. Author manuscript; available in PMC: 2023 Apr 1.

Published in final edited form as: Clin Gastroenterol Hepatol. 2021 May 29;20(4):766–775.e4. doi: 10.1016/j.cgh.2021.05.049

Table 2.

Linear regression coefficients, 95% confidence intervals, and p-values for the models of EEsAI PRO as outcome in all patients.

Model with eos/hpf as predictor	Coefficient [95% CI]	p-value^b
^a coefficient for 10-cell increase in max eos/hpf in non-dilated patients	2.69 (0.97, 4.40)	0.002
coefficient for 10-cell increase in max eos/hpf in patients dilated > 1 year prior to index endoscopy	0.78 (−1.48, 3.05)	0.494
coefficient for 10-cell increase in max eos/hpf in patients dilated ≤ 1 year prior to index endoscopy	−1.64 (−4.07, 0.79)	0.183
Model with EREFS as predictor	Coefficient [95% CI]	p-value
coefficient for unit increase in EREFS in non-dilated patients	1.88 (−0.14, 3.91)	0.068
coefficient for unit increase in EREFS in patients dilated > 1 year prior to index endoscopy	−2.30 (−5.03, 0.42)	0.097
coefficient for unit increase in EREFS in patients dilated ≤ 1 year prior to index endoscopy	−5.31 (−8.91, −1.71)	0.004

The coefficient represents the change in the predicted EEsAI for 10-cell increase in max eos/hpf in non-dilated patients. For a 10-cell increase in eos/hpf score, the predicted EEsAI PRO increased by 2.69 in non-dilated patients.

P-value is testing whether the slope of the regression line in each dilation group is different from zero.

Abbreviations: CI, confidence interval; eos/hpf, esophageal eosinophilia per high-power field; EEsAI, eosinophilic esophagitis activity index; EREFS, endoscopic reference score.