Objective
The purpose of this study was to describe the trends and determinants of cesarean section (CS) delivery rates in rural China.
Study Design
Data on rural primiparous women aged 15-49 years (n = 10,754) were obtained from 3 nationwide representative surveys in 1993, 1998, and 2003. The CS rate per 100 births and odds ratios by women’s background characteristics were calculated with the use of logistic regression.
Results
The CS rate increased from 1% in 1991 to 17% in 2002. After age adjustment, CS was most common among more educated women, who lived in Eastern China, who had high household income and health insurance, who used antenatal care, and who gave birth at a high-level hospital.
Conclusion
This development over the 10-year period may indicate very high CS rates in the near future; the epidemic of the use of CS that has been observed in urban China is likely to occur also in rural China. Further studies on the reasons and consequences of such excessive use of operative delivery are needed.
In recent years in many western countries, cesarean section (CS) delivery rates have risen beyond the recommended maximum level of 15%. For example, in the United States, the CS rate increased from approximately 21% in 1996 to 28% in 2003 2 and in Canada from 18% in 1993 to 26% from 2005-2006, which was an average of 21% in developed countries. In some developing countries, even higher rates have been found; for example, in Brazil the national average of 37%, and in Mexico the national average was 39%. On average, the nationwide rates in less and least developed countries have varied between 0.4% and 41%.
A large population-based study from urban China showed high and increasing CS rates from 18% from 1990-1992 to 40% from 1998-2002. Smaller local studies from urban cities have found similar increasing trends. No representative studies of CS rates in rural China have been published.
Among Chinese populations, CS rates are affected by many nonmedical factors (such as cultural issues, women’s personal and social characteristics and health insurance coverage). A potential reason is the financing mechanisms in the Chinese health care system that encourages health care providers to overuse drugs and other technology to generate the much-needed revenue. Much less is known of the reasons behind the rapid rise of CS rates. It has been proposed that high CS rates are associated with increasing rates of maternal requests for CSs and with concentrating births in large urban hospitals.
The aim of this study was to describe the trends and determinants of CS rates among primiparous women in rural China from 1991-2002 by the use representative national surveys.
Materials and Methods
The data that were used in this study were drawn from 3 National Household Health Interview Surveys that were conducted in 1993, 1998, and 2003 by the Centre for Health Statistics and Information (Ministry of Health). Four-staged (counties, townships, villages, and households) stratified cluster sampling was used to select the households. In the first stage, counties were categorized into 5 different strata based on K-Means clustering analysis of 10 indicators of socioeconomic development in the area. With the use of a probability proportional sampling method, 1 rural county was selected randomly within each province. In the second stage, the cluster sampling method was used to select 5 townships from each county. In the third stage, the probability proportional sampling method was used to select 2 villages within each township. In the fourth stage, 20% of households were selected randomly within villages. In rural areas, a total of 38,775 households in 1993, 40,238 households in 1998, and 40,212 households in 2003 were selected to receive surveys. In China, analysis of survey data does not require permission from an ethics committee. Data protection and confidentiality were taken into account by the Centre for Health Statistics and Information where all analyses were made and the permission to use the data was given.
For this study, married women who were having their first birth were chosen. The proportions of first-time mothers of all mothers were 51%, 64%, and 50% in 1993, 1998, and 2003, respectively. The numbers of interviewed women were 5426 in 1993, 2278 in 1998, and 3050 in 2003, which was a total of 10,754. The recall periods for the first birth were 2 years in 1993 (from June 1991 to May 1993) and in 1998 (from January 1996 to December 1997), but 5 years in 2003 (from September 1998 to December 2002). The sampling frame used by the Ministry of Health in China remained the same in all 3 surveys. However, the numbers of women interviewed in each survey varied because of migration of young women from rural to urban areas.
The 3 survey questionnaires had very similar structures and questions. The questionnaire consisted of several sections: general socioeconomic and demographic factors, perceived needs and demands for health care, the use and expenditure of health services for households and individual family members, and the reproductive health of the women 15-49 years old.
Women were asked to select 1 of 6 outcomes for how their pregnancy ended: natural delivery, elective induced delivery, CS, stillbirth, miscarriage, and induced abortion. Based on the review of literature and knowledge on the Chinese health care system, the following variables were selected to explore the determinants (which were classified by the designers of the surveys and are consistent in all 3 surveys): age, education (illiterate, secondary school, college and university), occupation (farmer; other; which included nonfarmers and government officials), household or family income (lowest, middle, highest income group), geographic region (Eastern, Western, Central China), birthplace (home, on road to the hospital, village-level health care facility, township or lower level hospital, maternal health care center, county or higher level hospital), having health insurance (government, labor, and basic medical insurance and cooperative medical scheme: yes, no), and the use of antenatal care during pregnancy (antenatal visit at least once, no visit).
The household income groups were formed by ranking individuals by their income level and then dividing the ranking into 3 categories, each of which contained one-third of the total number of households. The individual income is an estimate calculated by taking the reported gross household income and dividing it by the number of individuals in the household. The income ranges (Yuan) were: less than 1176 (1993), 2964 (1998), and 4398 (2003) for the low income group; 1176-2010 (1993), 2964-4770 (1998), and 4398-7880 (2003) for the middle income group; and 2010 (1993), 4770 (1998), 7880 (2003) and above for the high income group.
CS rates per 100 births were calculated by year (all available years: from 1991-1993 and 1996-2002) and by each survey (1993, 1998, and 2003) to study the trends over time. CS rates per 100 births were calculated by different background characteristics to find out the determinants that were associated with CS rates. The odds ratios of having CS by different background variables were calculated with the use of logistic regression, which was for each woman’s age.
Results
Table 1 shows the background characteristics of the women in the 3 surveys. Because of the sampling and categorization methods, the distribution by occupation, household income, and region were relatively similar in the 3 surveys. The education level increased over time. The mean age of primiparous women increased between the 1993 and 1998 surveys, but not between the 1998 and 2003 surveys (25.2, 27.9 and 26.7 years, respectively).
Characteristic | Survey | P value a | |||
---|---|---|---|---|---|
1993 (n = 5426) | 1998 (n = 2278) | 2003 (n = 3050) | Total (n = 10,754) | ||
Age | |||||
<25 | 59.7 | 43.2 | 27.4 | 47.1 | < .001 |
25-34 | 31.5 | 40.6 | 69.8 | 44.3 | |
35+ | 8.8 | 16.2 | 2.8 | 8.6 | |
Education | |||||
Illiterate/Primary school | 54.6 | 55.9 | 35.8 | 49.6 | < .001 |
Secondary school | 38.6 | 39.2 | 55.1 | 43.4 | |
College/University | 6.8 | 4.9 | 9.1 | 7.0 | |
Occupation | |||||
Farmer | 78.6 | 87.1 | 79.4 | 80.7 | < .001 |
Other | 21.4 | 12.9 | 20.6 | 19.4 | |
Household income | |||||
Lowest | 34.4 | 33.0 | 33.0 | 33.7 | .0098 |
Middle | 33.4 | 33.9 | 36.9 | 34.5 | |
Highest | 32.2 | 33.1 | 30.0 | 31.8 | |
Region | |||||
Western | 36.9 | 44.6 | 34.3 | 37.9 | < .001 |
Central | 25.9 | 21.7 | 25.2 | 24.8 | |
Eastern | 37.2 | 33.6 | 40.5 | 37.4 | |
Birth hospital | |||||
Township or lower level hospital | 2.3 | 17.4 | 28.2 | 12.9 | < .001 |
Maternal and child health center | 16.6 | 5.1 | 12.2 | 12.9 | |
County or higher level hospital | 20.2 | 26.1 | 29.5 | 24.1 | |
Home births and other b | 60.9 | 51.4 | 30.1 | 50.1 | |
Antenatal care | |||||
No | 68.7 | 15.0 | 9.2 | 40.5 | < .001 |
Yes | 31.3 | 85.0 | 90.8 | 59.5 | |
Health insurance | |||||
No | 85.3 | 99.3 | 90.1 | 89.6 | < .001 |
Yes | 14.7 | 0.7 | 9.9 | 10.4 | |
Total | 100 | 100 | 100 | 100 |