Skip to main content
Fig. 4 | BMC Infectious Diseases

Fig. 4

From: Filling gaps in notification data: a model-based approach applied to travel related campylobacteriosis cases in New Zealand

Fig. 4

Comparison of Bayesian and Multiple Imputation models regarding the mean and 95 % Credibility (Confidence) Intervals of regression coefficients for 10 % (Fig. 4a), 50 % (Fig. 4b), 65 % (Fig. 4c) and 80 % (Fig. 4d) missing data category as compared to the complete data on overseas travel status of campylobacteriosis cases (n = 44,285). Notes: (1) * Deprivation index (scale 1–10, 1 = least deprived and 10 = most deprived District Health Board; **proportion of DHB population under urban influence;*** Short term international travel per 100 residents of a DHB; ****a binary indicator variable to identify cases that were reported before or after 2006 poultry intervention period. (2) Complete cases: regression coefficients estimated from campylobacteriosis notification data with complete information on overseas travel. (3) The error bars indicate the 95 % confidence intervals (in Multiple Imputation models) and 95 % Credibility Intervals (in Bayesian models) of the regression coefficients

Back to article page