Methods
Premier data
We performed a retrospective cohort analysis of data from the Premier database ( www.premierinc.com ; Premier, Inc, Charlotte, NC) to obtain all maternal hospital records for deliveries that occurred from 2006-2012. The data include hospitalizations from in-patient, ambulatory, and emergency admissions in approximately 500 hospitals each year in the United States. These hospitals are chosen to provide a representation of hospitalizations across the United States. The Premier data can be purchased from Premier, Inc. All diagnosis and procedure codes in the Premier data were coded based on the International Classification of Disease, 9th version; the codes used for conditions in this study are listed in the Supplemental Table . We sought and obtained approval from the Institutional Review Board as an exempt protocol from Columbia University Medical Center, NY.
Placental abruption
A diagnosis of placental abruption was based on clinical symptoms that include vaginal bleeding accompanied with severe abdominal pain, uterine tenderness, or tetanic contractions. Severe placental abruption was defined as a delivery with an abruption accompanied by ≥1 of the following maternal, fetal, or neonatal complications. Maternal complications included disseminated intravascular coagulation, hypovolemic shock, blood transfusion, hysterectomy, renal failure, and in-hospital death. Fetal complications included nonreassuring fetal status, intrauterine growth restriction, or fetal death. Neonatal complications included neonatal death, preterm delivery, and small-for-gestational-age (SGA) births. Although the risk of some of the severe maternal morbidities, such as pulmonary edema or cardiomyopathy, are expected to be higher among pregnancies that are complicated by abruption, these conditions are not the typical complications after abruption; therefore, we do not consider these variables in the definition of severe abruption. Abruption cases that did not qualify as being severe were classified as mild abruptions.
Maternal morbidity profile
The primary endpoint was a composite morbidity outcome comprised of serious maternal complications that included pulmonary edema, acute respiratory failure, acute heart failure, acute myocardial infarction, cardiomyopathy, puerperal cerebrovascular disorder, coma, and amniotic fluid embolism. In addition, we also examined the associations between abruption and each of these serious maternal complications.
Clinical characteristics
We examined the rates of mild and severe abruption across patient characteristics. Maternal sociodemographic and behavioral characteristics included year of delivery (2006-2012), maternal age, single marital status, insurance status, and tobacco, drug, or alcohol use. Maternal comorbidities included hypertensive diseases (chronic hypertension, gestational hypertension, or preeclampsia/eclampsia), chronic renal disease, asthma, and congenital cardiac disease. Intrapartum and labor characteristics included premature rupture of membranes (at preterm or term gestations), anemia, intrapartum fever, polyhydramnios, oligohydramnios, and chorioamnionitis. SGA was used as a proxy for intrauterine growth restriction.
Statistical analysis
Two sets of log-linear regression models (with a Poisson distribution and a log-link function) were fit: the first model was to evaluate the maternal characteristics that are associated with mild and severe placental abruption; the second model was to estimate the association of serious maternal complications (morbidity profile) that are associated with births with mild and severe abruptions compared with births with no abruption and to compare serious maternal complications between severe vs mild (reference) abruptions. For evaluating risk factors for mild and severe abruptions, we first estimated the unadjusted rate ratio (RR) and 95% confidence interval (CI). From this analysis, we chose risk factors that had RRs either >1.2 or <0.8 for mild and severe abruption; risk factors that met this criterion were entered in the final multivariable log linear Poisson regression models from which we evaluated the associations.
RRs and 95% CIs were calculated for the composite serious maternal morbidity profile and for each severe maternal outcome individually. In this analysis, we adjusted for all maternal characteristics as potential confounding factors. All analyses were weighted based on the weights provided in Premier to generate national estimates.
Cohort composition
From 28,504,661 (weighted) singleton deliveries that were identified in the Perspectives database, records identified as male (n = 1308; unweighted, 236), twins and higher-order multiple births (n = 530,065; unweighted, 79,594) and women <15 or >59 years old were sequentially excluded (n = 32,688; unweighted, 5187). We additionally excluded women who received a diagnosis of placenta previa (n = 144,135; unweighted, 21,241). After all exclusions, the analysis cohort was composed of 27,796,465 (3,961,031 unweighted) women.
Results
In this cohort of 27,796,465 singleton births, the prevalence rates of mild and severe abruption were 3.1 and 6.5 per 1000, respectively (overall prevalence rate, 9.6 per 1000). The distribution of clinical characteristics among the 3 groups of nonabruption, mild abruption, and severe abruption is shown in Table 1 . Maternal age ≥35 years old, black race, cigarette smoking status, and the use of drugs or alcohol were associated with increased rates of abruption. Compared with nonabruption births, the prevalence rates of hypertensive disorders were increased among women with mild abruption but were substantially higher among women with severe abruption. Similarly, rates of premature rupture of membranes, polyhydramnios, and oligohydramnios were also relatively higher among severe abruption.
| Variable | No abruption, % | Mild abruption, % | Severe abruption, % |
|---|---|---|---|
| Maternal age, y b | |||
| <20 | 9.0 | 7.1 | 10.6 |
| 20-24 | 23.5 | 22.5 | 24.3 |
| 25-29 | 28.6 | 28.1 | 26.3 |
| 30-34 | 24.3 | 25.0 | 22.4 |
| 35-39 | 12.9 | 14.8 | 14.0 |
| 40-44 | 1.5 | 2.3 | 2.0 |
| ≥45 | 0.1 | 0.2 | 0.3 |
| Maternal race | |||
| White | 51.3 | 53.0 | 47.1 |
| Black | 12.6 | 13.0 | 19.7 |
| Hispanic | 9.9 | 8.6 | 8.3 |
| Other | 26.2 | 25.4 | 24.8 |
| Marital status | |||
| Married | 49.7 | 48.7 | 41.3 |
| Single | 37.6 | 39.3 | 46.6 |
| Unknown | 12.7 | 12.1 | 12.1 |
| Tobacco use | 4.7 | 7.6 | 10.2 |
| Alcohol use | 0.1 | 0.3 | 0.4 |
| Drug use | 0.2 | 0.7 | 1.0 |
| Insurance | |||
| Commercial | 52.1 | 48.7 | 43.7 |
| Medicare | 0.6 | 0.7 | 0.8 |
| Medicaid | 41.2 | 43.6 | 47.8 |
| Uninsured | 2.5 | 3.1 | 3.4 |
| Unknown | 3.6 | 3.9 | 4.2 |
| Hypertension status | |||
| Normotensive | 91.4 | 86.7 | 82.2 |
| Chronic hypertension | 1.8 | 2.5 | 3.2 |
| Gestational hypertension | 3.2 | 4.5 | 3.5 |
| Mild preeclampsia | 1.8 | 2.9 | 3.5 |
| Severe preeclampsia | 1.8 | 3.4 | 7.6 |
| Chronic renal disease | 0.2 | 0.2 | 0.5 |
| Asthma | 2.9 | 3.5 | 3.9 |
| Anemia | 9.8 | 15.1 | 23.9 |
| Congenital cardiac disease | 0.1 | 0.0 | 0.1 |
| Premature rupture of membranes | 3.5 | 4.5 | 8.8 |
| Intrapartum fever | 0.1 | 0.1 | 0.1 |
| Chorioamnionitis | 1.4 | 2.2 | 4.1 |
| Polyhydramnios | 0.8 | 1.0 | 1.2 |
| Oligohydramnios | 2.6 | 2.5 | 3.9 |
a Number (abruption rate per 1000): no abruption, 27,528,415; mild abruption, 86,917 (3.1); severe abruption, 181,133 (6.5)
b Mean ± standard deviation: no abruption, 27.7 ± 6.0; mild abruption, 28.3 ± 6.1; severe abruption, 27.7 ± 6.4.
The associations between the clinical characteristics and mild and severe placental abruption are shown in Table 2 . Several differences were found between mild vs severe abruption. For instance, compared with women 25-29 years (reference), maternal age ≤45 years showed stronger associations with mild abruption, whereas the risk among women ≥45 years old was higher among women with severe abruption. RRs were higher for severe rather than mild abruption for black race, single marital status, and tobacco use. The risk of severe abruption was substantially higher than mild abruption in relation to chronic hypertension (RR, 1.64 vs 1.35), mild preeclampsia (RR, 2.06 vs 1.69), and severe preeclampsia (RR, 4.21 vs 2.00). In contrast, the RRs of mild abruption were higher compared with severe abruption among women with gestational hypertension (RR, 1.47 vs 1.21). The RRs for severe abruption were higher than mild abruption in relation to premature rupture of membranes, anemia, polyhydramnios, oligohydramnios, and chorioamnionitis.
| Risk factors | Adjusted rate ratio (95% confidence interval) b | |
|---|---|---|
| Mild abruption | Severe abruption | |
| Maternal age, y | ||
| <20 | 0.70 (0.68, 0.72) | 0.97 (0.95, 0.99) |
| 20-24 | 0.89 (0.87, 0.91) | 0.95 (0.94, 0.97) |
| 25-29 | 1.00 (Reference) | 1.00 (Reference) |
| 30-34 | 1.10 (1.08, 1.13) | 1.11 (1.09, 1.12) |
| 35-39 | 1.23 (1.21, 1.26) | 1.28 (1.26, 1.30) |
| 40-44 | 1.58 (1.51, 1.65) | 1.47 (1.42, 1.52) |
| ≥45 | 0.70 (0.68, 0.72) | 0.97 (0.95, 0.99) |
| Maternal race | ||
| White | 1.00 (Reference) | 1.00 (Reference) |
| Black | 0.92 (0.90, 0.94) | 1.31 (1.29, 1.33) |
| Hispanic | 0.83 (0.81, 0.86) | 0.89 (0.88, 0.91) |
| Other | 0.94 (0.92, 0.96) | 1.01 (1.00, 1.02) |
| Single marital status | 1.06 (1.04, 1.08) | 1.20 (1.19, 1.22) |
| Tobacco use | 1.49 (1.45, 1.53) | 1.90 (1.87, 1.93) |
| Alcohol use | 1.86 (1.63, 2.13) | 1.78 (1.65, 1.92) |
| Drug use | 2.08 (1.91, 2.26) | 2.32 (2.21, 2.44) |
| Insurance | ||
| Commercial | 1.00 (Reference) | 1.00 (Reference) |
| Medicare | 1.14 (1.05, 1.24) | 1.10 (1.04, 1.16) |
| Medicaid | 1.21 (1.19, 1.23) | 1.19 (1.18, 1.21) |
| Uninsured | 1.43 (1.37, 1.49) | 1.56 (1.52, 1.60) |
| Hypertension status | ||
| Normotensive | 1.00 (Reference) | 1.00 (Reference) |
| Chronic hypertension | 1.35 (1.29, 1.41) | 1.64 (1.60, 1.69) |
| Gestational hypertension | 1.47 (1.42, 1.52) | 1.21 (1.18, 1.24) |
| Mild preeclampsia | 1.69 (1.63, 1.77) | 2.06 (2.01, 2.12) |
| Severe preeclampsia | 2.00 (1.92, 2.08) | 4.21 (4.13, 4.29) |
| Chronic renal disease | 0.73 (0.62, 0.86) | 1.35 (1.26, 1.45) |
| Asthma | 1.13 (1.09, 1.17) | 1.04 (1.02, 1.07) |
| Anemia | 1.59 (1.56, 1.63) | 2.45 (2.42, 2.47) |
| Premature rupture of membranes | 1.27 (1.23, 1.32) | 2.52 (2.48, 2.56) |
| Intrapartum fever | 2.13 (1.76, 2.57) | 1.34 (1.16, 1.55) |
| Polyhydramnios | 1.15 (1.07, 1.23) | 1.39 (1.33, 1.45) |
| Oligohydramnios | 0.96 (0.91, 1.00) | 1.45 (1.42, 1.49) |
| Chorioamnionitis | 1.50 (1.43, 1.58) | 2.42 (2.36, 2.48) |
a Associations for all factors listed in the Table 1 were adjusted with the use of the log-linear Poisson regression model
b Rate ratios for mild and severe abruption are each compared with the nonabruption group.
The morbidity profile and the rates of individual maternal complications in mild and severe abruption and the adjusted RRs for these complications are shown in Table 3 . Serious maternal complications occurred in 15.4 per 10,000 for nonabruption and in 33.3 and 141.7 per 10,000 in women for mild and severe abruption. After adjustment for confounders, compared with women without abruption, the RRs for maternal complications were 1.52 (95% CI, 1.35–1.72) in women with mild abruption and 4.29 (95% CI, 4.11–4.47) in women with severe abruption. RRs for many of the individual complications were increased moderately in women with mild abruption but were 2- to 7-fold higher among severe abruptions. In fact, the RR of the composite serious maternal complications in relation to severe abruption was 4.29 (95% CI, 4.11–4.47; Table 3 ). Similarly, the RRs for pulmonary edema (RR, 2.97; 95% CI, 2.68–3.29), acute heart failure (RR, 3.05; 95% CI, 2.78–3.36), and acute respiratory failure (RR, 7.00; 95% CI, 6.62–7.39) were all considerably higher in women with severe abruptions.
| Variable | Nonabruption (n = 27,528,415) | Mild placental abruption (n = 86,917) | Severe placental abruption (n = 181,133) | Severe vs mild abruption | ||
|---|---|---|---|---|---|---|
| Rate | Rate | Adjusted rate ratio (95% confidence interval) | Rate | Adjusted rate ratio (95% confidence interval) | Adjusted rate ratio (95% confidence interval) | |
| Composite maternal outcome | 15.4 | 33.3 | 1.52 (1.35–1.72) | 141.7 | 4.29 (4.11–4.47) | 3.47 (3.05–3.95) |
| Pulmonary edema | 2.8 | 7.2 | 1.60 (1.24–2.08) | 23.4 | 2.97 (2.68–3.29) | 2.40 (1.82–3.17) |
| Puerperal cerebrovascular disorders | 2.9 | 9.8 | 2.46 (1.97–3.08) | 16.5 | 2.72 (2.41–3.07) | 1.20 (0.92–1.55) |
| Acute heart failure | 4.1 | 5.7 | 0.93 (0.69–1.25) | 27.5 | 3.05 (2.78–3.36) | 4.20 (3.08–5.74) |
| Acute myocardial infarction | 0.2 | — | — | 2.7 | 7.56 (5.51–10.38) | — |
| Cardiomyopathy | 3.4 | 7.4 | 1.48 (1.13–1.92) | 15.2 | 2.12 (1.87–2.41) | 1.68 (1.26–2.26) |
| Acute respiratory failure | 5.7 | 13.0 | 1.62 (1.33–1.96) | 88.9 | 7.00 (6.62–7.39) | 5.47 (4.48–6.68) |
| Amniotic fluid embolism | 0.4 | — | — | 5.1 | 10.56 (8.42–13.24) | — |
| Coma | 0.1 | — | — | 1.9 | 7.04 (4.83–10.25) | — |
a Rates are expressed per 10,000
b Associations were adjusted for the factors listed in Table 1 with the use of the log-linear Poisson regression model.
The RRs for serious maternal complications among severe abruption compared with mild abruption was 3.47 (95% CI, 3.05–3.95). The associations were considerably stronger for virtually all maternal complications for severe abruption rather than for mild abruption.
Rates of mild and severe abruption between 2006 and 2012 and the corresponding rates of serious maternal complications in relations to abruption are shown in the Figure . Rates of mild and severe abruption were fairly constant during the study period. Although the maternal complication rate among births with no abruption was stable between 2006 and 2012, the rate of complications for mild abruption dropped between 2006 and 2008 and then leveled off thereafter. In contrast, the rate of serious complications for severe abruption remained fairly stable between 2006 and 2010, and increased sharply thereafter.
Results
In this cohort of 27,796,465 singleton births, the prevalence rates of mild and severe abruption were 3.1 and 6.5 per 1000, respectively (overall prevalence rate, 9.6 per 1000). The distribution of clinical characteristics among the 3 groups of nonabruption, mild abruption, and severe abruption is shown in Table 1 . Maternal age ≥35 years old, black race, cigarette smoking status, and the use of drugs or alcohol were associated with increased rates of abruption. Compared with nonabruption births, the prevalence rates of hypertensive disorders were increased among women with mild abruption but were substantially higher among women with severe abruption. Similarly, rates of premature rupture of membranes, polyhydramnios, and oligohydramnios were also relatively higher among severe abruption.
