Fear of COVID-19 among patients with prior SARS-CoV-2 infection: A cross-sectional study in Estonian family practices

Abstract

Background

Fear of coronavirus disease (COVID-19) has been associated with significant health effects.

Objectives

To assess COVID-19 fear and investigate factors associated with higher fear among COVID-19 survivors over 6 months after infection.

Methods

Cross-sectional study using multistage sampling (family practices within the highest 5th percentile of numbers of SARS-CoV-2 infected patients and random sample of patients within these practices) performed from March 15 to 17 July 2021. Adult patients with a laboratory-confirmed history of COVID-19 were recruited for a self-administered 79-item questionnaire including demographics, self-rated health, physical activity, COVID-19 characteristics, severity and the fear of COVID-19 Scale (FCV-19S). Comorbidity data were extracted from Estonian Health Insurance Fund. Logistic regression models were used to evaluate factors associated with COVID-19 fear.

Results

Of 341 participants included, 60% were women, 24.2% were hospitalised due to COVID-19 and 22.2% had long COVID, 143 (42%) participants reported high levels of fear (cut-off FCV-19S >17.8). Higher fear was associated with being female (aOR 2.12, 95% CI 1.14–3.95), age ≥61 years (aOR 3.23, 95% CI 1.28–8.16), two-member-households (aOR 3.70, 95% CI 1.40–9.77) physical inactivity 6 months prior to COVID-19 (aOR 3.53, 95% CI 1.26–9.95), and symptom severity during acute COVID-19. Long COVID was not associated with higher COVID-19 fear (aOR 1.82 95% CI 0.91–3.63).

Conclusion

Almost half of participants reported COVID-19 fear more than 6 months after infection. Greater fear was associated with sociodemographic factors, physical activity prior to COVID-19 and COVID-19 symptom severity. There is a need to target this population to develop appropriate interventions.

Keywords:

• FCV-19S
• COVID-19 fear scale
• post covid
• primary care
• long covid

KEY MESSAGES

• COVID-19 fear is common among COVID-19 survivors even over 6 months after acute infection.

• Individuals with long COVID do not exhibit higher fear than those without.

• Higher fear levels are associated with being female, being older, living in two member households, pre pandemic physical inactivity and COVID-19 symptom severity.

Introduction

In March 2020, coronavirus disease 2019 (COVID-19) was declared a global pandemic by the World Health Organisation (WHO) [1]. Since then, mental health has deteriorated compared with pre pandemic trends [2]. In particular, COVID-19 survivors are believed to be at risk of a decrease in mental well-being [3]. Due to their communicable and invisible nature, infections are a unique disorder that causes fear among people [4]. Fear is an adequate reaction to a threat; at the same time, a high level of fear can be interpreted as an inadequate fear presentation due to its close relationship with anxiety [5].

Higher COVID-19 fear has been associated with both positive and negative health factors. Fear has enhanced cautious behaviour, such as social distancing and improved hand hygiene [6]. However, it has also been associated with factors that deteriorate mental health, such as anxiety [6,7], stress [7], depression [6,7] and perceptions of decreased quality of life [6–8]. Greater fear of COVID-19 has been associated with being female, older and lower income [7]. After acute infection, patients are at risk of developing long COVID syndrome [9].

There is a lack of studies focussing on fear among COVID-19 survivors who experienced varying degrees of COVID-19 severity and assessing whether long COVID is associated with increased fear. Therefore, we aimed to investigate the level of and factors associated with COVID-19 fear among adults with a history of SARS-CoV-2 infection. Our second aim was to explore whether long COVID contributes to fear.

Methods

Study design

This cross-sectional study was conducted among Estonian family practice (FP) patients with a history of SARS-CoV-2 infection 6–14 months before enrolling. The study adheres to the Helsinki declaration’s ethical principles for medical research. Ethical approval for the study was received from the Tartu University Ethics Committee for Human Studies (protocol number 357/M-27).

Recruitment and selection of study subjects

Multistage sampling was used, which meant first involving a convenience sample of FP patients (amongst the 25 FP within the highest 5th percentile of numbers of SARS-CoV-2 infected patients) and then including a random sample of COVID-19 survivors. The target sample size was 370 COVID-19 survivors. Assuming a 65% response rate, the total sample retrieved from the family practices was 585 participants.

Seventeen FP centres in Tallinn (capital), Saaremaa (the island with the most COVID-19 cases in 2020) and Tartu (the second largest city) consented to participate in the study. Figure 1 shows the total COVID-19 cases in Estonia in March 2020 and July 2021. Appropriate training on the practical and ethical aspects of the study was given to the staff of these centres. Recruitment of COVID-19 survivors took place between March 15 and 17 July 2021. Inclusion criteria were as follows: (i) 18 years or older; (ii) laboratory confirmed COVID-19 (nasopharyngeal specimen positive for SARS-CoV-2 RNA) in the period from February 25 to 14 September 2020; (iii) resident of Estonia and a patient of a FP physician enrolled in the study; (iv) Estonian or Russian language speaker; and (v) able to provide informed consent.

Figure 1. Total COVID-19 cases in Estonian regions. The left map shows data from 15 March 2020. The right map shows data from 17 July 2021 [10].

Participants were identified in the national communicable diseases surveillance database at the Estonian Health Board and linked with the FPs using personal identification codes [11]. Nearly all people living in Estonia with a valid residential permit are registered with an FP. The FP is the first place to contact regarding most health issues [12]. All individuals meeting the first three eligibility criteria were invited by FP staff via telephone, email, or direct contact to schedule a study appointment. Participants were thoroughly instructed on the study procedure and asked to sign a written consent form before completing the questionnaire.

Data collection

Self-administered questionnaire

Participants were asked to complete a multiple-choice questionnaire, including demographic information, pre acute and post acute COVID-19 health status, COVID-19 severity and COVID-19 fear [13,14]. The questionnaire was available in Estonian and Russian. The questionnaire was translated according to relevant guidelines [15]. The study visit took approximately 60 min. A €20 food shop gift card was given as a participant’s reward.

Estonian Health Insurance Fund data

Data on comorbidities preceding COVID-19 infection by three years were abstracted from the Estonian Health Insurance Fund (EHIF) database. The EHIF electronic administrative database contains personal information (sex, age), information on health care utilisation (date of service, type of service provided) and diagnostic codes based on the International Classification of Diseases, tenth revision.

Variables

A five-item self-rated instrument was used to measure participants’ self-rated health 6 months prior to infection. Responses were categorised into four groups (bad, n = 10, and very bad, n = 1, were merged into one group). Self-rated physical activity (defined as half an hour of activity leading to an increase in heart rate and respiration) during the past 6 months was categorised as follows: ≥2 times per week, once or less per week, inactive (sporadically during the year or never) and unable due to injury or illness. Data on experienced symptoms were requested for three points in time: 6 months prior to COVID-19, during COVID-19, and at the time of study participation. Data about self-reported symptoms at the time of study participation were used to identify patients with long COVID.

To characterise COVID-19, three measures were used: (i) self-reported physical state during acute COVID-19 (classified into three groups: usual activity, limited activity, which meant being able to carry out vital household tasks, and bedridden); (ii) treatment setting (hospitalised vs. not); and (iii) assessment of long COVID. For long COVID, the WHO clinical case definition was used. Long COVID is defined as persistent or new onset symptoms in individuals with confirmed or suspected COVID-19 for at least 2 months within 3 months from the onset of COVID-19 that cannot be explained by an alternative diagnosis and that affect daily functioning [9]. Onset of at least one new symptom in combination with a perceived health-related limitation on usual activities was classified as long COVID.

Comorbidity burden was calculated based on the data from the EHIF according to the Charlson Comorbidity Index (CCI) for classifying comorbidity severity [16,17]. A CCI score of 0 points was classified as no comorbidity, while a CCI score >0 was classified as comorbidity.

Outcome variables

To assess COVID-19 fear, the seven-item instrument Fear of COVID-19 Scale (FCV-19S) was used [14]. Each item was answered on a seven-point Likert-scale ranging from strongly disagree to strongly agree, and the total score ranged from 7 to 49 (the higher the score was, the greater the COVID-19 fear). Fear was categorised as high or low with the cut-off at the sample mean value. Participants scoring > 17.8 were categorised as having a high level of COVID-19 fear, whereas participants scoring ≤ 17.8 were categorised as having a low level of COVID-19 fear.

Statistical analysis

First, a descriptive analysis was performed. Numbers and percentages were used to describe categorical variables. Participants missing answers on ≥ 4 items on the FCV-19S (n = 6) were excluded from the analysis. Data for participants lacking fewer than three answers (n = 6, total missing answers n = 8) were imputed. Imputing was done by calculating the mean item value for each individual, which then replaced the missing value.

To assess differences between variables and mean FCV-19S score between those with respectively without high COVID-19 fear, we used t-tests for two-group analysis and analysis of variance (ANOVA) for multiple comparisons. To correct for multiple comparisons, Bonferroni correction was used. The significance level was set as <0.05.

For analysis of factors associated with fear, the FCV-19S score was dichotomised with the cut-off at the sample mean value. To examine associations with a high level of COVID-19 fear, we used bivariate and multivariate logistic regression, from which odds ratios (ORs) and adjusted ORs (aORs) with corresponding 95% confidence intervals (CIs) were estimated. For all statistical analyses, STATA version 17 was used.

Results

Characteristics of participants

The inclusion process is shown in Figure 2. In total, 341 participants met the inclusion criteria and completed enough answers on the FCV-19S.

Figure 2. Flowchart showing the study recruitment process. Abbreviations. n: number; FCV-19S: Fear of COVID-19 Scale.

Comparing the overall age and sex distribution of study participants with nonparticipants, we observed an overrepresentation of women (p < 0.01) and younger participants (mean age 51.8 (SD 17.8) vs. 61.3 (SD 26.0) years, p < 0.01). The baseline characteristics of the participants are presented in Tables 1 and 2. Most participants were female (60%), working (78.9%) and living in a city (79.9%). Few respondents (9.7%) had CCI ≥1. The mean age was 49.4 years (SD 15.3), ranging from 18 to 89 years. The mean time elapsed since COVID-19 infection was 276 days (SD 87.85 days, ranging from 63 to 449 days).

Table 1. Level of fear of study participants (n = 341) according to sociodemographic characteristics.

		High COVID-19 fear, n (%)		Bivariate analysis
Variable	Subgroup	Yes	No	OR (95% CI)	p Value
Sociodemographic characteristics
Sex	Male	45 (32.9)	92 (67.1)	1
	Female	98 (48.0)	106 (52.0)	1.89 (1.21–2.96)	0.006
Age	≤50 years	56 (32.0)	119 (68.0)	1
	51–60	34 (42.5)	46 (57.5)	1.57 (0.91–2.71)	0.105
	≥61	53 (61.6)	33 (38.4)	3.41 (1.99–5.85)	<0.001
Education*	<10 years	13 (56.5)	10 (43.5)	1
	10–12 years	79 (46.8)	90 (53.2)	0.68 (0.28–1.62)	0.381
	> 12 years	50 (34.0)	97 (66.0)	0.40 (0.16–0.97)	0.042
Nationality^†	Estonian	92 (40.9)	133 (59.1)	1
	Other	50 (43.5)	65 (56.5)	1.11 (0.71–1.75)	0.647
Household members^‡	1	17 (34.7)	32 (65.3)	1
	2	67 (52.3)	61 (47.7)	2.07 (1.04–4.09)	0.037
	≥3	58 (35.6)	105 (64.4)	1.04 (0.53–2.03)	0.909
Place of residence	City	111 (41.0)	160 (59.0)	1
	Rural	31 (45.6)	37 (54.4)	1.21 (0.71–2.06)	0.489
Children < 18 years in household	No	100 (45.1)	122 (54.9)	1
	Yes	39 (34.5)	74 (65.5)	0.64 (0.40–1.03)	0.065
Occupation^§	Working	101 (38.0)	165 (62.0)	1
	Other	40 (56.3)	31 (43.7)	2.11 (1.24–3.58)	0.006

Abbreviations. n: number; SD: standard deviation; OR: odds ratio; CI: confidence interval. Statistically significant values are marked in bold.

Explanation to footnotes: *Education was trichotomised into: < 10 years (primary school), 10–12 years (secondary and vocational school) and >12 years (higher or university education). ^†Nationality was defined as Estonian or Other, the majority of other classified themselves as Russian (n = 94). ^‡Household members were categorised according to the numbers of members, where 1 meant living alone. ^§ Occupation was classified as working (full time or part time) or other (unemployed, retired, maternity/paternity leave, sick leave due to chronic illness, studying, on leave, military service).

Table 2. Level of fear of study participants (n = 341) according to self-reported health status before COVID-19, self-reported COVID-19 severity and experiencing long-COVID.

		High COVID-19 fear, n (%)		Bivariate logistic regression
Variable	Subgroup	Yes	No	OR (95% CI)	p Value
Pre-COVID-19 health status
Comorbidities (EHIF)	None	124 (40.5)	182 (59.5)	1
	Any	18 (54.6)	15 (45.5)	1.76 (0.86–3.63)	0.125
Self-rated health status	Very good	32 (33.7)	63 (66.3)	1
	Good	62 (38.3)	100 (61.7)	1.22 (0.72–2.07)	0.461
	Average	43 (61.4)	27 (38.6)	3.14 (1.65–5.96)	<0.001
	Bad/very bad	4 (36.4)	7 (63.6)	1.13 (0.31–4.13)	0.859
Self-rated physical activity	≥2 times/week	83 (36.7)	143 (63.3)	1
	Once or less/week	30 (44.1)	38 (55.9)	1.36 (0.78–2.36)	0.273
	Inactive	21 (63.6)	12 (36.4)	3.02 (1.41–6.44)	0.004
	Unable due to injury or illness	8 (61.5)	5 (38.5)	2.77 (0.87–8.70)	0.084
Self-reported symptoms within 6 months before COVID-19
Fatigue	No	99 (37.8)	163 (62.2)	1
	Yes	43 (56.6)	33 (43.4)	2.15 (1.28–3.60)	0.004
Sadness	No	119 (40.2)	177 (59.8)	1
	Yes	22 (52.4)	20 (47.6)	1.64 (0.86–3.13)	0.137
Sleeping problems	No	81 (40.9)	117 (59.1)	1
	Yes	62 (43.7)	80 (56.3)	1.12 (0.72–1.73)	0.612
Anxiety	No	100 (40.0)	150 (60.0)	1
	Yes	41 (47.1)	46 (52.8)	1.34 (0.82–2.18)	0.247
Panic attacks	No	133 (41.2)	190 (58.8)	1
	Yes	9 (60.0)	6 (40.0)	2.14 (0.75–6.16)	0.157
Agitation/ restlessness or depression	None	97 (38.2)	157 (61.8)	1
	Sometimes/often	43 (51.8)	40 (48.2)	1.74 (1.06–2.87)	0.030
Pain or ailments	None	76 (35.9)	136 (64.1)	1
	Sometimes/often	63 (50.8)	61 (49.2)	1.85 (1.18–2.90)	0.008
Self-reported COVID-19 severity
Physical state during COVID-19	Usual activity	39 (29.1)	95 (70.9)	1
	Limited activity	49 (45.8)	58 (54.2)	2.06 (1.21–3.51)	0.008
	Bedridden	52 (53.6)	45 (46.4)	2.81 (1.63–4.86)	<0.001
Hospitalisation due to COVID-19	No	94 (36.6)	163 (63.4)	1
	Yes	48 (58.5)	34 (41.5)	2.45 (1.47–4.07)	<0.001
Long COVID	No	89 (37.2)	150 (62.8)	1.00
	Yes	35 (51.5)	33 (48.5)	1.79 (1.04–3.08)	0.036

Abbreviations: n: number; SD: standard deviation; OR: Odds Ratio; CI: Confidence Interval; EHIF: Estonian Health Insurance Fund. Statistically significant values are marked in bold.

COVID-19 fear among participants

COVID-19 fear among participants is presented in Figure 3. The mean score on the FCV-19S was 17.78 (SD 9.56, range 7–49). In total, 143 (42%) participants showed high fear. For the distribution of FCV-19S scores among subgroups, see Table A1 in the Appendix.

Figure 3. Mean fear scores on the COVID-19 fear Scale (FCV-19S) among individuals with a prior COVID-19 infection, Estonia 2021. Abbreviations. SD: standard deviation; n: number; 95% CI 95%: confidential interval.

Factors associated with COVID-19 fear

In total, 14 factors were significantly associated with COVID-19 fear in the bivariate analysis. Based on the multivariate analysis (presented in Table 3, the factors associated with higher fear were female sex (aOR 2.12, 95% CI 1.14–3.95), age ≥61 years (aOR 3.23, CI 1.28–8.16), household consisting of two members (aOR 3.70, 95% CI 1.40–9.77), being physically inactive (and not prohibited by any injury or illness) within 6 months prior to COVID-19 (aOR 3.53, 95% CI 1.26–9.95), and experiencing more physically debilitating symptoms during COVID-19 (usual activity vs. limited activity aOR 2.14, 95% CI 1.08–4.24; usual activity vs. bedridden aOR 2.66, 95% CI 1.21–5.87) during COVID-19.

Table 3. Multivariable analysis of association between demographic and self-reported clinical variables and fear among COVID-19 survivors followed in Estonian family practice.

		Multiple logistic regression
Variable	Subgroup	aOR (95% CI)	p Value
Sociodemographic characteristics
Sex	Male	1
	Female	2.12 (1.14–3.95)	0.018
Age	≤50 years	1
	51–60 years	1.13 (0.50–2.54)	0.772
	≥61 years	3.23 (1.28–8.16)	0.013
Education*	<10 years	1
	10–12 years	0.90 (0.26–3.06)	0.861
	>12 years	0.48 (0.14–1.73)	0.264
Nationality^†	Estonian	1
	Other	0.58 (0.30–1.12)	0.106
Household members^‡	1	1
	2	3.70 (1.40–9.77)	0.008
	≥3	2.17 (0.73–6.49)	0.164
Place of residence	City	1
	Rural	1.14 (0.57–2.28)	0.704
Children <18 years in household	No	1
	Yes	1.12 (0.47–2.69)	0.794
Occupation^§	Working	1
	Other	0.94 (0.46–1.93)	0.870
Comorbidities (EHIF)	None	1
	Any	0.56 (0.20–1.60)	0.282
Self-rated health status	Very good	1
	Good	0.84 (0.41–1.70)	0.626
	Average	1.48 (0.60–3.68)	0.394
	Bad/very bad	0.30 (0.05–1.99)	0.213
Self-reported physical activity	≥2 times/ week	1
	Once or less/ week	1.16 (0.58–2.31)	0.675
	Inactive	3.53 (1.26–9.95)	0.017
	Unable due to injury or illness	1.00 (0.19–5.16)	1.000
Self-reported symptoms within 6 months before COVID-19
Fatigue	No	1
	Yes	1.83 (0.85–3.94)	0.121
Sadness	No	1
	Yes	0.93 (0.33–2.61)	0.896
Sleeping problems	No	1
	Yes	0.70 (0.37–1.33)	0.279
Anxiety	No	1
	Yes	0.96 (0.42–2.20)	0.931
Panic attacks	No	1
	Yes	1.21 (0.30–4.93)	0.793
Agitation/restlessness or depression	None	1
	Sometimes/ often	2.48 (0.99–6.23)	0.053
Pain or ailments	None	1
	Sometimes/ often	1.10 (0.57–2.14)	0.772
Self-reported COVID-19 severity
Physical state during COVID-19	Usual activity	1
	Limited activity	2.14 (1.08–4.24)	0.029
	Bedridden	2.66 (1.21–5.87)	0.015
Hospitalisation due to COVID-19	No	1
	Yes	1.11 (0.48–2.55)	0.807
Long COVID	No	1
	Yes	1.82 (0.91–3.63)	0.091

Abbreviations: n: number; aOR: adjusted Odds Ratio; CI: Confidence Interval; EHIF: Estonian Health Insurance Fund.

Statistically significant values are marked in bold.

Explanation to footnotes: * Education was trichotomised into: < 10 years (primary school), 10–12 years (secondary and vocational school) and > 12 years (higher or university education). ^†Nationality was defined as Estonian or Other, the majority of other classified themselves as Russian (n = 94). ^‡Household members were categorised according to the numbers of members, where 1 meant living alone. ^§Occupation was classified as working (full time or part time) or other (unemployed, retired, maternity/ paternity leave, sick leave due to chronic illness, studying, on leave, military service).

Discussion

Main findings

In this study, we assessed fear among COVID-19 survivors and examined factors associated with a higher fear level, including whether having long COVID contributed to COVID-19 fear. A sizeable proportion (42%) of participants reported high COVID-19 fear for more than 6 months after acute infection. Sociodemographic characteristics were associated with higher fear as well as pre-COVID-19 physical activity and COVID-19 symptom severity. However, long COVID was not independently associated with higher fear.

Interpretation

In line with previous studies, we found that older age and being female were associated with higher levels of fear [7,8]. Elderly individuals are at higher risk for severe COVID-19 infection [18]. Fear in general has been shown to be greater among women than men, which might be due to differences in patterns of gender socialisation [19]. Additionally, women have suffered from worse overall mental health during the pandemic [20].

Low levels of self-reported physical activity were associated with higher levels of fear. Physical activity has been associated with fewer anxiety symptoms during the COVID-19 pandemic [21]. This suggests that primary health care workers should enquire about and encourage patients’ physical activity, as this may be a way to ease fear and improve patient health.

In contrast to prior studies that indicated higher fear among singles [22,23], we found that households consisting of two persons were associated with higher fear. Two-member households might be a surrogate marker for older age or other sociodemographic factors we did not account for. However, worry about loved ones becoming infected with COVID-19 is a predictor of fear, although it was not measured with the FCV-19S [24].

Patients suffering more severely during the acute infection phase (irrespective of being hospitalised for COVID-19) displayed higher levels of fear, which might be interpreted as an expression of adequate fear due to COVID-19 being a fatal disease [25]. Interestingly, we found no association between higher fear level and long COVID or hospitalisation. Previously, worry about COVID-19 has been found to increase the risk of long COVID [26]. Hospitalisation and long COVID could be seen as markers for disease severity and be thought to impact fear level. However, our study oversampled patients hospitalised because of COVID-19. This might inflate our sample-based estimate on COVID-19 fear. Nonetheless, patients’ physical state remained significant in multivariate analysis. This shows that even among patients not hospitalised, the infection considerably affected the health status. It might be that hospitalised patients felt better-taken care of and therefore were less worried.

The classification of COVID-19 fear has varied in previous studies: some have considered the answers of agree or strongly agree to any item on the FCV-19S as high fear [8], others have calculated fear levels with receiver operating analysis and others have, similarly to this study, used the mean value as the cut-off [27,28]. A meta-analysis reviewing fear with the FCV-19S globally found that fear levels were similar worldwide. However, Asian countries expressed the highest fear, whereas the Australian continent reported the lowest [29]. Even though the absolute difference was small in number, geographical, cultural and sociodemographic factors might contribute to differences in fear levels. Numbers of individuals experiencing fear also vary due to differences in measurement methods and periods during the pandemic. A better understanding of how fear changes over time and which levels of fear contribute to worsening mental health is needed.

Strengths and limitations

To our knowledge, this is the first study to target fear among COVID-19 survivors in FP. A strength of this study is the participant recruitment method reflecting the general population since almost everyone living in Estonia is registered as FP patient. However, the results might be biased by some individuals being more prone to non-participation (e.g. older individuals with poor health and men). The nonresponse bias and oversampling of patients hospitalised with COVID-19 might bias our high COVID-19 fear prevalence estimate. Using multiple regression analysis to explore factors associated with high fear mitigates some of this bias for the high fear correlates. We used a seven-point Likert scale to assess fear, which complicates comparisons between absolute figures on the scale, although it may add greater characterisation to the fear. The FCV-19S has not been validated for an Estonian population. However, the FCV-19S is a strong and valid instrument for assessing fear across different languages [30,31] including several European countries. Using this instrument allows for comparisons of fear levels with other countries.

In this study, we used self-reported health and symptoms as health status measures. Self-rated health has been proven to be a valid and efficient measure of physical and mental health across the European continent [32]. Self-reported symptoms are considered the golden standard for measuring subjective experiences since the information comes directly from the patient.

The study design is limited by its incapacity to assess temporality. However, we used the format ‘time-line follow-back’, which anchors the respondent in their symptoms and health status prior to COVID-19 and at the time of the disease. Furthermore, focusing on events before the infection might imply recall bias. No symptom scales were used to assess depression, anxiety, panic disorders and sleep problems in the study population. Therefore, the results do not necessarily reflect the actual prevalence of these disorders. Additionally, we used participants’ self-reported symptom burden before and after initial infection to classify long COVID. No clinical examination of the participants was conducted regarding the diagnosis. This might therefore be seen as a subjective measure and induce bias since it is based on patient reports. Another limitation is using the CCI, which was developed as a mortality predictor with strict comorbidity classification. However, it is commonly used and enables comparisons with prior studies.

Implications

As higher fear may affect and deteriorate overall mental health in the population, there is a need to identify vulnerable groups to target with correct information to help increase their health literacy. FP doctors are in a perfect position to screen fear and tackle mental health deterioration due to their role as healthcare gatekeepers and the broad spectrum of patients they see. A first step would be to ask patients whether they experience fear and assess whether they need psychological treatment to handle the fear. Previously, treatment with cognitive behavioural therapy through the internet is helpful in reducing anxiety [33], which could be a proper means of treating people with high fear during pandemics. In addition, the pedagogic aspect of the doctor’s role cannot be forgotten.

Conclusion

In this cross-sectional study involving 341 FP patients in Estonia with a historically positive SARS-CoV-2 PCR test, almost half reported great fear. Higher levels of fear were associated with sociodemographic factors, self-rated physical activity prior to COVID-19 and disease severity during COVID-19.

Authors contribution

AÜ, KS, RK and TM designed the study and the study questionnaire and performed data collection. AS and SV carried out the analysis. AS wrote the first draft. All authors participated in the interpretation of the results and in critically reviewing the manuscript.

Acknowledgements

We would like to express our deepest gratitude to the entire team of researchers from the Institute of Technology, University of Tartu. The paper and the research behind it would not have been possible without their substantial support. Alla Piirsoo and Marko Piirsoo contributed significantly to the study design, execution, acquisition of data, and analysis. We also thank Karolin Toompere for remarks on statistical analysis. We are also deeply grateful to all family care practicioners and patients for their participation in the study and for making the study possible.

Disclosure statement

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Additional information

Funding

This work was supported by the European Regional Development Fund (RITA 1/02-120).

Appendix

Table A1. Characteristics of study participants (n = 341) and COVID-19 fear scale.

Variable	Subgroup	n (%)	Mean (SD)	p Value
Sociodemographic characteristics
Sex	Male	137 (40)	15.92 (8.52)	p = 0.003
	Female	204 (60)	19.03 (10.03)
Age	≤50 years	175 (51)	15.13 (7.63)	p < 0.001
	51–60 years	80 (24)	18.54 (10.17)
	>61 years	86 (25)	22.46 (10.64)
Education*	<10 years	23 (6.8)	21.83 (11.91)	p = 0.015
	10–12 years	169 (49.9)	18.47 (10.16)
	>12 years	147 (43.5)	16.34 (8.2)
Nationality^†	Estonian	225 (66.2)	17.27 (8.96)	p = 0.184
	Other	115 (33.8)	18.73 (10.65)
Household members^‡	1	49 (14.4)	16.84 (9.95)	p = 0.030
	2	128 (37.7)	19.54 (10.38)
	≥ 3	163 (47.9)	16.66 (8.61)
Place of residence	City	271 (79.9)	17.84 (9.66)	p = 0.853
	Rural	68 (20.1)	17.60 (9.33)
Children < 18 years in household	No	222 (66.3)	18.54 (10.02)	p = 0.020
	Yes	113 (33.7)	15.99 (8.11)
Occupation^§	Working	266 (78.9)	16.65 (8.69)	p < 0.001
	Other	71 (21.1)	21.99 (11.36)
Pre-COVID-19 health status
Comorbidities (EHIF)	None	306 (90.3)	17.35 (9.21)	p = 0.013
	Any	33 (9.7)	21.73 (12.01)
Self-rated health status	Very good	95 (28.1)	14.88 (7.47)	p < 0.001
	Good	162 (47.9)	16.94 (8.61)
	Average	70 (20.7)	23.26 (11.81)
	Bad/Very bad	11 (3.3)	20.36 (10.73)
Self-reported physical activity	≥2 times/ week	226 (66.5)	16.88 (9.44)	p = 0.001
	Once or less/week	68 (20)	17.54 (8.06)
	Inactive	33 (9.7)	21.09 (8.73)
	Unable due to injury or illness	13 (3.8)	26.08 (15.17)
Self-reported symptoms within 6 months before COVID-19
Fatigue	No	262 (77.5)	16.61 (8.45)	p < 0.001
	Yes	76 (22.5)	21.87 (11.82)
Sadness	No	296 (87.6)	17.41 (9.14)	p = 0.074
	Yes	42 (12.4)	20.24 (12.07)
Sleeping problems	No	198 (58.2)	16.67 (8.05)	p = 0.010
	Yes	142 (41.8)	19.39 (11.19)
Anxiety	No	250 (74.2)	17.18 (8.65)	p = 0.052
	Yes	87 (25.8)	19.49 (11.76)
Panic attacks	No	323 (95.6)	17.50 (9.17)	p = 0.003
	Yes	15 (4.4)	24.87 (14.74)
Agitation/ restlessness or depression	None	254 (75.4)	16.88 (8.76)	p < 0.004
	Sometimes/often	83 (24.6)	20.39 (11.42)
Pain or ailments	None	212 (63.1)	16.07 (8.30)	p < 0.001
	Sometimes/often	124 (36.9)	20.26 (10.70)
Self-reported COVID-19 severity
Physical state during COVID-19	Usual activity	134 (39.6)	15.01 (7.97)	p < 0.001
	Limited activity	107 (31.7)	18.11 (9.44)
	Bedridden	97 (28.7)	20.93 (10.63)
Hospitalisation due to COVID-19	No	257 (75.8)	16.41 (8.69)	p < 0.001
	Yes	82 (24.2)	22.07 (10.96)
Long COVID	No	239 (77.9)	16.83 (9.06)	p = 0.059
	Yes	68 (22.2)	19.21 (9.29)

Abbreviations. n: number; SD: standard deviation; EHIF: Estonian Health Insurance Fund.

Statistically significant values are marked in bold.

Explanation to footnotes: *Education was trichotomised into: < 10 years (primary school), 10–12 years (secondary and vocational school) and > 12 years (higher or university education). ^†Nationality was defined as Estonian or Other, the majority of other classified themselves as Russian (n = 94). ^‡Household members were categorised according to the numbers of members, where 1 meant living alone. ^§Occupation was classified as working (full time or part time) or other (unemployed, retired, maternity/paternity leave, sick leave due to chronic illness, studying, on leave, military service).