Abstract
With the outbreak of the novel and rapidly spreading coronavirus disease 2019 (COVID-19), many extraordinary emergency measures have been taken to prevent and control spread of the virus (1-3). National restrictions varied from total lockdown to targeted quarantine and social distancing (4-6). In the Netherlands, specific governmental measures included stay-at-home orders when experiencing mild symptoms, social distancing (1.5 meter), closure of schools and public places, and canceling all public events. Despite drastic efforts, the World Health Organization officially declared the COVID-19 outbreak a pandemic on March 11, 2020 (7).
As the COVID-19 pandemic has put healthcare systems under unprecedented stress, urgent rearrangements of non–COVID-19–related health care has been of vital importance (6,8-11). To prioritize hospital capacity for critically ill COVID-19 patients, elective care was suspended as much as possible, and only emergency care and semi-urgent oncological procedures were continued (6,12-14). For breast cancer, surgical procedures were postponed when possible, various types of treatment protocols (chemo- and radiotherapy) were adapted, and follow-up appointments canceled, postponed, or transformed into (video)calls (9,15). Also, paramedical care and aftercare such as medical rehabilitation and psychological support was scaled down to a minimum, and national breast cancer screening programs were temporarily halted (12,16).
Delays and changes in breast cancer diagnosis, treatment, and follow-up protocols due to COVID-19 may potentially induce concerns about recurrence or survival (16). This, in combination with concerns about the new viral threat in general, could impair patients’ and survivors’ mental and emotional well-being. Previous literature showed that, in general, women with a history of breast cancer have an increased risk of anxiety and depression, and that social support is crucial for supporting quality of life (QoL) and mental health in patients previously or currently being treated for breast cancer (17-19). Measures of social distancing or lockdown may interfere with networks of support and have a negative impact on mental health and emotional functioning. However, to date, no previous studies have used patient-reported outcome scores (PROs) of breast cancer patients and survivors in the context of COVID-19.
The purpose of this study was to measure the impact of the COVID-19 pandemic on patient-reported QoL, physical, and psychosocial functioning in a prospective cohort of patients previously or currently being treated for breast cancer.
Materials and Methods
Study Design and Participants
The present study was conducted within the ongoing prospective multicenter Utrecht cohort for Multiple BREast cancer intervention studies and Long-term evaLuAtion (UMBRELLA) (20,21). Since 2013, the UMBRELLA cohort included patients aged 18 years or older, who were routinely referred by 6 hospitals in the Utrecht region to the Department of Radiation Oncology of the University Medical Centre Utrecht, the Netherlands. Inclusion criteria were histologically proven invasive breast cancer or ductal carcinoma in situ and the ability to understand the Dutch language (written and spoken). Prior to the first appointment with the radiation oncologist, breast cancer patients were routinely invited to participate in the UMBRELLA cohort. Participants provided informed consent for longitudinal collection of clinical data and PROs through paper or online questionnaires at regular intervals during and after treatment (before the start of radiation therapy [baseline], after 3 and 6 months, and each 6 months up to 10 years thereafter) (20). Clinical data, including patient, tumor, and treatment characteristics, were provided by the Netherlands Cancer Registry (NKR).
Details on pre- and during COVID-19 treatment protocols are presented in the Supplementary Methods (available online). The UMBRELLA study adheres to the Dutch law on Medical Research Involving Human Subjects (WMO) and the Declaration of Helsinki (version 2013). The study was approved by the medical ethics committee of the University Medical Centre Utrecht (NL52651.041.15, Medical Ethics Committee 15/165) and is registered on clinicaltrials.gov (NCT02839863).
Data Collection
On April 7, 2020, shortly after the introduction of the Dutch COVID-19 measures (March 12, 2020), an additional COVID-19–specific online survey was sent to all active UMBRELLA cohort participants who were enrolled between October 2013 and April 2020 with a known e-mail address. Diseased UMBRELLA participants were excluded, as well as those who were not responding to standard UMBRELLA questionnaires and those electing paper questionnaires. One reminder was sent on April 15, 2020. The survey included 2 standard UMBRELLA PRO questionnaires (European Organization for Research and Treatment [EORTC] core [C30] and breast cancer-specific [BR23] Quality of Life Questionnaire [QLQ] and Hospital Anxiety and Depression Scale [HADS]) and 2 additional questionnaires (De Jong-Gierveld Loneliness Scale and a COVID-19–specific questionnaire).
The cancer-specific Quality of Life Core Questionnaire (QLQ-C30) of the EORTC was used to assess global health-related QoL; physical, role, cognitive, emotional, and social functioning; dyspnea; insomnia; and financial difficulties (22). Patients’ and survivors’ future perspective was evaluated with the breast cancer–specific module (QLQ-BR23). For each subscale, a summary score was calculated according to the EORTC manual (23).
The HADS was used to assess symptoms of anxiety and depression (24). Patients and survivors with scores above 7 are at risk of having anxiety or depressive disorders (25).
Overall, emotional and social loneliness were assessed using the 6-item De Jong-Gierveld Loneliness Scale (26). Scores between 2 and 4 represent moderate loneliness, and a score above 4 represents severe loneliness (27). Scores above 2 on each of the 3-item subscales for emotional and social loneliness indicate emotional and social loneliness, respectively (27).
Additional questions were developed to assess presence of (symptoms resembling) COVID-19 and the impact of COVID-19 on healthcare consumption and expectations.
PROs during COVID-19 were compared with the most recent pre–COVID-19 PROs as obtained within UMBRELLA. Patients and survivors were excluded from comparative analyses when their most recent questionnaire was completed more than 2 years before the first confirmed COVID-19 diagnosis in the Netherlands (February 27, 2020).
Clinical data was obtained from electronic patient records and from the UMBRELLA dataset as retrieved from NKR and included age at cohort enrollment, body mass index, smoking status, self-reported highest educational level, surgical treatment, axillary treatment, (neo-)adjuvant radiation therapy and systemic treatment, currently receiving active treatment, and pathological T- and N-stage (American Joint Committee on Cancer [AJCC] 7th edition).
Read more Updates in BRCA testing for People of Ashkenazi Jewish Ancestry
Statistical Analysis
Frequencies, proportions, and means with standard deviations or medians with interquartile ranges (IQRs), as appropriate, were used to describe patient and clinical characteristics, COVID-19–related questions, and PROs.
To measure the impact of COVID-19 on PROs using complete case analyses (ie, only including patients and survivors who completely filled out PROs pre–COVID-19 and during COVID-19), most recent reported EORTC QLQ-C30/BR23 and HADS scores from prior to COVID-19 were compared with the PROs during COVID-19. Crude mean EORTC scores were compared with the paired samples t test and crude median HADS scores with the Wilcoxon signed-rank test.
To estimate whether the impact of COVID-19 on clinically relevant PROs varied with time since (active) treatment, patients and survivors were categorized into 4 groups (ie, active treatment; nonactive treatment and enrolled in UMBRELLA less than 24 months before the survey; nonactive treatment and enrolled 24-60 months before the survey; and nonactive treatment and enrolled more than 60 months before the survey). A linear mixed effect model for repeated measurements was used to measure the impact of COVID-19 on PROs. The model included a patient-specific random intercept, a random linear time effect to allow a patient-specific slope, and an interaction between months since diagnosis and time point. Two time points were included in the model: PROs prior to COVID-19 (ie, measurement of last follow-up pre–COVID-19) and PROs during COVID-19. A random slope was chosen to account for a difference in time effect for each patient. To correct for potential confounders, age (linear) was included as fixed variable in the model in the nonactively treated group. In the actively treated group, further adjustment was performed for type of radiotherapy and type of surgery. Only fixed effects without missing cases were included in the model. Changes in PROs due to COVID-19 were reported as mean differences with 95% confidence intervals (CI).
All reported P values were 2-sided, and P values less than .05 were considered statistically significant. Statistical analyses were performed using IBM Statistical Package for Social Sciences software, version 25 (SPSS; IBM Corp, Armonk, NY).
Results
Study Population
Between October 2013 and April 2020, 3239 patients previously or currently being treated for breast cancer were enrolled in UMBRELLA (Figure 1). Of all UMBRELLA participants, 1595 met the inclusion criteria for the present study and were sent the extra COVID-19 survey, of whom 1051 patients and survivors (65.9%) responded. Mean age was 56 (SD = 9.8) years, and median time since diagnosis was 24 (IQR = 6-42) months (Table 1). Most patients and survivors (55.7%) were treated for a stage 1 tumor and received breast-conserving surgery (77.4%); 51 patients (4.9%) were receiving active treatment (chemo- and/or radiotherapy) for breast cancer (Table 2). Median time between completion of the most recent pre–COVID-19 questionnaire and the COVID-19 survey was 4 months (IQR = 3-6).
Physical and Psychosocial Well-being During COVID-19
In total, 1 (0.1%) survivor had confirmed COVID-19 infection, and 100 (9.5%) patients and survivors indicated to have been possibly infected as they experienced symptoms of fever, without having been tested for the virus. Of all patients and survivors, 27.2% (n = 286) felt that the COVID-19 measures affected their current treatment, care or aftercare, and 23.8% (n = 250) felt that these measures were likely to affect their care or aftercare in the future (Table 2).
Almost one-third (31.1%, n = 327) reported a higher threshold to contact their general practitioner because of the COVID-19 outbreak, and 162 patients and survivors (15.4%) indicated to be less likely to contact their breast cancer physician. Family and friends were contacted less easily by 87 patients and survivors (8.3%). Most patients and survivors (n = 983, 93.5%) were not worried or were a little bit worried about their financial situation as a result of COVID-19 (Table 2).
During the pandemic, 48.0% (95% CI = 45.0% to 51.1%; n = 505) reported loneliness, of whom 38.9% (95% CI = 36.0% to 41.9%; n = 409) reported moderate feelings of loneliness, and 9.1% (95% CI = 7.5% to 11.0%; n = 96) reported severe loneliness (Table 3). A total of 40.0% (95% CI = 35.7% to 44.4%; n = 202) felt socially lonely, and 78.4% (95% CI = 74.6% to 81.9%; n = 396) felt emotionally lonely.
PROs Before and During COVID-19
For 1022 patients and survivors (97.2%), pre- and during COVID-19 EORTC scores (Figure 2) could be compared, and for 942 (89.6%), pre- and during COVID-19 HADS scores could be compared. Overall, mean scores for the EORTC subdomains QoL, physical, and role functioning statistically significantly improved during COVID-19 (Table 4). Mean scores for the EORTC subdomain emotional functioning worsened statistically significantly from 82.6 (18.7) to 77.9 (17.3; P < .001). Also, median HADS total score and depression score deteriorated statistically significantly during COVID-19 (Table 3).
In the subgroup of actively treated patients, there was a strong and statistically significant drop in social functioning of 16.0 points during COVID-19 after adjustment for age, type of radiotherapy, and type of surgery in mixed model analysis (77.3 [95% CI = 69.4 to 85.2] to 61.3 [95% CI = 52.6 to 70.1]; P = .002; Table 5).
Among the nonactively treated patients and survivors, age-adjusted analyses showed that emotional functioning worsened statistically significantly in all groups, whereas physical functioning improved statistically significantly in all groups (Table 5). QoL, role, and social functioning improved statistically significantly in nonactively treated patients and survivors who were enrolled in UMBRELLA for less than 24 months (Table 5).
PROs Before and During COVID-19 Among Patients and Survivors With Suspected COVID-19 Infection
Within the subgroup of 100 patients and survivors with suspected COVID-19 infection, we found statistically significant worsening in mean [SD] QoL (75.8 [18.2] to 65.9 [18.7]; P < .001) and emotional functioning (81.6 [21.5] to 74.8 [19.9]; P < .001) and an increase in dyspnea score [SD] (11.3 [20.5] to 19.0 [23.8]; P = .008) when compared with pre-COVID-19. EORTC scores for cognitive, physical, role, and social functioning, as well as insomnia and financial and future perspectives, remained stable. Median total HADS score and median HADS anxiety score [IQR] increased slightly (7 [4-11] to 8 [5-13]; P < .001; and 4 [3-7] to 5 [3-7]; P = .15, respectively). Median HADS depression score [IQR] increased from 2 [1-5] to 4 [ 2-7] (P < .001).
Read more Cancer fatigue: Why it occurs and how to cope
Discussion
The COVID-19 pandemic has a substantial impact on patients previously or currently being treated for breast cancer. One in 3 reported to be less likely to contact their general practitioner, and 15.4% indicated to be less likely to contact their breast cancer physician because of barriers induced by COVID-19 restrictions. In patients actively receiving treatment, social functioning decreased substantially, and in patients and survivors who were no longer receiving active treatment, deterioration of emotional functioning was observed. Loneliness was reported by almost half of all patients and survivors.
The high proportion who indicated experiencing a higher barrier to contact their general practitioner (31.1%) is in line with the upsetting findings of the Dutch cancer registry (NKR), who reported a nationwide decrease up to 40% in cancer diagnoses during COVID-19 (16). Jones and colleagues (28) (United Kingdom) also expressed their concerns about patients potentially feeling higher barriers to consult a general practitioner for nonspecific symptoms. Patients might experience (moral) dilemmas, including concerns about wasting the physician’s time, insufficient capacity, or exposure to the virus in a healthcare institution (15). Moreover, an average referral drop of 37% to all medical specialties was observed in the Netherlands during the outbreak (29). Fortunately, most patients and survivors (84.6%) did not report a higher barrier to reach out to their breast cancer physician. Thus, although the threshold to contact a general physician was higher in a concerning proportion of patients and survivors, the pandemic seemed to have minimal impact on interaction with the breast cancer team once diagnosed. This does highlight the importance of creating public awareness about the risk potential delays in seeking medical help could cause, aiming to lower barriers for patients and survivors to contact their general practitioner when experiencing nonbreast cancer–related symptoms (16).
In addition to the decrease in cancer diagnoses during the pandemic, a decrease in mental health also raises concerns about future disease outcomes. Previous studies showed that lower levels of anxiety and depression, as well as social support, seem to positively affect treatment adherence (30,31).
Among patients who were receiving active breast cancer treatment during the COVID-19 pandemic, a major clinically and statistically significant decrease in social functioning was observed (Table 5) (32). The mean score approached the threshold score for clinical importance of 58 (33). One explanation for this decrease could be that these patients and survivors were more careful regarding social interaction. Recent publications underlining the risk of COVID-19–related adverse events in cancer patients might have amplified concerns about contracting the virus (9,15,34-36).
All nonactively treated patients and survivors showed a statistically significant reduction in emotional functioning. This domain assesses anxiety, depression, and general distress through questions about feeling tense, worrying, feeling depressed, and feeling irritable (37). The deterioration in this domain is likely attributable to COVID-19, because we know from pre–COVID-19 measurements in the same UMBRELLA cohort that emotional functioning generally increases over time [previously shown in Supplementary Data (available online) by Gregorowitsch and colleagues (38)]. Also, the median score for depression worsened statistically significantly during COVID-19. Concerns about the new viral threat might have enhanced overall uncertainty in individuals. Different types of coping mechanisms could play a role here; lower tolerance of uncertainty is related to higher appraisal of a health threat and higher levels of emotion-focused coping strategies (39). A previous study showed that, during the 2009 H1N1 viral outbreak, emotion-focused coping was related with increased levels of depression (39).
Interestingly, despite the deterioration in emotional functioning in all nonactively treated patients and survivors, there was a statistically significant increase in global QoL, role functioning, social functioning, and physical functioning. For QoL, physical and social functioning, the increases did not reach the minimal clinically important difference scores (10, 7, and 7, respectively) (32). Although cautiousness is therefore advised when interpreting these results, they do emphasize that, regardless the COVID-19 pandemic, PROs did not deteriorate in these domains. This may partly be explained by the fact that these PROs generally tend to increase over time since diagnosis (38). In our study, however, median time between pre- and during COVID-19 PROs was only 4 months. Furthermore, the shared crisis may put patients’ and survivors’ perceived QoL in relation to their disease in a different perspective and may even accelerate reconceptualization of their QoL (40). The observed increase in social and role functioning suggests that patients and survivors reported an increased ability to fulfill responsibilities associated with occupational and/or family roles. Governmental measures encouraging working from home and prohibiting social events (ie, less social obligations) in times of social distancing or lockdown may also play a role.
The impact of the COVID-19 pandemic seemed worse among patients and survivors with suspected COVID-19 infection. The largest effect was seen in a substantial drop in QoL and an increase in symptoms of dyspnea and depression. This is in line with the findings of our Italian colleagues (41), who reported worsening of QoL in 44.1% of patients 60 days after the onset of the first COVID-19 symptoms (n = 143). This may partly be explained by the fact that this subgroup will have likely experienced a more profound period of self- or imposed isolation.
In general, experiencing health problems can induce loneliness and vice versa (42). Moreover, a recent meta-analysis showed that breast cancer patients and survivors might be particularly vulnerable for feelings of loneliness, anxiety, and depression when compared with patients without prior cancer (19). Social isolation measures to fight the COVID-19 pandemic might enhance these feelings. This study showed that 48.0% felt lonely, of whom the majority felt emotionally lonely. Loneliness is not a parameter that was captured routinely in the UMBRELLA cohort, so there were no pre–COVID-19 measurements on loneliness. Furthermore, because we did not include a noncancer control group, it remains unclear whether patients previously or currently being treated for breast cancer are more severely affected by COVID-19 than the general population. Nonetheless, the reported proportion of 48.0% loneliness is substantially higher than the reported 34% in the general Dutch population in 2019 (pre–COVID-19), as measured by the same loneliness scale (Statistics Netherlands; n = 7398) (43). Also, our proportion of patients and survivors feeling lonely was substantially higher than the reported 30%-35% cancer patients in a study conducted pre–COVID-19 (44). Moreover, our percentage of patients and survivors feeling severely lonely was strikingly high; almost 10% compared with 0%-2% in other studies performed among cancer patients (42). Thus, even though we cannot rule out other contributing factors, the higher proportion feeling lonely in this breast cancer population is likely due to the COVID-19 measures.
Read more Breast Cancer Team Names: 650+ Catchy and Cool Names
The results of this study underline the magnitude of the impact of a major health crisis on the psychosocial well-being of breast cancer patients and survivors. With the high survival rates of breast cancer patients, mental health has become an integral focus of supportive treatment. However, a barrier to e-mental health still exists (11,45). Triggered by the ongoing pandemic, several e-mental health initiatives were developed in the Netherlands (ie, facilitation of video consultation or the Red Cross COVID-19 helpline for mental care during isolation or quarantine). However, at the time of this survey, breast cancer patients and survivors did not have regular access to digital mental health services. Especially in times when face-to-face contact is not an option and when the global need for psychological and/or peer support is rising because of the viral threat, efforts are needed to rapidly implement e-mental health screening programs and digital psychological interventions. Only by adapting to the new circumstances will we be able to treat both ongoing and emerging mental health care conditions due to COVID-19 and prevent long-term problems (11,46). Considering that a second wave of COVID-19 or another future outbreak with similar impact is probable (15,47), and now that the pandemic seems to be lingering, one would hope that the current experiences during this pandemic may serve as a turning point in the adoption and acceptance of successful e-mental health applications (45,48).
A limitation of this study is the fact that only 51 patients and survivors (4.9%) in our cohort received active treatment during COVID-19, causing relatively wide 95% confidence intervals in this group. As a consequence, the results of this study mainly show the impact of COVID-19 on breast cancer survivors, whereas the impact of COVID-19 might be more severe for newly diagnosed patients who experienced adjusted treatment protocols. However, as breast cancer survivors account for the largest group of cancer survivors in high-income countries (19), and a clinically significant proportion of breast cancer survivors still consider themselves patients 5-15 years postdiagnosis (49), we believe these results are valuable for an important part of the general population. Second, this study measured the impact of COVID-19 6 weeks after the start of COVID-19. Therefore, it is unclear whether our results represent short-term or longer-lasting effects. However, previous literature on the 2009 H1N1 viral threat showed that psychological effects can persist until 30 months after the outbreak (39). Last, we could not compare the impact of COVID-19 on patients previously or currently being treated for breast cancer to the impact on a healthy reference population. Nonetheless, these results contain important clinically relevant information that may be used to improve understanding the needs and experiences of breast cancer patients and survivors amid these exceptional times. An important strength of this study is that the UMBRELLA cohort provided a unique opportunity to longitudinally compare PROs during COVID-19 with PROs before COVID-19 in an identical population in a representative population of breast cancer patients and survivors (20).
In conclusion, COVID-19 is having a substantial impact on breast cancer patients and survivors. Emotional functioning deteriorated in nonactively treated patients and survivors following the COVID-19 pandemic, 1 in 2 patients and survivors reported loneliness that was moderate or severe, and 1 in 3 reported to be less likely to contact their healthcare providers. In actively treated patients, social functioning decreased substantially.
Funding
The authors received no financial support for the research (including study design; data collection, analysis, interpretation of data; and writing of the report), authorship, and/or (the decision to submit the article for) publication.
Notes
Role of the funder: Not applicable.
Disclosures: The authors have no conflicts of interest to disclose.
Author contributions: The corresponding author (HM Verkooijen) confirms that she had full access to all the data in the study and had final responsibility for the decision to submit for publication. Each author has contributed significantly to, and is willing to take public responsibility for, the following aspects of the study: Design: CA Bargon, MCT Batenburg, LE van Stam, DR Mink van der Molen, W Maarse, N Vermulst, IE van Dam, F van der Leij, EJP Schoenmaeckers, MF Ernst, IO Baas, T van Dalen, R Bijlsma, DA Young-Afat, A Doeksen, HM Verkooijen. Data acquisition: CA Bargon, MCT Batenburg, LE van Stam, DR Mink van der Molen, A Doeksen, HM Verkooijen. Analyses: CA Bargon, MCT Batenburg, LE van Stam, DR Mink van der Molen, DA Young-Afat, HM Verkooijen. Interpretation: CA Bargon, MCT Batenburg, DR Mink van der Molen, DA Young-Afat, HM Verkooijen. Drafting: CA Bargon, LE van Stam. Critical revision: CA Bargon, MCT Batenburg, LE van Stam, DR Mink van der Molen, W Maarse, N Vermulst, IE van Dam, F van der Leij, EJP Schoenmaeckers, MF Ernst, IO Baas, T van Dalen, R Bijlsma, DA Young-Afat, A Doeksen, HM Verkooijen.
Acknowledgments: We would like to thank Janet van Dasselaar for her support with the clinical data management and Rosalie van den Boogaard for support with institutional review board approval. We are greatly indebted to the medical ethics committee of the UMC Utrecht for expedited ethical review of the protocol.
Data Availability
The data underlying this article will be shared upon reasonable request to the corresponding author.
