Table 3 Summary of patient characteristics found to be significantly associated with total transitions by time period prior to death
| Â | Association of predictor to total transitions* | ||
|---|---|---|---|
| Â | Time periods when transitions experienced | ||
| Â | PCP registration to death | Four weeks prior to death | Two weeks prior to death |
Total transitions experienced(Mean; Standard Deviation [SD]) | 5903 (1.5, SD 1.8) | 2446 (0.6, SD 0.8) | 1654 (0.4, SD 0.6) |
Effect of patient characteristic on total transitions | Â | Â | Â |
Gender (vs male) | Â | Â | Â |
   Female gender associated with fewer transitions |
|
|
|
Age (vs < 65 years) | Â | Â | Â |
   Older age associated with fewer transitions |
| Â | Â |
Location of death (vs home) | Â | Â | Â |
   Death occurring in the inpatient PCP unit or acute hospital setting associated with more transitions |
|
|
|
   Death occurring in a long term care facility associated with fewer transitions |
|
|
|
Year of death (vs 1998) | Â | Â | Â |
   Patients dying between 1999 and 2002 associated with fewer transitions |
|
|
|
Diagnoses (vs other disease, no cancer) | Â | Â | Â |
   A cancer diagnosis associated with more transitions |
| Â | Â |
   A cancer diagnosis associated with fewer transitions |  |
|
|
Primary reason for referral to PCP (vs home consultation) | Â | Â | Â |
   Referral for pain or other symptom control associated with more transitions |  |
|
|
- *
Multivariate regression analyses indicate a significant association at the 0.05 level of association