06.05.2020

Development and Initial Validation of the COVID Stress Scales

Diagnostic Psychiatry
Taylor S et al
J Anxiety Dis

Main result

High level of anxiety and depressive symptoms for 28% and 22% of the total population respectively.
According to the PHQ-4 scale: 54% of subjects had normal levels of anxiety and depression, 23% of mild symptoms, 13% of moderate symptoms, and 10% of severe symptoms.

Exploratory factor analysis: parallel analysis revealed a 5-dimensional solution instead of 6 predicted. Inter-correlation between dimensions ranges from 0.29 to 0.49. The 5 dimensions identified are:
<P>
1. fear of COVID-19 dangerousness and contamination 

2. fear of socio-economic consequences

3. COVID-19 related xenophobia

4 COVID-19 associated compulsive checks

5. COVID-19 related symptoms of traumatic stress.
<P>
Maintenance of 12 items instead of 6 for the dimension grouping "fear of the dangerous nature of the pandemic and contamination", allowing those two aspects to be separated in future studies.

Confirmatory factor analysis: good quality of estimate with RMSEA = 0.050 (95% CI: 0.049-0.051), SRMR = 0.042 and CFI = 0.93.
A robustness analysis with 2-factor confirmatory factor analysis with simultaneous adjustment on both populations shows a good quality of estimation (RMSEA = 0.050, SRMR = 0.053 and CFI = 0.92) and therefore a reproducible factor structure between these two countries.

Internal coherence: good to excellent depending on population and size. All scales were cross-correlated, indicating consistent COVID-19 stress syndrome.

Convergent validity: favorable because the correlations between CSS scales (COVID-19 Stress Scales) and pre-COVID personality traits of anxiety and obsession-compulsion were significant and ranging from medium to important.

Discriminant validity: all correlations favored good discriminant validity

  • the correlations between the 5 scales of the SSC and social desirability were significant due to the number of patients, but of small importance (from -0.14 to -0.05).
  • The correlations with current anxiety were more important than with current depression.
  • The scale of xenophobia related to COVID-19 was more correlated with the scale of global xenophobia than with the scale of medium distress.

Takeaways

The COVID-19 Stress Scales seem to be useful for studying mental disorders associated with COVID-19 and future pandemics. They could also help identify patients requiring psychological assistance and targets for public health interventions. However, those scales should be further validated in other populations.

Strength of evidence Moderate

- Pre-proof study (-)
- Based on an online self-administered questionnaire (-)
- Scale created from two geographically close and culturally relatively close populations (-)
- Adjustment of existing scales (-)
- Alpha risk correction (+)
- No validity evaluation via the use of DSM-5 diagnostics e.g. (-)
- No comparison to other methods evaluating COVID-19-associated anxiety unpublished at the time of the study (-)

Objectives

Develop and evaluate screening scales for measuring mental disorders associated with COVID-19 pandemic and future pandemics: the COVID-19 Stress Scales.

Method

Scales constructed from data from two populations, collected between 21/03/2020 and 01/04/2020 and sampled to be representative of the source population in terms of age, sex, ethnicity, socio-economic status and geographical region:
  • 3479 adults living in Canada 
  • 3375 adults living in the United States.

Subjects have an average age of 49 years (18 to 94 years), 53% are men and 79% have reached higher education.

The survey included demographic questions, including current anxiety and depression, pre-pandemic personality traits, and COVID-19-associated mental disorders (over the previous 7 days, assessed on a 5-point scale).
For the latter, pre-identification of 6 domains via literature and expert opinion:

  1. fears of COVID-19 dangerousness (14 items)
  2. fears of contamination sources (8 items)
  3. COVID-19-related xenophobia (7 items)
  4. fears of COVID-19 personal socio-economic consequences (10 items)
  5. COVID-19-related checks (7 items) traumatic stress symptoms (7 items)

Validation scales: use of personality trait measurements to assess convergent and discriminant validity as well as social desirability bias.

  • Patient Health Questionnaire-4 (PHQ-4): measure of current anxiety and depression
  • Short Health Anxiety Inventory (SHAI): measure of health-related anxiety, regardless of physical health status, over the last 6 months
  • Obsessive and Compulsive Symptoms Assessment Questionnaire (OIC-R): use of two sub-scales for checking and cleaning
  • Xenophobia scale (XS): measure of negative attitudes towards immigrants
  • Short version of the Marlowe Crowne Social Desirability Scale (MCSD-SF): measure of tendency to respond

Scale Construction and Assessment Procedures: Separate results for both countries to determine robustness.

  • An exploratory factorial analysis was carried out for each country and for each of the scales corresponding to the 6 pre-identified domains.

For each scale, a parallel analysis was conducted.
For each scale, select the 6 items with the highest demands, for each of the two countries.
Exploratory factor analysis of the 36 items selected (6 domains * 6 items) on the Canadian population followed by assessment of the stability of the multifactorial structure by confirmatory factor analysis on the US population.
Quality of fit indices: SRMR (Standardized-Root-Mean-Square Residual), RMSEA (Root-Mean-Square Error of Approximation) and CFI (Comparative Fit Index).
Internal consistency was assessed using Cronbach's alpha coefficient and multiple correlation analyses were conducted.
Adjustment of the alpha risk significance threshold to 0.01 to take into account the multiplicity of analyses.

Read paper

bibliovid.org and its content are bibliovid property.

Legal Notice