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

Table 3 Summary of logistic regression analysis for variables predicting missing indicator (1 = missing overseas travel information, 0 = otherwise) to test the validity of Missing At Random assumption (n = 116721)

Coefficients	Estimate	Std. Error	Pr(>\|z\|)
(Intercept)	−8.757	0.089	<0.001
Urban^a	2.992	0.103	<0.001
DepIndex^b	0.525	0.006	<0.001
Travel Rate^c	0.081	0.001	<0.001
Age (5–19)	0.154	0.027	<0.001
Age (20–59)	0.033	0.023	0.145
Age (60+)	−0.142	0.027	<0.001
Summer	0.014	0.018	0.443
Autumn	−0.002	0.021	0.94
Winter	0.035	0.021	0.085
Male	0.153	0.014	<0.001
Intervention^d	0.345	0.016	<0.001

Keys: ^aProportion of DHB population under urban influence; ^bDeprivation index (scale 0–10, 0 being least deprived and 10 being most deprived DHB; ^cShort term international travel per 100 residents of a DHB; ^dA binary indicator variable to identify pre and post 2006 intervention. Age (<5), Spring, and Female sex are reference categories

ISSN: 1471-2334