The relative impact of barriers to care among military health services personnel: exploring differences using context specific scenarios

Table 4 Imputed logistic regression modelling: Physical health scenarios with and without intentions

Model Variables	Pneumonia		(with intentions)		Back Injury		(with intentions)
Model Variables	aOR	SE	aOR	SE	aOR	SE	aOR	SE
Individual characteristics
Female gender	1.58†	0.37	1.49	0.36	1.44	0.38	1.40	0.38
Age	0.73*	0.10	0.73*	0.11	1.15	0.18	1.19	0.19
Prefer accessing care in English	1.22	0.420	1.03	0.37	1.29	0.55	1.09	0.48
JNCM	0.42**	0.12	0.37**	0.11	0.90	0.28	0.83	0.27
Core trade	0.93	0.22	0.84	0.21	1.11	0.30	1.07	0.30
Stationed in rural location	0.74	0.21	0.71	0.21	1.55	0.47	1.55	0.49
Location related issues	1.41	0.39	1.30	0.37	1.22	0.36	1.14	0.35
Poor self-reported health	0.63	0.33	0.65	0.35	1.57	0.74	1.70	0.84
Accessed care in the past	0.41*	0.11	0.50*	0.14	0.40**	0.12	0.47*	0.15
Past experiences and expectations	2.28*	0.6	2.01*	0.63	1.65	0.53	1.44	0.47
Weak general intentions			3.39***	0.82			3.28***	0.91
Barrier Factors
CAF HCP identity	0.72	0.23	0.66	0.22	1.07	0.36	1.03	0.35
Discomfort accessing care at work	0.76	0.23	0.79	0.25	0.79	0.28	0.80	0.29
Conflict with career goals	1.58	0.43	1.44	0.40	1.82†	0.57	1.67	0.53
Staffing and workload resources	1.93*	0.55	1.62	0.48	2.29*	0.67	1.93†	0.67
Knowledge and abilities	1.23	0.38	1.28	0.41	0.91	0.31	0.93	0.32
Organizational and social support	1.17	0.44	1.22	0.48	0.90	0.34	0.95	0.38
Treatment preferences	1.38	0.37	1.24	0.34	1.48	0.46	1.31	0.41
Concerns about confidentiality	0.95	0.29	1.05	0.33	1.02	0.33	1.10	0.37
Model constant	0.77	0.51	0.57	0.39	0.05***	0.04	0.03***	0.02
Model Descriptors
n	530		530		503		503
Average RVI^a	0.0238		0.0278		0.0266		0.0320
Largest FMI^b	0.0961		0.1044		0.1684		0.1696
Model F Test	F( 18.869979.5) = 3.45		F(19,674,789.5) = 3.34		F(18,690,448.7) = 2.84		F(19,515,145.5) = 4.45
p	< 0.0001		< 0.0001		< 0.0001		< 0.0001

^aM ≥ 100 × FMI, in all models where M (number of imputations) is sufficient
^b The closer the RVI is to zero, the less effect missing data have on the variance estimate, in this case all RVI are low, JNCM- Junior non-commissioned member; ***p < 0.001, **p < 0.01, *p < 0.05, † p ≤ 0.06
Model variables are listed in a column and the adjusted odds ratio (aOR) estimate for not accessing care with standard error (SE) are reported in separate columns for the two physical health vignettes: pneumonia and back injury

ISSN: 1472-6963