Table 3. Acquisition of New Functional Limitations Before and After Sepsis Among Survivors, by Functional Class at Baseline.

Results of latent growth curve regression with individual-level fixed effects, controlling for all time-invariant characteristics of the patient. The within-patient R² were 0.25 for the no limitation group, 0.37 for those with mild/moderate baseline limitations, and 0.45 for those with severe baseline limitations. Confidence intervals are in parentheses. The absence of association would be indicated by the acquisition of 0 new functional limitations.

	Functional Class at Baseline
	No Limits	Mild/Moderate Limits	Severe Limits
	n = 269	n = 195	n = 159
Before Sepsis (per year)	−0.020 (−0.046,0.086)	0.11 (0.01,0.21)	0.84 (0.73,0.92)
p-value	p = 0.545	p = 0.027	p < 0.001
Effect of Sepsis	1.57 (0.99,2.15)	1.50 (0.87,2.12)	0.04 (−0.74,0.81)
p-value	p < 0.001	p < 0.001	p = 0.927
After Sepsis (per year)	0.19 (−0.03,0.41)	0.51 (0.24,0.77)	0.16 (−0.19,0.50)
p-value	p = 0.093	p < 0.001	p = 0.369

Interpretive Example: Patients with mild/moderate limitations at baseline were acquiring 0.11 new limitations per year prior to severe sepsis. They acquired 1.50 new limitations at the time of their hospitalization for severe sepsis. In addition, each year after sepsis, they acquired 0.51 new limitations per year, a statistically significant increase relative to their rate before sepsis. (See also Figure 3.)