Author + information
- Received April 28, 1999
- Revision received January 17, 2000
- Accepted March 30, 2000
- Published online August 1, 2000.
- Ian C Huggon, MDa,*,
- Andrew C Cook, BSca,
- Nigel C Smeeton, C Stat∗,
- Alan G Magee, MRCPa and
- Gurleen K Sharland, MDa
- ↵*Reprint requests and correspondence: Dr. I. Huggon, Fetal Cardiology, 15th Floor Guy’s Tower, Guy’s Hospital, St. Thomas Street, London SE1 9RT, United Kingdom
We sought to establish the outlook for fetuses diagnosed with atrioventricular septal defect (AVSD) prenatally and its relation to additional cardiac, extracardiac and chromosomal abnormalities.
Prediction of likely outcome of AVSD presenting prenatally is complicated by the wide variation in associated features.
Computerized records from 14,726 pregnancies referred to a fetal cardiology center were reviewed retrospectively. Pathological reports, postnatal records, follow-up inquiries and review of echocardiographic video recordings supplemented analysis of the records for all those with AVSD.
Atrioventricular septal defect was confirmed in 301 fetuses. Eighty-six (39%) of the 218 with known karyotype had trisomy 21, and 21/218 (10%) had other chromosome abnormalities. Right isomerism occurred in 37/301 (12%) fetuses, left isomerism in 62 (20%), mirror image atrial arrangement in 2 (1%), and 200 (67%) had usual arrangement. Atrioventricular septal defect occurred without any other intracardiac abnormality in 155 fetuses (51%). Extracardiac abnormalities and nonkaryotypic syndromes were evident in 40 fetuses (13%, confidence interval [CI] 9.5–17.1%). Uncomplicated cardiac anatomy was significantly associated with the presence of karyotype abnormality (p < 0.0001). Parents opted for termination of pregnancy in 175/298 (58.5%). For the continuing pregnancies, Kaplan-Meier estimates for live birth, survival past the neonatal period and survival to three years were 82% (CI 75.3–88.9%), 55% (CI 46.0%–64.3%) and 38% (CI 27.1–48.6%), respectively. Fetal hydrops and earlier year of diagnosis were independent variables with adverse influence on survival.
Despite some improvements in the outlook for AVSD diagnosed prenatally, the overall prognosis remains considerably poorer than that implied from surgical series. The detection of associated cardiac and extracardiac abnormalities is important in order to give the best indication of the likely outcome when counseling parents.
Atrioventricular septal defect (AVSD) is characterized by a common atrioventricular junction and central septal deficiency. Appropriate counseling of parents after prenatal diagnosis of AVSD requires a realistic estimate of the prognosis for the individual fetus. It is well recognized that the spectrum of congenital heart disease diagnosed prenatally is different from that presenting postnatally. More severe cases of a condition, both in terms of the cardiac lesion itself and of associated extracardiac abnormalities, are over-represented in fetal series, and prognosis derived from postnatal surgical series is not generally applicable to those diagnosed prenatally. This is especially the case with AVSD because it is a particularly heterogeneous condition. We report data from a large series of fetuses with AVSD seen at our institution. We describe the outcome of the group as a whole and also intracardiac and extracardiac features that may affect the prognosis for an individual fetus.
Details of pregnancies referred to the fetal cardiology department at Guy’s Hospital have been collated in a computerized database. Included in the database are details of the mother and the pregnancy, the reason for specialist fetal cardiac referral, diagnosis at initial and any further ultrasound scans and the outcome of the pregnancy, including karyotypic and autopsy or postnatal diagnosis. All pregnancies in which a definitive diagnosis of AVSD was made, either prenatally or subsequently, were identified for further study. Additional information was sought from autopsy reports and postnatal records, telephone inquiries and retrospective review of video recordings of scans.
Only pregnancies in which there was a final diagnosis of AVSD were included in the analyses of survival and of associated anatomical and karyotypic abnormalities of this lesion. The study period, based on the date of initial consultation, was from the beginning of January 1987 to the end of December 1997. The starting date was chosen because pregnancies with a diagnosis of AVSD before this study period were reported in a previous paper from this department (1). That series is contiguous with, and does not overlap, the current series. The study was closed to new follow-up data on May 1, 1998, but follow-up times for survivors were calculated up to the last follow-up available rather than up to the closing date of the study.
Details of the cardiac anatomy were determined from prenatal and postnatal echocardiograms or other postnatal investigations, from surgical findings or at autopsy. Based on a detailed anatomical diagnosis, each heart was classified according to basic features as follows. The atrial arrangement in the fetus was determined from atrial appendage morphology or indirectly from the venous anatomy, as usual atrial arrangement, left isomerism, right isomerism or mirror image atrial arrangement. The relative size of the ventricles was classified as approximately equal in size, right ventricle substantially larger than the left (dominant right) or left ventricle substantially larger than the right (dominant left). The presence or absence of pulmonary stenosis or atresia was determined from autopsy or Doppler findings or was inferred from pulmonary arteries being significantly smaller than the aorta. The arterial connection was classified as concordant, discordant, double outlet right ventricle or single outlet. The heart rate was classified as normal or bradycardia.
The fetal karyotype determined prenatally, postnatally or after termination of pregnancy was recorded whenever a result was available or otherwise recorded as unknown. Extracardiac abnormalities and syndromes that were detected prenatally or only became apparent postnatally were noted. Markers for chromosome abnormality without other clinical significance, such as choroid plexus cysts, were not regarded as extracardiac abnormalities for the purposes of this analysis. Recognized features of visceral heterotaxy, such as asplenia and minor degrees of gut malrotation occurring in the context of isomerism were also excluded for the purposes of this analysis.
The presence of hydrops fetalis was noted. Isolated pericardial effusion and early nuchal edema were specifically excluded from this category.
For comparison of subgroups within the whole population, confidence intervals (CIs) were calculated using the method of proportions (2). Confidence intervals, given in brackets, are the 95% limits. Associations between subgroups were analyzed using the chi-squared test, and trends with time were examined using the Mann-Whitney U test. Survival in continuing pregnancies was subject to Kaplan-Meier analysis, with censoring of surviving individuals at their last known follow-up date and with an additional event-time to describe prenatal spontaneous deaths. Differences in survival between separate subgroups were compared using separate Kaplan-Meier analyses and tested for statistical significance using the log-rank test. Many characteristics were not independent but significantly correlated with each other. Therefore, a multivariate analysis was also performed using Cox’s proportional hazards model (3) with forward entry of variables in order to establish significant independent variables affecting survival. Statistical analyses were performed using Statview for Windows version 4.53 (Abacus Concepts, Inc., Berkeley, California) and Confidence Interval Analysis, Version 1.1 (British Medical Journal). In all statistical tests p < 0.05 was regarded as significant.
A final diagnosis of AVSD was made in 301 of the 14,726 pregnancies referred to the department during the study period. The characteristics of these 301 are summarized in Figure 1. They accounted for 17% (301/1727) of structurally abnormal hearts seen. The principal indication for specialist referral was suspicion of congenital heart disease on a routine scan in 277 pregnancies. Thirty-seven of these also had additional high risk factors for congenital heart disease evident at referral. The reasons for referral in the remaining 24 pregnancies were extracardiac fetal abnormality (12), family history alone (5), arrhythmia alone (3), fetal nuchal edema (3) and drug exposure (1). The mean gestational age at the time of the first or only specialist cardiac scan was 22.3 weeks (standard deviation 4.9, range 15 to 39 weeks).
Karyotype results were available for 218 fetuses. In a further three fetuses, a specific karyotype abnormality (trisomy 18 or XO) was strongly suspected at autopsy on morphological grounds, but postmortem karyotyping failed for technical reasons. A definitive karyotype result was not available in 83 fetuses, either because parents declined the test, because of technical failure or because of failure to trace the results of tests performed in other centers. An abnormal karyotype result was obtained in 107 fetuses and strongly suspected in an additional three. Thus, 107 (49.1%) of the 218 with a known karyotype had abnormal chromosomes (CI 42.4% to 55.7%). The incidence of chromosome abnormality in the series as a whole was at least 107/301 (35.5%, CI 30.1% to 41%), assuming that there were no abnormalities among those with undetermined karyotype. The most common karyotype abnormality was trisomy 21, which accounted for 86 (80.4%) of all 107 abnormal karyotypes identified. Thirteen fetuses (12.1% of abnormal karyotypes) were recognized as having trisomy 18, and four (3.7%) had trisomy 13. The other abnormal karyotypes identified were 48XYY +21, 8q-, unbalanced translocations 14, 21 and unknown fragment to 12 and a 22q11 microdeletion.
Additional cardiac abnormalities
A final diagnosis of usual atrial arrangement was made in 200 (66%, CI 61.1% to 71.8%) of 301 fetuses. Of the remainder, 62 had left atrial isomerism, 37 had right atrial isomerism, and 2 had inverse atrial arrangement. Dominance of the right ventricle was a feature in 61 (20%, CI 15.7% to 24.8%) and dominance of the left ventricle in 11 (4%, CI 1.8% to 6.4%). Concordant ventriculoarterial connection occurred in 221 fetuses (73% CI 68.4% to 78.4%), discordant arterial connection in 6 (1.9% CI 0.74% to 4.3%), single outlet in 2 (0.6% CI 0.08% to 2.4%) and DORV in 72 (24% CI 19.1% to 28.7%) (of which 11/72 had anteriorly positioned aorta). Thirty-nine fetuses (13% CI 9.2% to 6.8%) had evidence of pulmonary atresia or stenosis. There were 155 (52% CI 45.8% to 57.1%) fetuses in whom AVSD was an isolated cardiac defect, in that is occurred with usual atrial arrangement and normal arterial connections in the absence of ventricular dominance or pulmonary obstruction.
Abnormalities of the arterial connection were associated with isomerism. Double outlet right ventricle (DORV) was present in 26/37 (70.3%) with right isomerism and 23/62 (37.1%) with left isomerism but only 23/200 (11.5%) with usual atrial arrangement. A great artery relationship in which the aorta was the more anteriorly positioned of the great arteries (ventriculoarterial discordance, DORV with subpulmonary ventricular septal defect (VSD) or single outlet aorta from right ventricle) was particularly associated with right isomerism. Anterior aorta occurred in 11 (30%) of 37 with right isomerism but only 4/62 (6%) with left isomerism and 4/200 (2%) with usual atrial arrangement.
Data on the size of the ventricular component were not systematically recorded at the time of diagnosis and proved difficult to ascertain retrospectively.
Significant extracardiac abnormalities and syndromes were identified in 40 of 301 cases (13.3%, CI 9.5% to 17.1%). The range of extracardiac abnormalities was wide and included renal, gastrointestinal, pulmonary, neurological and skeletal abnormalities as detailed in Table 1. In 26 pregnancies, prenatal diagnosis of the extracardiac abnormality was complete. Fourteen pregnancies, however, were complicated by extracardiac anomalies not recognized prenatally. The latter included three pregnancies resulting in infants in whom a diagnosis of the CHARGE association was made postnatally (4). Features of this anomaly contributed to the early death of all three infants. Other abnormalities only recognized after birth included tracheoesophageal fistula, choanal atresia and anal atresia. In nine pregnancies, significant extracardiac abnormalities were associated with proven karyotype abnormalities, and karyotype abnormality was strongly suspected in one further fetus in which the culture failed.
Association of karyotype and other abnormalities
There was a statistically significant association between atrial isomerism and normal karyotype (p < 0.0001). Of the 52 fetuses with isomerism and known karyotype, only two, both with left isomerism, had karyotype abnormalities (3.8% CI 0.5% to 13.2%). This compares with 105 karyotype abnormalities among 165 fetuses with usual atrial arrangement and known karyotype (63.6% CI 56.3% to 71.0%). The karyotype abnormalities in left isomerism were a 22q11 microdeletion and translocation of an unidentified fragment to chromosome 12. Fetuses with an isolated defect had a higher incidence of karyotype abnormality (77/123, 62.5%) than those with more complex anatomy (30/95, 31.6%), and this difference was statistically significant (p < 0.0001) (95% CI for difference between proportions 18.4% to 43.7%). The latter relationship was only partly accounted for by the mutual association of isomerism, complex cardiac anatomy and low incidence of karyotypic abnormality, as there remained a significant difference (p = 0.039), even if only those with usual atrial arrangement were considered.
Outcome could not be traced in three pregnancies. Parents opted to continue with 123/298 (41% CI 35.7% to 46.9%) pregnancies and for termination in 175 (59% CI 53.1% to 64.3%). Of those continuing, 104/123 (85% CI 78.2% to 90.9%) pregnancies resulted in a live-born child, and 58 of these children (47% CI 38.3% to 56%) were alive at latest follow-up. The duration of postnatal follow-up was limited in many instances, and, therefore, Kaplan-Meier analysis was performed to give unbiased estimates of survival (Fig. 2). The duration of follow-up and survival was from the time of birth, but an additional event-time was added before birth to include stillbirths and intrauterine deaths. For those with the intention to continue with the pregnancy, the probability of live birth was 82% (CI 75.3% to 88.9%), and 55% (CI 46.0% to 64.3%) survived the neonatal period. By three years of follow-up, only 38% (CI 27.1% to 48.6%) were alive. Furthermore, based on small numbers with longer follow-up, there appears to be continuing significant hazard beyond this age.
Factors affecting outcome
A major influence on overall outcome is the proportion of parents opting for termination of the pregnancy. There was no statistically significant difference in the frequency of termination in the presence of uncomplicated cardiac anatomy (56%) compared with those with more complex anatomy 62% (95% CI for the difference between proportions +17.1 to −5.2%). In those with known karyotype, there was no significant difference in the frequency of termination between those with normal chromosomes (56%) and those with a chromosome abnormality (60%) (95% CI for the difference between proportions −17.2 to +9.1%). The termination rate was marginally higher in those without extracardiac anomalies (55.5%) than those with (53.8%) (95% CI for the difference between proportions −11.1 to −22.4%). None of these relationships between the indicators of severity and parents’ decisions to continue reached statistical significance. However, those with the particular combination of left isomerism and bradycardia did have a significantly higher termination rate (82%, 23/28) than the rest. The proportion opting to terminate the pregnancy had decreased to 54.8% in the latter part of the study period (1993 onwards) compared with 63.1% in the earlier part. A significant decrease in the proportion opting for termination with increasing year of diagnosis was demonstrated (Mann-Whitney U test, p = 0.031).
Figures 3, A to D show grouped Kaplan -Meier survival analyses for variables with a possible influence on survival, each considered in isolation, for all pregnancies with intention to continue. In each case, the log-rank test was used to test for statistical significance between datasets. Figure 3A shows survival curves with and without fetal hydrops, which are significantly different (p < 0.0001). Bradycardia, due to sinus node dysfunction or complete heart block, which often occurs together with fetal hydrops in fetuses with AVSD, was associated with similarly poor outcome (p < 0.012, chart not shown). Worse survival was associated with extracardiac abnormalities than without (Fig. 3B), but the difference was not statistically significant (p = 0.055). Survival for fetuses with isomerism is similar to that of fetuses with usual atrial arrangement (p = 0.56, chart not shown). Survival curves with and without equal-sized ventricles and pulmonary atresia or stenosis are shown in Figure 3, C and D. Disparity of ventricular size was associated with worse survival than equal sized ventricles (p = 0.23). There appeared to be a trend towards improved survival in fetuses with pulmonary atresia or stenosis in the early months of life, compared with those with a normal pulmonary outflow. However, there was no significant difference overall (p = 0.13). No clear difference in survival was evident for those with and without anteriorly positioned aorta (p = 0.95, chart not shown). Survival for those with and without chromosome abnormalities was very similar (not shown). There was, however, a nonsignificant trend for survival to be worse with unknown karyotype (p = 0.086). The reason for this association is that early death is predictive of failure to determine the karyotype rather than vice versa.
Some variables influencing survival, for example bradycardia and fetal hydrops, were closely related to each other, and, therefore, further analysis was performed to identify significant independent variables. Using Cox’s proportional hazards model with forward entry of variables to relate associated features to survival probability, the only statistically significant independent predictors of outcome were the presence of fetal hydrops at presentation and the year of presentation as shown in Table 2. The year of presentation might be expected to influence survival because of improvement in surgical results for comparable cases but also because of improved fetal diagnosis in referring centers increasing the proportion of surgically more favorable cases identified prenatally. The incidence of proven chromosomal abnormalities decreased in the last two years of the study from 61% in 1995 to 37% in 1997, but there has been considerable variation in the frequency of chromosomal abnormalities within individual years, and this change may be coincidental.
Figure 4 shows the results of a Kaplan-Meier survival analysis for fetuses with AVSD where there was intention to continue and where the cardiac abnormality was isolated and uncomplicated. Excluded from this analysis are fetuses with known karyotype abnormality, isomerism, extracardiac abnormality, unequal ventricles or abnormal arterial connections. Only 43 of the 123 continuing pregnancies fulfilled these criteria as having an isolated and uncomplicated AVSD. Even this selected group had a poor outcome with three-year survival of 55% (CI 36.7% to 74.1%).
Specific data on the occurrence and timing of surgical procedures was available for 88 of the 104 live-born infants. Nineteen of 43 deaths occurred within 30 days of a cardiac operation while 21 died without any cardiac surgery being performed, and there were three “late” deaths after surgery. Of the 18 survivors with anatomy considered unsuitable for biventricular repair, three have completed Fontan circulations, and seven have completed hemi-Fontan or Glenn shunts. Two surgical deaths were associated with a Fontan repair and none with a Glenn shunt. Of 27 survivors considered suitable for eventual biventricular repair, seventeen have already been corrected and 10, mainly with partial defects, remained unoperated at latest follow-up. Death after complete repair occurred more often when coarctation repair was required in addition, but this difference was not statistically significant (5/8 with coarctation vs 7/21 without, 95% CI for difference between proportions −0.1 to +0.68).
Atrioventricular septal defect is an important congenital cardiac abnormality, which can be diagnosed prenatally with considerable accuracy. Fetuses with this diagnosis do not constitute a homogeneous group but differ in their associations with additional congenital cardiac defects, visceral heterotaxy and chromosomal disorders. Despite routine “four-chamber view” screening as part of the routine ultrasound examination, many cases of AVSD remain undetected until after birth. Those cases diagnosed prenatally are likely to be biased, in favor of features that more readily bring them to the ultrasonographer’s attention. Bradycardia, fetal hydrops, marked ventricular dominance or the heart being on the opposite side of the body to the stomach may be recognized in circumstances where more subtle indications of AVSD might be overlooked.
The diagnosis of AVSD in the fetus should prompt not only a detailed examination of the rest of the heart but also of the fetus as a whole. The incidence of karyotypic abnormalities is as high as 62% in those with the least complicated cardiac anatomy. However, even in fetuses such as those with isomerism, where the incidence of chromosomal abnormality is much less, such abnormalities are sufficiently common that fetal karyotyping still needs to be carefully considered. It is of particular concern that, even when the fetal karyotype and detailed ultrasound anomaly scans are apparently normal, significant anomalies and genetic syndromes may become evident after birth, and parents must be counseled with this in mind.
Early reports of fetal diagnosis of congenital heart disease in general and of AVSD specifically have emphasised that the outlook for the fetus diagnosed in utero is considerably worse than would be expected from postnatal surgical series (5). This study confirms this observation but also indicates that there has been some improvement in prognosis both during the period of the study and compared with a similar earlier study from the same center. In the study of Machado et al. (1), which included all fetuses diagnosed with AVSD in this center before the period of this study, only 4 of the 29 fetuses studied were still alive at the time of publication. Two of those four surviving at the time of publication are since known to have died (6.8% crude survival). In the current series 57 (19%) of the 301 fetuses studied were alive at latest follow-up, though the duration of follow-up is shorter.
The poor prognosis of congenital heart disease diagnosed in utero is often attributed to the fact that more severe examples are likely to be detected in the fetus. However, even when the survival analysis was restricted to uncomplicated cases, the outlook in our fetal series remained considerably worse than reported surgical series for AVSD repair (6–8). Spontaneous intrauterine death is an additional hazard after prenatal diagnosis of AVSD. Most outcome studies report on the results from a surgical perspective and consider only those who had surgery or at least those considered for surgery. Furthermore, surgical series tend to consider a particular type of surgical treatment such as complete repair and, therefore, exclude those who were not suitable for biventricular repair and require alternative treatment strategies. A fetal series will include patients who never reach the cardiac surgeons’ attention, either because death occurred before surgery could be offered or unsuitability for surgery as a result of poor cardiac anatomy or severe extracardiac abnormalities.
Even in the time period of this series, there have been significant changes in many aspects affecting the outcome. Improvements in screening for fetal heart disease have led to the detection of increasing numbers of cases including less severe or less complex cases. Improvements in the surgical technique and the intensive care are likely to have contributed significantly to the improvement in outcome. In addition newer surgical approaches, such as the Norwood operation, can now be offered to the parents of infants with right ventricular dominance, which may be unsuitable for biventricular repair.
First trimester screening using nuchal translucency measurements has facilitated early diagnosis of major chromosome abnormality and has also been shown to be an effective means of screening for heart defects in the absence of chromosomal abnormality (9–11). This technique came into widespread use towards the end of the study period and may account for a reduction in the proportion of AVSD with chromosomal abnormality in the last two years. In the future, it may significantly alter the spectrum of AVSD presenting to the fetal cardiologist by increasing prenatal diagnosis of heart defects but decreasing the number of undiagnosed karyotype abnormalities after the first trimester.
In predicting the likely outcome for any individual fetus with AVSD, careful consideration should be given to specific characteristics of the cardiac and noncardiac findings. Because of the heterogeneity of AVSD, prediction of outcome based solely on the group as a whole is not necessarily the most appropriate for a particular individual. Because the numbers within subcategories of AVSD are relatively small, statistical evidence for differences in outcome related to anatomical and other features is difficult to establish. However, until further data are available, it seems reasonable to base counseling on those differences that do seem to be apparent from the current data. Hydrops fetalis at presentation was significantly associated with poor outcome. Isomerism did not appear to have any independent influence on outcome. Extracardiac abnormality had a negative, but not significant, effect on survival. Some of these abnormalities had not been diagnosed prenatally and could not be considered at the time of the initial counselling, but the nature as well as the presence of the extracardiac abnormality is likely to be important in determining prognosis. The detection rate of such abnormalities might also improve with better experience of fetal medicine units.
Unequal-sized ventricles as a risk factor for survival is consistent with the findings of a postnatal study considering complete repair of complicated AVSD in the presence of usual atrial arrangement only (12). Some infants with unequal sized ventricles, deemed unsuitable for biventricular repair, will continue to be exposed to the long-term hazard associated with the establishment and maintenance of a Fontan-type circulation.
After postnatal presentation, the size of the ventricular component of the defect is considered to be a major determinant of outcome. Where the ventricular component is restrictive or absent, surgical repair may be safely deferred beyond the first months of life, and the outcome is better than that for defects with a large ventricular component (13). In our study, information about the size and presence of a ventricular component of the defect was not documented in many cases. Of those fetuses known to have a small or absent ventricular component, half had complex cardiac anatomy, which may have prompted recognition of a cardiac problem, and it is likely that most uncomplicated “partial” AVSDs are still unrecognized in fetal life. However, if screening continues to improve, it is likely to lead to increased recognition of these. It is important to recognize that fetuses in whom such a defect can be distinguished in utero may well have a significantly better outcome than that of AVSD diagnosed in fetal life generally. The observation that the size of the ventricular component may get smaller as the pregnancy advances indicates that prognostications based on the size of the ventricular component of the defect at presentation may sometimes prove unreliable (14).
Short duration of follow-up will bias the crude survival figures towards better survival. In the Kaplan-Meier analysis, observations on those alive at the latest follow-up are censored at that time interval and so provide information about the likelihood of survival up to that duration but do not bias the survival estimate beyond the time for which they were under observation. Comparison of survival curves for different subgroups has limitations, which are of particular importance when the duration of follow-up is limited. In this study, the number of children known to be alive beyond four years of follow-up is limited to seven. This is because, despite the relatively large number of affected pregnancies studied, the numbers for study are rapidly reduced because of terminations and early deaths for other reasons. Also, follow-up after diagnosis made close to the closing date is necessarily short, and follow-up data beyond the neonatal period is lacking for some children followed up at other centers. The survival curves for different subgroups differ in the slope and shape, and this necessitates some caution in the interpretation of the results. For example, in fetuses with evidence of pulmonary atresia or stenosis, the early mortality appears less than that of those without these features. However, in some of these infants definitive, high-risk surgery may have been deferred in favor of simpler shunt operations or conservative management, and the incidence of later deaths, associated with definitive surgery, within this subgroup may prove to be considerable with more extensive follow-up.
The prospect for survival of a fetus diagnosed with AVSD is one of several factors, which may be taken into account by parents in deciding whether or not to continue the pregnancy. In this study, we were unable to demonstrate statistically significant relationships between indicators of severity of the condition in a particular fetus and decisions of the parents to terminate the pregnancy or continue, except for left isomerism with bradycardia. Incomplete information on these indicators may have been available at the time that a decision was reached. However, it is our impression that parents’ personal beliefs and attitudes are usually a stronger influence on decision making than the details of the abnormality itself. Also, the quality of life of the survivors is an issue that is of major concern but for which detailed data are still not available.
Despite some improvements in the outlook for AVSD diagnosed prenatally, the overall prognosis remains considerably worse than that implied from surgical series. The detection of associated cardiac and extracardiac abnormalities is important in order to give the best indication of the likely outcome when counselling parents.
- atrioventricular septal defect
- confidence interval
- double outlet right ventricle
- ventricular septal defect
- Received April 28, 1999.
- Revision received January 17, 2000.
- Accepted March 30, 2000.
- American College of Cardiology
- Machado M.V.,
- Crawford D.C.,
- Anderson R.H.,
- Allan L.D.
- ↵Calculating confidence intervals for proportions and their differences. In: Gardner MJ, Altman DG, editors. Statistics With Confidence. 1st ed. London: British Medical Journal. 1989:28–33.
- Cox D.R.
- ↵Jones KL. Miscellaneous associations. Smith’s Recognizable Patterns of Human Malformation. 5th ed. Philadelphia: W.B. Saunders, 1997:668–70.
- ↵Alexi M, V, Ishino K, et al. Correction of complete atrioventricular septal defects with the double-patch technique and cleft closure [see comments]. Ann Thorac Surg 1996;652:519–24.
- Hyett J.A.,
- Perdu M.,
- Sharland G.,
- et al.
- ↵Redmond JM, Silove ED, De Giovanni, et al. Complete atrioventricular septal defects: the influence of associated cardiac anomalies on surgical management and outcome. Eur J Cardio-Thorac Surg 1996;10:991–5.
- Sharland G.K.,
- Qureshi S.A.