Table 2.
Negative binomial, linear, and logistic regressions assessing associations between QOD-NS and 90 day economic measures/medication use variables
| Economic Measures | Unadjusted β | Adjusted β | P-Value | Prevalence Ratio | CI |
|---|---|---|---|---|---|
| Days of missed productivity a | 0.037 | 0.031 | <0.001 | 1.032 | (1.016, 1.047) |
| Days of missed employment a | 0.050 | 0.034 | <0.001 | 1.035 | (1.020, 1.049) |
| Hours of missed employment due to physician visits b | 1.123 | 1.275 | 0.002 | - | (0.455, 2.096) |
| Average percent productivity b | −0.005 | −0.005 | 0.014 | - | (−0.008, −0.001) |
| Minutes spent in sinus care b | 0.561 | 0.355 | 0.015 | - | (0.069, 0.641) |
| Hours of child care b | −0.007 | −0.007 | 0.813 | - | (−0.065, 0.051) |
| Distance traveled to medical appointment b | 2.583 | 3.387 | <0.001 | - | (1.785, 4.990) |
| Disability Insurance c | 0.076 | 0.083 | 0.003 | 1.086 | (1.028, 1.147) |
| Medication usage | |||||
| Days on oral steroid a | 0.020 | 0.014 | 0.044 | 1.014 | (1.000, 1.028) |
| Days on oral antibiotic a | 0.002 | 0.002 | 0.766 | 1.002 | (0.991, 1.013) |
QOD-NS: Questionnaire of Olfactory Disorders
Models were adjusted for age, gender, nasal polyp status, depression, asthma, allergy, SNOT22, CT score, and endoscopy score in combinations specific to each economic variable.
Interpretation example for negative binomial regression: On average, for a one point decrease in QOD-NS, days of missed productivity increased by 3.2% and we are 95% confident that the true estimate is between 1.6% and 4.7%.
Interpretation example for linear regression: On average, for a one point decrease in QOD-NS, hours of missed employment goes up by 1.275 and we are 95% confident that the true estimate is between 0.455 and 2.096 hours.
Interpretation for logistic regression (Prevalence Odds Ratio): On average, for a one point decrease in QOD-NS, the odds of being on disability insurance increases by 8.6% and we are 95% confident that the true estimate is between 2.8% and 14.7%.