The occurrence of spontaneous twin pregnancies is variable across the world, ranging around 8–17 per 1,000 births.1 This incidence has been increasing in the last two decades mainly in high- and middle-income countries as a result of use of assisted reproductive technologies.1–4
Anemia; hyperemesis gravidarum; urinary tract infection; gestational diabetes; pulmonary edema; early preeclampsia; eclampsia; hemolysis, elevated liver enzymes, and low platelet count syndrome; preterm labor; placenta previa; premature rupture of membranes; and postpartum hemorrhage, among others, are more frequent in twin pregnancies.2–8 Some risk factors for maternal death in twin pregnancies have been previously reported: preeclampsia and eclampsia, abruptio placentae, cesarean delivery, and postpartum hemorrhage.5
Twin pregnancy is known to be associated with a higher risk of perinatal morbidity and mortality.5,8–10 However, its association with severe maternal morbidity had not yet been properly addressed. The concepts of potentially life-threatening conditions, maternal near miss, and severe maternal outcomes are relatively recent and considered important as an approach for better understanding of maternal mortality. The current World Health Organization (WHO) criteria for these conditions were used for the current study (Figure 1). The few existing related studies present some methodologic limitations mainly resulting from the small number of cases and inappropriate control groups resulting in data with insufficient statistical power to draw definitive conclusions.1–3
This current study aims to evaluate maternal complications including potentially life-threatening conditions, maternal near miss, and maternal death associated with twin pregnancy in the WHO Multicountry Survey on Maternal and Newborn Health. In addition, we sought to identify sociodemographic and obstetric characteristics associated with the occurrence of maternal morbidities among women with twin pregnancies and to identify which potentially life-threatening conditions are associated with twin deliveries in women with severe maternal outcomes (either maternal near miss or maternal death).
MATERIALS AND METHODS
This is a secondary analysis of the WHO Multicountry Survey on Maternal and Newborn Health, a cross-sectional study implemented in 359 institutions in 29 countries to estimate the prevalence of maternal near-miss cases, to evaluate the quality of maternal care, and to explore the use of the near miss concept in perinatal health. Methodologic details of the WHO Multicountry Survey on Maternal and Newborn Health have been previously published.11–13
In brief, each facility had a trained health professional who collected data daily in a manual questionnaire and then transcribed the data into an electronic format. The manual questionnaire was first reviewed by other researchers and pretested on a convenience sample of records and clinical settings, containing all the criteria recommended by the WHO. It was translated to the main language of each participating country by the country coordinator who was also responsible for training each facility coordinator and the study data collectors. A manual of operations containing instructions on the eligibility criteria, identification, sociodemographic and obstetrics characteristics, maternal complications, neonatal complications, and characteristics of deliveries was also available for the training of research staff.
Data were obtained from medical records of women with the following eligibility criteria: all women who gave birth during the data collection period in the participating centers and their newborns, all severe maternal morbidity cases admitted to the participating centers up to 7 days postpartum or postabortion, irrespective of gestational age and the type of delivery, and all cases of maternal death admitted to the participating centers up to 7 days postpartum or postabortion independently of gestational age and type of delivery. The data collected in the manual questionnaire were then entered onto a web-based data management system, which was developed to minimize the data entry errors and facilitate monitoring and quick resolution of queries and missing data. The data coordinating center regularly prepared lists of inconsistencies the facilities needed to answer to update the data system. If the data were unclear or missing during data collection, the facility medical staff were enlisted to clarify or obtain the missing data.11–13 The study was approved by the WHO Ethical Review Committee and by the relevant ethical clearance body in participating countries and institutions. Because the WHO Multicountry Survey on Maternal and Newborn Health was a study of anonymized data extracted from routine medical records (with no contact between data collectors and women), individual consent was not required.
Many procedures were performed to control the quality of the data collected: development of a manual of operations to minimize the need for judgment and interpretation by the data collectors, training workshop, review queries and missing data using specific reports included in the web-based data management system, and intraform validity crosschecks performed in addition to random crosschecks comparing medical records against form and electronic data.
Data collection took place from May 2010 to December 2011. At the end of the data collection period, 314,623 women with complete pregnancy information were included, of whom 23,015 (7.3%) had some potentially life-threatening condition and 2,538 (0.81%) had near-miss maternal morbidity; 486 (0.15%) women died.12 For this analysis, two groups were compared: twin and singleton pregnancies. For defining such groups in the study, from the total number of cases included in the database, the following conditions were excluded: pregnancies resulting in abortion or ectopic pregnancy, pregnancies resulting in neonates weighting less than 500 g, or in the absence of birth weight, those which ended before 22 weeks of gestation or missing data on when and the way the pregnancy ended, final mode of delivery or abortion, and total number of neonates delivered. Then, the cases of potentially life-threatening conditions, maternal near miss, and maternal death were directly identified in the database according to the WHO definition.14
Initially the occurrence of maternal outcomes was evaluated by country and its Human Development Index for twin and singleton pregnancies and then compiled by continent.15 The Human Development Index is an index published by United Nations Development Programme for 187 countries, which has the ability to emphasize the development of a country not only by economic growth, but, for instance, poverty, human security, and gender inequities. It is calculated using some parameters such as life expectancy at birth, years of schooling and expected years of schooling, and gross national income per capita, currently ranging from 0.392 (lowest for Chad) to 0.944 (highest for Norway).15 The countries are then classified into very high Human Development Index (0.800 or greater), high Human Development Index (0.700–0.799), medium Human Development Index (0.5500–0.699), and low Human Development Index (less than 0.550).
Statistical significance of differences between twins and singletons was assessed by χ2 tests, considering each health facility as the primary sampling unit and the country as the stratum. Then, twin pregnancies were compared with singletons with the estimated risk of any maternal complication using deliveries with no complications as reference and estimating the prevalence ratio adjusted by the cluster design effect only plus their correspondent 95% confidence intervals (CIs). Moreover, their correspondent health indicators as recommended by WHO were calculated, including the potentially life-threatening condition ratio per 1,000 live births, maternal near-miss ratio, severe maternal outcome ratio, maternal near miss:maternal death ratio, mortality index, and intrahospital maternal mortality ratio per 100,000 live births.14 Additionally, the maternal severity score and the maternal severity index were also calculated for both groups. They are both predictors of maternal mortality in women with organ dysfunction with high scores suggesting higher severity and higher likelihood of death.16 The between-group differences on these indicators were evaluated by χ2 tests. The majority of these indicators use the number of live births as the denominator. For singleton gestations, the exact number of live births was used. However, for twin pregnancies, an estimate of women with at least one live birth was used with the intent to not underestimate the occurrence of the complication.
Sociodemographic and obstetric characteristics possibly associated with twin deliveries were evaluated by the prevalence ratios adjusted by the cluster design effect and their correspondent 95% CIs. The analyses were performed excluding the cases with missing information considering they were not differently distributed between both groups of comparison (twins and singletons).
Potentially life-threatening conditions can be regarded as early signs of a more severe condition emerging and can therefore be used for screening and surveillance purposes. Thus, the next step of this analysis was to test whether the estimated risk of these conditions was higher for twin deliveries among women who had or did not have a severe maternal outcome. For that, the prevalence ratios adjusted for the cluster design effect were calculated. Additionally, using the isolated and joined criteria for organ dysfunction, the estimated risks of maternal near miss for twin deliveries were provided, using again the prevalence ratios adjusted by the cluster design effect plus their 95% CIs.
Finally, a Poisson multiple regression analysis was performed to identify the most important and significant factors independently associated with a severe maternal outcome. For that a regression model was built considering severe maternal outcomes as the outcome and all other variables as predictors, including the information on pregnancy being twin and singleton. The resultant prevalence ratios were therefore adjusted not only by the effect of the cluster design but also for all other predictors. The results were considered significant when estimated P values were <.05. All statistical procedures were performed using SPPS 16.0 and Stata 13 programs. We reported our results in accordance with the STrengthening the Reporting of OBservational studies in Epidemiology statement.17
Among the 318,534 women initially enrolled in the WHO Multicountry Survey, 312,867 women remained after the exclusion criteria were applied, 4,756 (1.5%) of them with twin pregnancies. Among these, 636 (13.4%) experienced potentially life-threatening conditions and 70 (1.5%) had maternal near miss; 18 (0.4%) died. Among the 308,111 singletons, 19,245 (6.2%) experienced potentially life-threatening conditions and 1,495 (0.5%) had a maternal near miss; 294 (0.1%) died (Fig. 2).
Table 1 shows the occurrence of potentially life-threatening conditions, maternal near miss, and maternal death by countries, region, and Human Development Index. In a few countries, there was no significant difference in the occurrence of severe maternal morbidity between singletons and twin pregnancies; however, the majority of these had relatively small numbers of maternal near misses and maternal death. When the countries were grouped according to their level of Human Development Index, significant differences in the occurrence of severe maternal outcomes for twin pregnancies were observed (P<.001). For instance, for countries with a high Human Development Index, 13.9% of twin pregnancies had potentially life-threatening conditions, 1.5% had a maternal near miss, and 0.1% had maternal death, whereas in singletons, 6.8% had potentially life-threatening conditions, 0.5% a maternal near miss, and less than 0.1% had maternal death. These differences persisted even across regions and continents (P<.001).
The estimated risks of maternal complications are presented in Table 2 and Figure 3. Twin pregnancies, compared with singleton pregnancies, had an approximately twofold increased risk of potentially life-threatening conditions, a threefold increased risk of maternal near miss and severe maternal outcomes, and a fourfold increase in the risk of maternal death. The evaluation of the health indicators potentially life-threatening condition ratio per 1,000 live births, maternal near miss ratio per 1,000 live births, severe maternal outcome ratio per 1,000 live births, and intrahospital maternal mortality ratio per 100,000 live births as recommended by WHO also identified worse results for twin pregnancies. The maternal severity score and the maternal severity index presented essentially the same results for twins and singletons. Some sociodemographic and obstetrics characteristics are shown to be associated with twin deliveries in Table 3: maternal age older 20 years, to have a partner, and multiparity. Higher maternal education appeared as protective factor for the occurrence of twin pregnancies. The elective cesarean delivery rate was more than twice as high in twin pregnancies compared with singletons (25.1% compared with 12.0%).
Table 4 features the analysis of diagnostic criteria for potentially life-threatening conditions, identifying postpartum hemorrhage and chronic hypertension as the criteria more strongly associated in both women with severe maternal outcomes and women without. Other obstetric hemorrhage, preeclampsia, eclampsia, human immunodeficiency virus-positive, severe anemia, malaria or dengue, and coincidental disorders were criteria significantly associated with twin deliveries in women without severe maternal outcomes. The analysis of estimated risks of maternal near miss based on organ dysfunction criteria, presented in Table 5, showed that all organ dysfunctions considered (cardiovascular, respiratory, coagulation or hematologic, uterine or hysterectomy, neurologic, hepatic, renal) were higher in twin pregnancies. Similarly, the occurrence of any organ dysfunction was threefold higher in twin pregnancies.
Finally, multiple regression analysis identified several factors independently associated with severe maternal outcomes (Table 6). As expected, severe maternal conditions including eclampsia, postpartum hemorrhage, severe anemia, ruptured uterus, preeclampsia, sepsis, and heart disease, among others, were highly associated with a severe maternal outcome. The only factor identified as protective for the occurrence of severe maternal outcomes was higher maternal education level. Additionally, twin pregnancy alone did not appear as a factor independently associated with severe maternal outcomes (P=.095).
Our findings suggest that women pregnant with twins have a threefold risk of dying during pregnancy, childbirth, or the first postpartum week. In addition, twin pregnancy is associated with a threefold risk of maternal near miss and a twofold risk of potentially life-threatening conditions. This is the first time these associations are shown, particularly in a very large, multicountry study. Irrespective of Human Development Index and regions, complications were always more frequent among twin pregnancies. The occurrence of any potentially life-threatening conditions, maternal near miss, or maternal death was at least twice as high among them. The lower the Human Development Index, the higher the number of maternal near miss and deaths.
We were unable to identify in MEDLINE any similar studies published after 1999 that used the WHO definitions for severe maternal outcomes among twin pregnancies (search terms “twin pregnancy” and “WHO criteria for maternal morbidity”). This may be the result of the fact that only approximately 6 years passed since they were standardized.14 One previous study used the concept of organ dysfunction and a definition of maternal adverse outcomes that did not capture a large number of critical severe maternal morbidities. Twin pregnancy doubled the risk of complications,9 whereas in our current study, twin pregnancy increased threefold the risk of severe maternal outcomes.
Adverse maternal outcomes showed to be higher in twin pregnancies, reinforcing the importance of special care during pregnancy and childbirth.2,3 We have no explanations why no major differences were found in outcomes for twin pregnancies when comparing countries with different Human Development Indices. Possibly the increased surveillance of multiple pregnancies, a condition considered high risk worldwide, has played a role. The same understanding can be applied for maternal severity score and for maternal severity index, which is a predictor of maternal mortality in women with organ dysfunction, high scores suggesting higher severity. For twin pregnancies, they are slightly higher than for singleton.12,16
The increased incidence of twin rates among women older than 35 years is well reported, resulting from physiologic endogenous ovarian hyperstimulation and higher use of techniques for assisted fertilization.1,18,19 Other factors associated are multiparity and lower socioeconomic conditions.1,18,20 Cesarean delivery is the most frequent mode of delivery for twin pregnancy, despite the lower risks of vaginal delivery when the first twin is in cephalic presentation.2,3,21–23 All these characteristics were also confirmed in the current study.
Obstetric hemorrhage, chronic hypertension, preeclampsia and eclampsia, postpartum hemorrhage, and severe anemia are all maternal complications known to be associated with twin pregnancy.3,8,9,23–25 A study on adverse maternal outcomes in multiple pregnancies found a higher risk for preeclampsia and postpartum hemorrhage, close to that observed in the present study.3 Only postpartum hemorrhage and chronic hypertension appeared associated with twin pregnancy among women with and without severe maternal outcomes. Postpartum hemorrhage is two to four times more common and hypertensive disorders are identified as a complication in 10–20% of twin pregnancies.2,3,5,23,24 Chronic hypertension alone is associated with preterm delivery and cesarean delivery, which may be associated with increased maternal morbidity.2,26
Organ dysfunction associated with maternal near miss is still not yet fully explored for twin pregnancies. The previous WHO study demonstrated that there is a higher occurrence of blood transfusion and admission to the intensive care unit and higher rates of severe maternal outcomes among twin pregnancies.9 This evidence confirms the findings from this study, which identified that the occurrence of any organ dysfunction was three times higher among twins.
There were a few limitations to our study such as limited data on use of assisted reproductive techniques and mainly the lack of information on the chorionicity for the twin pregnancies included, which could be associated with a series of other conditions leading to some adverse outcomes. Multivariate analysis confirmed the factors associated with severe maternal outcomes previously identified, including a protective effect of higher maternal education levels.27
Our findings suggest increased maternal mortality and morbidity associated with twin pregnancy. Twin pregnancies should be regarded as a very high-risk condition, requiring close surveillance and specialized maternal care. Strengthening health services and systems to deliver high-quality care to these women should be considered as part of strategies to end preventable maternal mortality and make every mother and child count.
1. Smits J, Monden C. Twining across the Developing World. PLoS One 2011;6:e25239.
2. Young BC, Wylie BJ. Effects of twin gestation on maternal morbidity. Semin Perinatol 2012;36:162–8.
3. Walker MC, Murphy KE, Pan S, Yang Q, Wen SW. Adverse maternal outcomes in multifetal pregnancies. BJOG 2004;111:1294–6.
4. Piccoli GB, Arduino S, Attini R, Parisi S, Fassio F, Biolcati M, et al.. Multiple pregnancies in CKD patients: an explosive mix. Clin J Am Soc Nephrol 2013;8:41–50.
5. Rao A, Sairam S, Shehata H. Obstetric complications of twin pregnancies. Best Pract Res Clin Obstet Gynaecol 2004;18:557–76.
6. Long PA, Oats JN. Preeclampsia in twin pregnancy—severity and pathogenesis. Aust N Z J Obstet Gynaecol 1987;27:1–5.
7. Buhling KJ, Henrich W, Starr E, Lubke M, Bertram S, Siebert G, et al.. Risk for gestational diabetes and hypertension for women with twin pregnancy compared to singleton pregnancy. Arch Gynecol Obstet 2003;269:33–6.
8. Rizwan N, Abbasi RM, Mughal R. Maternal morbidity and perinatal outcome with twin pregnancy. J Ayub Med Coll Abbottabad 2010;22:105–7.
9. Vogel JP, Torloni MR, Seuc A, Betrán AP, Widmer M, Souza JP, et al.. Maternal and perinatal outcomes of twin pregnancy in 23 low- and middle-income countries. PLoS One 2013;8:e70549.
10. Vogel JP, Holloway E, Cuesta C, Carroli G, Souza JP, Barrett J. Outcomes of non-vertex second twins, following vertex vaginal delivery of first twin: a secondary analysis of the WHO Global Survey on Maternal and Perinatal Health. BMC Pregnancy Childbirth 2014;14:55.
11. Souza JP, Gülmezoglu AM, Carroli G, Lumbiganon P, Qureshi Z; WHOMCS Research Group. The World Health Organization multicountry survey on maternal and newborn health: study protocol. BMC Health Serv Res 2011;11:286.
12. Souza JP, Gülmezoglu AM, Vogel J, Carroli G, Lumbiganon P, Qureshi Z, et al.. Moving beyond essential interventions for reduction of maternal mortality (the WHO Multicountry Survey on Maternal and Newborn Health): a cross-sectional study. Lancet 2013;381:1747–55.
13. Ganchimeg T, Morisaki N, Vogel JP, Cecatti JG, Barrett J, Jayaratne K, et al.. Mode and timing of twin delivery and perinatal outcomes in low- and middle-income countries: a secondary analysis of the WHO Multicountry Survey on Maternal and Newborn Health. BJOG 2014;121(suppl 1):89–100.
14. Say L, Souza JP, Pattinson RC; WHO Working Group on Maternal Mortality and Morbidity Classifications. Maternal near miss—towards a standard tool for monitoring quality of maternal health care. Best Pract Res Clin Obstet Gynaecol 2009;23:287–96.
15. United Nations Development Programme (UNDP). Human development report 2013. The rise of South: human progress in a diverse world. 2013. Avaliable at: http://hdr.undp.org
. Acessed October 14, 2013.
16. Souza JP, Cecatti JG, Haddad SM, Parpinelli MA, Costa ML, Katz L, et al.. The WHO maternal near-miss approach and the maternal severity index model (MSI): tools for assessing the management of severe maternal morbidity. PLoS One 2012;8:e44129.
17. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007;370:1453–7.
18. Bortolus R, Parazzini F, Chatenoud L, Benzi G, Bianchi MM, Marini A. The epidemiology of multiple births. Hum Reprod Update 1999;5:179–87.
19. Ananth CV, Chauhan SP. Epidemiology of twinning in developed countries. Semin Perinatol 2012;36:156–61.
20. Ibrahim I, Oyeyemi A, Obilahj A. Twin pregnancies in the Niger Delta of Nigeria: a four-year review. Int J Womens Health 2012;4:245–9.
21. Barrett JF, Hannah ME, Hutton EK, Willan AR, Allen AC, Armson BA, et al.. A randomized trial of planned cesarean or vaginal delivery for twin pregnancy. N Engl J Med 2013;369:1295–305.
22. Liu AL, Yung WK, Yeung HN, Lai SF, Lam MT, Lai FK, et al.. Factors influencing the mode of delivery and associated pregnancy outcomes for twins: a retrospective cohort study in a public hospital. Hong Kong Med 2012;18:99–107.
23. Nwankwo TO, Aniebue UU, Ezenkwele E, Nwafor MI. Pregnancy outcome and factors affecting vaginal delivery of twin at University of Nigeria Teaching Hospital, Enugu. Niger J Clin Pract 2013;16:490–5.
24. Bangal VB, Patel SM, Khairnar DN. Study of maternal and fetal outcomes in twin gestation at tertiary care teaching hospital. Int J Biomed Adv Res 2012;3:758–62.
25. Qazi G. Obstetric and perinatal outcome of multiple pregnancy. J Coll Physicians Surg Pak 2011;21:142–5.
26. Werder E, Mendola P, Männistö T, O'Loughlin J, Laughon SK. Effect of maternal chronic disease on obstetric complications in twin pregnancies in a United States cohort. Fertil Steril 2013;100:142–9.e1–2.
27. Tunçalp Ö, Souza JP, Hindin MJ, Santos CA, Oliveira TH, Vogel JP, et al.. Education and severe maternal outcomes in developing countries: a multicountry cross-sectional study. BJOG 2014;121(suppl 1):57–65.