Understanding the association of mass media with the timing of antenatal care in Ethiopia: an impression from the 2016 Ethiopia demographic and health survey

Figures & data

Table 1. Socio-demographic characteristics of mothers in the study in Ethiopia, EDHS 2016.

Variables	Weighted frequency (%)	Variables	Weighted frequency (%)
Age in 5 y group		Region
15–19	250.70(5.29)	Tigray	480.10(10.13)
20–24	974.25(20.55)	Afar	36.43(0.77)
25–29	1450.64(30.60)	Amhara	1091.32(23.20)
30–34	1037.63(21.89)	Oromia	1600.97(33.78)
35–39	655.10(13.82)	Somali	171.66(2.48)
40–44	272.02(5.74)	Benishangul	55.72(1.18)
45–49	99.73(2.10)	SNNPR	1109.82(23.41)
Cesarean section		Gambela	15.10(0.32)
No	4572.17(96.46)	Harari	13.30(0.28)
Yes	167.88(3.54)	Addis Ababa	190.56(4.02)
		Dire Dawa	29.07(0.61)
Educational level		Sex of child
No education	2555.04(53.90)	Female	2460.32(51.90)
Primary	1571.43(33.15)	Male	2279.73(48.10)
Secondary	386.69(8.16)
Higher	226.98(4.79)
Number of children		Wealth status
1	1111.48(23.45)	Poor	1718.22(36.25)
2–4	2083.30(43.95)	Middle	983.52(20.75)
5–8	1323.30(27.92)	Rich	204(43.00)
9–14	221.97(4.68)
Frequency of reading		Frequency of listening radio
Not at all	4264.10(89.96)	Not at all	3,139.08(66.22)
<1 week	352.76(7.44)	<1 week	785.96(16.58)
Once a week	123.19(2.60)	Once a week	815.02(17.19)
Frequency of watching TV		Use internet
Not at all	3540.74(74.70)	Never	4618.60(97.44)
<1 week	544.23(11.48)	Yes, last month	104.69(2.21)
Once a week	655.09(13.82)	Yes before last 12 month	16.76(0.35)
Place of residence		Previous Place of delivery
Urban	873.83(18.44)	Home	2580.94(54.45)
Rural	3866.22(81.56)	Health facility	2159.12(45.55)
Sex of the household leader		*Timing of 1st ANC
Male	4049.96(85.44)	Early	1549.74(32.69)
Female	690.11(14.56)	Late	2694.95(56.85)
		Very late	495.37(10.45)
Frequency of using internet
Not at all	4641.83(97.93)	Working status
≤ once a week	56.71(1.20)	Not working	2491.85(52.57)
Almost every week	41.51(0.88)	Currently working	2248.20(47.43)

Table 2. The relationship between early ANC initiation and mass media.

Very late initiation Antenatal care					95% CI
Variables	Coef.	Std. Err.	Z	p-value	Lower	Upper
Frequency of watching TV
Not at all	1.00	1.00	1.00	1.00	1.00	1.00
Less than once a week	0.41	0.10	3.98	.00	0.21	0.61
At least once a week	0.64	0.09	6.91	.00	0.46	0.82
Frequency of using the internet
Not at all	1.00	1.00	1.00	1.00	1.00	1.00
Less than or equal once a week	0.46	0.19	2.37	.02	0.08	0.84
Almost everyday	1.46	0.33	4.49	.00	0.82	2.10
Own mobile phone
No	1.00	1.00	1.00	1.00	1.00	1.00
Yes	0.59	0.08	7.83	.00	0.44	0.74
Model summary
Model	observation	LL(null)	LL(final)	Df	AIC	BIC
	4685	−4307.587	−4124.708	12	8273.416	8350.841
chi2(8) = 9.98 (p-value = .2666) – An insignificant test statistic indicates that the final model does not violate the proportional odds or parallel lines assumption

NB: application of autofit enabled us to have relatively similar coefficients per panel, which eased the interpretation.

Table 3. The association of mass media and socio-demographic factors on early ANC, EDHS 2016.

Very late initiation of antenatal care					95% CI
Variables	Coef.	Std. Err.	z	p-value	Lower	Upper
Birth order
1st	1.00	1.00	1.00	1.00	1.00	1.00
2–3	−0.62	0.20	−3.18	.00	−1.00	−0.24
≥4	−0.64	0.17	−3.73	.00	−0.97	−0.30
Place of residence
Urban	1.00	1.00	1.00	1.00	1.00	1.00
Rural	−0.02	0.19	−0.11	.91	−0.40	0.36
Place of delivery
Home	1.00	1.00	1.00	1.00	1.00	1.00
Health facility	0.70	0.13	5.48	.00	0.45	0.94
Religion
Orthodox	1.00	1.00	1.00	1.00	1.00	1.00
Protestant	0.02	0.14	0.14	.89	−0.26	0.30
Muslims	0.13	0.07	1.80	.07	−0.01	0.26
Other	−0.35	0.35	−1.01	.31	−1.03	0.33
Husband’s desire for children
Want the same	1.00	1.00	1.00	1.00	1.00	1.00
Wants more	−0.11	0.08	−1.36	.17	−0.26	0.05
Want fewer	−0.20	0.07	−2.83	.01	−0.35	−0.06
Total number of ever-born children
1 child	1.00	1.00	1.00	1.00	1.00	1.00
2–4 children	0.54	0.17	3.11	.00	0.20	0.88
5–8 children	0.38	0.16	2.35	.02	0.06	0.70
Maternal education
No education	1.00	1.00	1.00	1.00	1.00	1.00
Primary	0.17	0.08	2.25	.02	0.02	0.33
Secondary	0.35	0.13	2.79	.01	0.10	0.60
Higher	0.39	0.17	2.28	.02	0.05	0.73
Wealth index
Poor	1.00	1.00	1.00	1.00	1.00	1.00
Middle	0.17	0.09	1.90	.06	−0.01	0.35
Rich	0.52	0.15	3.53	.00	0.23	0.80
Model summary
Model	Observation	LL(null)	LL(model)	Df	AIC	BIC
	4320	−3979.362	−3747.679	31	7557.357	7754.859
Wald test of parallel lines assumption for the final model at p-value <.05 X^2 = 19.83 (p-value = .53) i.e. no violation of assumption by the final model

NB: During the parallel line testing at p < .05: the place of residence, place of delivery, religion, and wealth index violated the assumption. One can understand by just observing the variation of coefficients between very late and late ANC, which is one of the advantages of using autofit over other models. (1) Place of residence (p value = .01926); (2) Place of delivery (p value = .00021); (3) Religion (p value = .02000); (4) Wealth index (p value = .00255).

Figure 1. The distribution of timely initiation of antenatal care in Ethiopia: data from the EDHS 2016.

Data availability statement

The survey dataset used in this analysis is the third party data from the demographic and health survey website (www.dhsprogram.com) and permission to access the data is granted only for registered DHS data user.