Background
Adverse neonatal outcomes in multiple pregnancies have been documented extensively, in particular those associated with the increased risk of preterm birth. Paradoxically, much less is known about adverse maternal events. The combined risk of severe acute maternal morbidity in multiple pregnancies has not been documented previously in any nationwide prospective study.
Objective
The objective of the study was to assess the risk of severe acute maternal morbidity in multiple pregnancies in a high-income European country and identify possible risk indicators.
Study Design
In a population-based cohort study including all 98 hospitals with a maternity unit in The Netherlands, pregnant women with severe acute maternal morbidity were included in the period Aug. 1, 2004, until Aug. 1, 2006. We calculated the incidence of severe acute maternal morbidity in multiple pregnancies in The Netherlands using The Netherlands Perinatal Registry. Relative risks (RR) of severe acute maternal morbidity in multiple pregnancies compared with singletons were calculated. To identify possible risk indicators, we also compared age, parity, method of conception, onset of labor, and mode of delivery for multiple pregnancies using The Netherlands Perinatal Registry as reference.
Results
A total of 2552 cases of severe acute maternal morbidity were reported during the 2 year study period. Among 202 multiple pregnancies (8.0%), there were 197 twins (7.8%) and 5 triplets (0.2%). The overall incidence of severe acute maternal morbidity was 7.0 per 1000 deliveries and 6.5 and 28.0 per 1000 for singletons and multiple pregnancies, respectively. The relative risk of severe acute maternal morbidity compared with singleton pregnancies was 4.3 (95% confidence interval [CI], 3.7–5.0) and increased to 6.2 (95% CI 2.5–15.3) in triplet pregnancies. Risk indicators for developing severe acute maternal morbidity in women with multiple pregnancies were age of ≥ 40 years, (RR, 2.5 95% CI, 1.4–4.3), nulliparity (RR, 1.8, 95% CI, 1.4–2.4), use of assisted reproductive techniques (RR, 1.9, 95% CI, 1.4–2.5), and nonspontaneous onset of delivery (RR, 1.6, 95% CI, 1.2–2.1). No significant difference was found between mono- and dichorionic twins (RR, 0.8, 95% CI, 0.6–1.2).
Conclusion
Women with multiple pregnancies in The Netherlands have a more than 4 times elevated risk of sustaining severe acute maternal morbidity as compared with singletons.
The incidence of multiple pregnancies is increasing in many parts of the world. In The Netherlands, multiple pregnancy rates increased from 10.8 to 16.2 per 1000 births between 1980 and 2011. This result is among other factors attributed to delayed child-bearing and more frequent use of assisted reproductive techniques (ART).
Adverse neonatal outcomes in multiple pregnancies have been documented extensively, in particular those associated with the increased risk of preterm birth. Paradoxically, much less is known about adverse maternal events, although an increased incidence has been described for preeclampsia, anemia, nutritional deficiencies, cesarean delivery, and postpartum hemorrhage.
To our knowledge, the combined risk of severe acute maternal morbidity (SAMM) in multiple pregnancies has not been documented previously in any nationwide prospective study. Uncertainties with regard to the risk of SAMM and the possible effects of chorionicity on maternal morbidity are 2 important barriers to effective communication with women. The aim of this study was to assess the incidence of SAMM in multiple gestations in The Netherlands as well as risk indicators.
Materials and Methods
This study formed part of a 2 year nationwide prospective cohort study to assess SAMM during pregnancy, delivery, and puerperium in The Netherlands, called the LEMMoN study. Pregnant women were included from all 98 hospitals with a maternity unit, in the period Aug. 1, 2004, until Aug. 1, 2006. These 98 facilities comprised 8 tertiary care hospitals, 35 nonacademic teaching hospitals, and 55 general hospitals. Detailed information about the study and data collection was described previously.
Inclusion criteria for SAMM were categorized into 5 groups: intensive care unit (ICU) admission, uterine rupture, eclampsia, major obstetric hemorrhage (MOH, defined as ≥ 4 units of packed cells, peripartum hysterectomy or arterial embolization), and a miscellaneous group, containing rare cases of SAMM according to the treating obstetrician, which could not be included in any of the other groups.
Women who developed HELLP (hemolysis, elevated liver enzymes, and low platelet count) syndrome accompanied with liver hematoma or rupture were also included in the eclampsia category. It was possible for women to be included into more than 1 group of SAMM, but each woman was counted once in the overall analysis.
The management of multiple pregnancies and severe maternal complications in The Netherlands is guided through national (open access) guidelines. Maternal mortality cases were reported to the national Maternal Mortality Committee of the Dutch Society of Obstetrics and Gynecology by the attending obstetrician and added to our database. We excluded those cases in which the number of fetuses was not known.
First, we calculated the nationwide incidence of SAMM in multiple pregnancies using the number of multiple pregnancies in The Netherlands during the study period obtained from The Netherlands Perinatal Registry (PRN). The PRN is a national registration system specialized to monitor the quality of obstetric health care and gives insight in perinatal care process and outcome. For each SAMM inclusion group (ICU admission, uterine rupture, eclampsia, MOH, and maternal mortality), we also compared nationwide incidence between singleton and multiple pregnancies. Because of the heterogeneity of the miscellaneous group, no risk analysis was performed for this group.
Second, to identify possible risk indicators, data from the LEMMoN study were compared with data extracted from PRN. The following parameters were used for this comparison: age, parity, method of conception, onset of labor, and mode of delivery. All national reference data from PRN were corrected for the LEMMoN study period.
In the analysis of risk indicators, women with missing information were added to the same category as was usually done in the PRN reference database. For example, in the PRN database women with missing data for parity (n = 1) and method of conception (n = 20) were included into the nulliparous and spontaneous groups, respectively, because these are the most likely outcomes.
To analyze the possible effect of chorionicity, we used Weinberg’s differential rule (rate of dichorionic twins = 2 times the unlike-sex twins) to estimate the rate of mono- and dichorionic twins in The Netherlands. ART was defined as in vitro fertilization (IVF) or intracytoplasmic sperm injection (ICSI). The numbers for mode of delivery were calculated by counting 1 method for each fetus. Because of the absence of individual case-linked data from the PRN, a multivariable analysis and therefore adjustment for possible confounders was impossible.
Finally, we compared multiple and singleton pregnancies within our (LEMMoN study) cohort of women with SAMM. Compared characteristics included the following variables: age, body mass index, socioeconomic status (SES), parity, ethnicity, method of conception, induction of labor, mode of delivery, gestational age, birthweight, blood loss, number of packed cells, and hospital admission. We also analyzed the distribution of indications for ICU admission and cause of MOH within the 2 main inclusion groups of the SAMM cohort.
Statistical analysis
Relative risks (RRs) with 95% confidence intervals (CIs) were calculated where appropriate. Reference groups were depicted and are shown in the legend below the tables. For example, age was compared < 25 years vs ≥ 25 years, ≥ 35 years vs < 35 years and ≥ 40 years vs < 40 years. Differences in characteristics between women with singleton and multiple pregnancies were tested with a χ 2 test or independent Student t test. A statistical analysis was performed using SPSS statistics, version 20.0 (SPSS, Chicago, IL).
Ethical statement
This study was exempt from institutional review board approval because we used only anonymous data from the LEMMoN study and the PRN that cannot be related to any individual. The LEMMoN study was centrally approved by the Medical Ethics Committee of Leiden University Medical Center (P04-020; March 8, 2004).
Results
During the study period, there were 365,594 deliveries in The Netherlands: 358,369 singleton (98.0%) and 7225 multiple pregnancies (2.0%). In that period 2274 of 2352 monthly notification cards (98 hospitals, 24 months) (97%) were returned. A total of 2552 cases with SAMM were reported. Twelve cases (0.5%) were excluded because of an unknown number of fetuses. Of the remaining 2540 cases, there were 2338 singleton (92.0%) and 202 multiple pregnancies (8.0%), of which there were 197 twins (7.8%) and 5 triplets (0.2%). There were no higher-order pregnancies. Of the 202 women with multiple pregnancies, 153 (75.9%) were included with MOH, 70 (34.5%) with ICU admission, 21 (10.3%) with eclampsia, 18 (8.9%) in the miscellaneous group, and 3 (1.5%) with uterine rupture ( Supplementary Figure ).
The overall incidence of SAMM was 7.0 per 1000 deliveries, 6.5 per 1000 for singletons and 28.0 per 1000 for multiple pregnancies. The relative risk was 4.3 (95% CI, 3.7–5.0) for multiple pregnancies compared with singleton pregnancies (ie, 4.3 [95% CI, 3.7–4.9] for twins and 6.2 [95% CI, 2.5–15.3] for triplets) ( Table 1 ). The comparison of the incidence of the 4 SAMM conditions (MOH, ICU admission, eclampsia, uterine rupture) and maternal mortality for multiple and singleton pregnancies are shown in Table 2 . Compared with singletons, the occurrence of SAMM was higher in each group for multiple pregnancies, except for uterine rupture. The case fatality rate of SAMM in multiple pregnancies was 4 of 202 (2.0%) and 44 of 2338 (1.9%) for singletons.
| SAMM | The Netherlands | RR (95% CI) | |
|---|---|---|---|
| Total deliveries | 2540 | 365,594 | |
| Singleton | 2338 (92.0) | 358,369 (98.0) | Reference |
| Multiple | 202 (8.0) | 7225 (2.0) | 4.3 (3.7–5.0) |
| Twin | 197 (7.8) | 7102 (1.94) | 4.3 (3.7–4.9) |
| Triplet | 5 (0.2) | 123 (0.03) | 6.2 (2.5–15.3) |
| The Netherlands | Multiple | Singleton | RR (95% CI) |
|---|---|---|---|
| n = 7225 | n = 358,369 | ||
| ICU admission | 70 (1.0) | 771 (0.2) | 4.5 (3.5–5.8) |
| Uterine rupture | 3 (0.04) | 215 (0.1) | 0.7 (0.2–2.2) |
| Eclampsia | 21 (0.3) | 217 (0.1) | 4.8 (3.1–7.5) |
| MOH | 153 (2.1) | 1449 (0.4) | 5.2 (4.4–6.2) |
| Maternal mortality | 4 (0.1) | 44 (0.01) | 4.5 (1.6–12.5) |
The risk for a woman with multiple pregnancy in The Netherlands to develop SAMM as compared with the general Dutch pregnant population was elevated for women aged 40 years or older (RR, 2.5, 95% CI, 1.4–4.3), nulliparous women (RR, 1.8, 95% CI, 1.4–2.4), women who conceived through ART (RR, 2.0, 95% CI, 1.4–2.7), and women who did not deliver spontaneously (induced labor or elective cesarean delivery) (RR, 1.6, 95% CI, 1.2–2.1). No significant difference in SAMM was found between mono- and dichorionic twins (RR, 0.8, 95% CI, 0.6–1.2). The rate of cesarean delivery compared with women with multiple pregnancies in the general Dutch population (51.1% vs 37.0%) indicates that cesarean delivery is also associated with SAMM (RR, 1.8, 95% CI, 1.5–2.2) in multiple pregnancies ( Table 3 ).
| Multiple | SAMM (n = 202) | The Netherlands (n = 7225) a | RR (95% CI) |
|---|---|---|---|
| Age, y | |||
| < 25 | 7 (3.5) | 481 (6.7) | 0.5 (0.2–1.1) b |
| ≥ 35 | 64 (31.8) | 1958 (27.1) | 1.3 (0.9–1.7) c |
| ≥ 40 | 15 (7.5) | 226 (3.1) | 2.5 (1.4–4.3) d |
| Parity | |||
| 0 | 128 (63.4) | 3533 (48.9) | 1.8 (1.4–2.4) e |
| 1 | 51 (25.2) | 2553 (35.3) | |
| ≥ 2 | 23 (11.4) | 1140 (15.8) | |
| Chorionicity | |||
| Monochorionic | 49 (24.9) f | 2034 (30.0) f | 0.8 (0.6–1.2) g |
| Dichorionic | 138 (70.1) f | 4754 (70.0) f | |
| Conception | |||
| Spontaneous | 136 (67.3) | 5233 (72.4) | |
| IVF/ICSI | 52 (25.7) | 1018 (14.1) | 2.0 (1.4–2.7) h |
| Other i | 14 (6.9) | 974 (13.5) | |
| Onset of labor | |||
| Spontaneous | 78 (38.6) | 2030 (28.1) | |
| Induced/CD | 124 (61.4) | 5195 (71.9) | 1.6 (1.2–2.1) j |
| Deliveries | n = 409 | n = 13,843 | |
| Mode of delivery | |||
| Vaginal | 200 (48.9) | 8729 (63.0) | |
| Spontaneous | 157 (38.4) | 7429 (53.7) | |
| VE/forceps | 34 (8.3) | 984 (7.1) | |
| Breech extraction | 9 (2.2) | 316 (2.3) | |
| CD | 209 (51.1) | 5114 (37.0) | 1.8 (1.5–2.2) k |
| Elective | 84 (40.2) | 2778 (53.3) | |
| Emergency | 125 (59.8) | 2336 (45.7) | 1.7 (1.3–2.3) l |
a National reference values from the PRN (exact period)
b Age < 25 years vs ≥ 25 years
f Percentage of twins included in relative risk calculation for chorionicity
g Monochorionic vs dichorionic
i Intrauterine insemination, artificial sperm insemination, ovulation induction. and other reproductive techniques
The comparison of the 202 women with multiple pregnancies to the 2338 singleton pregnancies within our SAMM cohort showed that women with multiple pregnancies had a higher mean age (32.3 vs 31.5 years, P < .05), were more often nulliparous (63.2% vs 49.0%, P < .001), were less likely to have a scarred uterus (7.4 vs 19.8%, P < .001), and had more often conceived through ART (36.3% vs 5.1%, P < .001), and labor had more often been induced (56.9% vs 46.3%, P < .001). They also had a lower gestational age at delivery (34.9 vs 37.8 weeks, P < .001), were less likely to deliver spontaneously (32.7% vs 44.7%, P < .01), had more blood loss (2673 vs 2262 ml, P < .05), received a higher number of packed cells (5.4 vs 4.3 units, P < .01), were hospitalized longer (12.9 vs 8.4 days, P < .001), and their neonates had a lower mean birthweight (2306 vs 3104 g, P < .001) as compared with singleton pregnancies ( Table 4 ).
| Characteristic | Multiple (n = 202) | Singleton (n = 2338) | P value |
|---|---|---|---|
| Mean age, y | |||
| Unknown (n = 26) | 32.3 (4.7) | 31.5 (5.0) | .025 |
| Mean prepregnancy BMI, kg/m 2 | |||
| Unknown (n = 874) | 24.5 (5.1) | 24.8 (5.4) | .484 |
| SES | |||
| Low | 47 (27.0) | 654 (32.0) | .254 |
| Modest | 88 (50.6) | 906 (44.4) | |
| High | 39 (22.4) | 481 (23.6) | |
| Unknown (n = 171) | |||
| Parity | |||
| 0 | 127 (63.2) | 1137 (49.0) | < .001 |
| 1 | 51 (25.4) | 818 (35.2) | |
| 2 | 21 (10.4) | 242 (10.4) | |
| ≥ 3 | 2 (1.0) | 125 (5.40) | |
| Unknown (n = 29) | |||
| Ethnicity | |||
| Native Dutch | 154 (76.6) | 1711 (74.2) | .451 |
| Immigrant | 47 (23.4) | 595 (25.8) | |
| Unknown (n = 5) | |||
| Previous CD | |||
| Unknown (n = 57) | 15 (7.4) | 462 (19.8) | < .001 |
| Method of conception | |||
| Spontaneous | 116 (63.7) | 1865 (94.9) | < .001 |
| IVF/ICSI | 52 (28.6) | 68 (3.5) | |
| Other a | 14 (7.7) | 33 (1.7) | |
| Unknown (n = 404) | |||
| Mean gestational age, wks | |||
| Unknown (n = 166) | 34.9 (4.6) | 37.8 (4.5) | < .001 |
| Start of labor | |||
| Spontaneous | 78 (38.6) | 1240 (53.1) | < .001 |
| Induced | 115 (56.9) | 1082 (46.3) | |
| Elective CD | 9 (4.5) | 15 (0.6) | |
| Unknown (n = 1) | |||
| Mode of delivery b | |||
| Spontaneous c | 157 (38.4) | 1083 (46.3) | .001 |
| VE/forceps | 34 (8.3) | 279 (11.9) | |
| Breech extraction | 9 (2.2) | 11 (0.5) | |
| CD | |||
| Elective | 209 (51.1) | 964 (41.2) | |
| Emergency | 84 (20.7) | 205 (8.8) | |
| Unknown (n = 1) | 125 (30.6) | 759 (32.4) | |
| Mean blood loss, mL | |||
| Unknown (n = 275) | 2673 (2488) | 2262 (2111) | .015 |
| Mean number of packed cells, n | |||
| Unknown (n = 75) | 5.4 (5.1) | 4.3 (5.0) | .003 |
| Mean birthweight, g | |||
| Unknown (n = 233) | 2306 (765) | 3104 (1045) | < .001 |
| Mean hospital stay, d | |||
| Unknown (n = 110) | 12.9 (11.3) | 8.4 (9.4) | < .001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
