pronounced than our survey results indicate.

Reviewer #3:

We greatly appreciate the comments of reviewer #3, which clarify aspects of important open study design and reporting points. Thank you!

How were participants recruited and followed longitudinally?

Line 299: Is the study truly “population-based” (i.e., people with certain characteristics are more likely to self-select into the survey such as access to information about this study)?

We thank the reviewer for this comment and have added a sentence on the characteristics of our sample. Yes, it is ‘population-based’, because the study design allows the generalizability of our findings to the Swiss population. Study participants have been sampled from an online-panel whose members have been actively recruited using random probability sampling based on national landline telephone directories and random digit dialing of mobile phone numbers. Of course, response probability as well as sample attrition might be correlated to personal characteristics such as online-affinity. Even if the survey sample is prone to selection bias, the sampling process was defined in the whole Swiss population and with the construction of calibration weights we address potential biases which would affect the generalizability of our findings.

Line 49: For the loneliness/isolation measure, was this only asked among people older than 65 years?

Yes, this question was only asked among individuals older than 65 years.

The initial sample from March 2020 consisted of approximately 2,000 participants. Although retention was overall very good, some individuals dropped out during the course of the first 11 waves. Therefore, we decided to replenish the sample while maintaining representativeness. The new participants included in wave 12 were recruited from the same online panel using the same sampling strategy.

Line 182: Could you quantify the level of missingness?

Was multiple testing addressed?

We thank the reviewer for this important question. Our study aim was to investigate a (joint) period effect across different regions in Switzerland. Thus, we aim to investigate a (stratified by region) null hypothesis of no period effect for different study outcomes. However, as stated in our answers to the comments of reviewer #1, we decided not to specify an a priori type-I error of, say 5%. We believe that this a misleading approach given our study design. Given that a multiple comparison adjustment corrects a p-value conditioning under the null hypothesis, this ‘adjusted’ p-value is only valid under the null hypothesis given the true survey sampling mechanism and statistical model. This can of course be justified in a more ‘controlled’ setting (‘best’ example would be a randomized controlled trial with a priori defined interim analysis), but which was not possible for this survey design. Even under a multiple comparison adjustment, we would report our results in the same way, independent of whether an ‘adjusted’ p-value is below or above a certain (adjusted) threshold.

304: typo “varyiance”

Thank you for spotting this typo.

Line 318: What is the purpose of this section “The course of the epidemic situation in Switzerland?” It does not refer to the data/results and seems out of place.

We aimed to describe where the pandemic in Switzerland emerged (Ticino) and progressed (to areas of Zurich and Basel) and how quickly. We thought that this will lead to a better understanding of the incidence pattern in Switzerland, and also where cantonal mitigation measures were selectively implemented over time. We believe that this might explain some of our findings, because it includes important cultural aspects. We reformulated the title of this section to make our intention more explicit: Regional differences in the course of the epidemic.

What are some of the limitations that come with having a dichotomous yes/no for all the outcome variable

A potential limitation is the loss of information from categorical variables. Despite choosing ‘suitable’ cut-offs (for example, combining very poor and poor self-assessed health status), we lose relevant information from the other categories. On the other hand, dichotomization allows us to express estimates in term of percentages, which is more easily understandable than estimates from a Likert-scale. We think that the loss of information is justified by the gain in clarity.

We thank the Editor and the reviewers for their helpful feedback which substantially improved our manuscript.

Marc Höglinger (and Co-authors)

PLoS One. doi: 10.1371/journal.pone.0256253.r003

Decision Letter 1

Barbara Guidi

6 Jul 2021

PONE-D-21-02911R1

Health and social behaviour through pandemic phases in Switzerland: Regional time-trends of the COVID-19 Social Monitor panel study

PLOS ONE

Dear Dr. Höglinger,

==============================

ACADEMIC EDITOR: Please insert comments here and delete this placeholder text when finished. Be sure to:

Indicate which changes you require for acceptance versus which changes you recommend
Address any conflicts between the reviews so that it's clear which advice the authors should follow
Provide specific feedback from your evaluation of the manuscript

Please ensure that your decision is justified on PLOS ONE’s publication criteria and not, for example, on novelty or perceived impact.

For Lab, Study and Registered Report Protocols: These article types are not expected to include results but may include pilot data.

==============================

Please submit your revised manuscript by Aug 20 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

We look forward to receiving your revised manuscript.

Kind regards,

Lucinda Shen

Staff Editor

on behalf of

Barbara Guidi

Academic Editor

PLOS ONE

Journal Requirements:

Additional Editor Comments (if provided):

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

Reviewer #2: All comments have been addressed

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

Reviewer #1: The paper has two main problems that the authors did not address:

First, the statistical method applied inappropriately; second, the statistical interpretation of the results is incorrect.

1- For investigating the pattern of change during the time, the statistical model must have the ability to investigate the change of outcome during the time, but the logistic regression that was applied in this study is only able to estimate the outcome (odds ratio in this paper) for the specific time point and cannot compare the odds ratios in different time points to show whether the pattern of change exists or not; for showing the pattern of change during the time the longitudinal study should be applied.

2- Frequentist statistics uses P-value as a measure of probability to show that the effect exists and the observed effect is statistically meaningful; judging about the existence of effect is not subjective and cannot be shown without P-value (or confidence interval), so reporting p-value is not an optional choice.

Of course, there are some drawback to frequentist inference including the p-value approach to hypothesis testing, but it’s the only method for frequentists to test their hypothesis; The ASA statement is informing researchers to be aware of problems related to p-value and warn them to be cautious about the interpretation of the results; but without having the alternative to p-value, it cannot be omitted from the results; if a researcher does not accept the frequentist approach and their use of p-value for hypothesis testing, he/she can apply another branch of statistics, called Bayesian inference, that addresses some limitation of frequents approach including the use of p-value; otherwise, if the frequentist method is used to analyze the data, their rules must be followed exactly.

The Authors referred to a checklist they used to organize their results in order to justify their reported outcome, however, the checklist only mentioned the key points that must be reported in observational studies and does not explain the details of the application of the statistical method and interpretation of the results; for applying the statistical method correctly and vigorously, the authors need to refer to statistical textbooks.

Reviewer #2: Thank you for the good work in addressing my comments which were addressed to my satisfaction. I believe this manuscript will make an important contribution to the ongoing and rapidly developed COVID19 literature.

Reviewer #3: This manuscript adds to the literature health and social behavior changes associated with COVID-19 and provides a unique angle by examining differences in adherence by language regions. The authors addressed reviewer comments appropriately in the manuscript. All data are fully available online without restriction.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

PLoS One. 2021 Aug 25;16(8):e0256253. doi: 10.1371/journal.pone.0256253.r004

Author response to Decision Letter 1

22 Jul 2021

*** For a better readable form of our response, see the "response to reviewers" file. ***

Re: “Health and social behaviour through pandemic phases in Switzerland: Longitudinal analysis of the COVID-19 Social Monitor panel study”

Winterthur, Juli 22 2021

Dear Prof. Dr. Guidi

We thank you for the opportunity to submit a revised version of our manuscript to PLOS One. Please find below our point-by-point answers to the comments of reviewer #1. Reviewers #2 and #3 had no further comments which need to be addressed in the revised manuscript.

Based on the comment of reviewer #1 we changed the statistical analysis approach towards a hierarchical regression model which accounts for repeated measurements within individuals. Our previous analysis strategy used a cluster-robust approach for addressing repeated measurements which was criticized by reviewer #1. Compared to our previous analyses we found evidence for an interaction effect for the study outcomes poor quality of life (p<0.001), depressive mood (p=0.005), no physical activity (p=0.04), besides the already reported adherence to mitigation measures (p<0.001), from adjusted hierarchical logistic regression models. These findings do not change our previous discussion or interpretation and can likely be explained by the correlation structure between repeated measurements.

Further reviewer #1 required to add p-values for analyses. We added p-values from adjusted regression models (S7 Table).

We updated our discussion about socioeconomic factors with a recently published study which investigated the inverse care law in Switzerland (doi.org/10.1016/S2468-2667(21)00160-2).

We hope that the Editor and reviewer #1 are satisfied with our newly revised manuscript.

Yours sincerely

Marc Höglinger, PhD (on behalf of all co-authors)

PONE-D-21-02911R1

Health and social behaviour through pandemic phases in Switzerland: Regional time-trends of the COVID-19 Social Monitor panel study

PLOS ONE

[ No additional revision points has been raised by the Editor. ]

6. Review Comments to the Author

Reviewer #1: The paper has two main problems that the authors did not address:

First, the statistical method applied inappropriately; second, the statistical interpretation of the results is incorrect.

We appreciate the thoughtful feedback and time efforts of reviewer #1.

Our previous model approach is mentioned in classical statistical textbooks for longitudinal analyses like, for instance, Fixed Effects Regression Methods for Longitudinal Data Using SAS from Paul Allison (ISBN 978-1-59047-568-3), Chapter 3.3. Estimation of Logistic Models for Two or More Observations Per Person. In our previous analysis strategy, we accounted for repeated measurements on individuals with a cluster-robust approach.

Based on the comments from reviewer #1, we changed our modeling strategy to a hierarchical logistic regression model which accounts for repeated measurements on the same individual. That is, we account for nested repeated measurements within individuals while investigating a common fixed effect over periods (Reference: Models for Discrete Longitudinal Data from Molenberghs and Verbeke, doi: 10.1007/0-387-28980-1).

We note that an odds ratio inherently has a reference group, i.e., one compares the odds of the presence of the study outcome of a specific variable category (say, time period 3) with the odds of the reference category (say, time period 1). In our case the reference period is the first period as defined in the manuscript. Thus, our odds ratio estimates are not only for one specific time point – as reviewer #1 says - but compares subsequent time periods with the reference period.

We added to the statistical section that we defined a 5% alpha level as statistically significant and reported p-values in S7 Table. All other p-values for interactions and time effects have already been reported.

As mentioned in the previous rebuttal letter, we based our reasoning to focus on effect size estimation (rather than p-value based assessments of significance) on different statistical textbooks (Essential Medical Statistics by Kirkwood and Sterne, ISBN 978-0-865-42871-3), “Likelihood and Bayesian Inference” from Held and Sabanés Bové (ISBN 978-3-662-60792-3), and “Modern Epidemiology” form Lash, VanderWeele, Haneuse, Rothman (ISBN 978-3-662-60792-3).

We thank reviewer #2 for the helpful comments and feedback, and highly appreciate the time efforts for the review. We note that we changed the statistical analysis approach towards a hierarchical regression model which accounts for repeated measurements within individuals. Our previous analysis strategy used a cluster-robust approach for addressing repeated measurements. Further, we updated our discussion about socioeconomic factors with a recently published study which investigated the inverse care law in Switzerland (doi.org/10.1016/S2468-2667(21)00160-2).

Thank you. We believe that your feedback substantially improved our manuscript, and we highly appreciate the time efforts for the review. We note that we changed the statistical analysis approach towards a hierarchical regression model which accounts for repeated measurements within individuals. Our previous analysis strategy used a cluster-robust approach for addressing repeated measurements. Further, we updated our discussion about socioeconomic factors with a recently published study which investigated the inverse care law in Switzerland (doi.org/10.1016/S2468-2667(21)00160-2).

We thank the Editor and the reviewers for their helpful feedback which substantially improved our manuscript.

Marc Höglinger (and Co-authors)

Attachment

Submitted filename: Letter_R2.docx

Click here for additional data file.^{(24.2KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0256253.r005

Decision Letter 2

Barbara Guidi

4 Aug 2021

Health and social behaviour through pandemic phases in Switzerland: Regional time-trends of the COVID-19 Social Monitor panel study

PONE-D-21-02911R2

Dear Dr. Höglinger,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Barbara Guidi

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #2: No

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #2: Yes

**********

6. Review Comments to the Author

Reviewer #2: My comments had been addressed to my satisfaction. As before, I believe this paper will make an important contribution to this rapidly developing knowledge. Thank you

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: No

PLoS One. doi: 10.1371/journal.pone.0256253.r006

Acceptance letter

Barbara Guidi

11 Aug 2021

PONE-D-21-02911R2

Health and social behaviour through pandemic phases in Switzerland: Regional time-trends of the COVID-19 Social Monitor panel study

Dear Dr. Höglinger:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Barbara Guidi

Academic Editor

PLOS ONE

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Table. The domain, the source and the original question from the survey questionnaire.

(DOCX)

Click here for additional data file.^{(14.2KB, docx)}

S2 Table. Description of missing values.

(DOCX)

Click here for additional data file.^{(15.3KB, docx)}

S3 Table. Underlying results for Figs 3–5.

(DOCX)

Click here for additional data file.^{(68.8KB, docx)}

S4 Table. p-values from likelihood ratio test for period effect, by language region.

(DOCX)

Click here for additional data file.^{(15.2KB, docx)}

S5 Table. p-values from likelihood ratio test for interaction effect between language regions and mitigation periods from unadjusted and adjusted models.

(DOCX)

Click here for additional data file.^{(16.3KB, docx)}

S6 Table. Results for from adjusted logistic regression models.

(DOCX)

Click here for additional data file.^{(49.6KB, docx)}

S1 Text. Statistical methods for calibration weights.

(PDF)

Click here for additional data file.^{(287.4KB, pdf)}

S2 Text. Reproducible analysis example.

(PDF)

Click here for additional data file.^{(175.9KB, pdf)}

S3 Text. Codebook.

(PDF)

Click here for additional data file.^{(36.5KB, pdf)}

S1 Data

(CSV)

Click here for additional data file.^{(1.6MB, csv)}

Attachment

Submitted filename: Letter_R2.docx

Click here for additional data file.^{(24.2KB, docx)}

Data Availability Statement

All relevant data are within the manuscript and its Supporting Information files.

[pone.0256253.ref001] 1.Salathé M, Althaus CL, Neher R, Stringhini S, Hodcroft E, Fellay J, et al. COVID-19 epidemic in Switzerland: on the importance of testing, contact tracing and isolation. Swiss Med Wkly. 2020. doi: 10.4414/smw.2020.20225 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref002] 2.Flaxman S, Mishra S, Gandy A, Unwin HJT, Mellan TA, Coupland H, et al. Estimating the effects of non-pharmaceutical interventions on COVID-19 in Europe. Nature. 2020;584: 257–261. doi: 10.1038/s41586-020-2405-7 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref003] 3.Haug N, Geyrhofer L, Londei A, Dervic E, Desvars-Larrive A, Loreto V, et al. Ranking the effectiveness of worldwide COVID-19 government interventions. Nat Hum Behav. 2020;4: 1303–1312. doi: 10.1038/s41562-020-01009-0 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref004] 4.Abel T, Hofmann K, Schori D. Social and regional variations in health status and health behaviours among Swiss young adults. Swiss Med Wkly. 2013. doi: 10.4414/smw.2013.13901 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref005] 5.Jürgens V, Ess S, Phuleria H, Früh M, Schwenkglenks M, Frick H, et al. Tobacco-related cancer mortality: projections for different geographical regions in Switzerland. Swiss Med Wkly. 2013. doi: 10.4414/smw.2013.13771 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref006] 6.Chatelan A, Beer-Borst S, Randriamiharisoa A, Pasquier J, Blanco J, Siegenthaler S, et al. Major Differences in Diet across Three Linguistic Regions of Switzerland: Results from the First National Nutrition Survey menuCH. Nutrients. 2017;9: 1163. doi: 10.3390/nu9111163 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref007] 7.Spoerri A, Egger M, von Elm E. Mortality from road traffic accidents in Switzerland: Longitudinal and spatial analyses. Accid Anal Prev. 2011;43: 40–48. doi: 10.1016/j.aap.2010.06.009 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref008] 8.Zürcher K, Zwahlen M, Berlin C, Egger M, Fenner L. Trends in influenza vaccination uptake in Switzerland: Swiss Health Survey 2007 and 2012. Swiss Med Wkly. 2019. doi: 10.4414/smw.2019.14705 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref009] 9.Riesen M, Konstantinoudis G, Lang P, Low N, Hatz C, Maeusezahl M, et al. Exploring variation in human papillomavirus vaccination uptake in Switzerland: a multilevel spatial analysis of a national vaccination coverage survey. BMJ Open. 2018;8: e021006. doi: 10.1136/bmjopen-2017-021006 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref010] 10.Airhihenbuwa CO, Iwelunmor J, Munodawafa D, Ford CL, Oni T, Agyemang C, et al. Culture Matters in Communicating the Global Response to COVID-19. Prev Chronic Dis. 2020;17: 200245. doi: 10.5888/pcd17.200245 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref011] 11.Moser A, Carlander M, Wieser S, Hämmig O, Puhan MA, Höglinger M. The COVID-19 Social Monitor longitudinal online panel: Real-time monitoring of social and public health consequences of the COVID-19 emergency in Switzerland. Donnelly TT, editor. PLoS One. 2020;15: e0242129. doi: 10.1371/journal.pone.0242129 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref012] 12.Molenberghs G, Verbeke G. Models for Discrete Longitudinal Data. New York: Springer-Verlag; 2005. doi: 10.1007/0-387-28980-1 [DOI] [Google Scholar]

[pone.0256253.ref013] 13.Carle AC. Fitting multilevel models in complex survey data with design weights: Recommendations. BMC Med Res Methodol. 2009;9: 49. doi: 10.1186/1471-2288-9-49 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref014] 14.R Team Core. R: A language and environment for statistical computing, Vienna, Austria. R Foundation for Statistical Computing; 2020. Available: https://www.r-project.org/ [Google Scholar]

[pone.0256253.ref015] 15.Lumley T. Analysis of Complex Survey Samples. J Stat Softw. 2004;9. doi: 10.18637/jss.v009.i08 [DOI] [Google Scholar]

[pone.0256253.ref016] 16.Odone A, Delmonte D, Scognamiglio T, Signorelli C. COVID-19 deaths in Lombardy, Italy: data in context. Lancet Public Heal. 2020;5: e310. doi: 10.1016/S2468-2667(20)30099-2 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref017] 17.Wirz-Justice A, Graw P, Kräuchi K, Wacker HR. Seasonality in affective disorders in Switzerland. Acta Psychiatr Scand. 2003;108: 92–95. doi: 10.1034/j.1600-0447.2003.00127.x [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref018] 18.Hoffart A, Johnson SU, Ebrahimi O V. Loneliness and Social Distancing During the COVID-19 Pandemic: Risk Factors and Associations With Psychopathology. Front Psychiatry. 2020;11. doi: 10.3389/fpsyt.2020.589127 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref019] 19.Lesser IA, Nienhuis CP. The Impact of COVID-19 on Physical Activity Behavior and Well-Being of Canadians. Int J Environ Res Public Health. 2020;17: 3899. doi: 10.3390/ijerph17113899 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref020] 20.Wang Y, Kala MP, Jafar TH. Factors associated with psychological distress during the coronavirus disease 2019 (COVID-19) pandemic on the predominantly general population: A systematic review and meta-analysis. Murakami M, editor. PLoS One. 2020;15: e0244630. doi: 10.1371/journal.pone.0244630 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref021] 21.Richard A, Faeh D, Rohrmann S, Braun J, Tarnutzer S, Bopp M. Italianity is associated with lower risk of prostate cancer mortality in Switzerland. Cancer Causes Control. 2014;25: 1523–1529. doi: 10.1007/s10552-014-0456-5 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref022] 22.Schmidlin K, Spoerri A, Egger M, Zwahlen M, Stuck A, Clough-Gorr K. Cancer, a disease of aging (part 1)–trends in older adult cancer mortality in Switzerland 1991–2008. Swiss Med Wkly. 2012. doi: 10.4414/smw.2012.13637 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref023] 23.Schmidlin K, Spoerri A, Egger M, Zwahlen M, Stuck A, Clough-Gorr K. Cancer, a disease of aging (part 2)–risk factors for older adult cancer mortality in Switzerland 1991–2008. Swiss Med Wkly. 2012. doi: 10.4414/smw.2012.13607 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref024] 24.Spoerri A, Egger M, Von Elm E. Mortality from road traffic accidents in Switzerland: Longitudinal and spatial analyses. Accid Anal Prev. 2011;43: 40–48. doi: 10.1016/j.aap.2010.06.009 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref025] 25.Spoerri A, Zwahlen M, Panczak R, Egger M, Huss A. Alcohol-selling outlets and mortality in Switzerland-the Swiss National Cohort. Addiction. 2013;108: 1603–1611. doi: 10.1111/add.12218 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref026] 26.Bu F, Steptoe A, Fancourt D. Who is lonely in lockdown? Cross-cohort analyses of predictors of loneliness before and during the COVID-19 pandemic. Public Health. 2020;186: 31–34. doi: 10.1016/j.puhe.2020.06.036 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref027] 27.Groarke JM, Berry E, Graham-Wisener L, McKenna-Plumley PE, McGlinchey E, Armour C. Loneliness in the UK during the COVID-19 pandemic: Cross-sectional results from the COVID-19 Psychological Wellbeing Study. Murakami M, editor. PLoS One. 2020;15: e0239698. doi: 10.1371/journal.pone.0239698 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref028] 28.Hämmig O. Health risks associated with social isolation in general and in young, middle and old age. Ginsberg SD, editor. PLoS One. 2019;14: e0219663. doi: 10.1371/journal.pone.0219663 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref029] 29.Hajek A, De Bock F, Wieler LH, Sprengholz P, Kretzler B, König H-H. Perceptions of Health Care Use in Germany during the COVID-19 Pandemic. Int J Environ Res Public Health. 2020;17: 9351. doi: 10.3390/ijerph17249351 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref030] 30.Kutikov A, Weinberg DS, Edelman MJ, Horwitz EM, Uzzo RG, Fisher RI. A War on Two Fronts: Cancer Care in the Time of COVID-19. Ann Intern Med. 2020;172: 756–758. doi: 10.7326/M20-1133 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref031] 31.Iyengar K, Mabrouk A, Jain VK, Venkatesan A, Vaishya R. Learning opportunities from COVID-19 and future effects on health care system. Diabetes Metab Syndr Clin Res Rev. 2020;14: 943–946. doi: 10.1016/j.dsx.2020.06.036 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref032] 32.Reichert M, Sartelli M, Weigand MA, Doppstadt C, Hecker M, Reinisch-Liese A, et al. Impact of the SARS-CoV-2 pandemic on emergency surgery services—a multi-national survey among WSES members. World J Emerg Surg. 2020;15: 64. doi: 10.1186/s13017-020-00341-0 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref033] 33.Vogel AC, Schmidt H, Loud S, McBurney R, Mateen FJ. Impact of the COVID-19 pandemic on the health care of >1,000 People living with multiple sclerosis: A cross-sectional study. Mult Scler Relat Disord. 2020;46: 102512. doi: 10.1016/j.msard.2020.102512 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref034] 34.Granata T, Bisulli F, Arzimanoglou A, Rocamora R. Did the COVID-19 pandemic silence the needs of people with epilepsy? Epileptic Disord. 2020;22: 439–442. doi: 10.1684/epd.2020.1175 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref035] 35.Leocani L, Diserens K, Moccia M, Caltagirone C. Disability through COVID‐19 pandemic: neurorehabilitation cannot wait. Eur J Neurol. 2020;27. doi: 10.1111/ene.14320 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref036] 36.Sabatello M, Burke TB, McDonald KE, Appelbaum PS. Disability, Ethics, and Health Care in the COVID-19 Pandemic. Am J Public Health. 2020;110: 1523–1527. doi: 10.2105/AJPH.2020.305837 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref037] 37.Panczak R, Luta X, Maessen M, Stuck AE, Berlin C, Schmidlin K, et al. Regional Variation of Cost of Care in the Last 12 Months of Life in Switzerland. Med Care. 2017;55: 155–163. doi: 10.1097/MLR.0000000000000634 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref038] 38.Scheuter C, Wertli MM, Haynes AG, Panczak R, Chiolero A, Perrier A, et al. Unwarranted regional variation in vertebroplasty and kyphoplasty in Switzerland: A population-based small area variation analysis. Kou YR, editor. PLoS One. 2018;13: e0208578. doi: 10.1371/journal.pone.0208578 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref039] 39.Bähler C, Rapold R, Signorell A, Reich O, Panczak R, Blozik E. Regional differences in healthcare costs at the end of life: an observational study using Swiss insurance claims data. Int J Public Health. 2020;65: 969–979. doi: 10.1007/s00038-020-01428-w [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref040] 40.Wei W, Gruebner O, von Wyl V, Dressel H, Ulyte A, Brüngger B, et al. Exploring geographic variation of and influencing factors for utilization of four diabetes management measures in Swiss population using claims data. BMJ Open Diabetes Res Care. 2020;8: e001059. doi: 10.1136/bmjdrc-2019-001059 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref041] 41.Stoller N, Wertli MM, Zaugg TM, Haynes AG, Chiolero A, Rodondi N, et al. Regional variation of hysterectomy for benign uterine diseases in Switzerland. Garzon S, editor. PLoS One. 2020;15: e0233082. doi: 10.1371/journal.pone.0233082 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref042] 42.Wertli MM, Schlapbach JM, Haynes AG, Scheuter C, Jegerlehner SN, Panczak R, et al. Regional variation in hip and knee arthroplasty rates in Switzerland: A population-based small area analysis. Shah TI, editor. PLoS One. 2020;15: e0238287. doi: 10.1371/journal.pone.0238287 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref043] 43.Ulyte A, Wei W, Dressel H, Gruebner O, von Wyl V, Bähler C, et al. Variation of colorectal, breast and prostate cancer screening activity in Switzerland: Influence of insurance, policy and guidelines. Van Hemelrijck M, editor. PLoS One. 2020;15: e0231409. doi: 10.1371/journal.pone.0231409 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref044] 44.Giezendanner S, Bretschneider W, Fischer R, Diaz Hernandez L, Zeller A. The ecology of medical care in Switzerland: prevalence of illness in the community and healthcare utilisation in Switzerland. Swiss Med Wkly. 2020. doi: 10.4414/smw.2020.20221 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref045] 45.Nittas V, von Wyl V. COVID-19 and telehealth: a window of opportunity and its challenges. Swiss Med Wkly. 2020. doi: 10.4414/smw.2020.20284 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref046] 46.Vokinger KN, Nittas V, Witt CM, Fabrikant SI, von Wyl V. Digital health and the COVID-19 epidemic: an assessment framework for apps from an epidemiological and legal perspective. Swiss Med Wkly. 2020. doi: 10.4414/smw.2020.20282 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref047] 47.Plümper T, Neumayer E. Lockdown policies and the dynamics of the first wave of the Sars-CoV-2 pandemic in Europe. J Eur Public Policy. 2020; 1–21. doi: 10.1080/13501763.2020.1847170 [DOI] [Google Scholar]

[pone.0256253.ref048] 48.Wells K, Lurgi M, Collins B, Lucini B, Kao RR, Lloyd AL, et al. Disease control across urban–rural gradients. J R Soc Interface. 2020;17: 20200775. doi: 10.1098/rsif.2020.0775 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref049] 49.Panczak R, Galobardes B, Voorpostel M, Spoerri A, Zwahlen M, Egger M, et al. A Swiss neighbourhood index of socioeconomic position: development and association with mortality. J Epidemiol Community Heal. 2012;66: 1129–1136. doi: 10.1136/jech-2011-200699 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0256253.ref050] 50.Anderson G, Frank JW, Naylor CD, Wodchis W, Feng P. Using socioeconomics to counter health disparities arising from the covid-19 pandemic. BMJ. 2020; m2149. doi: 10.1136/bmj.m2149 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref051] 51.Hämmig O, Gutzwiller F, Kawachi I. The contribution of lifestyle and work factors to social inequalities in self-rated health among the employed population in Switzerland. Soc Sci Med. 2014;121: 74–84. doi: 10.1016/j.socscimed.2014.09.041 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref052] 52.Moser A, Panczak R, Zwahlen M, Clough K, Spörri A, Stuck A, et al. What does your neighbourhood say about you? A study of life expectancy in 1.3 million Swiss neighbourhoods. J Epidemiol Community Health. 2014;68: 1125–1132. doi: 10.1136/jech-2014-204352 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref053] 53.Rimmer A. Covid-19: Tackling health inequalities is more urgent than ever, says new alliance. BMJ. 2020; m4134. doi: 10.1136/bmj.m4134 [DOI] [PubMed] [Google Scholar]

[pone.0256253.ref054] 54.Bethlehem J. Selection Bias in Web Surveys. Int Stat Rev. 2010;78: 161–188. doi: 10.1111/j.1751-5823.2010.00112.x [DOI] [Google Scholar]

[pone.0256253.ref055] 55.Kreitchmann RS, Abad FJ, Ponsoda V, Nieto MD, Morillo D. Controlling for Response Biases in Self-Report Scales: Forced-Choice vs. Psychometric Modeling of Likert Items. Front Psychol. 2019;10. doi: 10.3389/fpsyg.2019.02309 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Health and social behaviour through pandemic phases in Switzerland: Regional time-trends of the COVID-19 Social Monitor panel study

André Moser

Viktor von Wyl

Marc Höglinger

Roles

Abstract

Background

Methods

Results

Discussion

Introduction

Objectives and research hypotheses

Methods

Study population

Data source

Table 1. Schematic overview of COVID-19 Social Monitor survey waves.

Mitigation periods

Table 2. Overview of mitigation periods and implemented mitigation measures.

Study outcomes

Variables of interest and confounding variables

Statistical methods

Ethics statement

Results

Study population

Fig 1. Number of survey participants, by language region.

Table 3. Survey population baseline characteristics and COVID-19 incidence, by language region.

COVID-19 incidence, by mitigation period and language region

Fig 2. New COVID-19 cases per day and 100,000 inhabitants, by mitigation period and language region.

Study outcomes

Fig 3. Proportion of study outcomes for the domains of health, loneliness/isolation and physical activity, by mitigation period, language region and for the whole of Switzerland.

Fig 5. Proportion of study outcomes for the domain of adherence to mitigation measures, by mitigation period, language region and for the whole of Switzerland.

Fig 4. Proportion of study outcomes for the domain of health care use, by mitigation period, language region and for the whole of Switzerland.

Fig 6. Results from adjusted hierarchical logistic regression models for the study outcome domains of health, loneliness/isolation and physical activity, by mitigation period and language region.

Fig 8. Results from adjusted hierarchical logistic regression models for the study outcome domain of adherence to mitigation measures, by mitigation period, language region and for whole Switzerland.

Fig 7. Results from adjusted hierarchical logistic regression models for the study outcome domain of health care use, by mitigation period, language region and for the whole of Switzerland.

Discussion

Summary of main findings

Regional differences in the course of the epidemic situation

Changes in health and social behavior

Health care use during the Coronavirus pandemic

Geographical and socioeconomic factors

Strengths and limitations

Conclusion

Supporting information

Acknowledgments

Data Availability

Funding Statement

References

Decision Letter 0

Barbara Guidi

Roles

Author response to Decision Letter 0

Decision Letter 1

Barbara Guidi

Roles

Author response to Decision Letter 1

Decision Letter 2

Barbara Guidi

Roles

Acceptance letter

Barbara Guidi

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases