Table 3.
Predictors of Change in Log-Total Costsa
| N = 434 | ||
|---|---|---|
| Parameter | Estimate | Standard error |
| Intercept | 4.01 | 0.75 |
| Control | −0.15 | 0.23 |
| Intervention | Ref | . |
| Pre-baseline cost (log) | −0.42 ***b | 0.05 |
| Charlson/Deyo | 0.03 | 0.02 |
| Patient age | −0.00 | 0.01 |
| Caregiver age | −0.00 | 0.01 |
| Spouse caregiver (No) | −0.06 | 0.14 |
| Spouse caregiver (Yes) | Ref | . |
| Days between diagnosis and intervention | 0.00 | 0.00 |
| PDC region (South) | 0.07 | 0.22 |
| PDC region (North) | Ref | . |
| ADLsc | 0.05** | 0.02 |
| Behavioral problems | 0.02 | 0.02 |
| Cognitive impairment | −0.03* | 0.01 |
| Priority group (1) | 0.32 | 0.17 |
| Priority group (2–6) | 0.34*** | 0.10 |
| Priority group (7,8) | Ref | . |
aWe used generalized linear mixed model regression with change in log cost as the dependent variable. Individuals were nested within the PDC site, with site treated as a random variable and nested within the intervention arm
b p ≤ 0.05 (*), p ≤ 0.01 (**), p ≤ 0.001 (***)
c ADLs activities of daily living