From: The development and preliminary validation of a Preference-Based Stroke Index (PBSI)
Steps | Population Source | Age (mean/sd) | Gender (men/women) (%) | Barthel score of 100 (%) | SF-36 PCS (mean/sd) | SF-36 MCS (mean/sd) |
---|---|---|---|---|---|---|
Item generation | Baseline data from cohort study [30] | Â | Â | Â | Â | Â |
 | Stroke subjects (n = 493) | 70/12 | 61/39 | 57% | 42/10 | 51/9 |
 | Control subjects (n = 442) | 65/12 | 33/67 | 93% | 45/11 | 52/9 |
Item selection | Mailed survey | Â | Â | Â | Â | Â |
 | Stroke subjects (n = 91) | 69/11 | 71/29 | 76% | 49/8 | 52/8 |
Pilot test | Mailed survey | Â | Â | Â | Â | Â |
 | Stroke subjects (n = 68) | 72/12 | 53/47 | 65% | 45/12 | 49/11 |
Elicitation of preference weights | Face-to-face interviews | Â | Â | Â | Â | Â |
 | Stroke subjects (n = 32) | 68/11 | 75/25 | Not evaluated | Not evaluated | Not evaluated |
 | Caregivers (n = 28) | 59/20 | 22/78 |  |  |  |
Validation | Baseline and 6 month data from randomized control trial [Mayo et al, unpublished work] | Â | Â | Â | Â | Â |
 | Stroke subjects (n = 91) | 69/15 | 64/36 | 75% | 43/12 | 50/11 |