Level and Predictors of Trauma-symptoms Among Health Workers and Public Service Providers During the COVID-19 Outbreak

Hypotheses/Research questions:

H1: Health employees and public service providers working directly with COVID-19 will have higher PTSD-symptoms, anxiety, depression and health anxiety, compared to health workers and public service providers working indirectly with the virus.

H2: Presence of a psychological disorder, anxiety and depression will predict PTSD symptoms. Those reporting having a diagnosis will reveal higher PTSD-symptoms. Higher levels of anxiety and depression will be associated with higher PTSD-symptoms.

H3: Emotional support, burnout, health-anxiety, worry about job and economy, interpersonal problems and metacognition will predict PTSD-symptoms controlling for direct vs. indirect exposure to trauma, psychological diagnosis, demographic variables (age and gender, living with a partner, living with children, anxiety and depression. Higher level of burnout, health-anxiety, worry about job and economy, interpersonal problems and metacognition will be associated with higher PTSD-symptoms. Higher emotional support will be associated with lower PTSD-symptoms.

Participants were asked to fill out a set of validated questionnaires including demographic variables, psychological symptoms and PTSD-symptoms. Some questionnaires are given as a whole, whereas other questions includes theoretically-driven selections of items from validated questionnaires. This study is part of a 'The Norwegian COVID-19, Mental Health and Adherence project". The study involves a Cross-sectional survey design. Health personnel and public service providers where systematically targeted through various channels: e.g., primarily through systematically reaching out to all hospitals in Norway and inviting health personnel to participants, through contacting the Associations of all major health-worker groups (i.e., Norwegian Medical Doctors Association; Norwegian Nurses Association, Norwegian Psychologists Association; and other associations related to health-workers) through national TV, national, regional and local radio stations, and national, regional and local newspapers in the different regions of the country, as well as Facebook groups in the different regions of the country. Additionally, a random selection of those qualifying as health-care workers were randomly targeted on Facebook by a Facebook Business algorithm. Public service providers were recruited trough Facebook. Politicians where also systematically contacted, with all political parties being contacted and subsequently sending an e-mail with the survey to their members.

Data collection started during the time-period with the strictest and equal number of government-initiated non-pharmacological interventions (NPI's) in Norway, and data collection was stopped once these NPI's were modified or new information about NPI's were added. The data include one directly identifiable variable (contact information) for participants in accordance to the General Data Protection Regulation (GDPR) law in EU, which is to give the participants the opportunity to have their data deleted upon request. Data are thus kept on a safe server belonging to the University of Oslo and is accessed first following de-identification. The project outline and study plan was also registered upon application to the Norwegian "Regional Ethical Committee" 10 days before data collection, a committee which evaluate the rational for data collection and hypotheses as well as evaluate the ethical aspects of the study before allowing the data to be collected.

Measures

PCL-5, PHQ-9; GAD-7; demographic variables (gender; age; relationship/marital status; living with children and number of children; employment status; education level); situational variables related to Covid-19, worry about own health and worry about economy/work, emotional support, interpersonal problems, burnout, metacognitions and whether one has a psychological diagnosis or not. The outcome variable is PCL-5. Predictors are specified in the hypothesis-section.

Indices:

Given Acceptable Cronbach's alpha (above 0.7) the investigators combined four items that measure health anxiety and fear of death related to COVID-19. Furthermore two items will be combined to represent worry about work and economy. Two items on emotional support were combined. On the CAS1 four items were combined into the subscale "positive metacognition" and four items were combined into the subscale "negative metacognition". Interpersonal problems consisted of 13 items selected from the IIP-64. Burnout was measured with a single item

Inference:

Given the large sample size in this study, the investigators pre-defined our significance level:

p < 0.01 to determine significance

Sample size 'Norwegian COVID-19 and Mental Health and Adherence Project' involves multiple studies, where some involve a Complex Systems (Network analysis approach). These multivariate analyses require large samples and power analysis was conducted accordingly. Following power analysis guidelines by Fried \& Cramer (2017), it is recommended that the number of participants be at the very least three times larger than the number of estimated parameters. However, more conservative recommendations by Roscoe (1975) for multivariate research, recommends sample size that is ten times larger than the number of estimated parameters. Thus, with the more conservative estimates by Roscoe, an optimal sample size included 1900.

Statistical models:

Descriptive statistics with frequency tables including N, mean and SDs or median and interquartile range depending on the data will be presented. A cut-off of 31 on the PCL-5 will be used together with the DSM-5 diagnostic guidelines applied to the PCL-5 to categorize participants as fulfilling the PTSD symptom criteria or not. Participants indicating scores of 2 or above on at least one of five re-experiencing symptoms, one of two avoidance symptoms, two of seven symptoms of negative alterations in cognition and mood and two of six arousal symptoms were classified as fulfilling the PTSD symptom criteria. Subclinical PTSD will be assessed by specifying how large part of the sample which fulfil parts of the PTSD-criterias. The results will be benchmarked against national and international studies.

Independent sample t-tests or Mann-Whitney U tests, depending on level of skewness, will be used to compare the PCl-5 scores of those working direct vs. indirect with COVID-19 patients. A One-way ANOVA or Kruskal-Wallis test, depending on level of skewness, will be conducted to exploratory investigate differences between the different types of health-workers (e.g. medical doctor and clinical psychologists) and public service providers (e.g., politicians, social security). If there is significant differences between the groups the investigators will either use a post hoc test, Tukey (HSD) or Dunn-Bonferroni, depending on wether a parametric or non-parametric test has been used.

Two multiple regression analyses will be conducted with PCL-5 as the dependent variable. Specific predictors of the different regression are listed in the hypothesis section. Multicollinearity and other assumptions will be checked; if the multicollinearity assumption is violated (if VIF > 5 and Tolerance < 0.2.

Exploratory analyses:

Exploratory: Examine the differences in levels of trauma, anxiety, depression and health anxiety among different health workers, public service providers and demographic subgroups.

The investigators may also do other exploratory analysis that not yet thought of at the time of pre-registration. Such analysis will be explicitly stated as exploratory.

Note that this project outline, study plan and analysis was registered upon application to the Regional Committees for Medical and Health Research Ethics (REC) and Norwegian Centre for Research Data.