The prevalence of psychological disorders among cancer patients during the COVID‐19 pandemic: A meta‐analysis

Abstract Purpose We aimed to assess the prevalence rate (PR) of depression, anxiety, posttraumatic stress disorder (PTSD), insomnia, distress, and fear of cancer progression/recurrence among patients with cancer during the COVID‐19 pandemic. Methods Studies that reported the PR of six psychological disorders among cancer patients during the COVID‐19 pandemic were searched in PubMed, Embase, PsycINFO, and Web of Science databases, from January 2020 up to 31 January 2022. Meta‐analysis results were merged using PR and 95% confidence intervals, and heterogeneity among studies was evaluated using I 2 and Cochran's Q test. Publication bias was examined using funnel plots and Egger's tests. All data analyses were performed using Stata14.0 software. Results Forty studies with 27,590 participants were included. Pooled results showed that the PR of clinically significant depression, anxiety, PTSD, distress, insomnia, and fear of cancer progression/recurrence among cancer patients were 32.5%, 31.3%, 28.2%, 53.9%, 23.2%, and 67.4%, respectively. Subgroup analysis revealed that patients with head and neck cancer had the highest PR of clinically significant depression (74.6%) and anxiety (92.3%) symptoms. Stratified analysis revealed that patients with higher education levels had higher levels of clinically significant depression (37.2%). A higher level of clinically significant PTSD was observed in employed patients (47.4%) or female with cancer (27.9%). Conclusion This meta‐analysis evaluated the psychological disorders of cancer patients during the COVID‐19 outbreak. Therefore, it is necessary to develop psychological interventions to improve the mental health of cancer patients during the pandemic.


| INTRODUCTION
The coronavirus disease 2019 (COVID- 19) pandemic poses a serious threat to public health globally and is a time of unprecedented psychosocial disorder for many people. 1,2 Due to the uncertainty of COVID-19, the psychology and spirit of individuals have been affected. 3 People are worried about becoming infected and how long the pandemic will last. Dong et al. 4 reported potential causes for the increased psychological problems in the general population during this pandemic. The virus has an uncertain incubation and may be accompanied by asymptomatic transmission, causing additional anxiety and fear among the public. The World Health Organization (WHO) declared that the prevalence of psychological disorder in the general population has risen dramatically worldwide and will become a global burden. 5 Psychological disorder is common in cancer patients and is associated with poor health outcomes, 6 and approximately 30%-50% of cancer patients suffer from psychological distress. 7 Owing to the immunosuppression induced by cancer and the treatment, patients with cancer are more susceptible to COVID-19 infection; meanwhile, cancer patients may exhibit a higher risk of death if they infected. 8 In addition, the reduction in services and delayed or missed counseling and treatment due to the shortage of medical resources has further adverse impacts on the mental health of cancer patients. 9,10 Meanwhile, loneliness caused by social distancing further affect the emotional well-being. 11 Evidence suggests that the COVID-19 pandemic exacerbates the psychological disorder of cancer patients. 12 Bargon 13 compared the psychological states of breast cancer patients and survivors before and during the outbreak and found that emotional functioning deteriorated in these patients, and loneliness increased in nearly half of them. Taken together, illness and COVID-19 are dual challenges for cancer patients. Maintaining a stable psychological state is key to ensuring that patients can receive effective cancer treatment. Given the vulnerability of cancer patients to psychological disorders, their psychological changes during COVID-19 infection need to receive increased attention from caregivers and social organizations. 14 Individuals affected by this pandemic may experience psychological distress, such as anxiety, stress, depression, insomnia, and suicidal behavior. 15 A previously published meta-analysis revealed the prevalence of depression and anxiety among patients with cancer during the COVID-19 outbreak. 16 However, the impact of COVID-19 on other psychological disorder in cancer patients has not been systematically studied. Therefore, we performed an updated metaanalysis to determine the prevalence rate (PR) of psychological disorders in cancer patients during the COVID-19 outbreak, including anxiety, depression, posttraumatic stress disorder (PTSD), insomnia, distress, and fear of cancer progression/recurrence. Our research will help support the development of policy interventions to mitigate psychological issues among cancer patients during COVID-19 pandemic.

This meta-analysis was performed in accordance with the Preferred
Reporting Items for Meta-Analyses (PRISMA) guidelines and was registered with PROSPERO (CRD42022308459). All analyses were based on previous published studies, thus no ethical approval and patient consent are required for this meta-analysis.

| Data source and retrieval strategy
Candidate studies were searched from PubMed, Embase, PsycINFO, and Web of Science databases from 1 January 2020 and 31 January 2022, without language restrictions. The following search algorithms were applied: ("depression" OR "distress" OR "stress" OR "anxiety" OR "post-traumatic stress symptoms" OR "post-traumatic stress disorder" OR "burnout" OR "psychological") AND ("neoplasm" OR "cancer" OR "tumor" OR "tumour") AND ("COVID-19" OR "SARS-CoV-2" OR "severe acute respiratory syndrome coronavirus 2"). The retrieval strategy was adjusted according to the characteristics of each database (Tables S1-S4). Further, to obtain more potential studies that could be used for meta-analysis, we manually searched the literature and reference lists of relevant reviews and included studies.

| Selection criteria
The following criteria for inclusion were applied: (1) cancer patients were pathologically diagnosed or treated in the hospital; (2) or studies reported the PR of at least one mental health outcomes: anxiety, depression, PTSD, insomnia, distress, and fear of cancer progression/recurrence during COVID-19 pandemic; (3) the cut-off value for the mental health status of patients evaluated by the Patient Health Questionnaire-9 (PHQ-9), Impact of Event Scale-Revised (IES-R), Insomnia Severity Index (ISI), and other scales was reported; and (4) research designs were cross-sectional or cohort studies.
The exclusion criteria were as follows: (1) mental status scores in patients with cancer were reported as mean � SD, not PR, and (2) nontreatise literature such as letters, reviews, and comments. Furthermore, for duplicate publications or multiple articles with the same data, we only included articles with the most complete information.

| Data extraction and quality assessment
Two researchers independently extracted information from each study, including the first author, publication year, research area, socio-demographic information (gender, age, sample size, marital status, education, and employment status), study type, type and stage of cancer, measurement scales, and cut-off value. After data extraction, inconsistencies were resolved through discussion. ZHANG ET AL.

-1973
The Joanna Briggs Institute (JBI), containing nine items, was used to perform a bias risk assessment. 17 Each item can be judged as "yes," "no," and "unclear or not applicable," corresponding to "low risk," "high risk," and "unclear risk." In brief, a study with at least one item at "high risk" is defined as "high" risk of bias; studies with at least three items at "unclear risk" are defined as "unclear" risk of bias; the remaining studies are regarded as "low" risk of bias.

| Statistical analysis
The prevalence of each mental health outcome in cancer patients was evaluated using PR with a 95% confidence interval (CI). P < 0.05 and/or I 2 > 50% represented significant heterogeneity between studies, and a random-effects model was used to merge the effect size. p ≥ 0.05 and I 2 ≤ 50% represented no heterogeneity, and a fixed-effects model was applied. To explore the source of heterogeneity, a subgroup analysis was conducted according to several variables (area, scale, risk of bias, cancer type, gender, marital status, education level, and employment status). Publication bias was examined using funnel plots and Egger's tests. 18 All metaanalyses were performed using the Stata14.0 software (Stata Corp, College Station).

| Study selection
A flowchart of the search results is shown in Figure 1. A total of 920, included in this meta-analysis. The summarized psychological health status is presented in Table S5.

| Characteristics of each included study
The detailed characteristics of the 40 included studies are presented in
Among these, patients with head and neck cancer had the highest PR for depression (74.6%, 95% CI: 0.658, 0.825). However, we found no statistically significant differences in the subgroup analysis by area or risk of bias (P > 0.05, Figure S1A and S1B).

| Anxiety
A total of 34 studies reported the anxiety prevalence for cancer patients; the pooled PR of anxiety was 31.3% (95%: 0.254, 0.375, Figure 3A), and the heterogeneity among studies was significant (I 2 = 98.975%, P < 0.001). Significant results were found in the subgroup analysis using different scales (P < 0.01, Figure 3B) Figure 4B).
Significant results were observed in a subgroup analysis by cancer type (P < 0.01, Figure 4C) and risk of bias (unclear risk vs. high risk, P = 0.001, Figure 4D). Nevertheless, no significant results were found in the subgroup analysis by area (P > 0.05, Figure S3).

| Insomnia
Five studies used the ISI scale to assess PR in patients with insomnia.
A meta-analysis showed PR of insomnia among patients with cancer   Figure S5).
Moreover, significant results were observed in the subgroup analyses classified by differences in scale, area, cancer type, and risk of bias Figures 7B-E). significantly associated with higher PR of PTSD than male (17.9%, P < 0.01, Figure S8A); employed patients (47.4%) had an observably higher rate of PTSD than unemployed patients (37.7%, P < 0.01, Figure S8B).  patients with head and neck cancer had the highest PR for depression and anxiety. Moreover, a stratified analysis revealed that patients with higher educational levels were more prone to depression; employed patients or women with cancer might tend to experience higher levels of PTSD.

| DISCUSSION
In this study, we observed a high level of psychological disorder among cancer patients during the pandemic. The majority of cancer patients suffer from fear of cancer progression/recurrence. During the early pandemic, the government recommended postponing nonemergency cancer surgery and routine cancer screening, resulting in a higher proportion of patients delaying or missing health care services. 57 It has been indicated that the COVID-19 pandemic can exacerbate the fear of disease progression or recurrence in patients with cancer due to access restrictions on follow-up and treatment, imposed isolation restrictions, and the possibility that the healthcare systems becomes overworked. 58 The vast majority of patients concern about the impact of delays on treatment and long-term health. According to recent studies, the fact that most patients reported high fear of recurrence was due to concerns about lack of access to medical services, which were completely limited during the COVID-19 pandemic. 59 The level of distress in cancer patients has also increased due to treatment delays. 60 Moreover, changes in treatment have led to concern and fear of disease recurrence in cancer patients, as well as increased levels of depression and anxiety. 61 Therefore, with the continued spread of COVID-19, more attention should be paid to its potentially harmful effects on the mental health of this particular population. However, only 1.6% of cancer patients sought psychological counseling during COVID-19. 51 We recommend that oncology clinics provide the necessary and timely mental health screening for cancer patients; accordingly, policymakers should develop personalized psychological care plans for cancer patients.
In this analysis, we used different scales to assess the PR of depression and anxiety, and the results were statistically significant. are simplified from items in it. 64 A previous study indicated that HADS-D significantly underestimated depression in prostate cancer patients compared to PHQ-9, and SDS showed a similar trend. 65 Moreover, HADS could be a better option for depression assessment than SDS in patients with lung cancer. 66 The State-Trait Anxiety Inventory (STAI) is a self-report measure of the severity of anxiety symptoms; HADS screens for clinically significant symptoms of anxiety in patients with medical conditions. 67 Taken together, these different scales may have an impact on the detection of depression or anxiety in patients with cancer, which is consistent with our findings.  72 Thus, to discern this information, the educational level of the patients is crucial. Yang et al. 22 indicated that patients with higher education had better awareness of cancer, especially in the context of the COVID-19 pandemic. In addition, they have a strong ability to identify and process uncertain information and do not panic blindly, thereby avoiding or reducing anxiety and depression. However, we obtained inconsistent results, which should be confirmed in future studies with larger sample sizes. In general, stressors affect women more than men at the population level.
Compared to men, the prevalence of PTSD and anxiety was increased in women among cancer and non-cancer patients. 73 We also observed that women were more vulnerable to PTSD during this pandemic. COVID-19 was with a major impact on the global economy economic levels had a greater impact on employed patients than on the unemployed. This phenomenon was also observed in this metaanalysis.

| CLINICAL IMPLICATION
This meta-analysis has several advantages and practical implications.
First, a large number of studies with large sample sizes were included in the analysis. Second, a merged meta-analysis revealed that cancer patients had varying degrees of mental health problems. Oncologists may ignore the psychological problems of patients when formulating cancer treatment plans. Thus, our meta-analysis suggests that a comprehensive assessment of the prevalence of psychological disorder is necessary before providing optimal care to cancer patients in clinical practice. 76 We call for the need to develop psychological interventions for cancer patients to improve their quality of life and reduce their levels of mental problems. 77 Third, the methodological quality of the included studies was high, and the results of the metaanalysis were reliable.

| STUDY LIMITATIONS
This meta-analysis has some limitations. First, a significant heterogeneity was observed among the included studies. This may be caused by the type of cancer and different measuring scales or cutoff values. Second, there may be interactive effects among some clinical factors such as cancer stage, education level, and work status.
Nevertheless, due to limited statistical methods, it is difficult to explore the source of heterogeneity and the impact of these factors on the results through quantitative analysis. Third, significant publication bias was observed in some variables (e.g., depression and anxiety), which may underestimate the prevalence of mental health issues. In future meta-analyses, stricter criteria, such as a limited evaluation scale and cancer type, should be adopted when selecting the included articles.

| CONCLUSION
Our meta-analysis revealed the PR of depression, anxiety, PTSD,