Cleary-Goldman, Jane MD*; Morgan, Maria A. PhD†; Robinson, Julian N. MD‡; D'Alton, Mary E. MD*; Schulkin, Jay PhD†
In 1961, Kurt Benirschke1 pleaded with obstetricians to determine the chorionicity of all twin placentas at delivery. In 1993, a similar petition was issued by Fisk and Bryan,2 with the modification that the diagnosis be established in utero by ultrasound examination. The goal was to emphasize the significance of monochorionicty in obstetric management.
Since these publications, there has been a significant increase in the incidence of multiple gestation secondary to assisted reproductive technology and other factors.3 Perinatal mortality is 6 times higher in twins compared with singletons.4 Preterm birth, growth restriction and discordance, and complications secondary to monochorionicity predispose these pregnancies to adverse outcomes.5 In 2001, 1.6% of singletons versus 11.8% of twins were delivered before 32 weeks of gestation.6 For both triplets and quadruplets, the risk of delivery before 28 weeks of gestation may be as high as 14%.7 Furthermore, an association between multiple birth, specifically monochorionicity, and neurological impairment has been established.8 In this study, we assessed current knowledge and practice patterns pertaining to the management of multiple gestation. Identification of knowledge gaps in this area may lead to the promotion of education with the aim of reducing morbidity in this patient population.
MATERIALS AND METHODS
In January 2003, questionnaires were mailed to 1,146 American College of Obstetricians and Gynecologists (ACOG) Fellows and Junior Fellows. Subjects were selected randomly, by computer, from the pool of practicing ACOG physicians from the United States, Puerto Rico, and Canada who had not received a survey within the past 2 years. The questionnaire was developed with medical specialists and pilot-tested. It recorded demographic details, professional experience, knowledge of multiple pregnancy, and an evaluation of training. Institutional review board exemption was obtained from the Columbia University Medical Center. All nonrespondents received a second mailing approximately 4 weeks after the first. A final mailing was sent out approximately 6 weeks later. Questionnaires returned by April 21, 2003, were included in this study.
The majority of questions in the demographic section and all of the questions in the knowledge, practice style, and education sections offered a range of multiple-choice answers. There were questions on demographic information relating to characteristics of the individual and his/her practice. There were questions on availability of ultrasonography and on the use of consultations for the management of multiple gestations. Didactic knowledge, management styles, and professional education regarding multiple pregnancies also were assessed. A copy of the questionnaire is available on request. An addendum contained 2 unrelated pilot questions for possible future studies.
The sample size was selected based on similar studies conducted by ACOG, where reliable effects have been found by using sample sizes of approximately 1,000 recipients, a response rate of approximately 50%, and a comparable proportion of responding obstetricians. Preliminary power analyses indicated the minimum number of responses needed to ensure significant effect sizes was approximately 100.9
The data were de-identified and analyzed by using a personal computer–based software package (SPSS; SPSS Inc, Chicago, IL). Descriptive statistics were computed for the measures used in the analyses, which are reported as mean ± standard error of the mean. The Student t test was used to compare group means of continuous variables. Differences on categorical measures were assessed by using the χ2 test. Group differences on ordinal measures were assessed by using the Mann–Whitney U test. We used the Spearman rho coefficient for correlations that included an ordinal measure. All analyses were tested for significance by using an alpha of .05.
Fifty-one percent (589/1,146) of surveys were returned. There were respondents from every state of the United States except Delaware, North Dakota, and Wyoming, as well as respondents from the District of Columbia, Puerto Rico, Canada, and overseas military installations. Respondents and nonrespondents did not differ by age (P = .232), and men and women did not differ in their response rate (P = .286). Seventy-three percent (430/589) of respondents managed obstetric patients whereas 27% (159/589) did not; 71.5% of men and 75.1% of women returning the survey practiced obstetrics (Table 1).
The analysis of data was limited to those 430 respondents who practiced obstetrics. The mean age of those practicing obstetrics was significantly lower than those not practicing obstetrics (44.46 ± 0.45 years versus 53.94 ± 0.91 years; P < .001). Respondents practicing obstetrics had been in practice for a mean of 12.08 ± 0.43 years and a median of 10 years, with a range of 1 to 36 years. Men had been in practice significantly longer than women (15.28 ± 0.59 years versus 7.96 ± 0.47 years; P < .001), and men were significantly older than women (47.79 ± 0.61 years versus 40.21 ± 0.52 years; P < .001).
The majority of responding obstetricians were in group practice (54.8%), 18.5% were in solo practice, 12.2% were in faculty practice, 11.7% were in a multispecialty group, and 2.8% worked for a health maintenance organization. The mean estimated number of deliveries per obstetrician in the most recent complete calendar year was 143.89 ± 4.5. The mean estimated percent of deliveries that were of multiple gestations was 4.66 ± 0.36%, although the most common estimate was 2%.
Thirty-eight physicians (8.84%) indicated that they were maternal–fetal medicine specialists. The vast majority of maternal–fetal medicine specialists indicated that their training and experience made them well-qualified to manage a variety of types of multiple gestation, ranging from 100% for dichorionic twins to 92.1% for triplets and 81.6% for quadruplets or greater. Of the obstetricians who were not maternal–fetal medicine specialists, 94.5% felt well qualified to manage dichorionic twins. They otherwise ranged from 78.6% for monochorionic diamniotic twins, to 19.3% for monoamniotic twins, to 9.8% for triplets. Regarding seeking subspecialty assistance, 62% of these practitioners would manage a twin pregnancy without any participation from a perinatologist, whereas only 4.9% would manage triplets independently (Table 2). The majority of maternal–fetal medicine specialists felt their residency training was comprehensive (43.2%) or adequate (48.6%) regarding multiples. General obstetrician–gynecologists were more likely to say that their training was adequate (65.4%) rather than comprehensive (21.5%).
Self-ratings of how well general obstetrician–gynecologists felt their training and experience qualified them to manage multiples corresponded with their tendency to seek the input of a maternal–fetal medicine specialist. Regarding the management of dichorionic diamniotic twins, those physicians who felt somewhat qualified were more likely to seek the input of maternal–fetal medicine specialist in any capacity for a twin gestation than were those who felt well qualified (somewhat qualified,73.7%, versus well qualified, 36.0%; P < .005). Regarding management of monochorionic diamniotic twins, a larger proportion of those who felt somewhat qualified sought the input of a maternal–fetal medicine specialist than those who felt well qualified (51.5% versus 34.2%; P < .05). In monochorionic monoamniotic twins, the proportion of physicians seeking any input from a maternal–fetal medicine specialist did not differ significantly by qualification ratings (35.9% versus 35.4%; P = .44). The vast majority of physicians would seek the input of maternal–fetal medicine specialist for the management of triplets (95%) whether they felt well qualified, somewhat qualified, or not qualified.
The management of multiples requires knowledge of issues particular to this type of pregnancy. Physicians were asked 12 knowledge-based questions pertinent to the management in which chorionicity was an element of the question. Four hundred twenty-two physicians answered at least 1 knowledge question correctly. Four physicians answered all 12 questions correctly. The mean number of questions answered correctly was 6.84 ± 0.11 or 57% (median and mode, 7). Test scores correlated negatively with the number of years in practice (P < .001) and positively with the percent of deliveries that were multiples (P < .01). Physicians who answered correctly at least 75% of the knowledge questions reported a significantly higher percent of deliveries that were multiples than did those who scored less than 75% (P < .001). Physicians with higher knowledge scores were marginally but not significantly less likely to seek the input of a maternal–fetal medicine specialist for twins than were those with lower knowledge scores (higher scorers 29.3% [n = 22] versus lower scorers 40.3% [n = 116]; P = .082).
The majority of practitioners knew the basics of chorionicity (Table 3); however, their knowledge on how to use this information was less than anticipated. Fifty-eight percent stated that the optimal time to diagnose chorionicity with ultrasonography was in the first trimester, whereas 34.3% indicated the second trimester. Seventeen percent of respondents indicated that advanced maternal age in patients with dichorionic twins is 33 years of age or older at estimated date of delivery, 67.4% indicated an age of 35 years, and 9.6% indicated an age of 32 years. Thirty-eight percent of physicians indicated that multifetal and selective reduction with potassium chloride is always contraindicated in monochorionic pregnancies, 19.2% thought it was conditional, and 34.2% were unsure. When asked to indicate a percent regarding the risk of neurologic sequelae in a surviving twin after intrauterine fetal demise of a cotwin, 55.4% of respondents were unsure, whereas 17.3% indicated somewhere between 10% and 25%.
Physicians were asked several questions about their practice patterns. The majority (84.7%) reported that an ultrasound machine was readily available in their office, whereas 15.3% referred their patients to another facility for all ultrasound examinations. Forty-eight percent of practitioners performed a first-trimester ultrasound examination in all pregnancies, whereas 51.8% did so only in the presence of a specific indication. Physicians who had an ultrasound machine available in their office were more likely to perform ultrasound examinations on all first-trimester pregnancies than were physicians for whom an ultrasound machine was not readily available (52.7% versus 22.6%; P < .001). Regardless of the availability of an ultrasound machine, 86% of practitioners attempted to establish amnionicity and chorionicity of twins by a first-trimester ultrasnography.
Analysis of the elective interventions for multiple pregnancy management revealed that 97.1% of practitioners do not use elective cerclage for twins or triplets and that 95.2% do not use elective long-term prophylactic tocolysis for either twins or triplets. Few practitioners prescribe home uterine activity monitoring. Approximately 6% use this monitoring for triplets, whereas 4.6% of practitioners prescribe this monitoring for both twins and triplets. Forty-six percent of obstetricians reported that they recommend bed rest for multiple pregnancies: 29% for twins or greater and 17% for triplets or greater. Of those who do prescribe bed rest, 29.4% do so at 24 weeks of gestation or less, 33.3% do so at 28 weeks of gestation, and 100% do so by 34 weeks of gestation. Regarding diet, only 46.6% of respondents recommended that patients with multiples consume 300 more calories per day than patients with singletons, whereas 31.5% suggested 450 to 600 additional calories, and 3% suggested 750 to 900 extra calories. Approximately 19% of practitioners suggested less than 150 additional calories daily.
With regards to delivery management, 69.1% perform breech extractions for the vaginal delivery of a second twin in the breech presentation. Of these practitioners, 41.2% do so at 34 weeks of gestation and greater, 28% at 32 weeks of gestation and greater, 4.2% at 30 weeks of gestation and greater, and 8.7% at 28 weeks of gestation and greater. Eighteen percent of obstetricians perform breech extractions for a breech-presenting second twin at all gestational ages. When analyzing these data by the estimated fetal weight, 35% perform breech extractions at 1,500 g and greater, 27.3% at 2,000 g and greater, and 18.5% do so at 2,500 g and greater. Fifteen percent of practitioners would do so regardless of estimated fetal weight.
The tendency to perform breech extraction for the vaginal delivery of a breech second twin did not differ substantially based on the number of years in practice. Approximately 65% of those in practice for 19 to 36 years reported performing breech extractions, whereas approximately 72% of those in practice between 5 to 10 years also perform breech extraction. Maternal–fetal medicine specialists were more likely to perform breech extractions than general obstetrician–gynecologists (84.2% versus 67.7%; P < .05).
Physicians were asked to indicate how they stay informed about advances in multiple-pregnancy care. Publications from ACOG were selected most frequently (86.5%) as an important source of information by general obstetrician–gynecologists, whereas maternal–fetal medicine specialists (94.6%) were more likely to select journals (Table 4).
In 2003, ACOG and the American Association of Pediatrics published a document on neonatal encephalopathy and cerebral palsy with the intent of increasing understanding of adverse neurologic outcomes. Multiple pregnancy, specifically monochorionicity, was implicated as contributing a disproportionate number of cases of poor neurologic sequelae.8 Concomitantly, Hankins et al10 published the results of a survey exploring obstetricians’ knowledge about neonatal encephalopathy and cerebral palsy, which demonstrated that practicing obstetricians have knowledge gaps in this subject. These 2 publications prompted us to devise this questionnaire assessing knowledge of multiple pregnancy, specifically chorionicity, in a random sample of currently practicing obstetricians.
The majority of respondents considered their training on multiples to be adequate and felt qualified to manage both dichorionic and monochorionic diamniotic twins without a maternal–fetal medicine specialist's input or, to a much lesser extent, with a maternal–fetal medicine specialist's consultation. The performance on the portion of the questionnaire assessing knowledge related to chorionicity was less than anticipated: only 4 physicians answered all 12 questions correctly; the average score was 57% correct. Physicians scored worse the further they were removed from their training. The more experience they had with multiples, the better they scored on the knowledge portion of the survey. Respondents’ knowledge score did not appear to relate to the tendency for general obstetricians to seek the input of a maternal–fetal medicine specialist when managing multiples. However, self-rating of the adequacy of their training appeared related to their tendency to obtain subspecialty input: practitioners who felt somewhat qualified to manage multiples were more likely to seek the input of a maternal–fetal medicine specialist than practitioners who felt well qualified.
Knowledge regarding the impact of chorionicity on management was less than expected. Although the majority of practitioners knew the basic definitions pertaining to chorionicity (Table 3), just more than half knew that the optimal time to diagnose chorionicity via ultrasonography was in the first trimester.11,12 Few realized that invasive prenatal diagnosis should be offered to patients aged 33 years or older with dichorionic twins at estimated date of delivery, the reason being that the second-trimester risk of Down syndrome in a 33-year-old woman with dichorionic twins is the same as that of a woman with a singleton at 35 years of age.13 Only a third understood that multifetal and selective reduction with potassium chloride is contraindicated in monochorionic pregnancies.14 Furthermore, more than half of respondents were unsure of the risk of neurologic sequelae in a surviving twin after intrauterine fetal demise of a cotwin.15
Physician understanding of how chorionicity affects outcomes is important when caring for multiples. Single intrauterine fetal demise and discordances in growth, amniotic fluid volume, and/or anomalies have different connotations in dichorionic than in monochorionic twins. In monochorionic pregnancies, a discordant lethal anomaly leading to single intrauterine fetal demise can have devastating repercussions on the surviving cotwin secondary to placental anastomoses.16 The risk of multicystic encephalomalacia may be greater than 20%.17–19 Delivering a monochorionic twin pregnancy at risk of a single intrauterine fetal demise preterm may be indicated to avoid permanent neurological complications in the uncomplicated cotwin. If the same scenario occurred in a dichorionic twin pregnancy, the main risk to the surviving cotwin would be increased risk for preterm labor and delivery.20
For the most part, responding obstetricians practice in accordance with current physician education materials.21,22 More than 89% of respondents do not use aggressive prophylactic measures, such as elective cerclage and elective long-term tocolytics. The majority of respondents do not rely on home uterine-activity monitoring. In contrast, almost half of respondents routinely prescribe bed rest for multiples, which has not been proven effective for decreasing rates of preterm delivery but may decrease the incidence of low birth weight neonates.21–23 Fewer than half realize that patients with multiples may benefit from consuming an additional 300 calories per day.21,22 Regarding breech extraction for a second twin, two thirds of nonperinatologists perform this procedure, with the majority doing so at estimated fetal weights of 1,500 g or more, which is consistent with current ACOG practice strategies.21,22
The use of first-trimester ultrasonography reflects the current controversy over routine ultrasound examination. Forty-eight percent of responding obstetricians perform an ultrasound exmaination on all pregnancies during the first trimester, whereas 52% perform first-trimester ultrasonography for specific indications, consistent with current ACOG recommendations. According to ACOG, the value of routine screening to promote early diagnosis of multiples is subject to debate.21 Because the knowledge of chorionicity may have an impact on management, it is hoped that all patients with risk factors for multiples, such as those undergoing assisted reproduction technology procedures, receive an early scan. If knowledge of chorionicity is needed and ultrasonography is unable to show chorionicity later in gestation, zygosity studies remain an option.24
Limitations of this study include the fact that the response rate was only 51%, and all respondents could not be included in the analysis because they all did not practice obstetrics. As a result, data were obtained from and our findings are based on only 430 practicing obstetricians. Although we acknowledge that our study may be flawed secondary to a nonresponse bias, we believe that our findings are reliable. The typical response rate in these survey studies is approximately 35–60%. Thus, for this type of study, our response rate was at the higher end of expected participation. Previous similar studies have demonstrated that reliable results can be obtained and the minimum number of responses needed to ensure significant effect sizes is approximately 100.9 Furthermore, the responses were geographically diverse and derived from physicians from different practice types, reflecting the impact of physician location and practice type on management patterns.
The management of multiple pregnancy typifies the concept of medical practice evolving with time. Assisted reproduction technology has dramatically increased the number of multiples, resulting in greater numbers of physicians caring for these patients. Increasing numbers of patients give additional insight into the natural history of these pregnancies. This additional knowledge may influence management guidelines. Given that monochorionicty may be linked to poor neurologic outcome and that physicians may lack knowledge about monochorionicity as well as about neonatal encephalopathy and cerebral palsy, the implications of this study are compelling.8,10 Better physician understanding of the physiology and pathophysiology of multiple gestation may result in an even greater number of physicians caring for patients according to current guidelines that may lead to decreased morbidity in this patient population. This survey revealed that more than 75% of physicians consider ACOG publications an important source of information regarding advances in multiple pregnancy care. Additional educational materials and efforts to promote awareness about the complexity of multiple gestations may be warranted in the hopes of decreasing morbidity in this patient population. In addition, the role of the maternal–fetal medicine specialist in these potentially complicated pregnancies requires further clarification as the majority of general obstetricians manage these patients independently.
1. Benirschke K. Accurate recording of twin placentation: a plea to the obstetrician. Obstet Gynecol 1961;18:334–47.
2. Fisk NM, Bryan E. Routine prenatal determination of chorionicity in multiple gestation: a plea to the obstetrician. Br J Obstet Gynaecol 1993;100:975–7.
3. Luke B, Nugent C, van de Ven C, Martin D, O'Sullivan MJ, Eardly S, et al. The association between maternal factors and perinatal outcomes in triplet pregnancies. Am J Obstet Gynecol 2002;187:752–7.
4. Bajoria R, Kingdom J. The case for routine determination of chorionicity and zygosity in multiple pregnancy [review]. Prenat Diagn 1997;17:1207–25.
5. D'Alton ME, Simpson LL. Syndromes in twins [review]. Semin Perinatol 1995;19:375–86.
6. Martin JA, Hamilton BE, Ventura SJ, Menacker F, Park MM, Sutton PD. Births: final data for 2001. Natl Vital Stat Rep 2002;51:1–102.
7. Devine PC, Malone FD, Athanassiou A, Harvey-Wilkes K, D'Alton ME. Maternal and neonatal outcome of 100 consecutive triplet pregnancies. Am J Perinatol 2001;18:225–35.
8. American College of Obstetricians and Gynecologists and American Academy of Pediatrics. Neonatal encephalopathy and cerebral palsy, defining the pathogenesis and pathophysiology. Washington, DC: ACOG; 2003.
9. Wilkins-Haug L, Hill L, Schmidt L, Holzman GB, Schulkin J. Genetics in obstetricians’ offices: a survey study. Obstet Gynecol 1999;93:642–7.
10. Hankins GD, Erickson K, Zinberg S, Schulkin J. Neonatal encephalopathy and cerebral palsy: a knowledge survey of Fellows of The American College of Obstetricians and Gynecologists. Obstet Gynecol 2003;101:11–7.
11. Benirschke K. The biology of the twinning process: how placentation influences outcome [review]. Semin Perinatol 1995;19:342–50.
12. Finberg HJ. The “twin peak” sign: reliable evidence of dichorionic twinning. J Ultrasound Med 1992;11:571–7.
13. American College of Obstetricians and Gynecologists. Prenatal diagnosis of fetal chromosomal abnormalities. ACOG Practice Bulletin 27. Washington, DC: ACOG; May 2001.
14. Golbus MS, Cunningham N, Goldberg JD, Anderson R, Filly R, Callen P. Selective termination of multiple gestations. Am J Med Genet 1988;31:339–48.
15. Eglowstein M, D'Alton ME. Intrauterine demise in multiple gestation: theory and management. J Matern Fetal Med 1993;2:272–5.
16. Bajoria R, Wigglesworth J, Fisk NM. Angioarchitecture of monochorionic placentas in relation to the twin-twin transfusion syndrome. Am J Obstet Gynecol 1995;172:856–63.
17. Fusi L, Gordon H. Twin pregnancy complicated by single intrauterine death: problems and outcomes with conservative management. Br J Obstet Gynecol 1990;97:511–6.
18. Nicolini U, Poblete A. Single intrauterine death in monochorionic twin pregnancies [review]. Ultrasound Obstet Gynecol 1999;14:297–301.
19. Pharoah PO, Cooke RW. A hypothesis for the aetiology of spastic cerebral palsy: the vanishing twin [review]. Dev Med Child Neurol 1997;39:292–6.
20. Carlson N, Towers C. Multiple gestation complicated by the death of one fetus. Obstet Gynecol 1989;73:685–9.
21. American College of Obstetricians and Gynecologists. Special problems of multiple gestation. ACOG Educational Bulletin 253. Washington, DC: ACOG; 1998.
22. American Academy of Pediatrics and The American College of Obstetricians and Gynecologists. Guidelines for perinatal care. Washington, DC: ACOG; 2002.
23. Crowther CA. Hospitalization and bed rest for multiple pregnancy (Cochrane Review). In: The Cochrane Library, 2001. Oxford: Update Software.
24. Norton ME, D'Alton ME, Bianchi DW. Molecular zygosity studies aid in the management of discordant multiple gestations. J Perinatol 1997;17:202–7.