in the model at a lower significance level than that of the main factors. Equal variance

was established for all values except for the SIS-16 measure for the entire group.

Therefore t-tests were deemed appropriate and equal (pooled) p-values were used except

for the SIS-16 with the Satteerthwaite unequal variance p-value used. After controlling

for age, there were significant differences in SIS measured mobility and participation

between those who succeeded and failed to advance to the next ambulation classification

(p < 0.05). SIS ADL/IADL approached significance at 0.0524. When controlling for age

and stratifying by baseline gait speed, those with initial gait velocity of < 0.4 m/s

demonstrated significant differences in SIS ADL/IADL, mobility, social participation,

and SIS-16 measures between those who succeeded and failed (p < 0.05) (Table 3-5). In

contrast, when age was controlled for and initial gait speed was between 0.4 m/s and 0.8

m/s, only participation scores were significantly different between those who succeeded

and those who failed to progress to the next ambulation classification.

Table 3-3. Proportions of success/fail for transitioning to next ambulation classification
 between baseline and 3 months
 Initial Gait Velocity in m/s All
 Success Fail
 Total n (%) < 0.4 (Household) 12 (63%) 7(37%)
 0.4-0.8 (Limited Community) 17 (38%) 28 (62%)
 _Total 29 (45%) 35 (55%)

Table 3-4. Three month outcomes for entire sample
 Success Failure p value p value^
 Sample Size 29 35
 SIS ADL/IADL 77.8 (14.1) 69.6 (19.0) 0.0589 N/A
 SIS Mobility 77.6 (2.2) 65.5 (16.1) 0.0015 0.0359A
 SIS Participation 68.2 (8.6) 52.9 (20.6) 0.0031 N/A
 SIS 16 81.3 (11.2) 71.6(16.5) 0.0076 0.0844A
 ^p-value after adjusting for age when age is significantly different at the < 0.100 level