Wen-Wei Cai, MD, James S. Marks, MD, MPH, Charles H. C. Chen, PhD,
You-Xien Zhuang, MD, Leo Morris, PhD, and Jeffrey R. Harris, MD, MPH
and in terms of the variables listed above. weight, maternal age and education, and tions were few and were randomly distrib-
To determine the factors affecting cesarean birth order in the 3 periods. In 1960 through uted among the respondents.
section, we employed logistic regression 1979, before the economic reform and 1- A considerable difference in the
analysis to estimate how maternal occupa- child family policy, the cesarean rate was cesarean section rate by birth order in the
tion, income, and method of medical pay- not significantly associated with any of the first period disappeared in the second and
ment affected the cesarean section rate in variables. But after the economic reform third periods. Thus, the variable of birth
1988 through 1993. and the 1-child policy in the last 2 periods, order did not affect the upward trend for the
the rate was significantly related to all vari- cesarean rate.
ables except birth order. Table 2 examines the cesarean section
Results The cesarean section rates with and rate in the most recent period, 1988
without complications presented in Table 1 through 1993, by maternal occupation,
Univariate Analysis must be interpreted with caution: (1) the income, residence, and medical payment.
number of women who reported having These variables are not included in Table 1
Since 1960, there has been a substan- complications was small, (2) women with because they are more closely indicative of
tial upward trend in the rate of women complications may have failed to report the women's status in the most recent time
whose most recent delivery was by them owing to failure to recall or lack of period, rather than in the earlier time
cesarean section. Table 1 shows that the awareness of the event. Thus, the cesarean periods.
overall rate of cesarean section deliveries section rates shown in the table could The table shows that during 1988
increased from 4.7% in 1960 through 1979 have been higher for women with complica- through 1993, the rates of cesarean sections
to 9.3% in 1980 through 1987 and to 22.5% tions and lower for women without compli- were highest among women with nonagri-
in 1988 through 1993. The table also shows cations. However, such a bias might not cultural occupations and those with the
the cesarean section rate according to self- seriously affect the following logistic regres- highest family incomes. Suburban resi-
reported pregnancy complication, birth- sion analysis since the cases with complica- dents had a much higher rate of cesarean
TABLE 1-Percentage of Women Who Had a Cesarean Section, by Year of Their Most Recent Live Birth and by Selected
Variables: Minhang District, Shanghai, China (n = 1959)
Last Live Birth
1960 through 1979 1980 through 1987 1988 through 1993
% of Women No. Women % of Women No. Women % of Women No. Women
Who Had in Variable Who Had in Variable Who Had in Variable
Cesarean Category Cesarean Category Cesarean Category
Section Section Section
Total 4.7 621 9.3 884 22.5 454
Self-reported complication
at prenatal checkupsa
Yes 2.9 34 25.5 47 62.5 24
No 5.3 491 8.5 826 20.2 425
P .549 <.001 <.001
Birthweight, kg
1.30-2.49 5.0 20 5.5 31 20.0 15
2.50-2.99 5.0 100 8.6 162 24.3 74
3.00-3.49 5.7 299 7.9 392 18.6 210
3.50-3.99 3.2 158 8.9 259 21.1 128
4.00+ 2.3 44 25.0 40 55.6 27
P .720 .010 .001
Maternal ageb
18-23 3.7 108 5.5 363 12.1 231
24-25 4.6 223 6.1 245 27.1 92
26-27 5.9 169 15.4 149 31.3 67
28-43 4.2 119 18.3 126 43.8 64
P .829 <.001 <.001
Maternal education
Primary or less 4.2 354 9.4 224 8.2 49
High school 4.6 219 8.1 555 20.3 315
Junior college or higher 8.3 48 15.2 52 37.8 90
P .450 .069 <.001
Birth order
1st birth 6.5 292 9.2 813 22.5 423
2nd or higher 3.0 329 9.9 71 22.6 31
P .069 .879 .977
a1 12 women who did not have a prenatal checkup were excluded.
bWomen of unknown maternal age were excluded.
778 American Joumal of Public Health May 1998, Vol. 88, No. 5
Shanghai Cesarean Sections
May 1998, Vol. 88, No. 5 American Journal of Public Health 779
Cai et al.
780 American Joumral of Public Health May 1998, Vol. 88, No. 5