Placental abruption: assessing trends in risk factors over time

To evaluate changes in the independent contribution of different risk factors for placental abruption over time. In this retrospective nested case–control study, trends of change in ORs for known risk factors for placental abruption occurring in three consecutive 8-year intervals were compared. A univariate assessment of factors associated with placental abruption and two multivariable logistic regression models were constructed to identify independent risk factors for placental abruption. Trends of change in the incidence and specific contribution of various risk factors were compared along the study time-period. During the study period, 295,946 pregnancies met the inclusion criteria; of these, 2170 (0.73%) were complicated with placental abruption. Using logistic regression models, previous cesarean delivery, in vitro fertilization (IVF) pregnancy, hypertensive disorders, polyhydramnios, and inadequate prenatal care were recognized as independent risk factors for placental abruption. While the relative contribution of IVF pregnancy and polyhydramnios to the overall risk for abruption decreased over the course of the study, previous cesarean delivery became a stronger contributor for placental abruption. In our study, a change over time in the specific contribution of different risk factors for placental abruption has been demonstrated.


Introduction
Placental abruption is classically defined as partial or complete separation of implanted placenta from the uterine wall before delivery of the baby. It occurs in about 0.8-1% of pregnancies, depending on the population studied and diagnostic criteria with increasing incidence over the years [1][2][3][4]. Placental abruption is frequently associated with perinatal morbidity and mortality [5]. Multiple difficulties, such as altering definition and no united universal classification method together with lack of known etiology or predictive markers, make placental abruption a worldwide challenge and still a "big unknown" [3].
Numerous pre-existing conditions are associated with placental abruption. At least 50 different risk factors for placental abruption have been reported, such as preterm premature rupture of membranes, gestational hypertension, pre-eclampsia, and previous cesarean delivery [5,6]. Smoking, pre-eclampsia, and history of previous placental abruption were demonstrated as being the strongest risk factors for placental abruption [3]. As with the incidence of placental abruption, the prevalence of some of its risk factors can also be affected by demographical and epidemiological trends.
Recently, a cohort analysis described different trends in the rates of placental abruption between the US and five countries from Europe (Sweden, Denmark, Norway, Spain, and Finland). While in the US, the abruption rate has been stable since 2000, in all other countries, declining rates were seen. The most dramatic decline was observed in Denmark. Changes in smoking prevalence partially explained the period effect in the US and Sweden [7]. Other studies also demonstrated shifts in specific risk factors for placental In memory of Dr. David Yohay, a beloved physician and teacher who passed away while working on this study. abruption such as cesarean delivery rates. In addition, changes in maternal characteristics, such as older maternal age, reduced parity, high pre-pregnancy weight, and increased weight gain during pregnancy may influence placental abruption incidence over time [8].
As placental abruption is often an unpredictable and lifethreatening event, knowledge about its risk factors enables to offer monitoring and prevention measures. Understanding the changes in risk factors as well as the changes in their independent contribution to placental abruption in a growing, developing population could have a significant impact in managing potential risks. Hence, the aim of this study was to evaluate changes in the independent contribution of different risk factor of placental abruption over time.

Setting
This retrospective study was conducted at the Soroka University Medical Center (SUMC), a tertiary medical center and the only hospital in the southern part of the country, occupies 65% of the country's territory. SUMC provides medical services for the entire population of the region (approximately 1.22 million people). Therefore, the study was based on non-selective population data. The institutional review board, in accordance with the Helsinki declaration, approved the study (IRB number 0155-19-SOR).

Study population
All singleton pregnancies of women with and without a placental abruption who delivered between the years 1988-2014 at the SUMC were included in the study. Women with multiple pregnancies and offspring with congenital malformations or chromosomal abnormalities were excluded from the analysis.

Data collection methods
Data were retrieved from the SUMC's perinatal database, which contains information documented directly after delivery by the attending obstetrician. Data include information on maternal demographics, clinical characteristics and perinatal outcomes. The information is captured from the patient's medical records and coded according to the ICD-9 codes by trained secretaries.

Clinical diagnosis of placental abruption
Placental abruption was clinically defined as the premature detachment of an implanted placenta from the uterine wall before the delivery of the fetus [1]. The diagnosis was made by the attending staff during or following the delivery. In some cases, the diagnosis was confirmed by pathological examination. Nevertheless, as abruption is considered a clinical diagnosis, only some cases of acute abruptions demonstrated histologic confirmation [9]. Clinical definitions are presented in Table 1 [10]. Maternal exposure, with or without placental abruption, as well as all other clinical characteristics were identified using ICD-9 codes, with ICD-9 code 641.2 for placental abruption.

Definitions
Inadequate prenatal care was defined as less than three visits to any prenatal care facility during pregnancy [13]. Polyhydramnios was defined as amniotic fluid index of more than 24 cm [14].
Gestational diabetes Gestational diabetes mellitus (GDM) was defined as carbohydrate intolerance resulting in abnormally high blood sugar level with onset or first recognition during pregnancy [11]. Hypertensive disorders were divided to gestational hypertension, pre-eclampsia, and chronic hypertension. Gestational hypertension was defined as a systolic blood pressure of 140 mm Hg or more or a diastolic blood pressure of 90 mm Hg or more, or both, on two occasions at least 4 h apart after 20 weeks of gestation in a woman with a previously normal blood pressure [12]. Preeclampsia was defined as a new onset of increased blood pressure (≥ 140/90 mm Hg) which occurs after 20 weeks of gestation with proteinuria, edema, or both [26]. Chronic hypertension was defined as hypertension diagnosed before conception or within the first 20 weeks in pregnancy [24]. Large for gestational age was defined as birth weight at or above the 90th percentile for gestational age [15].

Statistical analysis
Statistical analysis was performed with the SPSS package (SPSS, Chicago, IL).
The initial analysis was performed using descriptive statistics (mean, SD, graphs), followed by advanced analytical statistics using various parametric tests. Continuous variables with normal distribution were presented as mean ± standard deviation and compared between the study groups using t test. Continuous variables which are not normally distributed were presented as median with interquartile range and their statistical analysis was performed using the Mann-Whitney test. Categorical variables were presented in counts and percentages and their statistical analysis was performed using Chi-square or Fisher's exact test when appropriate.
Two different logistic regression models were constructed to identify independent risk factors associated with placental abruption at the different time intervals. To identify trends in the individual contribution of each risk factor, the ORs of each independent risk factor was compared between three consecutive 8-year intervals from 1988 to 2014 (T1-1988-1996; T2-1997-2005; T3-2006-2014). The ORs and their 95% confidence interval (CI) were calculated for each chosen risk factor. A p value of < 0.05 was considered statistically significant.

Results
During the study period, 295,946 singleton pregnancies met the inclusion criteria of these, 2170 (0.73%) were complicated by placental abruption.
Rates of placental abruption during the follow-up period are displayed in Fig. 1. Between 1988 and 2000, the incidence of placental abruption significantly increased from 0.6 to 1.0% and then significantly decreased back to 0.5% in 2014 (p < 0.005). Table 2 presents maternal characteristics and perinatal outcomes of women with and without placental abruption. Women with placental abruption were older with higher parity order compared to women in the control group. Rates of previous cesarean delivery (20.4% vs 12.3%, 442 vs 36,238, p < 0.001), IVF pregnancy (5.5% vs 2.7%, 120 vs 7910, p = < 0.001), recurrent pregnancy loss (8.0% vs 5.1%, 174 vs 15,124, p < 0.001), and inadequate prenatal care (13.2% vs 9.2%, 287 vs 26,898, p < 0.001) were higher among women with placental abruption compared to women without placental abruption. Accordingly, abruption was associated with higher rates of hypertensive disorders (13.3% vs 5.4%, 289 vs 15,957, p < 0.001), polyhydramnios (6.1% vs 3.4%, 132 vs 9925, p < 0.001), and premature rupture of membranes (PROM) ( Table 3 presents a logistic regression model, including all variables found to have a statistically significant association with placental abruption in the univariate analysis. In our population, maternal age, grandmultiparity, previous cesarean delivery, recurrent pregnancy loss, IVF pregnancy, inadequate prenatal care, hypertensive disorders, polyhydramnios, and PROM were demonstrated to be independent risk factors for placental abruption. Delivery of a large for gestational age neonate was noted as a protecting factor for placental abruption. Looking at the different time-periods of our study (Table 3), only previous caesarean delivery, IVF pregnancy, hypertensive disorders, polyhydramnios, and inadequate prenatal care were independently associated with placental abruption in all three time-periods. Delivery of a large for gestational age neonate was consistently protective factor for abruption in all three time-periods. To improve our analysis, we built another logistic regression model (Table 4) for the prediction of placental abruption including only the variables that were found to be significant in the first multivariate analysis in all three timeperiods. While previous cesarean delivery became a stronger contributor to the risk for abruption over the course of the study period, the relative contribution of IVF pregnancy and polyhydramnios to the overall risk for abruption decreased. Hypertensive disorders and inadequate prenatal care demonstrated mixed trends over the periods of the study. Radar chart analysis was used to elucidate the trends in the ORs for independent risk factors for abruption along the study period (Fig. 2). Table 5 compares the rates of the independent risk factors for abruption found in the course of the time-period in women with abruption. Rates of previous cesarean delivery demonstrated a rising trend, while that of polyhydramnios and hypertensive disorders decreased. All other rates of other risk factors did not change significantly along the time intervals.

Discussion
Our study found a significant increase in the incidence of placental abruption between 1988 and 2000, rising from 0.6% in 1988 to a peak of 1.0% in 2000 (p < 0.005). This was followed by a significant decrease in the incidence to 0.5% in 2014 (p < 0.005). In accordance with our study, Ananth et al. found that the rate of placental abruption among singleton births in the United Sates between 1979 and 2001 increased from 0.81% in 1979-1981 to 1.0% in 1999-2001 [2]. Other studies from the US, Israel, Taiwan, and Norway demonstrated similar trends [2,16]. A study from Finland showed a decrease in the incidence of placental abruption between the years 1980-2005, from 0.49 to 0.34% [17]. Differences in the incidence of placental abruption between studies may be explained by differences in the prevalence of different risk factors or differences in diagnostic criteria.
We have found five significant risk factors for placental abruption that remained consistently significant in the multivariate analyses in all three time-periods. These included previous cesarean delivery, IVF pregnancy, hypertensive disorders in pregnancy, polyhydramnios, and inadequate  prenatal care. Large for gestational age was found as a protective factor for abruption in all three time-periods. While previous cesarean delivery became a stronger contributor to the risk for abruption over the course of the study period, the relative contribution of IVF pregnancy and polyhydramnios to the overall risk for abruption decreased. Hypertensive disorders and inadequate prenatal care demonstrated mixed trends.
Concurrent with our study, the association between placental abruption and prior cesarean delivery has been demonstrated in previous studies [4,18]. Cesarean section rates are rising worldwide with reported rates increasing from 11.2 to 25% in Europe, from 22.3 to 32.3% in North America, and from 4.4 to 19.5% in Asia, between the years 1999 and 2014 [14]. Studies suggest that the uterine low segment adjacent scar may lead to impaired placental implantation. This can lead to impaired placental perfusion, resulting in inadequate blood flow to the intervillous space, and to increased risk of abruption. In addition, ligation of uterine vessels at the time of cesarean section may increase the risk of abruption. Therefore, the association between placental abruption and previous cesarean section may be biologically possible [19].
Compatible with other studies [20], our study also found an independent association between placental abruption and IVF pregnancy. Although the specific mechanism of the association is not fully understood, it may be explained by differences in placental formation. When the formation of the chorion is initiated in vitro, an inherent difference in the nature of the placenta itself may predispose the patient to develop placental-associated morbidities during gestation [20]. The use of IVF is increasing worldwide, and between 1995 and 2016, the number of ART procedures performed in the United States has more than tripled [21]. A report from 49 countries for the year 2000 demonstrated an increase of 5.5% in IVF treatments [22]. Data collected from 14 European countries showed that the pregnancy rates per aspiration and per transfer were higher in 2014 than in 2013, 35.8% versus 34.5%, respectively [23].
In agreement with our study, the association between hypertensive disorders of pregnancy and placental abruption has been well documented in previous studies [3,4,24]. Research on a large Canadian population demonstrated an increased incidence of pre-eclampsia from 26.4 per 1000 deliveries in 1989, to 50.6 in 2012 [25]. A study in Norway, between the years 1967 to 2003, shows an increase in the incidence of pre-eclampsia from 8% in 1967-1978 to nearly 20% in 1991-2003 [26]. Uteroplacental under perfusion or ischemia that may be a result of poor placentation in early pregnancy may be a common mechanism that links hypertensive disorders of pregnancy with placental abruption [24].
Another risk factor for placental abruption found in our study was polyhydramnios, which was described in previous studies as a prominent risk factor for placental abruption [14,16]. The association between polyhydramnios and placental abruption may be explained mechanically; an acute reduction in the uterine volume and intrauterine pressure can lead to a disruption of the utero-placental interface and eventually to placental abruption [16].
Inadequate prenatal care was found to be an independent risk factor for placental abruption in the current study, an association that has been recognized in previous studies [27]. While in industrialized countries, more than 95% of women undergo at least one prenatal consultation during pregnancy, rates of inadequate prenatal care are much higher in developing countries [28]. Indeed, adequate prenatal care has been shown to decrease vulnerabilities to adverse perinatal outcomes and childbirth-related death and disabilities [27]. As adequate prenatal care enables to treat pre-existing medical conditions, pregnancy complications associated with placental abruption such as chronic hypertension and pre-eclampsia could be detected early and managed appropriately.
Delivery of an LGA neonate was found to be a protective factor from abruption in all three time-periods. Hypertensive disorders, reflecting chronic processes associated with vascular dysfunction and ischemic placental disease, which were found to be in strong association with placental abruption [4] are associate with intra uterine growth restriction and delivery of a small-for-gestational age infant [29]. This may explain the protective association that was found between delivery of an LGA neonate and placental abruption.
Our study's main strength is its utilization of a large cohort, enabling exploration of long-term trends in risk factors for placental abruption. Through the study spanned over a long period (more than 2 decades), we were able to determine the independent risk factors in each separate time-period and to understand the trends of change during the entire span of the study. However, our study's main limitation is its retrospective design, which suggests association rather than causation. The fact that there is not a uniform diagnostic definition of placental abruption that is widely accepted likely caused a lack of accuracy in the incidence rates of placental abruption. Lack of information regarding other known risk factors for placental abruption such as substance abuse or previous placental abruption narrow our understanding of the full spectrum of risk factors for abruption. The observed changes might be due to many factors, including changes in registration. In addition, although the study population was 295,946, placental abruption is a rare complication, and therefore, the study outcomes can be limited by the relatively small sample size.
Another limitation is lack of information regarding other known risk factors such as nulliparity, SGA and rates of different gestational ages of preterm births (< 37, < 34, < 32 weeks of gestation). Analyses of those factors would have improved our understanding of the progression of placental abruption and increase our ability to prevent it.
In conclusion, in our study, we found various independent risk factors for placental abruption. Although several risk factors are known, there is still more to reveal about the strength of the association and the varying trends. In addition, a special attention should be given to women with these risk factors to prevent this morbid complication. Even though placental abruption is relatively rare, the consequences may be severe and further research regarding its risk factors is required. There may be room in future studies to choose a more specific study or control group. Nevertheless, our data represent population-based non-selective data which represents the real world more accurately.
Author contributions ES: Manuscript writing, data collection and management and manuscript editing. YB: data collection and management. RR: data collection and management. AYW: project development, data collection and management, and manuscript editing. GPG: project development, data collection and management, and manuscript editing.
Funding The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.