Social distancing during the COVID-19 pandemic and health-related quality of life among esophageal cancer survivors

Abstract

Purpose

The purpose was to investigate whether social distancing during the COVID-19 pandemic reduced health-related quality of life (HRQL) in esophageal cancer patients and if so, to identify factors related to the HRQL changes.

Methods

A prospective Swedish nationwide study of patients who undergone surgery for esophageal cancer between 2013 and 2019. Telephone interviews were conducted 5 weeks and 13 weeks after the introduction of social distancing recommendations. The participants responded to a few scales and items from the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire C30. Regression models adjusted for potential confounders were used to examine mean score differences (MSD) with 95% confidence intervals (CI) between compliance with the recommendations and HRQL.

Results

In total, 134 individuals participated in the study. At 5 weeks, a reduction in role function was seen for individuals who fully (MSD −8, 95% CI: −16 to 0) and to a large extent (MSD −19, 95% CI: −26 to −12) complied with the recommendations. Less fatigue (MSD 9, 95% CI: 3–15) was also detected for this group. Being female, elderly, having comorbidities, living in a region with higher COVID-19 incidence, living in a villa, and being considered as an at-risk person were related to changes in role function. At 13 weeks, reductions in role function continued for those who fully complied with the recommendations (MSD −10, 95% CI: −19 to −1), but the risk variables were of less importance. Improvements in fatigue were no longer detected. Global quality of life, emotional function, or insomnia remained stable over time.

Conclusions

This study indicates that individuals who undergone surgery for esophageal cancer and fully or to a large extent complied with the recommendations experienced reductions in role function, but not in global quality of life or emotional function, during the COVID-19 pandemic.

Keywords:

• Esophageal neoplasm
• cancer survivorship
• HRQL
• isolation

Background

The COVID-19 pandemic has had a profound impact on society all around the world. Governments across the world have promoted social distancing as a part of the strategy to cope with the COVID-19 outbreak. In Sweden, the response to the pandemic has been partly based on voluntary actions rather than enforce a nationwide lock-down. The strategy has been to protect groups at risk of becoming seriously ill with the virus and keep the numbers of infected people as low as possible to ensure that hospitals can continually provide care for COVID-19 patients. To reach these goals, the Swedish Public Health Agency has imposed recommendations of social distancing for the population in general, such as to keep an arm’s length distance from others, to avoid non-essential domestic transport, and work from home if possible [1]. More restrict recommendations, for example, to stay at home with the exception of outdoor walks, to keep socially distanced from other people and avoid meeting family members from other households, avoid all environments with gatherings, such as shopping, public spaces, and public transport, were declared for those considered at increased risk for becoming seriously ill if infected, for example, individuals above 70 years, those with comorbidities, or having immunosuppressive disease or treatments, for example, cancer treatment, [1,2]. A particularly vulnerable group of cancer patients is those diagnosed with esophageal cancer. Esophageal cancer carries a poor prognosis despite curatively intended extensive treatment with oncological therapy in combination with surgery. These individuals might be more likely to need to socially distance because of frail health and intensive adjuvant therapy [3–5].

There is a thin line between social distance and social isolation. Previous studies on COVID-19 have shown that social distancing has been associated with inactivity, smoking, alcohol abuse, unhealthy diet, depression, poor social skills, psychological distress, and lower life satisfaction, especially in the elderly [6–8]. Currently, research on the effect of social distancing is limited [9] and little is known about how the pandemic and related restrictions such as home isolation and social distancing affect health-related quality of life (HRQL) in patients with cancer. Such information could contribute to the policy-making and tailored care for people at high risk of being infected. Therefore, the purpose of this study was to investigate whether social distancing during the COVID-19 pandemic reduced HRQL in a cohort of esophageal cancer patients in Sweden and if so, to identify variables related to the HRQL changes.

Methods

Study design

This nationwide prospective study includes patients who underwent surgery for esophageal cancer between 2013 and 2019 in Sweden. Patients who survived 1 year after surgery and had no cognitive impairments are included in the Esophageal Surgery on Cancer patients – Adaptation and Recovery (OSCAR) study [10]. In brief, consenting patients were followed up regarding HRQL, symptoms and functions during a 5-year period. The study was approved by the Regional Ethical Review Board in Stockholm, Sweden (diary number: 2013/844-31/1) including an amendment for this particular study (2020-01755). Written consents were obtained from all participants before the inclusion.

Participants

All patients included in the OSCAR study, who had participated in at least one wave within 1 year of the COVID-19 outbreak, were invited to participate in the study.

Data collection

In April 2020, 5 weeks after the Swedish Public Health Agency recommendation of social distancing was released, all included patients were contacted by telephone for an interview. A follow-up phone call was performed 8 weeks after the first call, that is 13 weeks after the recommendation was officially issued (Figure 1). A few questions from the questionnaire European Organization for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire (QLQ) Core 30 [11] were selected to make the telephone interview short, relevant, and specific in order to minimize the burden on the patients. The selection of items was decided in collaboration with a patient research partnership group. Some study-specific questions regarding compliance with recommendations of social distancing, their living situation, type of living, and if they considered themselves to be at high risk of becoming seriously ill (if exposed to the virus), were included in the interview. A question about satisfaction with information on how to protect from the virus, with the four response alternatives: ‘not at all’, ‘a little’, ‘quite a bit’, and ‘very much’ was also asked. Data about marital status and geographical residence were collected from Swedish national registries [12–14]. Comorbidity data were retrieved from the Swedish Patient Register and Swedish Cancer Register [14] at inclusion in the OSCAR study. A self-reported update was obtained from the patient questionnaire response at study inclusion. Comorbidity was defined as hypertension, cardiovascular disease, diseases treated with immunosuppressive therapy, and pulmonary disease (apart from asthma). Data on the number of COVID-19 cases per region were collected from The Public Health Agency [1]. The participants were categorized into two groups depending on if they lived in a region with a high or low incidence of COVID-19 cases measured for each region the week before calls were made. The regions with 25% highest incidence were defined as high incidence regions and the remaining regions were categorized as low incidence regions.

Figure 1. Timeline for inclusion and follow-up.

Exposures

The study exposure was social distancing assessed with the question ‘To what extent have you followed the recommendations issued by the Swedish Public Health Agency, for example, to avoid socializing with relatives and grandchildren, crowding, and/or grocery shopping?’. There were four response alternatives: ‘fully’, ‘to a large extent’, ‘to a little extent’, and ‘not at all’. The individuals were categorized into three groups depending on their reply – ‘to a little extent’ and ‘not at all’ – were merged into ‘not at all/a little’.

Outcomes

The main outcomes were HRQL changes between before the recommendations of social distancing, 5 weeks and 13 weeks after the recommendations were issued, assessed with specific items from the EORTC QLQ C-30 [11] for global quality of life (2 items), role function (2 items), emotional function (4 items), fatigue (3 items) and insomnia (1 item). There are four response alternatives for each question: ‘not at all’, ‘a little’, ‘quite a bit’, and ‘very much’. The HRQL scores are transformed into a scale ranging from 0 to 100, where a higher score for symptoms implies more problems and a higher score for functions represents a better function. HRQL data (EORTC-QLQ C30) before the recommendations of social distancing, were collected from the latest follow-up of the OSCAR study (within one year of the COVID-19 pandemic outbreak in Sweden).

Statistical analyses

Descriptive statistics were presented as counts (n) and proportions (%). Logistic regression models were used to assess associations between social distancing, socio-demographic variables, and HRQL aspects and presented as mean score differences (MSD) with 95% confidence intervals (CI). Changes over time in global quality of life, role function, and emotional function were calculated as HRQL at the time of assessment – HRQL before the recommendations. HRQL before the recommendations was included in the calculation to show the change (improvement/deterioration). In regression models, adjustments were made for age (continuous variable), sex (men/women), comorbidity (0 or ≥1), an indicator of different time points for the interviews (continuous variable), years since surgery (continuous variable), and tumor histology (adenomcarcinoma/squamous cell carcinoma). To avoid multiple testing, statistical significance was tested (at 5% level of significance) only if the MSD was found to be clinically relevant. According to evidence-based interpretation guidelines, clinically relevant MSDs were denominated as small, medium, and large [15]. All data management and statistical analyses were conducted by a senior statistician (AJ) with expertise in HRQL analyses using SAS version 9.4 (Cary, NC).

Results

Among the 189 included patients, 134 persons participated in the telephone interview at both times, which corresponds to a response rate of 71%. Eleven individuals (6%) declined participation at the 5-week or 13-week assessments and 44 (23%) did not answer the phone call within the 5-day intervals. Patient characteristics were similar between responders and non-responders (data not shown).

Characteristics

Details of the 134 participants are shown in Table 1. As expected in a population of esophageal cancer survivors, most individuals were elderly (mean age of 71 years), male (78%) with adenocarcinoma (83%). Among the participants, 44% stated that they fully complied with the recommendation of social distancing, while 49% complied to a large extent and 7% did not follow the recommendations (Table 1). The majority of individuals who fully complied with the recommendations were cohabitants (73%), lived in a villa (66%), and considered themselves as at-risk persons (80%) (Table 1). Compared to the 5-week assessment, fewer individuals (−7%) stated that they fully complied with the recommendations at the 13-week assessment. Instead, a larger proportion answered that they complied with the recommendations to a large extent (+4%) or not at all/a little (+3%) (Supplementary Table 1).

Table 1. Patient and demographic variables of the 134 included patients categorized into compliance to the recommendations at the 5-week assessment.

	Compliance to recommendations
	All patients	Always/fully	To a large extent	Not at all/a little
Characteristics	Numbers (%)	Numbers (%)	Numbers (%)	Numbers (%)
Total numbers	134 (100)	59 (44)	66 (49)	9 (7)
Age
Current age in years (SD)	71 (8)	72 (7)	68 (8)	75 (4)
Sex
Men	105 (78)	50 (85)	47 (71)	8 (89)
Women	29 (22)	9 (15)	19 (29)	1 (11)
Comorbidity
0	84 (63)	40 (68)	40 (61)	4 (44)
≥1	50 (37)	19 (32)	26 (39)	5 (56)
Tumor histology
Adenomcarcinoma	111 (83)	50 (85)	53 (80)	8 (89)
Squamous cell carcinoma	23 (17)	9 (15)	13 (20)	1 (11)
Years since surgery (SD)	3.0 (1.5)	3.4 (1.5)	3.1 (1.5)	4.1 (1.7)
Region of COVID-19 incidence
High (>75 percentile)	50 (37)	26 (44)	23 (35)	1 (11)
Low (<75 percentile)	84 (63)	33 (56)	43 (66)	8 (89)
Living situation
Live alone	34 (25)	16 (27)	15 (23)	3 (33)
Co-habitant	100 (75)	43 (73)	51 (77)	6 (67)
Type of living
Apartment/care home	41 (31)	20 (34)	20 (30)	1 (11)
Villa	93 (69)	39 (66)	46 (70)	8 (89)
Considered as an at risk person
Yes	91 (68)	47 (80)	38 (58)	6 (67)
No	43 (32)	12 (20)	28 (42)	3 (33)

SD: standard deviation.

Compliance to recommendations of social distancing and HRQL

Five weeks after the recommendations of social distancing were issued, reductions in role function were seen for individuals who complied fully (MSD −8, 95% CI: −16 to 0) (small reduction) and to a large extent (MSD −19, 95% CI: −26 to −12) (medium reduction) (Table 2). Also, less fatigue (MSD 9, 95% CI: 3–15) (small improvement) was seen for individuals who fully followed the recommendations (Table 3). At thirteen weeks, role function was still reduced (MSD −10, 95% CI: −19 to −1) (small reduction) for individuals who fully complied with the recommendations, but improvements in fatigue could no longer be detected (Table 2). No reductions or improvements were seen in global quality of life, emotional function, or insomnia at either time point. Results from the crude analyses can be found in Supplementary Tables 2 and 3.

Table 2. Socio-demographic variables and HRQL changes for functions with comparisons between before recommendations, 5 weeks and 13 weeks after recommendations of social distancing with adjustments for age, sex, comorbidity, indicator of different time points for the interviews, years since surgery, and tumor histology, presented as mean scores (MS) and mean score differences (MSD) with 95% confidence intervals (CI).

	Global QOL				Role function			Emotional function
Person characteristics	Before recommendations	5 weeks		13 weeks	Before recommendations	5 weeks	13 weeks	Before recommendations	5 weeks	13 weeks
	MS (95% CI)	MSD (95% CI)		MSD (95% CI)	MS (95% CI)	MSD (95% CI)	MSD (95% CI)	MS (95% CI)	MSD (95% CI)	MSD (95% CI)
Compliance
Always/fully	72	1		0	78	−8^a	−10^a	84	−1	−3
	(66 to 77)	(−4 to 7)		(−5 to 7)	(71 to 84)	(−16 to 0)	(−19 to −1)	(79 to 89)	(−5 to 4)	(−8 to 2)
To a large extent	76	−2		−2	88	−19^b	−6	87	−2	1
	(71 to 81)	(−7 to 2)		(−7 to 3)	(82 to 94)	(−26 to −12)	(−13 to 1)	(82 to 92)	(−6 to 3)	(−3 to 6)
Not at all/a little	77	−9		3	85	−9	−13	82	3	7
	(64 to 90)	(−21 to 3)		(−7 to 2)	(69 to 102)	(−27 to 10)	(−28 to 2)	(70 to 94)	(−8 to 14)	(−2 to 17)
Age
Age <70	72	0		2	83	−12^a	−6	84	0	3
	(66 to 77)	(−5 to 5)		(−3 to 7)	(76 to 90)	(−20 to −5)	(−14 to 2)	(79 to 89)	(−5 to 4)	(−2 to 7)
Age ≥70	76	−2		−3	84	−15^b	−10^a	87	−1	−2
	(71 to 81)	(−7 to 2)		(−8 to 2)	(77 to 90)	(−22 to −8)	(−17 to −3)	(82 to 91)	(−6 to 3)	(−6 to 3)
Sex
Men	73	1		2	87	−7^a	−4	85	2	3
	(68 to 77)	(−3 to 6)		(−2 to 7)	(82 to 93)	(−13 to 0)	(−11 to 3)	(81 to 90)	(−2 to 6)	(−1 to 7)
Women	77	−3		−4	84	−24^c	−9	87	−4	−2
	(69 to 84)	(−10 to 3)		(−11 to 3)	(75 to 93)	(−34 to −14)	(−19 to 1)	(80 to 93)	(−10 to 2)	(−8 to 4)
Comorbidity
0	73	−2		−4	83	−13^a	−10^a	84	−1	−1
	(69 to 78)	(−7 to 3)		(−9 to 0)	(78 to 89)	(−20 to −6)	(−17 to −3)	(79 to 88)	(−5 to 4)	(−5 to 4)
≥1	74	−1		4	83	−15^b	−5	87	−1	1
	(68 to 80)	(−7 to 4)		(−2 to 9)	(76 to 90)	(−23 to −7)	(−14 to 3)	(81 to 92)	(−6 to 4)	(−4 to −6)
Region with COVID-19 incidence
Higher incidence ≥75%	75	−1		−1	84	−18^b	−11^a	82	−1	0
	(69 to 81)	(−7 to 4)		(−5 to 4)	(76 to 91)	(−27 to−10)	(−17 to −4)	(76 to 87)	(−6 to 5)	(−4 to 3)
Lower incidence <75%	74	−1		−1	83	−12^a	0	87	−1	3
	(69 to 78)	(−6 to 3)		(−9 to 6)	(77 to 89)	(−18 to −5)	(−11 to 11)	(83 to 91)	(−5 to 3)	(−4 to 10)
Living situation
Live alone	71	−3		−4	83	−14^a	−9	85	−3	−6^a
	(64 to 78)	(−9 t 4)		(−10 to 3)	(74 to 92)	(−24 to −4)	(−18 to 1)	(79 to 91)	(−9 to 3)	(−11 to 0)
Co-habitant	75	−1		1	84	−13^a	−8^a	86	0	3
	(71 to 80)	(−5 to 4)		(−4 to 5)	(78 to 89)	(−20 to −7)	(−14 to −2)	(82 to 90)	(−4 to 4)	(−1 to 7)
Type of living
Apartment/Care home	69	1		2	77	−9^a	−3	83	0	0
	(62 to 75)	(−5 to 7)		(5 to 8)	(69 to 84)	(−18 to 0)	(−12 to 6)	(78 to 89)	(−5 to 6)	(−6 to 5)
Villa	77	−3		−2	87	−16^b	−11^a	86	−1	1
	(72 to 81)	(−7 to 2)		(−6 to 2)	(81 to 92)	(−23 to−10)	(−18 to −5)	(82 to 90)	(−5 to 3)	(−3 to 5)
Considered as an at risk person
Yes	70	0		−3	79	−17^b	−11^a	82	−2	−2
	(66 to 75)	(−5 to 5)		(−6 to 3)	(74 to 85)	(−24 to−10)	(−17 to −4)	(78 to 86)	(−6 to 2)	(−6 to 2)
No	81	−4		1	90	−7	−5	92	1	4
	(75 to 87)	(−10 to 2)		(−4 to 7)	(82 to 97)	(−16 to 1)	(−13 to 3)	(86 to 97)	(−4 to 7)	(−1 to 9)

^aSmall clinically relevant and statistically significant mean score difference; ^bMedium clinically relevant and statistically significant mean score difference; ^cLarge clinically relevant and statistically significant mean score difference; Statistical significance= p < 0.05.

Table 3. Socio-demographic variables and HRQL changes for symptoms with comparisons between before recommendations, 5 weeks and 13 weeks after recommendations of social distancing with adjustments for age, sex, comorbidity, indicator of different time points for the interviews, years since surgery, and tumor histology, presented as mean scores (MS) and mean score differences (MSD) with 95% confidence intervals (CI).

	Insomnia			Fatigue
	Before recommendations	5 weeks	13 weeks	Before recommendations	5 weeks	13 weeks
Person characteristics	MS (95% CI)	MSD (95% CI)	MSD (95% CI)	MS (95% CI)	MSD (95% CI)	MSD (95% CI)
Compliance
Always/fully	25	5	0	31	9^a	6
	(17 to 32)	(−3 to 13)	(−9 to 9)	(24 to 37)	(3 to 15)	(−1 to 13)
To a large extent	19	−3	−5	27	3	3
	(12 to 26)	(−10 to 5)	(−12 to 3)	(21 to 33)	(−2 to 9)	(−2 to 9)
Not at all – a little	21	4	−1	25	7	3
	(4 to 39)	( −14 to 23)	( −16 to 14)	(9 to 41)	(−7 to 21)	(−9 to 15)
Age
Age < 70	26	3	−3^a	29	8	5
	(18 to 33)	(−5 to 10)	(−11 to −4)	(22 to 36)	(−2 to 14)	(−1 to 10)
Age ≥70	18	−1	−3	28	4	3
	(11 to 25)	(−8 to 7)	(−10 to 4)	(22 to 34)	(−2 to 9)	(2 to 9)
Sex
Men	18	−1	−5	27	8	5^a
	(12 to 24)	(−7 to 6)	(−11 to 2)	(21 to 32)	(−3 to 13)	(1 to 11)
Women	25	2	−1	29	3	2
	(16 to 35)	(−8 to 12)	(−11 to 9)	(21 to 38)	(−5 to 11)	(−5 to 10)
Comorbidity
0	23	4	0	30	6^a	2
	(16 to 29)	(−3 to 11)	(−7 to 6)	(24 to 35)	(0 to 11)	(−3 to 7)
≥1	20	−2	−5	27	5	6^a
	(12 to 27)	(−10 to 6)	(−14 to 2)	(20 to 34)	(−1 to 11)	(0 to 13)
Region with COVID-19 incidence
Higher incidence ≥75%	23	−3	−3	28	6	4
	(15 to 32)	(−12 to 6)	(−9 to 4)	(21 to 36)	(−1 to 12)	(0 to 9)
Lower incidence <75%	21	3	−5	28	6^a	2
	(15 to 27)	(−4 to 10)	(−17 to 6)	(23 to 34)	(1 to 11)	(−6 to 11)
Living situation
Living alone	18	−2	−11^b	30	6	4
	(8 to 27)	(−12 to 8)	−20 to −1	(21 to 38)	(−1 to 13)	(−3 to 11)
Co-habitant	23	2	0	28	6^a	4
	(17 to 29)	(−4 to 9)	(−6 to 7)	(23 to 33)	(1 to 10)	(−1 to 9)
Type of living
Apartment/care home	24	2	−7	31	10^b	5
	(15 to 32)	(−7 to 11)	(−14 to 3)	(23 to 39)	(4 to 17)	(−2 to 12)
Villa	21	1	−1	27	3	3
	(15 to 27)	(−6 to 7)	(−8 to 5)	(22 to 33)	(2 to 8)	(−2 to 8)
Considered as an at risk person
Yes	22	2	−4	31	6^a	2
	(16 to 28)	(−5 to 8)	(−11 to 3)	(5 to 37)	(1 to 11)	(−3 to 8)
No	21	−1	−1	24	5	6^a
	(13 to 30)	(−9 to 8)	(−9 to 7)	(16 to 32)	(−2 to 11)	(0 to 12)

^aSmall clinically relevant and statistically significant mean score difference; ^bMedium clinically relevant and statistically significant mean score difference; ^cLarge clinically relevant and statistically significant mean score difference; Statistical significance = p < 0.05.

Person characteristics and HRQL functions

At the 5-week assessment, being female (MSD −24, 95% CI: −34 to −14), aged ≥70 (MSD −15, 95% CI: −22 to −8), having ≥1 comorbidity (MSD −15, 95% CI: −23 to −7), living in a region with higher COVID-19 incidence (MSD −18, 95% CI: −27 to −10), living in a villa (MSD −16, 95% CI: −23 to −10), and being considered as an at-risk person (MSD −17, 95% CI: −24 to −10) predicted large or medium deteriorations in role function (Table 2).

At the 13-week assessment, age ≥70 (MSD −10, 95% CI: −17 to −3), living in a region with higher COVID-19 incidence (MSD −11, 95% CI: −17 to −4), living in a villa (MSD −11, 95% CI: −18 to −4), and being considered as an at-risk person (MSD −11, 95% CI: −17 to −4) remained related to reductions in role function. However, the reductions were small (Table 2).

Person characteristics and symptoms

At the 5-week assessment, having no comorbidity (MSD 6, 95% CI: 0–11), living in a low COVID-19 incidence region (MSD 6, 95% CI: 1–11), cohabitating (MSD 6, 95% CI: 1–10), living in an apartment (MSD 10, 95% CI: 4–17) and considered as an at-risk person (MSD 6, 95% CI: 1–11) indicated small or medium improvements in fatigue (less fatigue) (Table 3).

Satisfaction with information from the Swedish Public Health Agency

At 5 weeks, the majority of participants (n = 126, 96%) were very or to a large extent satisfied with the provided information and advice about how to protect themselves against virus transmission. At 13 weeks, 92% (n = 122) were still highly satisfied with the information provided.

Discussion

This study indicates that the individuals who had undergone surgery for esophageal cancer and fully or to large extent complied with the Swedish recommendations of social distancing experienced reductions in role function, but not in global quality of life or emotional function. Being female, elderly, having comorbidities, living in a region with higher COVID-19 incidence, living in a villa, and being considered as an at-risk person were related to the HRQL changes in role function. However, the HRQL changes reduced with time to small or no deteriorations.

The major strengths of this study were the nationwide cohort design with a high participation rate and the repeated measures during the first three months of the COVID-19 pandemic with a first assessment before the recommendations of social distancing. However, there are a number of limitations that must be kept in mind when interpreting the results. Despite a participation rate of 71%, selection bias through nonparticipation cannot be entirely precluded. We considered the last reported postoperative HRQL results as baseline data for HRQL before the COVID-19 outbreak. The time point for this assessment was related to the time of surgery and therefore varied within the year between the participants. To counteract this potential bias, the number of years since surgery was included as a covariate in the regression analyses. A previous study indicated that HRQL may increase during the summer months [16]. The second assessment of this study was conducted in June and may therefore induce a risk of underestimation of true results.

The EORTC QLQ-C30 is a well-validated questionnaire assessing HRQL in cancer patients [11] and might not optimally capture the consequences of social isolation. Also, using selected items from the questionnaire might have negatively influenced the validity. The pragmatic approach where selected variables that previously had been identified as associated with social isolation, was discussed and agreed with the patient research partnership group in order to not increase the burden of questionnaires. Despite the nationwide and population-based design and high participation rate, the number of patients was limited, and a larger sample would have reduced the risk of type II errors.

Quarantine, social isolation, and social distancing have previously been shown to increase the risk for symptoms of depression and posttraumatic stress [9]. Studies from other countries during the pandemic indicate a general decrease in well-being and decline in physical and mental health for people who are socially distancing [17,18], a finding that is not confirmed in our study. Global quality of life and emotional function among esophageal cancer survivors did not seem to be affected by this situation. A voluntary approach to social isolation, such as the Swedish strategy with no public regulations other than the prohibition of gatherings of more than 50 people, might lower the risk of developing depression symptoms and posttraumatic stress [19]. Clear communication about the logic behind the decision, how to act in a given situation, and the altruistic benefit for society at large [9] have also been shown to be of importance in mitigating psychosocial distress [19]. A study including 846 individuals from Wuhan city and Zhengzhou city in China during the COVID-19 pandemic, showed that higher appraisals of risk communication early in the outbreak reduced the susceptibility of emotional response [20]. Adequate information may mitigate speculations and rumor spreading [21]. The Swedish Health Agency has emphasized the importance of providing extensive information and having a dialogue with the public. The current state has continuously been communicated at press meetings and all types of media. In this study, most of the participants were highly satisfied with the information about the pandemic and subsequent recommendations. The result may have improved their ability to cope with the situation and possibly contributed to maintaining the level of emotional function.

Role function is a measure of societal disposition in the workplace and everyday life. Deterioration in role function has previously been found to be associated with psychological distress [22] and unmet physical and daily needs (related to the fact that you are not able to do the things that you usually appreciate doing) among cancer patients [23]. In our study, the decrease in role function might indicate a high level of compliance with social distancing and the recovery could possibly be explained by a tendency to disregard the recommendations or that an adaptation effect with customization to new living conditions have taken place. With time, a better understanding of how the virus spreads and customization of social interaction and daily chores might change accordingly.

In our study, females experienced the largest decrease in role function. This result is in line with a German cohort study including 15,000 citizens, showing that younger women, in general, felt lonely, indicating a greater need for social interaction for females [24]. An Australian study of 5741 individuals, investigating the importance of social support in elderly people, showed that women had larger social networks than men [25]. Also, a meta-analysis review concluded that women scored higher on sociotropy than men [26], meaning that women to a higher degree than men tend to prioritize maintaining positive social relationships.

During the COVID-19 lock-down in Israel, a significant decrease in physical activity was reported among the general population [18]. For cancer patients, physical activity and exercise are important and have been associated with improvements in several symptoms, including fatigue [27,28]. The results of this study also suggest less fatigue while living under social distancing recommendations for subgroups of individuals. One study on the Canadian general population indicated that physical activity while enduring social distancing recommendations during the COVID-19 pandemic improved well-being, especially for people who were less active before the pandemic [17]. However, larger studies are needed to investigate the relationship between physical activity and the effects of social distancing.

To date, pandemics have been rare and studies on the effect of social distancing and self-isolation on the individual are therefore also sparse. However, the world is changing into becoming more globalized with increasing intercontinental contacts. Thus, the risk of future pandemics will probably also increase. Understandably, the focus of today’s research is mostly on how to prevent or reduce the spread of the disease. But, if we can also understand how recommendations and advice affect risk groups, we will have a better preparedness to take care of these individuals in the aftermath of the pandemic.

In conclusion, this study indicates that survivors of esophageal cancer who complied with social distancing recommendations during the COVID-19 pandemic, experienced reductions in role function, but not in global quality of life or emotional function. Some characteristic variables were related to the HRQL changes. However, these changes were dissolved with time. The negative effect might be ameliorated by informative advice about the disease.

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Author contributions

Conception and design: Anna Schandl, Kalle Mälberg, Asif Johar and Pernilla Lagergren; Collection and assembly of data: Kalle Mälberg and Asif Johar; Data analysis: Asif Johar; Interpretation of results and manuscript writing: All authors; Final approval of manuscript: All authors

Acknowledgements

We would like to thank all participants of the study for sharing their experiences with us and the members of the patient research partnership group for their invaluable contribution and comments throughout the development of the publication.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

Funding for the OSCAR study was provided by the Swedish Cancer Society [14 0323], the Swedish County Council [LS 2015-1198] and the Cancer Research Foundations of Radiumhemmet [141223]. The study sponsors had no role in the design of the study, data collection, analysis or interpretation of the results, the writing of the manuscript or the decision to submit the manuscript for publication.