Table 3.
Multivariate logistic regression with backward elimination predicting cognitive complaints vs. no cognitive complaints 1.
| Variable | Odds Ratio | Wald | B | 95% Confidence Interval (Lower Bound) | 95% Confidence Interval (Upper Bound) | p-Value |
|---|---|---|---|---|---|---|
| Peak COVID symptom score | 1.321 | 6.622 | 0.279 | 1.069 | 1.634 | 0.010 |
| PHQ-9 score | 1.363 | 5.604 | 0.309 | 1.055 | 1.760 | 0.018 |
| RBANS Attention | 0.912 | 4.679 | −0.092 | 0.840 | 0.991 | 0.031 |
| Trails A | 0.890 | 3.967 | −0.116 | 0.794 | 0.998 | 0.046 |
| Age | Removed by backward stepwise (conditional) elimination | |||||
| Number of Medical Comorbidities | ||||||
| Appt. 1 COVID symptom score | ||||||
| Chalder score | ||||||
| PCL-5 Score | ||||||
| GAD-7 | ||||||
| RBANS immediate memory score | ||||||
| RBANS Visuospatial | ||||||
| RBANS Language | ||||||
| Trails B | ||||||
| Stroop CW | ||||||
| MoCA | ||||||
1 In order to determine predictors of cognitive complaints, a backward conditional variable selection logistic regression model was developed utilizing all variables with significant differences between the CC group and the NC group. The procedure excluded 12 of the variables and included Chalder in the final model, although it was not statistically significant. The model, including the remaining four predictors (severity of acute COVID-19 illness, PHQ-9 score, RBANS Attention, and Trails A), was significant (−2log likelihood = 37.884, Chi-Square = 53.088, df = 5, p < 0.001).