Recruiting migrant workers in Australia for Public Health surveys: how sampling strategy make a difference in estimates of workplace hazards

Table 2 Odd ratios for psychosocial (S1 and S2) and carcinogen exposure (S2 only) by sampling sources

Sample type	Perceived exposure to 3 or more psychosocial adversities				Exposed to at least 1 of 10 workplace carcinogens^a
Sample type	OR (95% CI)	p	aOR^b (95% CI)	p	OR (95% CI)	p	aOR^b (95% CI)	p
Non probability (Recall)	1.00		1.00		1.00		1.00
Probability (EWP based random sampling)	1.31 (0.99, 1.73)	0.06	1.3 (0.92, 1.85)	0.14	0.59 (0.47, 0.74)	< 0.001	0.67 (0.5, 0.89)	0.006
Study phone
Landline S1 (Sample broker)	1.00		1.00
Landline S2 (EWP based random sampling)	1.29 (0.94, 1.76)	0.12	1.5 (0.8, 2.82)	0.20	1.00		1.00
Mobile S2 (Sample broker)	1.04 (0.69, 1.58)	0.84	1.26 (0.62, 2.54)	0.52	1.77 (1.4, 2.22)	< 0.001	1.54 (1.16, 2.05)	0.003
Sample source
Original surname sample (landline)	1.00		1.00		1.00		1.00
Suburb based sample (landline)	1.22 (0.86, 1.75)	0.27	1.41 (0.95, 2.09)	0.09	0.78 (0.6, 1.01)	0.06	1.41 (0.95, 2.09)	0.09
Sample broker (mobile)^c	0.81 (0.6, 1.11)	0.20	0.86 (0.57, 1.28)	0.45	1.63 (1.29, 2.07)	< 0.001	1.45 (1.09, 1.93)	0.01
Recall (both)	0.72 (0.25, 2.08)	0.54	0.74 (0.25, 2.17)	0.59	1.14 (0.71, 1.83)	0.58	1.11 (0.59, 2.08)	0.74

^aExposure to carcinogens reported only from Survey 2
^bModels were adjusted for age, sex, area of residence, education, country of birth, weekly hours worked, whether employed full or part-time and occupation
^cThe same sample broker was used to provide sample for both S1 and S2 but only provided mobile numbers for S2

ISSN: 1756-0500