| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH |
From the Department of Obstetrics and Gynecology, and School of Public Health, University of Illinois at Chicago, Chicago, Illinois.
Address reprint requests to: Melissa Gilliam, MD, MPH, University of Illinois at Chicago, Department of Obstetrics and Gynecology, 820 South Wood Street MC 808, Chicago, IL 60612; E-mail: Mgilli2{at}uic.edu.
 |
ABSTRACT |
|---|
TOP
 ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS: A hospital-based, case-control study was conducted in which 316 multiparous women with placenta previa were identified. Controls consisted of 2051 multiparous women with spontaneous vaginal deliveries. Information on prior cesarean delivery was examined in three forms: as a dichotomous variable, as an ordinal variable, and as a set of three indicator variables for one, two, and three or more cesarean deliveries. Multivariable logistic regression modeling was used to obtain an adjusted estimate of this association.
RESULTS: Women with a prior cesarean delivery were more likely to have a placenta previa than those without (odds ratio [OR] 1.59, 95% confidence interval [CI] 1.21, 2.08). The likelihood of placenta previa increased as both parity and number of cesarean deliveries increased. Thus, the adjusted OR for a primiparous woman with one cesarean delivery was 1.28 (95% CI 0.82, 1.99). For a woman who has four or more deliveries with only a single cesarean delivery, the OR increases to 1.72 (95% CI 1.12, 2.64). This trend continues with greater parity and a greater number of cesarean deliveries such that the likelihood of placenta previa for a woman with parity greater than four and greater than four cesarean deliveries was OR 8.76 (95% CI 1.58, 48.53).
CONCLUSION: This study supports the association between prior cesarean delivery and placenta previa and demonstrates that the joint effect of parity and prior cesarean delivery is greater than that of either variable alone.
Placenta previa occurs when the placenta partially or completely occludes the internal cervical os. This condition is a major cause of third-trimester bleeding, postpartum hemorrhage, and maternal and neonatal morbidity and mortality. Placenta previa complicates approximately 4.8 per 1000 deliveries in the United States annually, and placenta previa is fatal to the mother in 0.03% of cases.1 Several risk factors for placenta previa exist including advanced maternal age, high parity, smoking, previous abortion, and prior cesarean delivery.2–5
Controversy exists over the epidemiology of placenta previa. Although many authors have commented on the relationship of placenta previa and cesarean delivery,6–8 the strength of this association remains in question. Studies of this relationship have not taken into account both the number of prior cesarean deliveries and the role of potential confounding variables. If the likelihood of placenta previa increases with a greater number of cesarean deliveries, this finding would support the idea of a causal relationship between prior cesarean delivery and placenta previa. Similarly, as the merits of elective repeat cesarean delivery are debated, possible consequences of this practice should be examined. The purpose of this study is to more accurately estimate the likelihood of placenta previa after multiple cesarean deliveries and to examine the effect of parity and other risk factors for placenta previa on this association.
|
MATERIALS AND METHODS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Information on prior cesarean delivery was examined in three forms: as a dichotomous variable "history of previous cesarean delivery" (yes, no), as an ordinal variable "number of previous cesarean deliveries," and as a set of three indicator ("dummy") variables for one, two, and three or more previous cesarean deliveries, with no prior cesarean deliveries serving as the reference group. Information on potential confounders/effect modifiers, including age, cigarette smoking during pregnancy, induced abortion, spontaneous abortion, race/ethnicity, second-trimester bleeding, third-trimester bleeding, and parity was also examined.
Univariate statistics for each variable, along with the bivariate association between each variable and placenta previa were examined. Single-factor stratified analysis was conducted to assess potential confounding or effect modification of the association between prior cesarean delivery and placenta previa. Multivariable logistic regression modeling was then used to obtain an adjusted estimate of this association.
The multivariable regression modeling was carried out separately for each of the three forms of the prior cesearean delivery variables. Importantly, interaction terms for parity and each of the three prior cesarean delivery variables were either statistically significant or close to statistically significant when included in these models (data not shown). This is not unexpected because the number of prior cesarean deliveries is by definition constrained by the total number of deliveries. To facilitate interpretation of our results, therefore, the multivariable modeling was repeated, this time stratified by parity. The final models for each parity group included age greater than 35, smoking, and prior abortion in addition to the various forms of prior cesarean delivery.
Repeating the modeling three times permitted an assessment of overall association as well as an assessment of trend. In particular, use of the ordinal variable for number of prior cesearean deliveries permits a test specifically for linear trend, whereas use of the indicator variables permits an examination of trend with no assumption about its shape. Results for the ordinal variable are reported for zero, one, two, three, and four or more prior cesarean deliveries, with corresponding parity categories of one, two, three, and four or more. Results for the indicator variables are reported for zero, one, two, and three or more prior cesarean deliveries, with corresponding parity categories of one, two, three or more. Fewer categories for the indicator variables were used to ensure adequate sample size in each category because very few women had four or more prior cesarean deliveries.
|
RESULTS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
).
|
). Table 3 shows this association stratified by parity and adjusting for age greater than or equal to 35, smoking, and prior abortion. The relationship between placenta previa and cesarean delivery for the dichotomous variable was significant at parity two, three, and four or greater, and increased with each level of parity.
|
|
shows this association after stratifying by parity and after adjustment. Here, the likelihood of placenta previa is significant and increasing as parity increases. In addition, the likelihood of placenta previa is significant and increasing as the number of prior cesarean deliveries increases within each parity group. For example, the adjusted OR for women of parity of four or more who had three prior cesarean deliveries was 5.09 (95% CI 1.41, 18.39). This value is greater than the ORs shown in Table 4 for women of the same parity who had fewer prior cesarean deliveries, and also greater than the ORs for women of lesser parity who had the same number of prior cesarean deliveries.
|
). Table 5 shows the adjusted OR for the association between prior cesarean delivery and placenta previa using the three indicator variables, stratified by parity. Again, the likelihood of placenta previa increased with a greater number of cesarean deliveries and also as parity increased. The adjusted ORs ranged from 1.28 (95% CI 0.82, 1.99) for primiparous women to 4.05 (95% CI 1.63, 10.10) for women of parity of three or greater with three or more prior cesarean deliveries. Thus, when using the indicator variables, the joint effect of parity and prior cesarean delivery also appears to be greater than the effect of either factor alone.
|
|
DISCUSSION |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In this study, by using a large and detailed database, we examined not only the relationship between one cesarean delivery and subsequent placenta previa, but also the contribution of each additional cesarean delivery to the development of placenta previa. In addition, the independent nature of this relationship was determined by adjusting for multiple known risk factors for placenta previa including age, smoking, and abortion. Using this methodology, we demonstrated the importance of the number of cesarean deliveries when assessing the likelihood of future placenta previa. When we calculated the likelihood of placenta previa using only a dichotomous variable, "any history of cesarean delivery," the OR was 1.59. Once we considered the actual number of cesarean deliveries, the strength of this association became apparent: the OR for the likelihood of placenta previa with three or more prior cesarean deliveries was 3.62.
In this study, we also demonstrated that the likelihood of placenta previa increases with greater parity independent of the number of prior cesarean deliveries. In other words, the association between cesarean delivery and placenta previa grows stronger as parity increases even if the number of cesarean deliveries stays the same. We demonstrated this relationship in two ways. First, with the use of an ordinal variable (Table 4) in logistic regression modeling we showed that the likelihood of placenta previa increases with each additional pregnancy. Second, because use of an ordinal variable imposes a linear relationship even when one does not truly exist, we also analyzed the data using indicator variables (Table 5) for the number of prior cesarean deliveries. The resulting ORs, using the indicator variables, reflect the actual experience of the women in our sample. Even using this assumption-free approach, a similar pattern for the association between prior cesarean delivery and placenta previa emerged, although it was somewhat less pronounced than when the ordinal variable was used. Again, the likelihood of placenta previa increased both across and within parity groups.
The use of database records raises questions as to the quality of the data, as review of all medical records was not possible. For example, chart abstractors might have missed cases of placenta previa even if this diagnosis was recorded in the medical record. Or, the medical record itself might have been in error, if bleeding occurred while in labor and a specific diagnosis of placenta previa was not made. However, there is no reason to believe that missed placenta previa cases were systematically different than the cases included in our study with regard to history of prior cesarean delivery. Change in diagnosis at the time of cesarean delivery might also occur, but as the abstractions were performed on postpartum charts, such changes in diagnosis would have been apparent at the time of abstraction.
One question that is raised and not answered in this study, or previous studies, is the role of prior placenta previa as a risk factor for repeat placenta previa. Women who have a history of placenta previa have an increased risk of placenta previa in a subsequent pregnancy.9 Information on the indication for previous cesarean delivery in our cases could not be obtained through review of medical records. Some women with prior cesarean delivery may actually be cases of placenta previa. We, therefore, raise this issue as a potential source of bias in this study.
This study supports the conclusions of previous studies showing an increased likelihood of placenta previa in women with prior cesarean delivery, and also shows the relationship of parity in the occurrence of placenta previa. Although one cesarean delivery does not significantly increase the likelihood of placenta previa in a primiparous woman, subsequent deliveries, whether vaginal or cesarean, and cesarean deliveries in particular, do increase the likelihood of future placenta previa. In fact, women with the combination of high parity and multiple repeat cesarean deliveries have the greatest likelihood of placenta previa.
The relationship between multiple prior cesarean deliveries and placenta previa is particularly important to consider given the renewed controversy regarding the benefits of a vaginal trial of labor after prior cesarean delivery.10 Repeat cesarean delivery has been associated with increased health care costs and maternal morbidity when compared with a vaginal trial of labor.11,12 As shown in this study, the increase in maternal morbidity caused by repeat cesarean delivery is not limited to immediate operative complications but extends throughout a woman’s reproductive life.
|
Footnotes |
|---|
We thank Kristine Stolti, MPH, for statistical assistance and Arden Handler, DrPH, for consultation and advice.
Received September 18, 2001. Received in revised form January 11, 2002. Accepted February 14, 2002.
|
REFERENCES |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2. Zhang J, Savitz DA. Maternal age and placenta previa: A population-based, case-control study. Am J Obstet Gynecol 1993;168:641–5.[Medline]
3. Handler AS, Mason ED, Rosenberg DL, Davis FG. The relationship between exposure during pregnancy to cigarette smoking and cocaine use and placenta previa. Am J Obstet Gynecol 1994;170:884–9.[Medline]
4. Barrett JM, Boehm FH, Killam AP. Induced abortion: A risk factor for placenta previa. Am J Obstet Gynecol 1981; 141:769–72.[Medline]
5. Miller DA, Chollett JA, Goodwin TM. Clinical risk factors for placenta previa-placenta accreta. Am J Obstet Gynecol 1997;177:210–4.[Medline]
6. McMahon MJ, Rongling L, Schenck AP, Olshan AF, Royce RA. Previous cesarean birth, a risk factor for placenta previa? J Reprod Med 1997;7:409–12.
7. To WWK, Leung WC. Placenta previa and previous cesarean section. Intl J Gynecol Obstet 1995;51:25–31.
8. Clark SL, Koonings PP, Phelan JP. Placenta previa/accreta and prior cesarean section. Obstet Gynecol 1985;66: 89–92.
9. Monica G, Lilja C. Placental previa, maternal smoking and recurrence risk. Acta Obstet Gynecol Scand 1995;74:341–5.[Medline]
10. Greene M. Vaginal delivery after cesarean section—Is the risk acceptable? N Engl J Med 2001;345:54–5.
11. Grobman WA, Peaceman AM, Socol ML. Cost-effectiveness of elective cesarean delivery after one prior low transverse cesarean. Obstet Gynecol 2000;95:745–51.
12. Flamm BL, Goings JR, Liu Y, Wolde-Tsadik G. Elective repeat cesarean delivery versus trial of labor: A prospective multicenter study. Obstet Gynecol 1994;83:927–32.[Medline]
This article has been cited by other articles:
|
|
 |
|
  A. K. Daltveit, M. C. Tollanes, H. Pihlstrom, and L. M. Irgens Cesarean Delivery and Subsequent Pregnancies Obstet. Gynecol., June 1, 2008; 111(6): 1327 - 1334. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. A. Grobman, R. Gersnoviez, M. B. Landon, C. Y. Spong, K. J. Leveno, D. J. Rouse, M. W. Varner, A. H. Moawad, S. N. Caritis, M. Harper, et al. Pregnancy Outcomes for Women With Placenta Previa in Relation to the Number of Prior Cesarean Deliveries Obstet. Gynecol., December 1, 2007; 110(6): 1249 - 1255. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Ball and K. Hinshaw The current management of vaginal birth after previous caesarean delivery Obstet Gynaecol (Lond), April 1, 2007; 9(2): 77 - 82. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. P. Walker, E. A. McCarthy, A. Ugoni, A. Lee, S. Lim, and M. Permezel Cesarean Delivery or Vaginal Birth: A Survey of Patient and Clinician Thresholds Obstet. Gynecol., January 1, 2007; 109(1): 67 - 72. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. M. Allen, C. M. O'Connell, and T. F. Baskett Cumulative economic implications of initial method of delivery. Obstet. Gynecol., September 1, 2006; 108(3): 549 - 555. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. M. Silver, M. B. Landon, D. J. Rouse, K. J. Leveno, C. Y. Spong, E. A. Thom, A. H. Moawad, S. N. Caritis, M. Harper, R. J. Wapner, et al. Maternal Morbidity Associated With Multiple Repeat Cesarean Deliveries. Obstet. Gynecol., June 1, 2006; 107(6): 1226 - 1232. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. K. Laughon, H. M. Wolfe, and A. G. Visco Prior Cesarean and the Risk for Placenta Previa on Second-Trimester Ultrasonography Obstet. Gynecol., May 1, 2005; 105(5): 962 - 965. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Wu, A. F. Hundley, and A. G. Visco Elective Primary Cesarean Delivery: Attitudes of Urogynecology and Maternal-Fetal Medicine Specialists Obstet. Gynecol., February 1, 2005; 105(2): 301 - 306. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Bahl, B. Strachan, and D. J Murphy Outcome of subsequent pregnancy three years after previous operative delivery in the second stage of labour: cohort study BMJ, February 7, 2004; 328(7435): 311. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Minkoff and F. A. Chervenak Elective Primary Cesarean Delivery N. Engl. J. Med., March 6, 2003; 348(10): 946 - 950. [Full Text] [PDF]
|
|
|
|
|
|
J. R. Scott Putting Elective Cesarean Into Perspective Obstet. Gynecol., June 1, 2002; 99(6): 967 - 968. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
