Table II.
Univariable and multivariable ordinal logistic regressiona
| Variable | Univariable |
Multivariable |
||
|---|---|---|---|---|
| OR (95% CI) | P value | OR (95% CI) | P value | |
| Gender | 2.20 (1.79-2.71) | <.001 | 1.39 (1.08-1.79) | .010 |
| COVID-19-related modifications of home routine | 1.75 (1.50-2.05) | <.001 | 1.38 (1.14-1.67) | .001 |
| COVID-19-related stressors | ||||
| Donning/doffing PPE | 3.87 (3.10-4.84) | <.001 | 1.81 (1.41-2.33) | <.001 |
| Worry about potential adverse patient outcomes due care delay | 2.68 (2.18-3.31) | <.001 | 1.47 (1.16-1.87) | .002 |
| Financial concerns | 3.13 (2.54-3.85) | <.001 | 1.90 (1.49-2.42) | <.001 |
| Summative hospital supportb | 0.84 (0.78-0.89) | <.001 | 0.83 (0.76-0.91) | .001 |
| Active coping strategies | ||||
| Use of emotional support | 1.41 (1.33-1.50) | <.001 | 1.09 (1.01-1.18) | .028 |
| Positive reframing | 1.16 (1.10-1.23) | <.001 | 0.88 (0.81-0.95) | .001 |
| Planning | 1.37 (1.29-1.45) | <.001 | 1.12 (1.03-1.21) | .008 |
| Avoidant coping strategies | ||||
| Self-distraction | 1.66 (1.55-1.78) | <.001 | 1.32 (1.21-1.43) | <.001 |
| Substance abuse | 1.53 (1.43-1.65) | <.001 | 1.16 (1.06-1.27) | .001 |
| Behavioral disengagement | 1.77 (1.63-1.93) | <.001 | 1.26 (1.14-1.39) | <.001 |
| Venting | 2.17 (1.99-2.37) | <.001 | 1.42 (1.28-1.59) | <.001 |
| Self-blame | 2.12 (1.93-2.32) | <.001 | 1.35 (1.21-1.51) | <.001 |
CI, Confidence interval; COVID-19, coronavirus disease 2019; OR, odds ratio; PPE, personal protective equipment.
A proportional odds cumulative logit model was used of factors associated with an increasing degree of self-reported anxiety (no, mild, moderate, and severe anxiety) by 1259 respondents who had completed the Generalized Anxiety Disorder 7-item survey. The model was built in two steps. First, all factors significant on univariable analysis (also listed in Supplementary Table II and Supplementary Fig 2 (online only), Supplementary Fig 3 (online only), Supplementary Fig 4 (online only), Supplementary Fig 5 (online only)) were considered for entry into the model using a stepwise selection process (with the exception of the active and avoidant coping strategies from the Brief Coping Orientation to Problems Experienced survey). Second, the variables selected in the first step were retained, and the active and avoidant coping strategies were included in a stepwise selection process.
Summative hospital support was created as a sum of the significant individual hospital support variables, including “operating room protocols,” “adequate personal protective equipment,” “preoperative testing of patients,” and “opportunities to interact with leadership,” because they all correlated. “Transparency from leadership about COVID-19 management and planning” was not included because the responses were nearly identical to those for “opportunities to interact with leadership.”