Advancing Sexual and Reproductive Health and Rights
 
International Perspectives on Sexual and Reproductive Health
Volume 38, Number 2, June 2012
DIGEST

Prevalence of Cesarean Delivery Rising Sharply Among Poor, Rural Women

The cesarean section rate rose from 3% to 39% in China from 1988 to 2008, and it was especially high among urban, wealthy and highly educated women, according to analyses of data from four cross-sectional surveys.1 However, the prevalence of cesarean delivery increased fastest among rural, poor and less educated women. For example, while the rate and relative risk of cesarean delivery more than tripled among urban women during the 20-year span, they increased by a factor of 15 among women living in rural areas.

Prior studies have found that China’s rate of nonmedically indicated cesareans has risen dramatically and is the highest in Asia. To explore factors associated with this increase, investigators analyzed data on 34,482 live births obtained from National Health Service surveys conducted in 1993, 1998, 2003 and 2008. The surveys used a four-stage sampling design that stratified regions by socioeconomic status (rural areas, for example, were divided into four subregions) to obtain a random sample of households throughout the country. Using a structured questionnaire, trained interviewers collected information on characteristics of all pregnancies and deliveries (e.g., number of antenatal visits, place and mode of delivery) that women aged 15 or older had had in the past 2–6 years (the interval varied across surveys). Data were weighted to ensure that findings were nationally representative. The investigators used modified Poisson regression analyses to calculate crude and adjusted relative risks for cesarean deliveries over time, controlling for household income (categorized into quartiles), access to health insurance, mother’s educational attainment, parity, maternal age and number of antenatal visits. Separate analyses were conducted for urban and rural areas.

Nationally, the cesarean section rate spiked from 3% in 1988 to 39% in 2008. The rate increased by more than 50 percentage points among urban women (from 10% to 64%), and by 11–36 percentage points among women in the four rural regions (e.g., from 0% to 11% in the poorest areas and from 3% to 39% in the wealthiest).

During the study period as a whole, rates of cesarean deliveries were elevated among women who were wealthy, were insured, had at least a college education and were having their first birth. For example, among urban women, cesarean deliveries accounted for 50% of births to those in the wealthiest quartile, compared with 16% of births to women in the poorest quartile; 34% of births to women with insurance, compared with 26% of those to uninsured women; and 48% of births to college-educated women, compared with 5% of those to women with no education.

Similar patterns occurred among rural women, though prevalence levels were lower. For example, cesarean deliveries accounted for 23% of births to women in the wealthiest quartile, but only 2% of those to women in the poorest; 19% of those to women with health insurance, but just 5% of those to uninsured women; and 30% of births to college-educated women, compared with only 3% of those to women who had no education. In addition, 16% of deliveries in the wealthiest rural region, but just 2% of those in the poorest, were by cesarean section.

Between the 1993 and 2008 surveys, urban women’s risk of having a cesarean more than tripled (crude relative risk, 3.6). The increase in risk was only slightly lower in multivariate analyses that adjusted for women’s characteristics (3.1), which suggests that in urban areas, population-level changes in education, income and health insurance explain only a small portion of the increase in cesarean deliveries. In rural regions, the unadjusted relative risk of cesarean delivery in the years preceding the 2008 survey was more than 15 times that in the years before the 1993 survey (15.5); in this case, the relative risk was notably lower when such factors as household income, health insurance and education were taken into account (7.2), suggesting that these characteristics partly explain the increased use of cesarean.

However, trend analyses examining interactions between women’s characteristics and survey year found that the likelihood of cesarean deliveries rose fastest among uninsured women, both in urban areas (8.3) and in rural ones (28.7), a finding that stands at odds with the hypothesis that the increased use of cesarean deliveries has been driven by providers seeking higher insurance reimbursements. Increases in rural areas were also especially high among the poorest women (12.1) and those who were uneducated (16.6).

The investigators note several study limitations. Births may have been underreported, especially those that had not been approved by the family planning system or that occurred among rural migrants who were temporarily away from home. Furthermore, the investigators’ definition of health insurance may not have captured the variation in coverage among insurance schemes; therefore, the analyses may have underestimated the strength of the association between health insurance and cesarean deliveries.

Nevertheless, the researchers conclude that a woman’s likelihood of having a cesarean delivery depends more on whether she lives in a wealthy or urban region, where appropriately trained providers and equipped hospitals are more accessible, than on her individual socioeconomic characteristics. “Hence,” they write, “supply side factors may be a more important determinant of caesarean section than [are] ability to pay or educational level.” In addition, the increase in the cesarean rate may reflect a “societal consensus about the safety and benefits of cesarean.” However, they note that information on the procedure’s safety—actual or otherwise—is scant in China and “urgently needed.”—A. Kott

REFERENCE

1. Feng XL et al., Factors influencing rising caesarean section rates in China between 1988 and 2008, Bulletin of the World Health Organization, 2012, 90(1): 30–39.