Table 3.

Health facility utilization among children under 5 years of age with fever, diarrhea, and respiratory illness: odds ratios from bivariate and multivariable logistic regression models

	Fever				Diarrhea				Respiratory illness
	Unadjusted odds ratio (95% CI)	P value	Adjusted odds ratio (95% CI)	P value	Unadjusted odds ratio (95% CI)	P value	Adjusted odds ratio (95% CI)	P value	Unadjusted odds ratio (95% CI)	P value	Adjusted odds ratio (95% CI)	P value
Age (12 vs. 6 months)	1.39 (0.86, 2.26)	0.44	1.02 (0.96, 1.09)	0.51	1.52 (0.85, 2.72)	0.37	1.09 (0.99, 1.21)	0.085	0.41 (0.22, 0.77)	0.002	0.29 (0.12, 0.71)	< 0.001
Male vs. female child	1.00 (0.73, 1.37)	0.99	0.96 (0.67, 1.38)	0.84	0.88 (0.60, 1.28)	0.50	0.85 (0.56, 1.29)	0.46	1.21 (0.89, 1.64)	0.23	1.41 (0.94, 2.11)	0.094
Respondent age (per 10 years)	0.78 (0.65, 0.95)	0.011	0.92 (0.73, 1.15)	0.45	0.77 (0.62, 0.97)	0.025	0.84 (0.63, 1.12)	0.23	0.84 (0.65, 1.09)	0.19	0.95 (0.69, 1.31)	0.76
Education (5 vs. 0 years)	2.08 (1.62, 2.67)	< 0.001	1.50 (0.97, 2.32)	0.006	2.14 (1.48, 3.08)	< 0.001	1.59 (1.00, 2.51)	0.048	1.88 (1.33, 2.66)	< 0.001	1.46 (0.86, 2.48)	0.16
Marital status		0.79		0.83		0.13		0.051		0.76		0.73
Married/common law (ref)	1		1		1		1		1		1
Divorced or separated	1.29 (0.47, 3.52)		1.04 (0.35, 3.09)		4.09 (0.67, 25.08)		4.63 (0.99, 21.59)		0.64 (0.07, 6.19)		0.38 (0.06, 2.31)
Single	0.95 (0.67, 1.34)		0.96 (0.65, 1.41)		0.88 (0.57, 1.36)		0.89 (0.54, 1.46)		0.80 (0.52, 1.24)		1.05 (0.61, 1.79)
Widowed	0.71 (0.31, 1.64)		0.66 (0.27, 1.61)		0.29 (0.07, 1.17)		0.23 (0.05, 1.00)		1.15 (0.27, 4.89)		1.43 (0.19, 11.09)
Understands Portuguese	2.01 (1.48, 2.72)	< 0.001	1.34 (0.91, 1.98)	0.14	1.63 (1.08, 2.47)	0.02	0.88 (0.54, 1.46)	0.63	1.52 (1.05, 2.21)	0.028	0.64 (0.36, 1.14)	0.13
Household size (per two members)	0.82 (0.69, 0.97)	0.018	0.81 (0.67, 0.98)	0.031	0.85 (0.68, 1.07)	0.18	0.84 (0.63, 1.12)	0.24	1.03 (0.81, 1.30)	0.81	1.09 (0.81, 1.45)	0.57
Transport to health facility		0.44		0.50		0.12		0.29		0.91		0.69
Bike/motorbike/car (ref)	1		1		1		1		1		1
On foot	1.44 (0.81, 1.61)		1.13 (0.79, 1.62)		1.41 (0.91, 2.19)		1.30 (0.80, 2.12)		1.02 (0.65, 1.61)		0.90 (0.53, 1.52)
Travel time to health facility (per 1 hour)	0.80 (0.64, 1.00)	0.047	0.82 (0.57, 1.17)	0.27	0.75 (0.66, 0.86)	< 0.001	0.81 (0.69, 0.94)	0.006	0.62 (0.43, 0.88)	0.008	0.64 (0.30, 1.38)	0.26
Rural vs. urban	0.60 (0.33, 1.11)	0.11	0.87 (0.47, 1.60)	0.65	0.66 (0.32, 1.36)	0.26	0.69 (0.30, 1.58)	0.38	0.82 (0.53, 1.27)	0.37	0.98 (0.58, 1.65)	0.94
District		0.87		0.57				0.69		0.027		0.19
Alto Molócuè (ref)	1		1		1	0.49	1		1		1
Morrumbala	0.90 (0.54, 1.50)		1.44 (0.80, 2.60)		0.60 (0.29, 1.24)		0.98 (0.48, 1.99)		0.51 (0.28, 0.94)		0.66 (0.32, 1.35)
Namacurra	1.14 (0.67, 1.94)		1.33 (0.76, 2.34)		0.79 (0.37, 1.71)		1.24 (0.55, 2.80)		0.66 (0.37, 1.15)		0.67 (0.32, 1.38)
Others	1.03 (0.66, 1.59)		1.34 (0.82, 2.16)		0.97 (0.55, 1.70)		1.35 (0.79, 2.30)		0.54 (0.34, 0.86)		0.52 (0.28, 0.95)
House has radio	1.60 (1.19, 2.15)	0.002	1.42 (0.98, 2.05)	0.060	1.64 (1.11, 2.41)	0.013	1.44 (0.99, 2.10)	0.058	2.10 (1.40, 3.13)	< 0.001	1.78 (1.17, 2.72)	0.007
Monthly income < 1,000 MZN	0.79 (0.53, 1.19)	0.26	0.86 (0.58, 1.27)	0.45	0.45 (0.25, 0.82)	0.008	0.50 (0.30, 0.85)	0.009	0.45 (0.27, 0.74)	0.002	0.75 (0.45, 1.25)	0.27
Delivered last child in a HCF	3.17 (2.16, 4.67)	< 0.001	2.69 (1.64, 4.40)	< 0.001	2.50 (1.59, 3.92)	< 0.001	1.93 (1.10, 3.37)	0.021	4.16 (2.44, 7.09)	< 0.001	4.43 (2.18, 9.01)	< 0.001
Decision-making: child health care				0.55				0.80		0.003		0.16
Men (ref)	1	0.65	1		1	0.29	1		1		1
Women	1.01 (0.60, 1.70)		0.91 (0.52, 1.62)		1.32 (0.71, 2.46)		1.08 (0.58, 1.99)		2.13 (1.08, 4.21)		2.01 (0.96, 4.21)
Both	1.17 (0.72, 1.92)		0.76 (0.41, 1.40)		1.54 (0.89, 2.65)		0.90 (0.43, 1.86)		2.96 (1.58, 5.58)		1.97 (0.89, 4.33)

CI = confidence interval; HCF = health-care facility.

1. There were 984 children included in the fever model. Missing data were multiply imputed for respondent age (6%), education (< 1%), Portuguese understanding (< 1%), transport (2%), travel time (28%), radio ownership (< 1%), monthly income (23%), delivery in health facility (36%), and decision-making (1%). Modeling without imputation would have resulted in 63% (620/984) casewise deletion.

2. There were 512 children included in the diarrhea model. Missing data were multiply imputed for respondent age (5%), Portuguese understanding (< 1%), transport (2%), travel time (31%), radio ownership (< 1%), monthly income (19%), delivery in health facility (28%), and decision-making (2%). Modeling without imputation would have resulted in 55% (283/512) casewise deletion.

3. There were 543 children included in respiratory illness model. Missing data were multiply imputed for respondent age (6%), Portuguese understanding (< 1%), transport (1%), travel time (31%), radio ownership (< 1%), monthly income (23%), delivery in health facility (29%), and decision-making (1%). Modeling without imputation would have resulted in 58% (325/556) casewise deletion.

a. In the adjusted model, missing values of covariates were accounted for using multiple imputation.

b. Because there was evidence (P < 0.10) that the relationship with log-odds of health facility utilization for fever is nonlinear, respondent education is fit using restricted cubic splines.

c. Because there was evidence (P < 0.10) that the relationship with log-odds of health facility utilization for respiratory illness is nonlinear, child age is fit using restricted cubic splines.