Skip to main content

Table 4 Predictors for being in the best tertile of PROMIS Fatigue T-scores

From: Subcutaneous immunoglobulin in primary immunodeficiency – impact of training and infusion characteristics on patient-reported outcomes

Predictor

Category

Logistic regression

Linear regression

OR

95% CI

p-value

Coefficient

95% CI

p-value

Confidence after traininga

1–5

1

 

0.01

0

 

0.02

6–7

1.95

1.16, 3.28

−0.24

−4.5, −0.4

Actual infusion time

≥2 h

1

 

0.01

0

 

0.06

< 2 h

1.80

1.14, 2.82

−1.8

−3.7, −0.1

  1. Fatigue – multivariate logistic regression and linear regression models calculated predictors for being in the best tertile of PROMIS Fatigue T-scores. PROMIS Fatigue scores were transformed to a 1–100 scale (0 = least fatigue, 100 = most fatigue), where respondents were grouped in T2 + T3 (intermediate/worst) if they had a fatigue score > 53 and in T1 (best) if they had a fatigue score ≤ 53
  2. CI confidence interval, OR odds ratio, PROMIS Patient-Reported Outcome Management Information System
  3. aPredictor on an anchored numeric scale from 1 to 7 (1 = not very confident and 7 = very confident). The logistic regression provides an OR which predicts the likelihood of that category falling into T1, where the higher number corresponds to better odds. The least squares regression considers scores on a continuous scale using the original 0–100 scale, where a lower coefficient implies a better fatigue score for that category. The least squares  model had an R2 = 2.5%, suggesting that factors examined were not strongly associated with a respondent’s PROMIS Fatigue score