A Growing Problem: Maternal Death and Peripartum Complications Are Higher in Women With Grown-Up Congenital Heart Disease

Transcription

1 ORIGINAL ARTICLES: SURGERY: The Annals of Thoracic Surgery CME Program is located online at To take the CME activity related to this article, you must have either an STS member or an individual non-member subscription to the journal. A Growing Problem: Maternal Death and Peripartum Complications Are Higher in Women With Grown-Up Congenital Heart Disease Tara Karamlou, MD, MS, Brian S. Diggs, PhD, Brian W. McCrindle, MD, MPH, and Karl F. Welke, MD, MS Pediatric Cardiothoracic Surgery, Congenital Cardiac Center, Seattle Children s Hospital, Seattle, Washington; Department of Surgery, Oregon Health & Science University, Portland, Oregon; Division of Cardiology, Hospital for Sick Kids, Toronto, Ontario, Canada; and Division of Pediatric Cardiac Surgery, Mary Bridge Children s Hospital, Tacoma, Washington Background. As patients with grown-up congenital heart disease (GUCH) increase, more women with GUCH will become pregnant. Heart surgeons may be involved in maternal GUCH care, yet the prevalence, characteristics, and outcomes for these women are unknown. We determined the national prevalence of GUCH parturients, their diagnostic makeup, and whether they have increased risk of peripartum complications, maternal or fetal death. Methods. We searched the Nationwide Inpatient Sample for women undergoing delivery in the United States between 1998 and 2007, with GUCH patients indicated by a code for congenital cardiac diagnosis complicating pregnancy, Patient and hospital characteristics were compared between women with and without GUCH. National estimates for maternal and fetal mortality, cardiac complications, induction, caesarean or surgically assisted birth, and preterm delivery were derived. Outcomes were compared between women with and without GUCH, and also within diagnostic GUCH subgroups. Results. A total of 39.9 million births occurred, 26,973 (0.07%) of which were GUCH. Mean age was 27 years for both groups. Most common congenital diagnoses included ventricular septal defect (VSD) in 15%, aortic stenosis or insufficiency in 13%, atrial septal defect in 13%, pulmonary stenosis in 4%, and tetralogy of Fallot in 2%. Stillborn delivery was equivalent among groups. Maternal mortality was 18-fold higher in GUCH parturients (0.09%) compared with women without GUCH (0.005%; p < 0.001). Complications were higher for GUCH parturients compared with age-matched women, including cardiac complications (2.3% vs 0.2%), induction (37% vs 33%), caesarean or surgically assisted birth (45% vs 35%), and preterm delivery (10% vs 7%), p < for all. A diagnosis of VSD was associated with the highest risk of maternal death and complications (p < 0.05 for all). More GUCH women delivered at teaching hospitals (58%) compared with women without GUCH (45%; p < 0.001). Conclusions. The GUCH parturients, especially those with VSD, have increased risk of mortality and peripartum complications compared with other age-matched women. Despite these risks, nearly 50% of GUCH patients deliver at nonteaching hospitals. Current national practice patterns for GUCH women are inadequate, and outcomes could be improved by education and proper triage of even relatively simple GUCH lesions such as atrial septal defect and VSD. Further studies that investigate risk-adjusted outcomes in a variety of care settings are necessary to resolve this complex issue. (Ann Thorac Surg 2011;92:2193 9) 2011 by The Society of Thoracic Surgeons We, and others, have shown that the number of patients with grown-up congenital heart disease (GUCH) is increasing [1 3]. The National Heart, Lung and Blood Institute Working Group on Research in Adult Congenital Heart Disease estimated 1 million patients with GUCH in 2005, with a growth rate of 5% Accepted for publication May 24, Presented at the Forty-seventh Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 31 Feb 2, Address correspondence to Dr Karamlou, Seattle Children s Hospital, Congenital Heart Center, Mail Stop G-0035, 4800 Sand Point Way NE, Seattle, WA 98105; tara.karamlou@seattlechildrens.org. per year [2]. It follows that more women with GUCH will become pregnant, and some authors have estimated that the proportion of pregnant women with heart disease who have an underlying congenital cardiac diagnosis will increase from 5% to 80% in the near future [3 9]. Congenital heart surgeons may be involved in maternal GUCH care, yet the prevalence, characteristics, and outcomes for these women are not fully characterized. We determined the national prevalence of GUCH parturients, their diagnostic makeup, and whether they have increased risk of peripartum complications, maternal or fetal death by The Society of Thoracic Surgeons /$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 2194 KARAMLOU ET AL Ann Thorac Surg A GROWING PROBLEM: WOMEN WITH GUCH 2011;92: Material and Methods The Nationwide Inpatient Sample (NIS) is a stratified, cross-sectional database that includes approximately 20% of all non-veterans Affairs hospital discharges in the United States. The NIS is managed under the Health Care Cost and Utilization Project of the Agency for Healthcare Research and Quality [10]. To ensure the representative nature of the database, the NIS is stratified by geographical region, hospital bed size, teaching status, urban versus rural location, and hospital ownership. The NIS database was searched for the individual years 1998 to 2007, selecting all women undergoing delivery. The NIS estimates were crosschecked to US population statistics to verify accuracy [11]. Women with GUCH were then identified by the presence of a congenital cardiac diagnosis code ( ), or by a specific NIS code (648.5) for congenital cardiac diagnosis complicating pregnancy. According to NIS definitions, this code is assigned whenever a diagnosis of congenital heart disease is present in a parturient, rather than indicating that a pregnancy or delivery complication per se occurred. The sensitivity and specificity of this code was cross-checked to determine whether an additional congenital code was listed for that discharge and the specific diagnosis, and was also evaluated for consistency using our previously published GUCH dataset [1]. Cardiac complications include any cardiac event related to anesthesia and delivery itself, as well as cardiac arrhythmias and cardiac arrest occurring anytime during hospital admission. All of our analyses use sampling weights provided within the NIS database to derive national estimates of the number of patients undergoing delivery, and the associated perinatal complications and mortality rates. Patient and hospital characteristics were compared between GUCH and non-guch patients. National estimates were derived for our primary and secondary outcomes. Outcomes were then compared between GUCH and non-guch groups, and also within diagnostic subgroups of GUCH. The primary outcomes were in-hospital maternal and fetal mortality, and our secondary outcomes were preterm delivery (defined as estimated gestational age 37 weeks), induction, caesarean or surgically-assisted delivery, hospital length of stay, and maternal cardiac complications. The C-indices and receiver operating characteristic curves were generated for all logistic regression models to provide insight into model discrimination qualities. The SAS software, version 9.1 (SAS Institute, Cary, NC) was used to fit the models to account for the survey design of the NIS, the potential clustering of outcomes within a hospital, and the decrease in clustering occurring over increasing time intervals. Results Patients We identified 39.9 million births. Of these, 26,973 (0.07%) were among GUCH parturients. Non-GUCH deliveries increased 26% over the study period, and, as expected, these numbers approximated the US national birth rate (Fig 1A). The GUCH deliveries, however, increased by 43% over the study period (Fig 1B). Regarding baseline demographics, mean age in both groups was similar at 27 years of age. The majority of GUCH parturients (57.2%) delivered at urban teaching hospitals, but nearly 10% of GUCH women delivered at rural hospitals. Diagnostic subgroups within the GUCH patients are shown in Table 1. Of patients with a defined congenital diagnosis, ventricular septal defect (VSD) (n 4,152, 15%), aortic valvar pathology (n 3,412, 12.7%), and ostium secundum atrial septal defect (ASD) (n 3,402, 12.6%) predominated. Mortality and Obstetrical Complications Estimated unadjusted maternal mortality was 18-fold higher among GUCH parturients (n 25; 0.09%) compared with non-guch parturients (n 2,119; 0.005%) (p 0.001). Stillborn prevalence was equivalent among GUCH patients (0.8%) compared with non-guch patients (0.7%). Maternal mortality among GUCH parturients was higher in urban teaching hospitals (n 20; 0.12%) compared with nonteaching institutions (n 5; 0.06%), but the difference was not statistically significant. The GUCH women had a significantly higher prevalence of all peripartum complications studied compared with non-guch women, including preterm delivery (10.2% vs 7.2%), induction (35.6% vs 32.5%), surgicallyassisted delivery (11.8% vs 7.9%), and caesarean birth (33.6% vs 27.2%) (p for all). Hospital length of stay was similar for both GUCH and nonguch women and was 2.5 days. A diagnosis of VSD was associated with a significantly higher prevalence of maternal death (n 10; 40%), preterm delivery (n 218; 11.2%), surgicallyassisted birth (n 233; 10.3%), and caesarean section (n 688; 11.2%), than other known diagnostic subgroups (p 0.05 for all). Maternal Cardiac Complications The GUCH women had a significantly higher prevalence of peripartum cardiac complications (n 615; 2.3%) compared with nonguch women (n 75,325; 0.2%) (p 0.00). A diagnosis of ASD was associated with a higher prevalence of cardiac complications than other known subgroups (n 59) but statistical significance was not reached. Cardiac complications among GUCH women were equivalent among teaching hospitals (2.6%) compared with nonteaching hospitals (2.1%). Comment We have shown that the number of GUCH parturients is increasing more rapidly than nonguch women during the last 10 years (43% vs 26%). The GUCH women continue to have higher mortality and a higher prevalence of obstetric and cardiac complications compared with non-guch parturients, and these risks are not mitigated by the current national practice patterns. Although congenital heart surgeons may not be the primary caregivers for GUCH parturients, it is critical that surgeons (as part of a multidisciplinary team of specialists caring for these patients) be informed re-

3 Ann Thorac Surg KARAMLOU ET AL 2011;92: A GROWING PROBLEM: WOMEN WITH GUCH 2195 Fig 1. (A) Histogram showing the estimated US deliveries among women without grown-up congenital heart disease (GUCH) over the study period. Deliveries increased by 26% over the 10-year period, with figures approximating the national birth rates. The vertical axis is deliveries per 1,000 for convenience. (B) Histogram showing the estimated US deliveries among women with GUCH over the study period. Deliveries increased by 43% over the 10-year period in this population. garding several important issues that are highlighted in our paper: (1) the mortality risk incurred by GUCH patients during pregnancy, because it is likely that women undergoing or contemplating repair of congenital heart disease (CHD) (or even parents of adolescent female children) may ask their surgeon about the risk of subsequent pregnancy; (2) risk and prevalence of cardiac complications during pregnancy; (3) whether maternal patients with GUCH should be cared for in a specialized setting; and (4) whether more frequent care by their cardiologist is indicated, and if so, at what time intervals. Moons and colleagues [6] recently queried 47 women with GUCH, using the Leuven knowledge questionnaire for congenital heart disease. These investigators showed that, while the majority of women with GUCH knew that they were at risk for deterioration during pregnancy, only about 50% of respondents were able to recognize the signs or symptoms indicating deterioration. Moons and colleagues demonstrated in a subsequent publication [7] that less than 50% of GUCH women regularly visited a cardiologist during their pregnancy. These data are concerning, because obstetricians, especially those in nonteaching hospital settings (where 42% of GUCH parturients deliver), are unlikely to be more educated regarding the risks of complex CHD on pregnancy than a congenital heart surgeon. These sentiments were echoed by Moons and colleagues [6], who asserted that... obstetricians, general practitioners, and midwives possibly have too little knowledge about congenital heart disease to know why visits with a cardiologist during different gestational states are vital for these patients. We [1, 8],

4 2196 KARAMLOU ET AL Ann Thorac Surg A GROWING PROBLEM: WOMEN WITH GUCH 2011;92: Table 1. Patient and Hospital Characteristics Patients Variable GUCH Non-GUCH p Value a separate complication, pointing out that a large-scale multicenter study would be required for accurate estimation of such serious and rare events. Our study is useful in that we have been able to demonstrate the national prevalence of maternal mortality among GUCH parturients, recognizing that this figure is an underestimation of the true scope of the problem because only those women reaching delivery were included. Additionally, we have previously shown that care of GUCH patients is delivered at a variety of care settings, and that many of these complex patients receive care outside of tertiary or quaternary medical centers [8]. The current study confirms our prior results, showing that nearly 42% of GUCH parturients deliver at rural or nonteaching hospitals, and therefore provides a contemporary snapshot of peripartum outcomes in GUCH women throughout the entire spectrum of clinical care settings. We were surprised that maternal mortality and cardiac and obstetric complications were independent of hospital type, yet considering that we could not adjust for either the acuity or complexity of maternal cardiac disease or the quality of prenatal care, we are reluctant to make a recommendation in this regard. One inference from this study, however, is that current national practice patterns for GUCH women are inadequate, and outcomes could be improved by proper triage of even relatively simple GUCH lesions such as ASD and VSD. Certainly, a follow-up study from a large state database that investigates risk-adjusted outcomes in the spectrum of care settings is necessary to resolve this complex issue. We chose to include GUCH parturients with a broad spectrum of diagnoses, including ASD and PDA, to capture outcomes for the entire population of GUCH women. Prior investigators have either combined all women with left-to-right shunts together, or excluded women with so-called uncomplicated or repaired congenital heart disease, hypothesizing that outcomes in these subgroups should be equivalent to non-guch women [12, 13 17]. However, Yap and colleagues [18] demonstrated, in their study of 243 pregnancies, that women with repaired and unrepaired ASDs had comparable prevalence of cardiac and obstetric complications, both of which were higher than the general population. Similarly, in another recent study, adverse cardiac, obstetric, and neonatal outcomes were highest in women with ASDs compared with other congenital diagnoses [16]. We found that a diagnosis of VSD, a relatively simple form of congenital heart disease, was associated with an increased risk of maternal and peripartum complications. However, from the current dataset we cannot determine whether the lesion was unrepaired or repaired, nor, if done, the timing of repair. It is possible, therefore, that these women had large or longstanding defects with concomitant pulmonary vascular disease. Unfortunately, hemodynamic (shunt-fraction, pulmonary vascular resistance) data and echocardiographic data are not available in the NIS. However, pulmonary hyperten- Age (mean SD, y) GUCH diagnosis VSD 4,152 (15%) AS/AI 3,412 (13%) ASD 3,402 (13%) Pulmonary 2,684 (10%) stenosis ToF 677 (3%) Coarctation of 594 (2%) aorta PDA 589 (2%) TGA 259 (1%) DORV 57 (0.2%) HLHS 30 (0.1%) PAVSD 23 (0.1%) TAPVR 15 (0.1%) Unknown 11,079 (41%) Insurance Private 15,850 (59%) 2,197 a (55%) Medicaid 9,461 (35%) 1,524 a (38%) Self-pay 662 (3%) 140 a (4%) Medicare 216 (1%) 17 a (0.4%) None 34 (0.1%) 10 a (0.2%) Other 696 (3%) 96 a (2%) Hospital type Urban-teaching 15,421 (57%) 1,793 a (45%) Urbannonteaching 8,853 (33%) 1,694 a (42%) Rural (10%) a (13%) a /10,000. AS/AI aortic stenosis or aortic insufficiency; ASD atrial septal defect; DORV double outlet right ventricle; GUCH grown-up congenital heart disease; HLHS hypoplastic left heart syndrome; PAVSD pulmonary atresia with ventricular septal defect; PDA patent ductus arteriosus; TAPVR total anomalous pulmonary venous return; TGA transposition of the great arteries; ToF tetralogy of Fallot; VSD ventricular septal defect. and others [9], have demonstrated the inadequacies of transition of care in the GUCH population, and that these inadequacies lead to important morbidity and mortality. It is therefore essential that opportunities for counseling and dissemination of knowledge regarding pregnancy risks not be lost with these patients, even if that counseling occurs during a preoperative visit many years remote from the realization of these risks. Our results build on those obtained from prior publications [3, 12 16). Though previous reports have elucidated that peripartum outcomes are worse in women with cardiac disease, most of these have been underpowered to analyze the prevalence of rare outcomes, such as maternal death, which is approximately 0.1% in our study. Drenthen and colleagues [14], in their thoughtful literature review of outcomes in 2,491 women with GUCH, were unable to report mortality as

5 Ann Thorac Surg KARAMLOU ET AL 2011;92: A GROWING PROBLEM: WOMEN WITH GUCH 2197 sion has previously been shown to be a strong predictor for adverse maternal cardiac events, with an odds ratio of 21.8 [15]. Ford and colleagues [15] demonstrated that patients with shunts had the highest incidence of preterm birth and small for gestational age infants. However, because these authors did not further characterize the shunt lesions, comparative analysis to our study is impossible. Aggarwal and colleagues [3] and Presbitero and colleagues [19] showed that cyanotic congenital heart disease has a worse prognosis than acyanotic heart disease. Pathophysiologic mechanisms for increased neonatal complications in the presence of maternal cyanosis are well defined [8 10], but the reasons for increased maternal complications in this group (exclusive of Eisenmenger syndrome) are less understood. Presbitero and colleagues hypothesized that perhaps polycythemia could lead to an increased incidence of thrombotic complications. Siu and colleagues [13] developed the Cardiac Disease in Pregnancy risk (CARPREG) index in 2001 from a prospective cohort of Canadian women with congenital and acquired heart disease. These investigators developed a multivariable model in which prior cardiac events or arrhythmia, poor functional class or cyanosis, left heart obstruction, or left ventricular systolic dysfunction were predictive of cardiac complications. The risk of a maternal cardiac event increased from 0% with none of the aforementioned predictors to 75% in the presence of more than one predictor. However, the CARPREG index was not specific to women with GUCH and also excluded any pregnancies that ended in miscarriage before 20 weeks. Khairy and colleagues [12] more recently calibrated the CARPREG index specific to GUCH parturients, retrospectively examining 53 women with GUCH. Model fit improved significantly with the inclusion of two additional risk factors, decreased subpulmonary ejection fraction and pulmonary regurgitation. However, the absence of maternal deaths in this study limited the model outcomes to cardiac and neonatal complications. Limitations Our study is limited by the retrospective and administrative nature of the NIS. Though the large sample size mitigates the influence of coding errors, we cannot exclude informative bias. Additionally, diagnosis codes in the NIS do not differentiate between GUCH parturients with repaired versus unrepaired lesions. Parity status of the mother, another potentially important confounder could not be abstracted as the NIS does not allow linkage of patient data across multiple hospital admissions. Most importantly, this study considerably underestimates the true influence of congenital heart disease in pregnancy for two important reasons. First, we studied only those GUCH women who presented for delivery, and therefore excluded all women whose heart disease precluded successful conception or continuation of gestation, either voluntarily or otherwise. Second, we were unable to capture important prenatal events (such as hospitalizations, procedures, etc) that may have negatively impacted GUCH parturients. There was also an important prevalence of unknown congenital diagnosis which limited our ability to quantify risk among diagnostic subgroups of GUCH. Finally, coding of delivery method as a complication is confounded because we could not ascertain whether a caesarean section or surgically assisted birth was related to the underlying cardiac disease or scheduled for obstetrical reasons. However, because both caesarean section and surgically assisted delivery are associated with substantially increased risk to both the mother (the UK National Health Service reported that maternal death rates were threefold higher for caesarean section than for vaginal delivery [20]) and the fetus, we feel that including these as complications is justified. Additional incremental risks of bowel obstruction from adhesions, placenta accreta, etc, associated with caesarean section, also must be weighed. Conclusions The national prevalence of GUCH parturients is increasing rapidly. Women with GUCH, especially those with VSD, have increased risk of mortality and peripartum complications compared with other agematched women. Despite these risks, nearly 50% of GUCH patients deliver at nonteaching hospitals. Current national practice patterns for GUCH women are inadequate, and outcomes could be improved by education and proper triage of even relatively simple GUCH lesions such as ASD and VSD. Further studies that investigate risk-adjusted outcomes in a variety of care settings are necessary to resolve this complex issue. References 1. Karamlou T, Diggs BS, Person T, Ungerleider RM, Welke KF. National practice patterns for management of adult congenital heart disease: operation by pediatric heart surgeons decreases in-hospital death. Circulation 2008;118: Williams RG, Pearson GD, Barst RJ, et al. Report on the National, Heart Lung, and Blood Institute Working Group on research in adult congenital heart disease. J Am Coll Cardiol 2006;47: Aggarwal N, Suri V, Kaur H, Chopra S, Rohila M, Vijayvergiya R. Retrospective analysis of outcome of pregnancy in women with congenital heart disease: single-centre experience from North India. Aus N Z J Obstet Gynaecol 2009;49: Somerville J. Near misses and disasters in the treatment of grown-up congenital heart patients. J R Soc Med 1997;90: Colman JM, Siu SC. Pregnancy in adult patients with congenital heart disease. Prog Paediatr Cardiol 2003;17: Moons P, Budts W, Costermans E, Huyghe E, Pieper PG, Drenthen W. Pregnancy-related health behavior of women with congenital heart disease: room for behavioral change interventions. Congenit Heart Dis 2009;4: Moons P, De Volder E, Budts W, De Geest S, Elen J, Waeytens K, Gewillig M. What do adult patients with congenital heart disease know about their disease, treatment, and prevention of complications? A call for structured patient education. Heart 2001;86:74 80.

6 2198 KARAMLOU ET AL Ann Thorac Surg A GROWING PROBLEM: WOMEN WITH GUCH 2011;92: Karamlou T, Diggs BS, Ungerleider RM, Welke KF. Adults or big kids: what is the ideal clinical environment for management of grown-up patients with congenital heart disease? Ann Thorac Surg 2010;90: Clarizia NA, Chahal N, Manlhiot C, Kilburn J, Redington AN, McCrindle BW. Transition to adult health care for adolescents and young adults with congenital heart disease: perspectives of the patient and the health care provider. Can J Cardiol 2009;9:e Agency for Healthcare Research and Quality. HCUP nationwide inpatient sample (NIS): Healthcare Cost and Utilization Project (HCUP): Available at Accessed January 11, Accessed November 12, Khairy P, Ouyang DW, Fernandes SM, Lee-Paritz A, Economy KE, Landzberg MJ. Pregnancy outcomes in women with congenital heart disease. Circulation 2006;113: Siu SC, Serner M, Colman JM, et al. Prospective multicenter study of pregnancy outcomes in women with heart disease. Circulation 2001;104: Drenthen W, Pieper PG, Roos-Hesselink JW, et al. Outcome of pregnancy in women with congenital heart disease. A literature review. J Am Coll Cardiol 2007;49: DISCUSSION DR WINFIELD J. WELLS (Los Angeles, CA): Tara, if I saw that one slide correctly, there were 40% of the patients with no known diagnosis. Why didn t you throw those out? DR KARAMLOU: Well, insofar as we looked at the influence of specific diagnosis on outcomes, in essence we did throw those patients out. We didn t throw them out in terms of being GUCH [grown-up congenital heart disease] parturients because we knew that they had a congenital diagnosis, we just did not know what the particular lesion was. So I think to be as unbiased as we could, looking at the entire population of GUCH parturients, we can t exclude patients who we knew had congenital heart disease but we didn t know the diagnosis. I did not mention in the talk, but the group of unknown diagnosis had a large number of event[s]; indeed, 40% of the deaths occurred in that group. As you can imagine, these were likely the patients with more complex lesions that the data abstractors could not assign into a discrete category. I don t think it would be fair to analyze the data knowing that those patients had congenital heart disease, but to examine outcomes by diagnosis, yes, we excluded them from that analysis. DR WELLS: I just wonder if, in the final manuscript, you might want to separate those out and bring that point out. DR KARAMLOU: I think it s a good point. DR HITENDU H. DAVE (Zurich, Switzerland): I was intrigued about the outcome of parturients with a diagnosis of VSD [ventricular septal defect]. If I understand right, these were patients who had already undergone a VSD repair or was the status of their VSD unknown? DR KARAMLOU: We don t know whether the lesions were repaired or unrepaired, only that they had a diagnosis of VSD. That is an important point, but there have been actually two papers in the literature looking at the influence in pregnancy of repaired and unrepaired defects, and shunt lesions in particular. And to be honest with you, the data were surprising but showed that even women who had repaired congenital heart disease fared worse than women without congenital heart disease. 15. Ford AA, Wylie BJ, Wasmonski CA, Simpson LL. Maternal congenital cardiac disease. Outcomes of pregnancy in a single tertiary care center. Obstet Gynecol 2008;112: Song YB, Park SW, Kim JH, et al. Outcomes of pregnancy in women with congenital heart disease: a single center experience in Korea. J Korean Med Sci 2008;23: Karamlou T, Diggs BS, Ungerleider RM, Welke KF. Adults or big kids: what is the ideal clinical environment for management of grown-up patients with congenital heart disease? Ann Thorac Surg 2010;90: Yap SC, Drenthen W, Meijboom FJ, et al. Comparison of pregnancy outcomes in women with repaired versus unrepaired atrial septal defect. BJOG 2009;116: Presbitero P, Somerville J, Stone S, Aruta E, Speigelhalter D. Rabajoli F. Pregnancy in cyanotic congenital heart disease. Outcome of mother and fetus. Circulation 1994;89: Caesarean section. NHS Direct. Available at: Accessed May 13, DR DAVE: So that s intriguing. But suppose you take the cohort of patients who had a successfully repaired VSD based on modern strategy with regards to timing, etc; considering that we imagine that they are essentially cured, have you any hypothesis, any guess, what could have been the cause? DR KARAMLOU: I think the majority of these patients probably had unrepaired VSD. I mean that s our suspicion from looking at the data. The NIS [Nationwide Inpatient Sample] does not contain physiologic variables in terms of Qp/Qs [pulmonary-tosystemic flow ratio], hemodynamics. We don t obviously have cath or echo data. But my suspicion, and again in the literature, pulmonary hypertension has an odds ratio of about 22 in terms of risk for peripartum events. So our feeling, looking at the data set, is that these women most likely had unrepaired VSD. This is pure speculation, though. DR DAVE: And, I suppose, again there is a restriction to the data of the newborns, but just in case, have you got an idea about the incidence of congenital heart defects in the newborns? DR KARAMLOU: No. Well, I mentioned that as a limitation. All we know from the data set is whether the child, or the baby actually, lived or died, but that s the extent that we can link the neonate to the mother. DR HAROLD LINDBERG (Oslo, Norway): Since preeclampsia is known to be a very significant contributor to maternal death, have you looked into if there are any more frequent episodes of preeclampsia in the GUCH patients? DR KARAMLOU: Yes, that s actually a very good point. We actually are looking at the data set and trying to figure out what the incidence is of other complications that we didn t look at in our original data set. DR JOHN E. MAYER (Boston, MA): Just one quick question that maybe would help get at some of the questions that have been raised. Does the data set include V-codes? DR KARAMLOU: Yes, the data set included V-codes.

7 Ann Thorac Surg KARAMLOU ET AL 2011;92: A GROWING PROBLEM: WOMEN WITH GUCH 2199 DR MAYER: So shouldn t all of those patients who had prior surgery for congenital heart defects have had a V-code for postsurgical status of some sort or another? DR KARAMLOU: No, unfortunately, unless it s specifically coded that they have had previous surgery or not previous surgery. We actually looked into this. This has been actually something we ve been working on for about 2 to 3 years, trying to come up with a way to identify the majority of women and exclude women that didn t have congenital cardiac diagnoses. And unfortunately, I d be happy to go through the data set with you, but we were not able to find a way to any more precisely figure out whether the women were repaired or unrepaired. But, happy to talk to you about it later.