Skip to main content
. 2024 Mar 6;19(3):e0295814. doi: 10.1371/journal.pone.0295814

Table 1. Bayesian multivariable logistic regression models of baseline evaluation associated with week 4 remission and additional therapy/colectomy for patients initially treated with IV steroids.

Remission, all patients Remission by Initial Treatment Additional Therapy/ Colectomy
Odds Ratio (95% CI) Total (N = 419) # 5-ASA (N = 135) Oral CS (N = 142) IV CS (N = 142) IV CS only (N = 142)
Model sample size (% of total N) n = 355 (85%) n = 132 (98%) n = 123 (87%) n = 142 (100%) n = 120 (85%)
Number of events (% of model n) 179 (50%) 73 (55%) 70 (57%) 57 (40%) 32 (27%)
Total Mayo score Mayo < 10: - Mayo < 10: Mayo < 11: Mayo ≥ 11:
1.85 (1.11, 3.12) 3.43 (1.25, 9.81) 2.35 (1.18, 4.80) 5.83 (1.99, 18.91)
Albumin per 1 g/dL 1.38 (0.99, 1.92) 2.11 (1.23, 3.73) - - 0.25 (0.09, 0.58)
Proctosigmoiditis 5.37 (1.93, 18.97) 3.15 (1.11, 10.66) - - -
Rectal biopsy eosinophil peak count /hpf Count > 32: - Count > 32: - Count ≤ 32:
1.74 (1.11, 2.74) 2.86 (1.29, 6.70) 7.63 (2.50, 27.20)
Relative rectal sparing 4.71 (1.92, 13.43) - 8.55 (1.85, 67.12) - -
Rectal biopsy surface villiform changes - - - - 3.29 (1.14, 9.96)
Model evaluation
AUC 0.69 (0.67, 0.70) 0.66 (0.64, 0.66) 0.70 (0.68, 0.71) 0.60 (0.60, 0.60) 0.86 (0.84, 0.87)
CV-AUC 0.67 (0.59, 0.72) 0.65 (0.59, 0.68) 0.69 (0.58, 0.73) 0.59 (0.54, 0.60) 0.83 (0.72, 0.91)
Sensitivity 0.62 (0.41, 0.76) 0.73 (0.38, 0.95) 0.77 (0.73, 0.90) 0.67 (0.00, 1.00) 0.58 (0.41, 0.69)
Specificity 0.66 (0.48, 0.85) 0.46 (0.20, 0.78) 0.54 (0.26, 0.64) 0.51 (0.00, 1.00) 0.90 (0.83, 0.97)
Positive predictive value 0.66 (0.60, 0.75) 0.63 (0.58, 0.70) 0.70 (0.62, 0.73) 0.46 (0.00, 0.49) 0.69 (0.58, 0.80)
Negative predictive value 0.63 (0.59, 0.67) 0.59 (0.49, 0.74) 0.65 (0.64, 0.67) 0.67 (0.00, 0.71) 0.86 (0.81, 0.89)

# Total N = number evaluable at week 4 and with no protocol violations. AUC = area under the curve. CV-AUC = 10-fold cross validation AUC