Original article:
Maternal diabetes mellitus: which views are essential for fetal echocardiography?
Smith RS, Comstock CH, Lorenz RP et al. Obstet Gynecol 1997; 90: 575-9.

Summary
The authors' objective in this study was to determine which fetal echocardiographic views are essential to detect as many cardiac defects as possible in the high-risk subgroup of fetuses of women with insulin-requiring diabetes.
The study group consisted of 223 pregnant women with insulin-requiring diabetes who underwent fetal echocardiography between Feb-ruary 1990 and May 1996. The fetal echocardi-ograms were retrospectively reviewed and the individual component views of the examinations analysed to detect fetal congenital heart defects. Women with multiple gestations or additional risk factors for congenital heart defects were excluded from the study.
The sensitivity, specificity, positive predictive value and negative predictive value of each view were calculated. Eleven (4.9%) congenital heart defects were identified, of which eight (3.6%) were conotruncal in origin. There were two false-negative and no false-positive diagnoses. When the four-chamber view and outflow tracts appeared normal, additional views such as the aortic or ductal arches did not yield any further diagnostic accuracy. However, the sensitivity of the four-chamber view for detecting a structurally abnormal heart increased from 73 to 82% with the addition of the aortic outflow tract.
The authors logically conclude that the four-chamber and outflow tracts are essential views for detecting the majority of cardiac defects in fetuses of women with insulin-requiring dia-betes.

Comment
This article has several merits that are worth emphasizing. First, it has the obvious merit of drawing more attention to this subgroup of high-risk pregnancies and confirming previous observations of an increased probability of cardiac malformations among this population. It is evident from the authors' data that fetal echo-7 cardiography should be advocated in this type of pregnancy, which must be considered at high risk of malformation.
Second, it is interesting for the practitioner in charge of the echocardiographic follow-up of such pregnancies, to know which views are mandatory to detect, with the maximum of sensitivity, the presence of fetal heart defects. The observation that the great vessels' view increased the diagnostic sensitivity of the technique from 73% for the four-chamber view alone to 82% when the two views were combined, makes it mandatory to visualize and identify both great vessels. Of interest also is the remark that it is not sufficient to visualize the aorta only, since the relationship of both vessels to one another is important, particularly for the identification of transposition of the great arteries (which was present in 28% of the fetuses identified as having cardiac defects). It is also necessary to identify the crossing of both great vessels as well as their respective ventriculo-arterial concordance. Incidentally, I was surprised by the presence in this series of two cases of corrected transposition of the great arteries, a diagnosis which is sometimes difficult to make.
Third, it is useful to know that additional views add very little diagnostic power to the procedure, thus allowing the echographer to be more confident in the diagnosis, even if each and every view required for a 'complete' examination (Table I) has not been obtained. This can undoubtedly shorten the duration of the exam-ination and will probably reduce the number of visits necessary to obtain 'reasonably complete' instead of 'fully complete' investigations of the fetal heart, particularly in the first-line follow-up of pregnancies by gynaecologists and obstetricians. This is clearly of major importance from a clinical point of view.
Some points, however, could, or should, have been considered even though they were not among the primary concerns of the article. First, the authors focused their attention on this subgroup of diabetic pregnancies because a sizeable number of women had to undergo several examinations (60 min for each one) for the clinical evaluation to be considered complete. These return visits were considered an extension of the initial scan, but the authors do not explain why the initial scan was incomplete in a larger number of diabetic women than in normal pregnancies. This issue would have been of interest to the practitioner facing this type of examination.
Second, it is disappointing that the authors did not comment on the presence of the diabetic cardiomyopathy described in numerous previous studies [1-7]. Even though this type of pathology was not one of the targets of the current study, the data were available in all the examinations considered complete, since all the ventricular walls were measured (Table I).

It is to be hoped that the authors will report in the near future their findings in this area because their number of observations is important and it would be interesting to know the ventricular thickness that can be considered normal, tolerable or abnormal among this fetal population. It would greatly enhance our knowledge of this entity and help us to better appreciate the effect of diabetes on the cardiac wall dimensions.
Third, it would have been of interest to consider the functional aspect of ventricular contraction among this population. It has recently been reported to be normal [8], but this observation deserves verification.
I would not like to give the reader the impression, from the above three remarks, that the study by Smith et al. is not of great value. It is indeed of primary importance for the prenatal echocardiographer and its quality is such that I would like still more information from the team that produced the work. What better mark of interest could any scientific article raise?

Reference
1. Rowland TW, Hubbell JP Jr, Nadas AS. Congenital heart disease in infants of diabetic mothers. J Pediatr 1973; 83: 815-20.
2. Way GL, Wolfe RR, Eshaghpour E et al. The natural history of hypertrophic cardiomyopathy in infants of diabetic mothers. J Pediatr 1979; 95: 1020-5.
3. Gutgesell HP, Speer ME, Rosenberg HS. Characterization of the cardiomyopathy in infants of diabetic mothers. Circulation 1980; 61: 441-50.
4. Gutgesell HP, Mullins CE, Gillette PC et al. Transient hypertrophic subaortic stenosis in infants of diabetic mothers. J Pediatr 1976; 89: 120-5.
5. Reller MD, Anderson CF, Shenker L. Fetal echocardiography: an atlas. New York: Alan R. Liss, 1988.
6. Fanaroff A, Veinne JC, Sivakoff M. [abstract]. International neonatal intensive care collegium, June 7-12 1987, Sassari, Italy.
7. Soyeur DJ, Schaaps J-P, Kulbertus HE. Le cœur du fœtus de mère diabétique. Inf Cardiol 1989; 12: 803-9.
8. Eidem J, Edwards J, Stokas A et al. Nongeometric assessment of ventricular function in the fetus: comparison of normal fetuses and fetuses of diabetic mothers [abstract]. J Am Coll Cardiol 1998; 31 (suppl C): 167C.

Summary and Comment:
D. Soyeur, Liège, Belgium

Authors' reply
Women with diabetes were chosen as the select high-risk group because many of them were obese with 'thick' maternal walls, thus hindering our ability to perform an adequate transabdominal examination. We realized that we were asking these patients to return for completion more often than others. This was at times an inconvenience for the patient. We therefore began to look at which views on our checklist actually contributed to the detection of a defect. Additionally, in the majority of cases, we measured and documented the maternal wall thickness in obese patients and commented if fetal position hindered the examination. In some cases there was overlap.
With regard to the presence of cardiomyopathy in the fetus, although we continue to document interventricular septum thickness and ventricular width, we have found cardiomyopathy to be less of a concern in our patient population due to the current intense prenatal diabetic control. Lastly, ventricular function is indeed an interesting area that we have not investigated.