Effect of alcohol consumption on breast cancer: probabilistic bias analysis for adjustment of exposure misclassification bias and confounders

Table 1 The bias parameters for Triangular, Beta and Logit-logistic distributions in control and case groups

	Bias parameters (95% CI)		Group	Triangular distribution: min; max; mode	Beta distribution: alpha; beta	Logit-logistic distribution: location; scale
Type of misclassification	Differential	Sensitivity 65% (41, 89)	Case group	0.41; 0.89; 0.65	9.21; 4.96	0.65; 0.0675
		Specificity 93% (80, 100)	Case group	0.80; 1; 0.93	46.74; 3.51	0.93; 0.0196
		Sensitivity 54% (42, 65)	Control group	0.42; 0.65; 0.54	42.05; 35.82	0.54; 0.0309
		Specificity 90% (85, 100)	Control group	0.85; 1; 0.90	30.22; 3.36	0.90; 0.0281
	Non-differential	Sensitivity 60% (49, 77)	Both groups	0.49; 0.77; 0.60	18.54; 12.36	0.60; 0.0478
	Non-differential	Specificity 92% (77, 100)	Both groups	0.77; 1; 0.92	39.87; 3.47	0.92; 0.0224

ISSN: 1471-2288