The number of pregnancies in which delivery is accomplished by a cesarean delivery has been steadily rising over the last decade.1 This increase is due both to a greater frequency of primary cesarean delivery as well as decreased probability of vaginal birth after cesarean delivery. Consequently, obstetricians are more likely to be caring for women who have had one, and in many cases more than one, cesarean delivery.
This change in the clinical characteristics of the obstetric population in the United States has implications for the frequency of obstetric morbidity. Using a large registry database, Silver et al2 documented that having had a prior cesarean delivery increases a woman's risk of obstetric complications. Indeed, the chances of obstetric complications, such as blood transfusion, hysterectomy, and endometritis, were more likely with each additional cesarean delivery that a woman had experienced. Also progressively more likely with each cesarean delivery was the chance of placenta previa, a finding that has been reported by others.2–5
Although the relationship between prior cesarean deliveries and placenta previa is well established, there has been relatively less investigation regarding the maternal and neonatal outcomes that occur once a placenta previa exists in the presence of a prior uterine scar, and whether adverse outcomes become progressively more likely with an increasing number of prior cesarean deliveries. This information is important to know so that women with a placenta previa can be appropriately counseled regarding their outcomes and physicians can be appropriately prepared for their deliveries. Thus, the aim of the present investigation was to estimate the associations between the number of prior cesarean deliveries and pregnancy outcomes among women with a placenta previa.
Between 1999 and 2002, investigators at 19 academic medical centers, belonging to the National Institute of Child Health and Human Development Maternal–Fetal Medicine Units Network, created a registry that included pregnancy outcomes of women who delivered at their institutions. During the first 2 years of the study, data on all cesarean deliveries were concurrently collected, whereas the remaining 2 years had data collection limited to those women with a prior cesarean delivery. Full details of the study design and technique of data collection have been previously described.6
This analysis concerns those women in the registry with a singleton gestation and a placenta previa. In this registry, the presence of a placenta previa was based upon the documentation in the intrapartum medical record of a “placenta previa.” The position of the placenta within the uterus (ie, posterior, anterior, etc.) as well as the type of previa (ie, marginal, partial, complete) was not recorded. Women with an antepartum stillbirth or whose number of prior cesarean deliveries was unknown were excluded. For women who met inclusion criteria, maternal demographic information as well as maternal and neonatal health outcomes were analyzed. Placenta accreta, increta, and percreta were diagnosed on the basis of pathologic findings, although clinical findings were used if hysterectomy was not performed. A composite adverse outcome variable was created for both maternal and perinatal adverse outcomes. The adverse maternal composite included any of the following: transfusion, hysterectomy, operative injury (cystotomy, ureteral injury, or bowel injury), coagulopathy, thromboembolic event, pulmonary edema, or death. The adverse perinatal composite included any of the following: respiratory distress syndrome, necrotizing enterocolitis, intraventricular hemorrhage grades 3 or 4, seizures, or death. All definitions of outcomes were prespecified in a manual of operations and made uniform across all centers.
Data were stratified according to the number of prior cesarean deliveries a woman had experienced. Differences in patient characteristics among the women with different numbers of cesarean delivery were evaluated with the Kruskal-Wallis test for continuous variables and χ2 test for categorical outcomes. The evaluation of whether an increasing number of cesarean deliveries were associated with an increasing risk of adverse outcome was performed using tests of trend. The Cochran-Armitage trend test was used for categorical variables and the Jonckheere-Terpstra trend test was used for continuous variables.7,8 The association of composite maternal and perinatal morbidity with the number of prior cesarean deliveries was also evaluated for the subgroup of women with a placenta previa who had a nonemergent scheduled delivery (ie, a delivery that was not precipitated by maternal bleeding).
Last, the composite measures of maternal and perinatal morbidity were further explored with multivariable analyses. These multivariable analyses were performed to control for possible confounding of the outcomes by differences in patient characteristics. Also, these analyses allowed the determination of whether a threshold number of cesarean deliveries was necessary before the risk of composite adverse outcomes began to increase. In these analyses, the composite outcome served as the dependent variable and the number of prior cesarean deliveries, entered as a multiple dichotomous variable (ie, “one prior cesarean delivery,” “two prior cesarean deliveries,” or “at least 3 prior cesarean deliveries”), served as an independent variable. Other independent variables that were considered for inclusion were the patient characteristics that, in univariable analysis, had been shown to be significantly different between the groups. The multivariable model was constructed by initially including all these variables and then sequentially removing those (other than the number of prior cesarean deliveries) with the lowest significant χ2 value, until only variables that were statistically significant remained. Odds ratios and 95% confidence intervals were calculated for the risk of adverse outcomes for each additional cesarean delivery, with the risk of a woman with no prior cesarean delivery serving as the referent.
For all statistical tests, nominal two-tailed P values are reported, with statistical significance defined as a P<.05. No adjustment was made for multiple comparisons. The SAS 8.2 software (SAS Institute, Cary, NC) was used for analysis. Approval for the study was obtained at the institutional review board of each participating institution.
In this registry of 70,442 cesarean deliveries, 900 women (1.3%) had a placenta previa. Twenty-one women had a multiple gestation, five women had an antepartum stillbirth, and six women did not have information regarding their number of prior cesarean deliveries; correspondingly, 868 pregnancies met inclusion criteria and were available for analysis. Among these women, the number of prior cesarean deliveries was as follows: 488 (56.2%) had none, 252 (29.0%) had one, 76 (8.8%) had two, 39 (4.5%) had three, 9 (1.0%) had four, 3 (0.35%) had five, and 1(0.12%) had nine. Given the infrequency of women with more than 3 prior cesarean deliveries, further analysis was performed by condensing results for women with 3 or more prior cesarean deliveries into a single stratum.
Demographic information for the 868 women, stratified by number of prior cesarean deliveries, is presented in Table 1. Women with greater numbers of prior cesarean deliveries were more likely to be older, of greater body mass index, and never to have been delivered vaginally. Also, classical uterine incisions were more common among women with two prior cesarean deliveries.
Table 2 contains the frequencies of maternal morbidity stratified by number of prior cesarean deliveries. Some outcomes, such as uterine atony, wound infections, thromboembolic events, and maternal deaths were not increasingly likely with greater numbers of cesarean delivery. However, multiple other adverse maternal outcomes occurred progressively more commonly as the number of prior cesarean deliveries increased. Of note, composite maternal morbidity increased to such an extent that women with three or more prior cesarean deliveries and a placenta previa had more than an 80% chance of incurring composite morbidity. Because hysterectomies and transfusions made large contributions to the “composite morbidity” measure, we also assessed the frequency of composite maternal morbidity that did not include these two outcomes. The results were similar, with maternal morbidity increasing with each additional cesarean delivery (1.6% compared with 7.1% compared with 31.6% compared with 38.5%, P<.001, Cochran-Armitage trend test).
The frequency of perinatal morbidity stratified by number of prior cesarean deliveries is presented in Table 3. There were no cases of hypoxic ischemic encephalopathy.
In all groups, the average gestational age at delivery was lower than that reported in pregnancies in the general population.7 Yet, in contrast to maternal morbidity, neither individual adverse perinatal outcomes nor the composite measure of adverse perinatal outcome were more likely as the number of prior cesarean deliveries increased.
When the data were analyzed only for those women who underwent nonemergent scheduled delivery in the pregnancy with the placenta previa, the associations between the number of prior cesarean deliveries and composite morbidity were similar to the overall cohort. As illustrated in Table 4, composite maternal morbidity significantly increased for each additional cesarean delivery that had been performed, and composite neonatal morbidity did not change with increasing number of prior cesarean deliveries. Results for individual maternal and neonatal morbidities also were similar to those for the overall cohort (data not shown).
Last, multivariable analysis was performed for the composite pregnancy outcomes. The odds ratios (ORs) and 95% confidence intervals (CIs) are in comparison with those women with no prior cesarean delivery, and are presented in Table 5. Even one prior cesarean delivery was sufficient to increase the risk of an adverse maternal outcome (a composite of transfusion, hysterectomy, operative injury, coagulopathy, venous thromboembolism, pulmonary edema, or death) from 15% to 23%, which corresponded, in multivariable analysis, to an adjusted odds ratio of 1.9 (95% CI 1.2–2.9). Moreover, the risk of composite maternal morbidity continued to increase significantly with each additional cesarean delivery. The only other independent variable that remained in this model was the number of prior vaginal deliveries: having had at least three prior vaginal deliveries also increased the risk of adverse maternal outcome (OR 2.2, 95% (CI) 1.3–3.9). With regard to composite perinatal morbidity, no association with the number of prior cesarean deliveries was revealed by the multivariable analysis. The only variables that were associated with the composite perinatal outcome were maternal age and body mass index at delivery. Both were inversely related to adverse perinatal outcome (OR 0.96, 95% CI 0.94–0.99; OR 0.96, 95% CI 0.93–0.99, respectively). Results were similar when “center” was added to the multivariable model. Of note, the presence of a prior classical incision was not significantly associated with adverse maternal or perinatal outcome.
As the frequency of cesarean delivery increases, complications and adverse outcomes related to the history of prior cesarean delivery have become increasingly recognized. The recent study of the association between cesarean delivery and blood transfusion by Rouse et al9 revealed that repeat cesarean deliveries are associated with an increased risk of transfusion in women with placenta previa. The analysis by Silver et al2 further revealed the extent to which a prior cesarean delivery is associated with an increased risk of complications such as blood transfusions requiring at least 4 units, operative injuries, and prolonged maternal hospital stays. Indeed, in that analysis, complications were not just increased by a history of prior cesarean delivery, but were progressively increased by each additional cesarean delivery a woman had received.
The analysis by Silver et al also revealed that each additional cesarean delivery is associated with an increased subsequent risk of incurring a placenta previa. This relationship between cesarean delivery and placenta previa has been previously described by multiple investigators.3–5 However, these other studies have used large administrative databases or have had relatively small sample sizes, and have not been able to ascertain the pregnancy outcomes of the patients with placenta previa. Aside from evidence that each prior cesarean delivery, in the setting of placenta previa, increases the risk of placenta accreta, the relationship between maternal and perinatal outcomes and the number of prior cesarean deliveries in women with a placenta previa remains uncertain.
In this study, we have analyzed only women with a placenta previa and determined the extent to which the number of prior cesarean deliveries is associated with adverse maternal and perinatal outcome. Although some complications, such as uterine atony or wound infection, were not found to have any association with the presence or number of prior cesarean deliveries, a relationship between prior cesarean deliveries and many other major complications was found. As noted by other authors, placenta accreta was progressively more common with each additional cesarean delivery2,3; other outcomes that also showed this significant trend included intraoperative procedures (uterine artery or hypogastric ligation), intraoperative complications (coagulopathy, bladder injury, hysterectomy), and postoperative complications (ileus, pulmonary edema). Moreover, there was no evidence that a certain number of cesarean deliveries were required before the risk of maternal complications began to rise, or that risks did not continue to increase beyond a certain number of cesarean deliveries. Multivariable analysis revealed that even one cesarean significantly increased composite maternal morbidity, which continued to rise with each additional cesarean delivery. Given that, in the analysis, we grouped women with at least three prior cesarean deliveries together, we cannot know if risks would continue to rise for each additional cesarean delivery greater than three.
In contrast to the findings for maternal outcomes, the findings for perinatal outcomes were more reassuring. Although one might theorize that perinatal risk would also increase with additional cesarean deliveries due to an increased risk of earlier and heavier bleeding, there was no evidence in our data that any adverse perinatal outcome was progressively associated with cesarean delivery history. This signifies that among women with a placenta previa, a history of prior cesarean delivery does not seem to worsen perinatal outcome. This does not imply that a prior cesarean delivery will not increase perinatal risks. Although adverse outcomes did not increase with each cesarean delivery, preterm delivery (and the corresponding perinatal morbidity) did increase in this population compared with a population without placenta previa,10 and it has been shown that prior cesarean delivery increases the risk of placenta previa.
Providers and their patients should be aware that even one prior cesarean delivery increases the risk of a complicated obstetric course. Also, these complications become not only more frequent, but in some cases, probable for women with a placenta previa and two or more prior cesarean deliveries. For women with a previa and two prior cesarean deliveries, more than one half will have a significant adverse outcome; for those women with a previa and three or more prior cesarean deliveries, fewer than one in five will be free of serious maternal complications. These data are important for the counseling of women with placenta previa and the preparation of physicians for their surgery.
1. Menacker F. Trends in cesarean rates for first births and repeat cesarean rates for low-risk women: United States, 1990–2003. Natl Vital Stat Rep 2005;54:1–8.
2. Silver RM, Landon MB, Rouse DJ, Leveno KJ, Spong CY, Thom EA, et al. Maternal morbidity associated with multiple repeat cesarean deliveries. Obstet Gynecol 2006;107:1226–32.
3. Clark SL, Koonings PP, Phelan JP. Placenta previa/accreta and prior cesarean section. Obstet Gynecol 1985;66:89–92.
4. Gilliam M, Rosenberg D, Davis F. The likelihood of placenta previa with greater number of cesarean deliveries and higher parity. Obstet Gynecol 2002;99:976–80.
5. Getahun D, Oyelese Y, Salihu HM, Ananth CV. Previous cesarean delivery and risks of placenta previa and placental abruption. Obstet Gynecol 2006;107:771–8.
6. Landon MB, Hauth JC, Leveno KJ, Spong CY, Leindecker S, Varner MW, et al. Maternal and perinatal outcomes associated with a trial of labor after prior caesarean delivery. N Engl J Med 2004;351:2581–9.
7. Agresti A. Categorical data analysis. New York (NY): John Wiley and Sons; 1990. p. 100–2.
8. Hollander M, Wolfe DA. Nonparametric statistical methods. New York (NY): John Wiley and Sons; 1973. p. 120–3.
9. Rouse DJ, MacPherson C, Landon M, Varner MW, Leveno KJ, Moawad AH, et al. Blood transfusion and cesarean delivery. Obstet Gynecol 2006;108:891–7.
10. Martin JA, Hamilton BE, Sutton PD, Ventura SJ, Menacker F, Kirmeyer S. Births: final data for 2004. Natl Vital Stat Rep 2006;55:1–101.
In addition to the authors, other members of the National Institute of Child Health and Human Development Maternal-Fetal Medicine Units Network are as follows:
Ohio State University—J. Iams, F. Johnson, S. Meadows, H. Walker
University of Alabama at Birmingham—J. Hauth, A. Northen, S. Tate
University of Texas Southwestern Medical Center—S. Bloom, J. Gold, D. Bradford
University of Utah—M. Belfort, F. Porter, B. Oshiro, K. Anderson, A. Guzman
University of Chicago—J. Hibbard, P. Jones, M. Ramos-Brinson, M. Moran, D. Scott
University of Pittsburgh—K. Lain, M. Cotroneo, D. Fischer, M. Luce
Wake Forest University—P. Meis, M. Swain, C. Moorefield, K. Lanier, L. Steele
Thomas Jefferson University—A. Sciscione, M. DiVito, M. Talucci, M. Pollock
Wayne State University—M. Dombrowski, G. Norman, A. Millinder, C. Sudz, B. Steffy
University of Cincinnati—T. Siddiqi, H. How, N. Elder
Columbia University—F. Malone, M. D'Alton, V. Pemberton, V. Carmona, H. Husami
Brown University—H. Silver, J. Tillinghast, D. Catlow, D. Allard
Northwestern University—A. Peaceman, M. Socol, D. Gradishar, G. Mallett
University of Miami, Miami, FL—G. Burkett, J. Gilles, J. Potter, F. Doyle, S. Chandler
University of Tennessee—W. Mabie, R. Ramsey
University of Texas at San Antonio—D. Conway, S. Barker, M. Rodriguez
University of North Carolina—K. Moise, K. Dorman, S. Brody, J. Mitchell
University of Texas at Houston—L. Gilstrap, M. Day, M. Kerr, E. Gildersleeve
Case Western Reserve University—P. Catalano, C. Milluzzi, B. Slivers, C. Santori
The George Washington University Biostatistics Center—E. Thom, S. Gilbert, H. Juliussen-Stevenson, M. Fischer
National Institute of Child Health and Human Development—D. McNellis, K. Howell, S. Pagliaro