| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
by the American Institute of Ultrasound in Medicine J Ultrasound Med 21:713-718 • 0278-4297
Routine Obstetric Sonography
1 Department of Radiology University of California San Francisco, California
Abbreviations: ACOG, American College of Obstetricians and Gynecologists • AIUM, American Institute of Ultrasound in Medicine • NIH, National Institutes of Health • RADIUS, Routine Antenatal Diagnostic Imaging With Ultrasound A 20-year-old, healthy, pregnant woman with a known normal last menstrual period sees her obstetric care provider in early pregnancy. Should second-trimester sonography be performed? The Clinical Problem
Diagnostic sonography has made a dramatic impact on the practice of obstetrics. It is difficult to think of an obstetric problem in which diagnostic sonography does not contribute to the solution. When one adds to this the long-term safety record of sonography for fetal imaging,1,2 it is little wonder that this technique is pervasive in modern obstetric practice. (A high degree of safety is established for exposures to spatial peak temporal average intensities of <100 mW/cm2. For general obstetric imaging, current sonographic systems are within this limit.) Indeed, the value and safety of the technique have resulted in a long list of indications for its use in pregnant women. The National Institutes of Health (NIH) published a list of indications for sonography in pregnant American women.3 The list (Table 1
Routine obstetric sonography has also become a social experience and an expectation in our society. The obstetric sonographic examination is attractive in a manner that is atypical of other medical procedures.4 Expectant mothers generally view the sonogram of their fetus as a positive experience. They do not ask their obstetricians if they need a sonogram, they ask when it will be scheduled. It is extraordinarily common to give sonographic images to the parents as a memento of the experience and not at all uncommon to provide videotapes of the examination. A recent study showed that 87% of patients show up for their examination accompanied by their partners.4 That does not happen when they are reporting for flexible sigmoidoscopy, nor are mementos sought. Rather than swim against this tide of enthusiasm, some insurance companies, health maintenance organizations, and capitated physician groups have started paying for routine sonography. (Kaiser Health Care of Northern California, the largest health maintenance organization in northern California, includes sonography in the care of all obstetric patients regardless of indication.) Other insurers compensate obstetricians who perform routine sonographic examinations in their offices by increasing the global delivery fee paid to these practitioners. Instead of asking the question regarding whether routine sonography should be performed, it may be more appropriate to ask whether it can be stopped even if it cannot be justified from a cost-benefit perspective. If the brakes are applied now, only the poor and poorly informed may be excluded. Evidence As early as 1980, the Federal Republic of Germany initiated a program of sonographic screening of pregnancies.5 All pregnant women were offered screening sonography of their pregnancies. Experts in many countries recognized the benefits of obstetric sonography when performed to answer clinical questions and wondered whether Germany had not taken the correct approach. Sonography had the potential to show many unsuspected problems, including erroneous menstrual dates, twins, placenta previa, and devastating anomalies of the fetus.6,7
In 1984, the NIH convened an expert panel to examine the evidence in favor of screening sonography. The panel published their results and noted that there was no convincing evidence that routine sonographic screening of low-risk women would improve perinatal morbidity or mortality.3 They developed the previously noted comprehensive list of indications for sonography (Table 1 In the early 1990s, the NIH funded just such a study.7 Importantly, many obstetricians in the United States had already made their own value judgment by this time and had instituted programs of screening in their practices. Indeed, it is fair to state that some practitioners had already concluded that if 1 sonogram during pregnancy was "good," then 2 sonograms were "better," and 3, 4, or 5 sonograms were "best." Because this was a classic self-referral situation, there was no fiscal restraint exerted.8,9 Thus, by the time the NIH study results were published, the issue had already taken on a major financial impact in American obstetrics. Any result other than an overwhelming endorsement of routine sonography was certain to meet with a backlash.
The study, which has come to be known as the Routine Antenatal Diagnostic Imaging With Ultrasound (RADIUS) study,3 included more than 15,000 low-risk women.7 These women underwent 2 sonographic examinations (1 in the second trimester and the other in the third trimester of pregnancy). The examination was sufficiently comprehensive to detect virtually all unsuspected twins, to correctly assign gestational age, to detect abnormally low or high amniotic fluid volumes, to discover abnormalities of the cervix and uterine wall, and to identify placenta previa. The sonography performed was not completely comprehensive with regard to evaluation for fetal anomalies, but the screen for fetal anomalies was more comprehensive than currently mandated by oversight agencies (American Institute of Ultrasound in Medicine [AIUM] and American College of Obstetricians and Gynecologists [ACOG]; Table 2
Although screening sonography showed virtually every type of unsuspected abnormality that one might anticipate (e.g., unsuspected twins, erroneous menstrual dating, growth disturbances, and anomalies), the evidence failed to disclose a benefit in either perinatal morbidity or mortality.7 The unambiguous finding is that routine sonography for low-risk women resulted in many more sonograms than in the control group (2.2 compared with 0.6). It also documented that 1 or more indications for sonography develop in 45% of low-risk women during their pregnancy. Many years ago, a large Scandinavian study suggested that screening sonograms may actually reduce the total number of sonograms obtained during pregnancy, because once a "problem" developed, it often required numerous additional sonograms to sort out the issue.12 This notion was debunked by the RADIUS study. Areas of Uncertainty Interestingly, ongoing controversy regarding the value of routine sonography has settled on the issue of fetal anomaly detection as a means of decreasing neonatal morbidity and mortality.5,13–15 Furthermore, it is conceded that for fetal anomaly identification to improve outcomes, it is often necessary for prospective parents to elect pregnancy termination of fetuses with severe anomalies.13,14 This is a simple but important shift in the timing of mortality, because serious fetal anomalies account for 25% of neonatal deaths and can lead to debilitating long-term disabilities at considerable societal cost. When the RADIUS study was published, the criticism leveled against the results was that the detection rate (sensitivity) for fetal anomalies was much lower than in previously published European studies.7,16–19 Analysis indicated that had the RADIUS study practitioners achieved comparable sensitivities for the detection of fetal anomalies, as in the more successful European studies, there would have been a demonstrable benefit to screening.13 Much debate was engendered. Were American sonologists poorly trained to detect anomalies, or were the European studies that showed better detection rates spuriously overoptimistic? The answer appears to be a combination of both. Crane et al19 found that the factors influencing these differences arose from 3 sources. These included (1) a higher rate of difficult-to-detect anomalies in the American study (e.g., atrial and ventricular septal defects), (2) a broader set of criteria for defining major anomalies and more complete postnatal ascertainment in the American study, and (3) the level of operator experience.19 Many anomalies are not readily apparent at birth, and the thoroughness of neonatal follow-up therefore has a major influence on the reported prevalence of congenital malformations. Likewise, there is typically an inverse relationship between the prevalence of anomalies in neonates and the reported sensitivity of obstetric sonographic screening.13,19 For example, a meta-analysis of 6 European studies showed sensitivity for sonographically detected anomalies of 51% compared with only 17% in the RADIUS clinical trial; however, the prevalence of major anomalies among neonates in the RADIUS trial was 60% to 71% higher.19 Operator experience is also a critical factor in assessing the efficacy of sonographic screening for fetal anomalies. The overall detection of anomalies among screened pregnancies in the RADIUS trial was 3 times higher than in the control group.7 However, the RADIUS study also showed that the experience of the sonologist had a measurable impact on anomaly detection. Highly experienced sonologists detected 35% of anomalies compared with only 13% by less experienced sonologists. Had the overall detection rate been 35%, the RADIUS study would have shown a benefit.
Once the discussion centers on the value of screening sonography in detecting and improving the outcome of fetuses with anomalies, many rational judgments can be made. The examination performed in the RADIUS study was more comprehensive with respect to anomaly detection than the anatomic survey currently recommended by the AIUM or ACOG (Table 2 Several ancillary issues also become important when anomaly detection becomes the focus of the examination. False-negative results will be diminished, but false-positive results will increase. Additionally, a "new" type of falsenegative result becomes important. This is the failure to detect the second or third anomaly present in the fetus. It is a good rule that the most difficult anomaly to detect is the second anomaly. A fetus with a congenital diaphragmatic hernia may be savable, but if the congenital diaphragmatic hernia is a manifestation of Fryn syndrome, then the child cannot be saved. Prospective parents armed with the proper information can make an informed choice about continuing the pregnancy, and, quite obviously, the decision may be different in this circumstance. Kaiser et al20 showed additional findings at autopsy in 51% of fetuses aborted because of an abnormality noted on prenatal sonography. They also recorded a 3% occurrence of failure to confirm the anomaly for which the termination was carried out. Finally, other common observations that potentially indicate an increased risk for Down syndrome, the so-called soft markers for Down syndrome, are also a source of great prenatal anxiety for pregnant women undergoing routine sonography.21 Importantly, the expectations of the medical community do not match those of prospective parents when it comes to sonographic detection of fetal anomalies. The parents consider that the examination is being performed to ensure that the "baby" is OK.4 It is clear that they are not nearly as interested in their placenta, amniotic fluid, cervix, uterus, or ovaries as they are in the sex and normalcy of their fetus. Furthermore, the concept that a routine or screening sonogram is "normal" implies to the parents that the fetus is normal. To them a normal sonogram provides assurance that they are carrying a normal little girl or boy fetus. Eighty-nine percent of women and 84% of their partners think the examination was performed to look for fetal anomalies.4 Guidelines In 1997, ACOG commissioned another extensive review of the literature and subsequently published guidelines for the use of sonography during pregnancy.13 The panel noted that the evidence was compelling that perinatal morbidity and mortality were not improved by routine sonograms in low-risk women. This was true despite a salutary effect on correcting erroneous dates, finding unsuspected twins, and identifying fetuses with anomalies. The panel analyzed data for improvement in outcome of anomalous fetuses, reduction of neonatal sepsis, lower prevalence of moderate intraventricular hemorrhages, and admission to neonatal intensive care units. Routine sonographic screening also failed to decrease the number of obstetric interventions, maternal hospital days, or cesarean delivery rates. ACOG and the AIUM have published guidelines for obstetric sonography.10,11 Neither of these organizations supports the use of routine sonographic screening of all pregnant women. Conclusions and Recommendations On the basis of evidence-based medicine, the following recommendations and conclusions can be made:
References
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
