Eight Months Into The COVID-19 Outbreak, Whether This Pandemic Still Psychologically Affected Convalescent Psychiatric Patients: A Cross-Sectional Study.

Background This study aimed to nd out eight months into the COVID-19 outbreak, how much psychological effect was experienced by convalescent psychiatric patients, and identify the potential risk and protective factors related to their psychological response. Methods Data from 136 convalescent psychiatric patients were collected. We used binary logistic regression analyses to identify participants’ SRQ score and the factors associated with it. Results There were 47 (34.5%) participants scored eight or higher. Working or marital status and annual income were signicantly associated with a higher score in SRQ. Similarly, sleep duration, interpersonal relationship, exercise time, special events, and general condition were linked to SRQ grade.

reported that COVID-19 infected 50 patients and 30 medical staff at Wuhan Mental Health Centre 10 . Most hospitals stopped regular family visit to protect patients who stayed in closed psychiatric wards from being contaminated.
One research revealed that severe mental disorder patients gave a higher anxiety response than patients with other mental disorder and healthy controls 11 . For patients who are in the recovery phase, their treatment continuity is under threat. Although quarantine and lockdown restrictions are crucial to containing the virus, these measures disrupted their communication with loved ones and restricted public mental health service access. According to a report, isolation from the family and the loneliness increased disease recurrence and mental health status deterioration 12 . We reasonably assume that convalescent psychiatric patients encounter no fewer obstacles than their counterpart in hospitals. This ongoing public health crisis reduced outpatient visits and hindered patients from obtaining timely medication. The previous study conducted by Gao et al. found that 70% of outpatients with emotional disorders had to postpone their mental health treatment, and had more negative expectations towards COVID-19 13 . However, there remains few research on investigating the in uence of COVID-19 pandemic on psychiatric patients in stable condition; and the published surveys were about the psychological impacts in the early stage of the outbreak. As this pandemic in China is now under control, the whole workforce and outpatient service are getting back to normal. Additional research is needed to improve understanding of the after psychological effects of infectious disease outbreaks.
Thus, this study has two aims. Firstly, we want to nd out that eight months into the COVID-19 outbreak, whether there is still a psychological effect of this pandemic on our convalescent psychiatric patients. The second aim is to identify the risk factors related to their psychological response.

Method
To investigate the psychological effects of the post-pandemic, we utilized a cross-sectional method. We sent an online questionnaire via social media (Wechat) to psychiatric patients from the data collected by Tongde Hospital of Zhejiang Province, Hangzhou, China. The psychiatric patients were recruited from 1 to 31 August 2020 and signed written informed consent. The authors state that all procedures contributing to this study comply with the relevant national and institutional committees' ethical standards on human experimentation. The Ethics Review Committee of Tongde Hospital of Zhejiang Province approved all procedures involving human subjects/patients. Inclusion criteria were: (1) aged 18 years old or above; (2) had been previously diagnosed by a psychiatrist as schizophrenia, schizo-affective disorder, bipolar disorder, major depression, anxiety disorder, or obsessivecompulsive disorder according to Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition or International Classi cation of Diseases 10th Revision; (3) able to understand and complete the survey and accept informed consent. Exclusion criteria included: (1) aged 18 below; (2) inability to understand or complete the survey; (3) presence of any neurological, cardiovascular, respiratory, endocrine, or in ammatory diseases; (4) suspected or con rmed cases of COVID-19.
General demographic information included participants' age, gender, educational level, career, nancial condition, marital status, and clinical pro le (previous diagnosis and medication compliance).
Questions related to COVID-19 have two parts. In part one, we asked the participants to answered six basic questions about COVID-19 itself, such as 'this virus is airborne and droplets from cough and sneezes could transmit the virus', and 'washing your hands with soap and water can kill the virus that may be on your hands'. Then 'fully aware' would be recorded for participants get all questions right, 'partially aware' for at least three correct responses, and 'poorly aware' for less than three. Next, participants were required to answer questions like 'do you have enough protective equipment'. In part two, we designed several additional questions about the psychological in uence of pandemic: sleep duration, interpersonal relationship, exercise time, and general state. Responses were graded from 1 ( very satis ed) to 5 ( not at all satis ed). Participants also needed to consider any special event, other than this pandemic, impacting them positively or negatively. Grade 1 means very positive impact, while grade 5 means very negative impact.
There were 20 questions in Self-reporting Questionnaire, intending to estimate if the participants displayed any mental distress due to the COVID-19 pandemic. The total SQR score was divided into the low-risk group (< 8) and high-risk group (≥ 8).
All the analyses were conducted using SPSS Statistics 24.0, and the level of signi cance was set at 5%. Descriptive statistics were used to summarize the variables, mean and standard deviation were used for continuous variables, while frequency and percentage were used for categorical variables. Inferential statistics, including independent sample t-test, and Mann-Whitney U test, was used to examine if there was any difference in the outcome variables.
We adopted a binary logistic regression analysis to examine the association between the outcome variables and the demographic variables.

General demographic information of participants
One hundred and thirty-six convalescent psychiatric patients completed the survey. The participants' mean age was 32.74 (SD=14.94), and 94 (69.12%) were female. Over 60% were not married, and 46.32% were patients who held an undergraduate degree. A total of 38.97% of the patients were working, and 66.18% had an annual income at less than 60,000 RMB. A larger number of the respondents had major depressive disorder (33.09%), followed by bipolar disorder (19.12%) and schizophrenia (11.65%), with the average duration of disease for 5.73 years. Further information about sociodemographic and clinical aspects refer to Table 1.  Regarding other sociodemographic characteristics, the independent t-test detected being younger was a risk factor for getting a higher score in SRQ, whereas the duration of disease was not.
In the next analysis step, we included all variable with statistically signi cant differences among the convalescent psychiatric patients in the binary logistic regression. The results demonstrated the robust relationship between SRQ score and personal annual income (OR=3.340,95%CI:1.135-9.831) or special events (OR=0.231, 95%CI:0.100-0.534).
Patients with less annual income and had negative life events were signi cantly more likely to endorse higher SRQ score. Furthermore, a higher score moderately correlated with sleep duration (OR=0.506,95%CI:0.259-0.987) and general conditions (OR=0.395, 95%CI:0.172-0.905). This result suggests that patients who reported to sleep less and thought they were in bad general conditions tend to be more concerned. Among these factors, sleep duration and general conditions connected with annual income, which indicated patients who earn more than 60,000 RMB each year sleep more and feel more content with their conditions.

Discussion
This study indicated that one-third of participants present with mental distress even when China was recovering from the COVID-19 pandemic. In the earlier stage of this pandemic, several studies conducted to estimate the level of mental distress induced by COVID-19 were inconclusive. Some research revealed that the symptoms of psychiatric patients worsened 1415 . Besides, as a vulnerable group of the population, they showed signi cantly more mental distress than people without psychiatric illnesses 16 . However, one study showed no noteworthy difference in the distress level between patients with severe mental illness and the general population 17 . In another study, anxiety disorder patients did not exhibit with marked psychological impact secondary to COVID-19 18 . According to Gao et al., 43.2% of patients with emotional disorder were affected, similar to our nding 13 . One explanation we proposed would contribute to this result: negative news like the virus might strike again during the upcoming winter trigger a series of emotional stress responses including anxiety and fear since these stable patients could easily access the Internet and media. Patients with psychiatric diseases tended to hold less positive anticipation for a future epidemic prospect 13 , resulting in worries about their routine medical consultant and drug compliance, which are critical in their convalescence, would be disrupted again.
According to our study, unmarried status correlated with a higher SRQ score. This result contrasted with those obtained in pre-existing mental disorder patients 19 . Convalescent psychiatric patients who remained single might have worse personal and more trouble maintaining relationships at baseline, making them easier to be affected. A prior national survey indicated that those who were single or widowed or divorced were less likely to seek mental health service 20 . Without timely help-seeking behaviour, convalescent psychiatric patients, in particular, who are single, might struggle more to cope with the accumulating adverse emotional effects induced by COVID-19. However, we found the marital status had no statistically signi cant effect on patients' SRQ score with an additional analysis conducted, which may partially because some of our participants were students. Age, another in uencing factor concerning anxiety caused by the pandemic 8 , had its role to play in increasing our patients' SRQ score. Compared with older age patients, younger patients had less experience and weaker psychological self-regulation ability when faced this health crisis might be the reason.
It is also showed that events that happened during this period affected patients' psychological distress level. In our study, we made no restriction on the types of the event, intending to focus on the psychological impact they might have on our participants. A large body of research had provided evidence that negative life events were partly responsible for the onset of mental disorder 2122 . Moreover, stressful life events and the rst appearance of neuroses and depression revealed a strong association 23  School children and adolescent, in particular, may feel more anxious because they had to learn in a completely different environment, which leads to sleep di culties 26 . Moreover, now the companies and schools reopened, convalescent psychiatric patients may need more time for adaption. Besides, working patients might be anxious about their nancial security with fears of a new recession following this pandemic. Times like this was prone to trigger negative emotion amongst patients who used to be in stable recovery, and their sleep duration could be affected. If the patients experienced symptoms uctuation, naturally, their self-perceived condition would be worse.
One impressive result of our study was annual income remarkably associated with self-report psychological impacts. The higher socioeconomic status convalescent psychiatric patients were in, the less worried they felt. Base on the income survey published in 2020, the average earnings for citizens from Zhejiang Province, where most of our participants lived, was 60,000 RMB 27 . Therefore, we set this amount as the point by which we dived participants into two groups, the high-income one and the low-income one. The COVID-19 outbreak affected global economies and inevitably interfered every aspect of human life. Primary sectors, secondary sectors, and tertiary sectors all suffered from heavy losses because of the implement border shutdown, travel control and strict home con nement measures 5 .
From an economic perspective, people in the low-income group might lack resources to cope with external risks.
Even when the COVID-19 pandemic seemed to get closer to an end, they may be concerned about the economic burden of the family and family members' well-being as well, which created signi cant challenges for their coping strategies. Their economic disadvantage, including decreased wages and job loss due to the post-pandemic effect, damaged the treatment continuance. Convalescent psychiatric patients from the low-income group might prefer to spend money on daily expenditure rather than medications if they had reduced personal disposable income. Also, they might not get su cient protective equipment because of nancial strain. Another reason would be that their unstable living conditions make them more susceptible to COVID-19 infection. Moreover, we found out that higher income was associated with longer sleep duration and more content feeling about the general condition in stable patients. Because sleep duration and general condition of the patients affected how they reacted to the pandemic, we rationally suggested that patients with higher income would report fewer problems on those two aspects and had a lower distress level.
This study has several limitations. First of all, since our research had a modest sample size of 136 participants and lacked a control group, caution is required in interpreting the results. Second, the way we used to collect data excluded convalescent psychiatric patients who have less access to digital technologies such as smartphones or the Internet and, therefore, our ndings may not adequately re ect trends seen throughout the whole targeted population. Third, we adopted self-reported psychometric tools that have common drawbacks of such instruments.
Finally, because the causal relationship indicated in our study was based on cross-sectional data, more research should be carried out to clarify the reason behind its existence.
To the best of our knowledge, this is one of the few studies that focused on convalescent psychiatric patients and whether COVID-19 pandemic still has psychological impacts on them.

Conclusions
According to our study, eight months into the COVID-19 pandemic, there was still more than a third of the participants reported they were psychologically affected by this widespread virus. Furthermore, this result was also related to less annual income. This nding stressed the need to offer comprehensive interventions and medical service for convalescent psychiatric patients during pandemic's peak and the necessity to provide closer follow-up and nancial support in the late stage of this health crisis. Future studies should be conducted with a longitudinal design to study the psychological impact in this group of patients over time.