Abnormally invasive placenta includes three pregnancy complications pertaining to invasion of the placental villi into myometrium (placenta accreta and increta) or uterine serosa (placenta percreta) and, in rare cases, involves invasion of the surrounding organs. Abnormally invasive placenta is associated with major maternal and neonatal morbidity and mortality and is a leading cause of obstetric hemorrhage.1,2
Incidence rate estimates range from 1.7 to 90 per 10,000, complicated by differences in case definition, and predominantly limited to institution-based retrospective case reviews.3–5 Incidence appears to be increasing,6 purportedly associated with the rise in cesarean delivery, which is a well-established risk factor.3,4,7 However, primiparous women have been reported to comprise up to 38% of cases, suggesting that there may be other important etiologic factors.2,8
The association between prior cesarean delivery and abnormally invasive placenta is thought to arise from the uterine scar, which may cause a localized failure of decidualization.9 Similarly, other gynecologic procedures that result in uterine scarring may be associated with abnormally invasive placenta. Few studies have examined its relationship with prior procedures, with most considering only curettage4,10 and limited by small sample size. The inclusion of multiparous women in previous studies makes it difficult to disentangle associations of abnormally invasive placenta with prior cesarean delivery, prior birth trauma, and prior gynecologic procedures.3,6,8
This study aims to examine whether prior gynecologic procedures are associated with increased risk of subsequent development of abnormally invasive placenta. We used population-level data to explore the association between prior procedures and the likelihood of abnormally invasive placenta diagnosis in primiparous women over a 10-year period. We hypothesized that women who have undergone invasive procedures of the uterus will be more likely to have pregnancies complicated by abnormally invasive placenta.
MATERIALS AND METHODS
We conducted a population-based data linkage study including primiparous women who delivered a liveborn or stillborn neonate(s) of at least 20 weeks of gestation in New South Wales, Australia, in public or private hospitals or at home during the period January 1, 2003, to December 31, 2012. New South Wales is Australia's most populous state, comprising approximately one third of the country's births.11 Data on maternal and neonatal characteristics, pregnancy, labor, and birth were obtained from the New South Wales Perinatal Data Collection, which records all births of at least 400 g birth weight or 20 weeks of gestation. Previous diagnoses and procedures and additional risk factors were obtained from the New South Wales Admitted Patient Data Collection from January 1, 2001, to December 31, 2012, allowing a minimum lookback period of 2 years with a maximum of 12 and a mean of 7.2 years. The Admitted Patient Data Collection is a census of all hospital admissions in New South Wales with diagnoses and procedures coded using the International Classification of Diseases, 10th Revision Australian modification and the Australian Classification of Health Interventions, 8th Edition, respectively. The data were linked by the New South Wales Centre for Health Record Linkage using probabilistic methods based on personal identifiers. Unique project identifiers were assigned to each mother and neonate, allowing for longitudinal linkage of hospitalizations before pregnancy and delivery. Rates of missed and incorrect links are less than 5 per 1,000.12,13
The outcome of interest was defined by the presence of the diagnosis code for morbidly adherent placenta (O43.2), a synonym for abnormally invasive placenta, which encompasses placenta accreta, increta, and percreta. The comparison group comprised deliveries in women with no record of abnormally invasive placenta. Multiple births (eg, twins) were counted as a single delivery. Women with a uterine anomaly were excluded from analysis.
Exposures of interest were invasive procedures including 1) gynecologic laparoscopy with instrumentation of the uterus; 2) hysteroscopy, including operative hysteroscopy; 3) curettage, including suction curettage and surgical termination; and 4) endometrial ablation. A list of the codes used to identify invasive uterine procedures is provided in Appendix 1, available online at http://links.lww.com/AOG/B49. Procedures were defined as having been performed before pregnancy if the procedure date preceded the date of delivery by more than 287 days (41 weeks).
Potential confounders considered were maternal age, socioeconomic disadvantage, whether the mother was born in Australia, and having used assisted reproductive technology in the year before delivery. Procedure codes used to identify assisted reproductive technology are provided in Appendix 1 (http://links.lww.com/AOG/B49). Maternal age groups were determined based on clinical relevance and quintiles of socioeconomic disadvantage were used following the Australian Bureau of Statistics.14 As a result of the etiologic nature of this study, independent factors predictive of abnormally invasive placenta in previous studies, but which are unlikely to be correlated with prior procedures, were not included as covariates. Placenta previa was not included as a covariate because of the strong likelihood that it is on the causal pathway from uterine scarring to developing abnormally invasive placenta.
A multivariable modified Poisson regression model with robust error variances was used to test whether the number of prior procedures was associated with an increased risk of a delivery complicated by abnormally invasive placenta. Potential confounders with P<.2 in univariate models were included in the full multivariable model. Backward selection was used to remove factors not significant at the 1% level from the final model, following Hosmer et al.15 Women with missing data for full model variables were excluded. The distribution of missing data is provided by variable in Appendix 2 (available online at http://links.lww.com/AOG/B49).
Sensitivity analyses were performed as follows: 1) considering a case definition of a primiparous woman with a diagnosis of abnormally invasive placenta and delivery by cesarean delivery as a proxy for more severe cases and 2) excluding women with a diagnosis of placenta previa to test the association between prior procedures and abnormally invasive placenta in the absence of placenta previa.
Statistical analyses were performed using SAS Enterprise Guide 6.1. Results are presented as relative risks with 99% CIs. Ethics approval for this study was granted by the New South Wales Population and Health Services Research Ethics Committee (2006/06/011).
There were 922,925 deliveries of neonates at at least 20 weeks of gestation during the period 2003–2012, involving 595,597 mothers. Primiparous women represented 381,499 deliveries. Of those, 865 deliveries were complicated by abnormally invasive placenta, with primiparous women representing 38% of all abnormally invasive placenta diagnoses. There were 724 women who had a uterine anomaly and were excluded from further analysis (with abnormally invasive placenta: n=11 [1.3%]; without abnormally invasive placenta: n=713 [0.2%]). The overall rate of abnormally invasive placenta in our study population was 22.4 per 10,000. Complete information on possible confounding factors was available for 378,303 deliveries (99.2%).
Women with abnormally invasive placenta were older and had a higher socioeconomic advantage (Table 1). Higher socioeconomic status was independently associated with having had a prior procedure and with a higher number of prior procedures (Appendix 3, available online at http://links.lww.com/AOG/B49). Women with lower socioeconomic status also had proportionally fewer hysteroscopies than women with higher socioeconomic status. Gestational ages were lower with abnormally invasive placenta, and there was a somewhat higher proportion of neonates born small for gestational age, although this was not statistically significant. Placenta previa, hypertension, multiple births, and female fetal sex were also positively associated with abnormally invasive placenta (Table 1). Median length of stay in the birth admission was 5 days for the abnormally invasive placenta group (interquartile range 4–7) compared with 4 days for the no abnormally invasive placenta group (interquartile range 2–5).
A total of 33,296 primiparous women had at least one procedure before pregnancy (8.7%; Table 2). Women with abnormally invasive placenta had higher rates of gynecologic laparoscopy, hysteroscopy, and curettage than those without (Table 2). No women with abnormally invasive placenta had a history of endometrial ablation, whereas 318 women (0.08%) without abnormally invasive placenta had undergone the procedure and therefore ablation is not included in overall procedure counts. One hundred fifty-two (17.8%) women with abnormally invasive placenta had at least one prior procedure compared with 33,144 (8.7%) women without abnormally invasive placenta (relative risk [RR] 2.3, 99% CI 1.8–2.8; Table 2). The median length of stay for the last procedure before pregnancy was 1 day (interquartile range 1.0–1.0) for both the abnormally invasive placenta and no abnormally invasive placenta groups. Maternal age at last procedure was higher for the abnormally invasive placenta group, and the interval between the last procedure and delivery was similar between the groups (maternal age in years: abnormally invasive placenta group median 33.0, interquartile range 28.3–36.6; no abnormally invasive placenta group 30.0, interquartile range 25.7–34.0; lead time in years: abnormally invasive placenta group median 1.5, interquartile range 1.1–2.2; no abnormally invasive placenta group median 1.5, interquartile range 1.1–2.5). Rates of gynecologic procedures differed according to socioeconomic disadvantage, with more advantaged women more likely to have had a gynecologic procedure, but this difference dissipated after adjustment for age (unadjusted: disadvantage quintile 2 vs 1, RR 0.93, 99% CI 0.89–0.97; disadvantage quintile 5 vs 1, RR 0.70, 99% CI 0.67–0.74; P<.01; age-adjusted: disadvantage quintile 2 vs 1, RR 1.02, 99% CI 0.98–1.06; disadvantage quintile 5 vs 1, RR 0.92, 99% CI 0.88–0.97; P<.01).
The absolute risk of abnormally invasive placenta for women with no history of procedures was 20.2 (95% CI 18.7–21.7) per 10,000 compared with 37.0 (95% CI 29.9–44.2), 74.5 (95% CI 48.7–100.3), and 149.9 (95% CI 78.6–221.2) per 10,000 for women with one, two, and three or more prior procedures, respectively. After adjustment for maternal factors, the risk of abnormally invasive placenta increased with increasing number of prior procedures with RRs ranging from 1.5 (one prior procedure, 99% CI 1.1–1.9) to 5.1 (three or more prior procedures, 99% CI 2.7–9.6) (Table 3). Demographic factors including higher maternal age, socioeconomic advantage, and being born in Australia were positively associated with abnormally invasive placenta (Table 3).
When the case definition was restricted to abnormally invasive placenta diagnosis and delivery by cesarean delivery, the results were similar to the main analysis. Risk ratios were generally slightly higher, particularly for the number of procedures and higher maternal age (Appendix 4, available online at http://links.lww.com/AOG/B49). Similarly, the results of the sensitivity analysis excluding women with placenta previa showed a negligible difference from the main analysis (Appendix 5, available online at http://links.lww.com/AOG/B49).
Our findings suggest that prior gynecologic uterine procedures may be etiologic factors for abnormally invasive placenta. The association is dose-dependent with a fivefold increase in risk for women with a history of three or more procedures. Previous studies have suggested an etiologic role for curettage and to a lesser extent other intrauterine procedures,3,4,8,10,16,17 and frequent endomyometrial injury has been demonstrated in women who have undergone vacuum termination or curettage.18 However, most previous studies did not compare rates of prior procedures among cases with baseline rates8,16,17 or found no significant differences.4,10 Potential reasons for differences in findings maybe lower numbers of cases (n=28 and n=62) in other studies4,10 and differences in ascertainment of prior procedures such as type and patient recall of such procedures. Fitzpatrick et al3 found an increased risk for previous uterine surgery, but they did not consider hysteroscopy or biopsy and included manual removal of the placenta, which may be indicative of previous abnormally invasive placenta. Wu et al6 found that a history of abortion or curettage was not associated with an increased risk of abnormally invasive placenta, but they acknowledge limitations with the accuracy of their data set. Our study uses a large, validated population-level data set to obtain a large cohort of primiparous women with abnormally invasive placenta and clearly demonstrates a strong association between a range of gynecologic procedures and the development of abnormally invasive placenta, with an increasing association of abnormally invasive placenta as the number of prior gynecologic procedures increases. It is plausible that the increasing rate of abnormally invasive placenta, demonstrated for these data2 and overseas,4,6,19 may be associated with an increase in uterine procedures, although this remains to be examined.
Consistent with previous studies, advanced maternal age,3,4,6,8 placenta previa,3,4,6,10 hypertension,20 and female fetal sex21 were significantly associated with abnormally invasive placenta, and multiple birth was also a significant predictor. Contrary to Hung et al,10 we found that the increased risk from prior procedures remained even when excluding cases of placenta previa.
Socioeconomic status and maternal country of birth were significantly associated with increased adjusted risk of abnormally invasive placenta. Creanga et al19 found that abnormally invasive placenta risk increased with higher socioeconomic advantage for primary but not repeat cesarean deliveries. The association between abnormally invasive placenta and socioeconomic status observed here is likely related to access to procedures, as indicated by our finding that people with greater socioeconomic advantage underwent more procedures. Furthermore, types of prior procedures differed with more disadvantaged women less likely to have a hysteroscopy, the group of procedures with the strongest association with abnormally invasive placenta. The finding that Australian-born women are at increased risk is consistent with studies on other rare conditions in pregnancy in New South Wales.22 The increased risk of abnormally invasive placenta among Australian-born women may partly reflect cultural differences in propensities to undergo gynecologic procedures before pregnancy or ultrasound screening during pregnancy,23 different access to care, or differential ascertainment of procedures for Australian and overseas-born women. Overseas-born women may have undergone procedures elsewhere before giving birth in New South Wales, which would not be detectable in our data. Another possible reason for different abnormally invasive placenta rates by country of birth may be country-specific differences in gynecologic surgical techniques.
The rate of assisted reproductive technologies in the year before birth was more than doubled for women with abnormally invasive placenta, but lack of a statistically significant association after adjustment in our study suggests confounding with maternal age. Our findings contrast with those of Esh-Broder et al,24 Thurn et al,8 and Fitzpatrick et al,3 but of these studies, two did not adjust for maternal age and other confounding factors,8,24 and the third included only five women with a history of in vitro fertilization.3
Our findings highlight the risk of abnormally invasive placenta in nulliparous women presenting for care in pregnancy after gynecologic procedures. Among primiparous women, the rate of abnormally invasive placenta was almost as high as that among all women (22.4 vs 24.8/10,000).2 Antenatal detection of abnormally invasive placenta is associated with lower morbidity.25,26 Pregnant women with a history of prior procedures should be advised to present for regular antenatal care; at the morphology ultrasound examination, careful placental assessment should occur to look for any features suggestive of abnormally invasive placenta. For women in whom there is uncertainty for abnormally invasive placenta on routine ultrasound examination, further investigations such as second opinion tertiary-level ultrasonography or follow-up magnetic resonance imaging may be considered.27,28
One of the strengths of this study is its use of routinely collected population data and the clinical diagnosis code for abnormally invasive placenta. The 10 years of complete population data for New South Wales yielded a large cohort of primiparous women with abnormally invasive placenta and a sufficient number of women with prior procedures to allow for adjustment for confounding factors.
The use of routinely collected data also imposes some limitations. An abnormally invasive placenta diagnosis cannot be further categorized into specific variants of accreta, increta, and percreta. However, previous studies suggest that at least three fourths of cases are likely to represent placenta accreta.4,29 Although the code has not been validated, we have demonstrated face validity elsewhere,2 and the results of the sensitivity analysis provide further support for this. We have demonstrated that procedures are generally well recorded in these data.30 Our hospital data have a minimum lookback period of 2 years. Intervals between intrauterine procedures and a woman's first birth can be lengthy, and we do not have a complete history for all women. However, this has likely under- rather than overestimated the relationship.
This study provides strong evidence that invasive gynecologic procedures are associated with an increased likelihood of developing abnormally invasive placenta. These insights are important in identifying women at risk to allow better management of pregnancies and reduce morbidity and mortality associated with this condition.
1. Oyelese Y, Smulian JC. Placenta previa, placenta accreta, and vasa previa. Obstet Gynecol 2006;107:927–41.
2. Baldwin HJ, Patterson JA, Nippita TA, Torvaldsen S, Ibiebele I, Simpson J, et al. Maternal and neonatal outcomes following abnormally invasive placenta: a population-based record linkage study. Acta Obstet Gynecol Scand 2017;96:1373–81.
3. Fitzpatrick KE, Sellers S, Spark P, Kurinczuk JJ, Brocklehurst P, Knight M. Incidence and risk factors for placenta accreta/increta/percreta in the UK: a national case-control study. PLoS One 2012;7:e52893.
4. Miller DA, Chollet JA, Goodwin TM. Clinical risk factors for placenta previa–placenta accreta. Am J Obstet Gynecol 1997;177:210–4.
5. Gielchinsky Y, Mankuta D, Rojansky N, Laufer N, Gielchinsky I, Ezra Y. Perinatal outcome of pregnancies complicated by placenta accreta. Obstet Gynecol 2004;104:527–30.
6. Wu S, Kocherginsky M, Hibbard JU. Abnormal placentation: twenty-year analysis. Am J Obstet Gynecol 2005;192:1458–61.
7. Hull AD, Moore TR. Multiple repeat cesareans and the threat of placenta accreta: incidence, diagnosis, management. Clin Perinatol 2011;38:285–96.
8. Thurn L, Lindqvist P, Jakobsson M, Colmorn L, Klungsoyr K, Bjarnadóttir R, et al. Abnormally invasive placenta—prevalence, risk factors and antenatal suspicion: results from a large population-based pregnancy cohort study in the Nordic countries. BJOG 2015;123:1348–55.
9. Jauniaux E, Jurkovic D. Placenta accreta: pathogenesis of a 20th century iatrogenic uterine disease. Placenta 2012;33:244–51.
10. Hung TH, Shau WY, Hsieh CC, Chiu TH, Hsu JJ, Hsieh TT. Risk factors for placenta accreta. Obstet Gynecol 1999;93:545–50.
11. Australian Institute of Health and Welfare. Australia's mothers and babies 2014—in brief. Perinatal Statistics Series no. 32. Cat no. PER 87. Canberra (Australia): AIHW; 2016.
12. Bentley JP, Ford JB, Taylor LK, Irvine KA, Roberts CL. Investigating linkage rates among probabilistically linked birth and hospitalization records. BMC Med Res Methodol 2012;12:149.
13. Centre for Health Record Linkage. CHeReL—quality assurance. Available at: http://www.cherel.org.au/quality-assurance
. Retrieved May 30, 2017.
14. Australian Bureau of Statistics. Australian social trends: health and socioeconomic disadvantage. Available at: http://www.abs.gov.au/AUSSTATS/[email protected]/Lookup/4102.0Main+Features30Mar+2010
. Retrieved June 13, 2017.
15. Hosmer D, Lemeshow S, May S. Applied survival analysis. New York (NY): John Wiley & Sons; 2008.
16. Jacques SM, Qureshi F, Trent VS, Ramirez NC. Placenta accreta: mild cases diagnosed by placental examination. Int J Gynecol Pathol 1996;15:28–33.
17. Pron G, Mocarski E, Bennett J, Vilos G, Common A, Vanderburgh L, et al. Pregnancy after uterine artery embolization for leiomyomata: the Ontario multicenter trial. Obstet Gynecol 2005;105:67–76.
18. Beuker JM, Erwich JJ, Khong TY. Is endomyometrial injury during termination of pregnancy or curettage following miscarriage the precursor to placenta accreta? J Clin Pathol 2005;58:273–5.
19. Creanga AA, Bateman BT, Butwick AJ, Raleigh L, Maeda A, Kuklina E, et al. Morbidity associated with cesarean delivery in the United States: is placenta accreta an increasingly important contributor? Am J Obstet Gynecol 2015;213:384.e1–11.
20. Usta IM, Hobeika EM, Abu Musa AA, Gabriel GE, Nassar AH. Placenta previa-accreta: risk factors and complications. Am J Obstet Gynecol 2005;193:1045–9.
21. Khong T, Healy D, McCloud P. Pregnancies complicated by abnormally adherent placenta and sex ratio at birth. BMJ 1991;302:625–6.
22. Chen JS, Roberts CL, Simpson JM, March LM. Pregnancy outcomes in women with rare autoimmune diseases. Arthritis Rheumatol 2015;67:3314–23.
23. Davies-Tuck ML, Davey MA, Wallace EM. Maternal region of birth and stillbirth in Victoria, Australia 2000–2011: a retrospective cohort study of Victorian perinatal data. PLoS One 2017;12:e0178727.
24. Esh-Broder E, Ariel I, Abas-Bashir N, Bdolah Y, Celnikier DH. Placenta accreta is associated with IVF pregnancies: a retrospective chart review. BJOG 2011;118:1084–9.
25. Warshak CR, Ramos GA, Eskander R, Benirschke K, Saenz CC, Kelly TF, et al. Effect of predelivery diagnosis in 99 consecutive cases of placenta accreta. Obstet Gynecol 2010;115:65–9.
26. Fitzpatrick KE, Sellers S, Spark P, Kurinczuk JJ, Brocklehurst P, Knight M. The management and outcomes of placenta accreta, increta, and percreta in the UK: a population-based descriptive study. BJOG 2014;121:62–70.
27. Comstock C, Bronsteen R. The antenatal diagnosis of placenta accreta. BJOG 2014;121:171–81.
28. Placenta accreta. Committee Opinion No. 529. American College of Obstetricians and Gynecologists. Obstet Gynecol 2012;120:207–11.
29. Wong HS, Cheung YK, Zuccollo J, Tait J, Pringle KC. Evaluation of sonographic diagnostic criteria for placenta accreta. J Clin Ultrasound 2008;36:551–9.
30. Lain SJ, Hadfield RM, Raynes-Greenow CH, Ford JB, Mealing NM, Algert CS, et al. Quality of data in perinatal population health databases: a systematic review. Med Care 2012;50:e7–20.