TABLE 2. Cesarean Delivery Rate by Year of Survey and Background Characteristics, Weighted Percentages

Characteristics	2002		2007		2012
Characteristics	%	P Value	%	P Value	%	P Value	Percentage Point Difference^a	P Value for Trend^b
*Overall Rate*	*18.2*		*20.1*		*30.3*
Age, y		.001		<.001		<.001		.53
15–19	14.5		20.3		22.2		7.7
20–24	14.6		11.1		25.2		10.6
25–29	15.5		16.9		28.0		12.5
≥ 30	20.5		24.0		32.8		12.3
Place of Residence		.33		.45		.12		.74
Urban	18.6		20.3		30.7		12.1
Rural	17.0		19.2		28.4		11.4
Geographic Region		.008		.18		.06		.37
Central	19.7		20.2		31.2		11.5
North	15.1		19.1		28.2		13.1
South	17.3		22.7		30.9		13.6
Mother's Education		.01		.001		<.001		.75
No education	22.0		24.5		23.5		1.5
Primary	23.1		24.9		40.2		17.1
Secondary	16.6		18.6		29.0		12.4
Higher	19.9		22.0		31.3		11.4
Partner's Education		.13		.44		<.001		.48
No education	23.1		20.7		28.8		5.7
Primary	16.9		21.5		33.0		16.1
Secondary	17.3		19.4		28.1		10.8
Higher	20.3		20.9		34.6		14.3
Employment		.02		<.001		<.001		.02
Unemployed	17.7		19.4		28.5		10.8
Employed	22.6		24.8		40.1		17.5
Place of Delivery		.004		.005		.02		.53
Public sector	16.8		19.0		29.4		12.6
Private sector	20.8		22.0		32.2		11.4
Wealth Index		.18		<.001		<.001		.47
Poorest	17.4		17.0		28.8		11.4
Poorer	16.3		17.1		25.4		9.1
Middle	18.1		19.2		33.3		15.2
Richer	20.0		24.0		31.0		11.0
Richest	21.1		26.5		34.0		12.9
Parity in Last 5 Years		<.001		.02		.01		.01
1	20.1		21.3		30.0		9.9
2	14.8		18.5		29.5		14.7
≥ 3	20.6		18.4		36.6		16.0
Age at First Birth, y		<.001		<.001		<.001		<.001
12–18	16.9		16.8		24.3		7.4
19–24	16.7		17.0		27.0		10.3
25–30	20.3		28.2		34.5		14.2
> 30	43.9		38.5		66.9		23.0
Birth Weight, g		<.001		<.001		<.001		.85
Normal (2500–3999)	16.3		18.1		28.4		12.1
Low/High (< 2500 or ≥ 4000)	24.2		27.1		36.6		12.4
Birth Multiplicity		<.001		<.001		<.001		<.001
Singleton birth	17.3		19.3		29.5		12.2
Multiple birth	60.0		59.3		65.7		5.7

↵a Percentage point difference between 2002 and 2012.
↵b P value for the trend was obtained from logistic regression models by using outcome and explanatory variables as continuous variables and including interaction terms between survey year and each tested explanatory variable for the years 2002 and 2012.