Study population
A retrospective online survey among nurses working in care homes was conducted between the 15th of November 2020 and the 28th of February 2021. The targeted time frame of the survey was between the 1st of March 2020 and the 30th of June 2020 in accordance with the first wave of the pandemic and the following ‘Lockdown’ in Germany. The survey is part of the cooperation project COVID-Heim, which aims to draw lessons from the pandemic for structural developments in the care home setting by combining various data sources in Germany. The sample consisting of 811 nurses was recruited online via social media. We invited all nurses working in care homes in Germany to participate in our online survey in relevant German facebook groups. The survey was provided by a secure web application for building and managing online surveys and databases – Research Electronic Data Capture System (REDCap). The questionnaire was opened 1,884 times and fully completed by 811 nurses. For the anonymous survey, we followed all relevant guidelines and regulations. The online survey was approved by the ethics committee of the Faculty of Medicine of the Charité – Universitäsmedizin Berlin (EA1/254/20).
Measures
Characteristics of participants and affiliated care homes
The questionnaire included the following sociodemographic characteristics of the participants: age, sex, qualification (certified nurse, geriatric nurse, nurse in training, health care assistant, other), employment status (permanent, temporary), and risk factors for COVID-19 infections (e.g., >60 years, cardiovascular disease, diabetes mellitus, immunodeficiency, chronic obstructive pulmonary disease) which was taken from the COVID-19 Pandemic Mental Health Questionnaire (CoPaQ) [17]. The characteristics of the affiliated care homes included questions regarding size of care home (small [1-50 residents], medium [51-100 residents], large [>100 residents]), focus of care (e.g. dementia, mental illness, palliative care) and COVID-19 cases residents/staff (no COVID-19 cases, 1-20 COVID-19 cases, 11-20 COVID-19 cases, >20 COVID-19 cases).
COVID-19-related measures
Working demands since the pandemic. To evaluate the change in working needs since the coronavirus pandemic, we asked nurses how working demands changed during the COVID-19 pandemic since 1st of March 2020, with three possible answer categories: 1) strongly/rather increased 2) did not change 3) rather/strongly decreased.
Satisfaction with COVID-19-related management of care home manager. We asked nurses whether they were satisfied with the COVID-19-related management of their care home manager between the time of 1st of March until 30th of June 2020 (e.g., implementation of protection and hygiene measures against the COVID-19, communication with staff); two answers were possible: 1) very/rather satisfied, 2) very/rather unsatisfied.
COVID-19-related burden at work. Nurses were asked to state to what extent the following possible COVID-19-related responsibilities at work were perceived as burden between 1st of March until 30th of June 2020 on a 4-point Likert scale ranging from 0) no burden to 3) extremely severe burden. Items were 1) The concern about the wellbeing of the residents; 2) Purchase and consumption of personal protection equipment; 3) Compliance with hygiene guideline of Robert Koch-Institute; 4) Concerns about COVID-19-infections among staff; 5) Concerns about COVID-19-infections among residents; 6) Implementation of protection measures (e.g., isolation of COVID-19-infected residents, contact precautions for relatives); 7) Small number of COVID-19-tests for residents; 8) Small number of COVID-19-tests for staff; 9) High expectations of relatives. The nine single items were adapted from Hower, Pfaff, and Pförtner [9]. We generated an index – ranging from 0 to 27 - by summing up all items answered. Higher results indicate more burden due to COVID-19 at work. The scale obtained a Cronbach's alpha score of 0.82 in our sample, which can be considered as ‘good’ [18].
COVID-19-related anxiety. Nurses were asked to state to what extent the following statements applied to them in the time from 1st of March until 30th of June 2020 on a 7-point Likert scale ranging from 0) does not apply at all to 7) fully applies. Items were 1) I was afraid to get infected with Corona; 2) I was afraid of the consequences of the Corona Pandemic on my life; 3) I was afraid of the consequences for my health if I got infected; 4) I was afraid of the consequences for the health of my relatives if I got infected; 5) I was afraid of the social consequences of Corona and 6) I was afraid of the economic consequences of Corona on my life. The six single items were adapted from Petzold and colleagues [19]. We generated an index - ranging from 0 to 36 - by summing up all item answers. Higher results indicate more COVID-19-related anxiety. The scale obtained a Cronbach's alpha score of 0.80 in the present sample, considered ‘good’ [18].
Stress, anxiety, and depression
Stress Scale (SDASS-21). Stress symptoms were measured with the Stress Scale of the Depression Anxiety Stress Scale (DASS-21). The subscale consists of 7 items and asks for stress criteria on a 4-point Likert scale ranging from 0) did not apply to me at all to 3) applied to me very much/most of the time. The final scores - ranging from 0 to 21 - were multiplied by two [20, 21]. A score of 14 or less is considered ‘normal,’ 15-18 ‘mild,’ 19-25 ‘moderate,’ 26-33 ‘severe,’ and 34 or greater is considered ‘extremely severe’ [22]. The DASS-21 is widely used to study depression, anxiety, and stress in the general population and health care workers [12, 23]. In the present sample, the validated German version of the stress scale of the DASS-21 obtained a Cronbach’s Alpha of 0.88.
Generalized Anxiety Disorder - 2 (GAD-2). General anxiety symptoms were measured with the ultra-short 2-item version of the 7-item scale GAD-7. It incorporates the first two questions of the GAD-7, which are critical components of every anxiety disorder [24]. The score ranges from 0 to 6. A cut-off value of ≥ 3 was suggested to detect possible clinically relevant levels of anxiety symptoms. The GAD-2 is widely used to screen for general signs of anxiety, and its psychometric properties are well documented [25]. In the present sample, the Cronbach’s alpha of the GAD-2 was 0.80.
Patients-Health-Questionnaire – 2 (PHQ-2). Depressive symptoms were measured with the ultra-short 2-item version of the 9-item scale PHQ-9 [26]. It incorporates the first two questions of the PHQ-9. The score ranges from 0 to 6. A cut-off value of ≥3 was suggested to detect possible clinically relevant levels of depressive symptoms. The PHQ-2 has been widely used during the COVID-19 pandemic [12, 13]. In the present sample, the Cronbach’s alpha of the PHQ-2 was 0.82.
Social Relations at Work
Social Relations at work were measured by 3 subscales of the validated Copenhagen Psychosocial Questionnaire (COPSOQ) [27]. We used ‘support at work (4-item scale)’, ‘feedback (2-item scale)’ and ‘sense of community (2-item scale)’ All items were scored on a 5-point Likert scale ranging from 0) never/hardly ever to 100) always. In our sample, the 3 subscales obtained Cronbach’s alpha scores of 0.83, 0.69, and 0.88, respectively.
Statistical analysis
Frequencies, percentages, means, and standard deviations for characteristics of participants and affiliated care homes were generated. Further, normal distribution was tested using the Shapiro-Wilk-Test. The Shapiro-Wilk-Test revealed that the data was not normally distributed. Therefore, relationships between Covid-19-related burden at work, stress, anxiety, depression, and categorical characteristics of participants and affiliated care homes were tested using the Mann-Whitney-U-Test or the Kruskal-Wallis-Test, respectively. We used Spearman correlation tests to test relationships between COVID-19-related burden at work, stress, anxiety, depression, and continuous variables. Finally, we used multiple regression analyses to determine factors associated with COVID-19-related burden at work, stress, anxiety, and depression. Analyses were performed using IBM SPSS Statistics for Windows, version 25.0 (IBM Corp., Armonk, NY). P values <.05 were considered to indicate statistical significance.