ABSTRACT
Ultrasound imaging has become integral to the practice of obstetrics and gynecology. With increasing educational demands and limited hours in residency programs, dedicated time for training and achieving competency in ultrasound has diminished substantially. The American Institute of Ultrasound in Medicine assembled a multi-Society Task Force to develop a consensus-based, standardized curriculum and competency assessment tools for obstetric and gynecologic ultrasound training in residency programs. The curriculum and competency-assessment tools were developed based on existing national and international guidelines for the performance of obstetric and gynecologic ultrasound examinations and thus are intended to represent the minimum requirement for such training. By expert consensus, the curriculum was developed for each year of training, criteria for each competency assessment image were generated, the pass score was established at or close to 75% for each, and obtaining a set of five ultrasound images with pass score in each was deemed necessary for attaining each competency. Given the current lack of substantial data on competency assessment in ultrasound training, the Task Force expects that the criteria set forth in this document will evolve with time. The Task Force also encourages use of ultrasound simulation in residency training and expects that simulation will play a significant part in the curriculum and the competency-assessment process. Incorporating this training curriculum and the competency-assessment tools may promote consistency in training and competency assessment, thus enhancing the performance and diagnostic accuracy of ultrasound examination in obstetrics and gynecology. Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd.
Subject(s)
Clinical Competence/standards , Gynecology/education , Obstetrics/education , Ultrasonography , Accreditation , Consensus , Curriculum , Gynecology/standards , Humans , Internship and Residency , Obstetrics/standards , Quality Assurance, Health Care , Ultrasonography/standardsABSTRACT
STUDY OBJECTIVE: The aim was to quantify the changes in placental blood flow and combined ventricular output with gestational age in normal human fetuses to determine the percentage of total cardiac output that placental blood flow represented, and whether this changed with age. DESIGN: Two dimensional echocardiographic images of the umbilical vein, the proximal aorta and proximal main pulmonary artery were obtained, and cross sectional areas calculated from vessel diameters. Doppler velocity signals were recorded from each vessel and digitised to obtain velocity-time integrals. Placental blood volume flow and combined ventricular output were calculated as the products of flow velocity time integrals and cross sectional areas of the umbilical vein, and of the great arteries respective. SUBJECTS: Subjects were 64 normal human fetuses aged between 20 and 42 weeks gestation. MEASUREMENTS AND MAIN RESULTS: Placental flow and combined ventricular output both increased exponentially with gestational age (r = 0.79, and r = 0.84; both p less than 0.001). Placental flow correlated linearly with combined ventricular output (r = 0.69; p less than 0.01) and comprised almost one third of total cardiac output throughout the second and third trimesters. CONCLUSIONS: These data describe the relationship between placental blood flow and combined ventricular output with age in the normal human fetus and provide a substrate by which placental insufficiency and resulting intrauterine growth retardation may be recognised early.
Subject(s)
Fetal Heart/physiology , Placenta/blood supply , Cardiac Output/physiology , Echocardiography , Fetal Heart/diagnostic imaging , Gestational Age , Humans , Regional Blood Flow/physiology , Umbilical Veins/physiologyABSTRACT
Early antenatal detection of intrauterine growth retardation (IUGR) may decrease the associated perinatal morbidity and mortality. A parameter based on sonographically measured femur length (FL) and abdominal circumference (AC), expressed as FL/AC X 100 and termed the FL/AC ratio, has recently been proposed by Hadlock et al as an age-independent predictor of IUGR. We studied 285 normal and 37 IUGR fetuses to verify that the FL/AC ratio is independent of gestational age (GA) and to assess its value as a predictor of IUGR. Our results confirm that the FL/AC ratio is age-independent above 20 weeks and that its mean value differs in normal (22.4 +/- 1.7) and IUGR (23.7 +/- 1.4) fetuses (P less than .01, t-test). Because of considerable overlap between these two groups, however, there is no cutoff value for the FL/AC ratio that yields both a high sensitivity and a high specificity, or that leads to a high positive predictive value. With a cutoff of 23.5, for example, the sensitivity is 56% and the specificity 74%, and, even assuming an IUGR prevalence rate of 10%, the likelihood of IUGR in a fetus with an FL/AC ratio above the cutoff is only 19%. We conclude that the FL/AC ratio, though an age-independent measure whose mean value differs in normal and IUGR fetuses, is not clinically useful as a predictor of IUGR.
Subject(s)
Fetal Growth Retardation/diagnosis , Ultrasonography , Abdomen/embryology , Female , Femur/embryology , Fetus/anatomy & histology , Gestational Age , Humans , PregnancyABSTRACT
The objective of this study was to analyze the improvements in film reading performance made by radiology residents during their first six months of training. Five first-year residents and eight radiologic technology students each interpreted two of three matched sets of 39 films under two conditions. One set's readings were reviewed by a staff radiologist, while the other's were unreviewed. Six months later, each observer read all three sets. After the first six months of training, residents improved their reporting of findings. There was less improvement in technologists' readings. Review and instruction by staff, both in the laboratory and daily work settings, appeared to contribute to improved performance. Accuracy of residents' final diagnosis did not improve significantly. We conclude that a training system in which residents' film interpretations are reviewed by staff can lead to improved resident performance. When studied in a longitudinal fashion, these improvements are detectable within six months. This teaching system is used in many radiology departments.
Subject(s)
Educational Measurement/methods , Faculty, Medical , Internship and Residency , Radiography/education , Humans , Technology, Radiologic/educationABSTRACT
Pulsed Doppler studies of left and right ventricular outputs were obtained over time in a hydropic fetus with erythroblastosis fetalis. Despite severe anemia, cardiac outputs were within the normal range and remained normal after in utero percutaneous intravascular transfusions, which reversed the hydrops. The measurement of cardiac output in utero provides direct evidence that high-output failure due to anemia is not the mechanism for hydrops in erythroblastosis fetalis.
Subject(s)
Cardiac Output , Erythroblastosis, Fetal/diagnosis , Fetal Heart/physiopathology , Prenatal Diagnosis , Ultrasonography , Adult , Erythroblastosis, Fetal/physiopathology , Female , Humans , PregnancyABSTRACT
OBJECTIVE: To determine the clinical significance of hyperechoic bowel seen sonographically in second-trimester fetuses. METHODS: Fifty fetuses (0.6%) with echogenic bowel were identified sonographically from a population of 8680 consecutive second-trimester fetuses over 21 months. The fetal bowel was considered hyperechoic if its echogenicity was similar to that of surrounding bone. Follow-up was obtained through medical record review. RESULTS: Twenty-nine of 50 fetuses (58%) were normal; eight (16%) were aneuploid, including six Down syndrome, one trisomy 13, and one Turner syndrome. All eight fetuses with aneuploidy had sonographic anomalies in addition to the echogenic bowel. Eight of 50 fetuses (16%) were growth-retarded, and five others (10%) had normal karyotypes but are still undelivered. Among the eight growth-retarded fetuses, there were five intrauterine or neonatal deaths, one elective abortion, and two survivors. In addition, the six fetuses with Down syndrome and echogenic bowel represented 12.5% of all second-trimester Down syndrome fetuses karyotyped in our laboratory during the study period. Combining results from the present study (six Down syndrome fetuses) with three studies from the literature (21 additional Down syndrome fetuses), a total of 27 fetuses with echogenic bowel and Down syndrome were identified, 11 (40.7%) of whom had no other sonographic findings. We calculate that if 1,000,000 second-trimester fetuses were scanned, 5105 would have hyperechoic bowel as the only finding, of whom 71 would have Down syndrome and 5034 would not. The risk of Down syndrome in fetuses with isolated hyperechoic bowel is, therefore, 71 in 5105 or 1.4%. CONCLUSION: The finding of isolated hyperechoic bowel in the second trimester should prompt genetic counseling and consideration of karyotypic analysis.
Subject(s)
Amniocentesis , Chromosome Aberrations/diagnosis , Fetal Diseases/diagnosis , Intestines/diagnostic imaging , Ultrasonography, Prenatal , Chromosome Aberrations/epidemiology , Chromosome Aberrations/genetics , Chromosome Disorders , Down Syndrome/diagnosis , Down Syndrome/epidemiology , Down Syndrome/genetics , Female , Fetal Diseases/epidemiology , Fetal Diseases/genetics , Follow-Up Studies , Humans , Karyotyping , Pregnancy , Pregnancy Trimester, Second , Risk FactorsABSTRACT
We undertook a study to determine whether partial hydatidiform mole could be distinguished from other cases of first-trimester missed abortion using ultrasound. Scans from 22 cases of pathologically proved partial hydatidiform mole and 33 cases of first-trimester missed abortion were independently reviewed by three radiologists, each unaware of the final pathologic diagnosis. Using a standard data form, each radiologist recorded the dimensions, shape, and contents of the gestational sac, the sonographic appearance of the decidual reaction/placenta and myometrium, and the presence or absence of adnexal cysts. The following two criteria were found to be significantly associated (P less than .05) with the diagnosis of partial mole: 1) ratio of transverse to anteroposterior dimension of the gestational sac greater than 1.5, and 2) cystic changes, irregularity, or increased echogenicity in the decidual reaction/placenta or myometrium. There was high interobserver correlation for both criteria, as measured by the kappa statistic. In 50% of the cases, either both or neither of these criteria were met. When both criteria were met, the frequency of partial mole was 87%; when neither criterion was met, the frequency of missed abortion was 90%. These results indicate that ultrasound can be of value in predicting a high likelihood of partial mole prior to curettage.
Subject(s)
Hydatidiform Mole/diagnosis , Ultrasonography , Uterine Neoplasms/diagnosis , Abortion, Missed/diagnosis , Animals , Diagnosis, Differential , Female , Hydatidiform Mole/pathology , Pregnancy , Uterine Neoplasms/pathologyABSTRACT
A randomized, double-blind study was performed on 16 women to compare the efficacy of daily subcutaneous (SC) injections of leuprolide acetate (LA; TAP Pharmaceuticals, North Chicago, IL) plus oral placebo tablets (group A, n = 7) with SC LA plus oral medroxyprogesterone acetate (The Upjohn Company, Kalamazoo, MI; group B, n = 9) in the treatment of leiomyomata uteri. Patients in group A had a significant reduction in uterine size from a pretreatment volume of 601 +/- 62 cm3 (mean +/- standard error) to a mean uterine volume of 294 +/- 46 cm3 at 24 weeks of therapy (P less than 0.01). Group B patients had a reduction in uterine volume from 811 +/- 174 cm3 to 688 +/- 154 cm3, which was not statistically significant. However, only one patient in group B experienced hot flashes, whereas six patients in group A had this symptom (P less than 0.01). Both groups demonstrated significant increases in mean hemoglobin concentrations, hematocrits, and serum iron levels at 24 weeks of therapy compared with pretreatment levels.
Subject(s)
Antineoplastic Agents/therapeutic use , Gonadotropin-Releasing Hormone/analogs & derivatives , Leiomyoma/drug therapy , Medroxyprogesterone/analogs & derivatives , Uterine Neoplasms/drug therapy , Adult , Cholesterol/analysis , Double-Blind Method , Drug Therapy, Combination , Female , Gonadotropin-Releasing Hormone/therapeutic use , Humans , Leiomyoma/blood , Leiomyoma/pathology , Leuprolide , Medroxyprogesterone/therapeutic use , Medroxyprogesterone Acetate , Middle Aged , Random Allocation , Uterine Neoplasms/blood , Uterine Neoplasms/pathologyABSTRACT
Eighteen premenopausal women with symptomatic leiomyomata uteri were enrolled in a stratified, randomized, double-blind, placebo-controlled study evaluating the efficacy of leuprolide acetate (LA) depot treatment before myomectomy. Stratification was based on pretreatment uterine volume (less than 600 cm3 versus greater than or equal to 600 cm3). Nine women received intramuscular (IM) depot LA 3.75 mg every 4 weeks for 12 weeks (group A); nine women received IM placebo with the same injection schedule (group B). All women underwent myomectomy within 4 weeks of their last injection. Mean total intraoperative blood loss was 213 +/- 44 mL (mean +/- standard error of the mean [SEM]) in group A and 302 +/- 43 mL in group B. When data from patients with large uteri (pretreatment uterine volumes of 600 cm3 or greater) were analyzed, mean total blood loss was 189 +/- 44 mL in group A and 390 +/- 20 mL in group B. These data suggest that leuprolide depot treatment before myomectomy may decrease intraoperative blood loss in women with large leiomyomata uteri.
Subject(s)
Leiomyoma/drug therapy , Uterine Neoplasms/drug therapy , Delayed-Action Preparations , Double-Blind Method , Estradiol/blood , Female , Follow-Up Studies , Gonadotropin-Releasing Hormone/therapeutic use , Hemorrhage/etiology , Hemorrhage/physiopathology , Hormones , Humans , Intraoperative Complications , Leiomyoma/pathology , Leiomyoma/surgery , Leuprolide , Placebos , Postoperative Complications , Premedication , Uterine Neoplasms/pathology , Uterine Neoplasms/surgery , Uterus/pathologyABSTRACT
Ultrasound plays an important role in the management of multiple gestations in several key areas: (1) diagnosing multiple gestation and determining fetal number; (2) determining placentation; (3) diagnosing complications; (4) guiding procedures; and (5) identifying fetal lie late in pregnancy. The information obtained from single or serial ultrasound examinations can lead to improved obstetric decision-making and thereby to improved outcome in these high-risk pregnancies.
Subject(s)
Pregnancy, Multiple , Ultrasonography , Abortion, Eugenic , Congenital Abnormalities/diagnosis , Female , Fetal Death/diagnosis , Fetofetal Transfusion/diagnosis , Fetus/anatomy & histology , Gestational Age , Humans , Placentation , Pregnancy , Pregnancy Complications/diagnosis , Twins , Twins, ConjoinedABSTRACT
When the presence of a specific disease is being considered, a diagnostic test can often help the physician to choose between subjecting the patient to an extensive workup (or treatment) and proceeding no further. Decisions concerning the use of a test require that three judgments be made: (1) Should the test be done? (2) Which test (if any) should be used if two or more are available? (3) If a test that can take on more than two values is performed, what is the correct cutoff point (or positivity criterion) that determines whether or not to proceed? This paper presents a mathematical technique to answer these questions, taking into account data concerning the patient (summarized as the prior, or pre-test, probability of disease), the test, and the workup (or treatment). The technique is presented in a graphical form that can be applied to any clinical situation in which the needed data are available.
Subject(s)
Diagnosis , Humans , Hypertension, Renovascular/diagnosis , Hypertension, Renovascular/diagnostic imaging , Mathematics , Models, Theoretical , Probability , Radioisotope Renography , UrographyABSTRACT
Diagnostic tests are typically used to help the physician select among available management options. When two or more tests are available, using them sequentially is potentially more efficient than simultaneously performing multiple tests, in that the former approach may allow the physician to perform fewer tests. In particular, we demonstrate that if two common conditions are met, any simultaneous strategy involving at least as many tests as management options can be replaced by a sequential strategy with the same outcome and a smaller expected number of tests. It follows that, in many clinical situations in which the benefits of performing fewer tests outweigh the costs that may result from delaying diagnosis, simultaneous strategies cannot be optimal. This result can decrease the number of diagnostic strategies that the physician or decision analyst needs to consider.
Subject(s)
Decision Making , Diagnosis , Humans , Time FactorsABSTRACT
Physicians frequently face decisions whether or not to undertake a treatment that carries an immediate risk but offers a potential cure. However, data directly comparing life expectancies of similar groups of patients with and without that treatment are rarely available and in fact could not ethically be obtained. We present an indirect mathematical technique for computing the two life expectancies and specifically apply this approach to evaluating surgery for patients with gastric carcinoma without known metastases. We demonstrate that surgery increases life expectancy for average men and women up to age 80 and does not alter life expectancy at age 90. Hence surgery is the appropriate treatment for patients who are either risk-neutral or risk-seeking in their attitudes comparing the short-term hazards with the long-term benefits of surgery.
Subject(s)
Gastrointestinal Neoplasms/surgery , Adult , Age Factors , Aged , Decision Making , Female , Humans , Life Expectancy , Male , Middle Aged , Mortality , Neoplasm Metastasis , Risk , Sex FactorsABSTRACT
We examined the decision whether to perform coronary angiography (followed by bypass surgery if appropriate findings are present) in middle-aged men who have chest pain and have undergone exercise tolerance testing (ETT). We developed a model of this decision that combines data from a variety of sources and selects the optimal strategy based on health outcome and, if desired, monetary cost. The analysis supports the following conclusions: for patients with nonspecific chest pain or atypical angina, the ETT provides useful information concerning the decision; furthermore, the number of millimeters of ST-segment depression above which angiography should be performed depends on coronary risk factors and pain severity. A normal ETT is insufficient evidence to exclude coronary angiography for patients with typical angina, provided that one is willing to expand resources for health benefits at levels comparable to those for other accepted medical practices. If monetary considerations are excluded, the preceding statement concerning ETT and angiography also holds for patients with atypical angina and for those with nonspecific pain and advanced risk factors. These last two conclusions suggest that ETT is not useful in guiding management decisions concerning coronary angiography in patients at high enough risk of coronary artery disease on the basis of symptoms and risk profile.
Subject(s)
Coronary Angiography , Coronary Disease/diagnosis , Cost-Benefit Analysis , Pain/diagnosis , Thorax , Angina Pectoris/diagnosis , Coronary Artery Bypass , Decision Making , Exercise Test , Humans , Male , Middle Aged , Risk , SoftwareABSTRACT
A computer program has been developed to aid in diagnostic and therapeutic decisions concerning a patient with chest pain. It provides an analysis tailored to the individual patient, in that the data used in the analysis depend on specific patient characteristics. The user can elect to examine all stored data values (probabilities, quality-adjusted life expectancies, and monetary costs), and to alter any of them. Decisions at three stages in the patient workup are considered: prior to any diagnostic test, following an exercise tolerance test, and following coronary angiography. The results of the analysis can be displayed in several tabular and graphical formats. In addition, the program can carry out a Monte Carlo simulation (or probabilistic sensitivity analysis) to determine the effect of uncertainty in the data on the stability of the choice of optimal strategy.
Subject(s)
Computers , Coronary Disease/diagnosis , Diagnosis, Computer-Assisted , Microcomputers , Coronary Disease/therapy , Humans , Male , Middle Aged , SoftwareABSTRACT
This paper examines the implications of occult fecal blood loss in patients taking aspirin (at least 2 grams daily). Although such patients do have a somewhat higher probability of colonic carcinoma than do members of the general population, their risk is far lower than that of patients who have gastrointestinal blood loss when not taking aspirin. This difference in risk exists because aspirin itself can provoke occult blood loss in stool. Patients who manifest gastrointestinal blood loss while taking aspirin can be separated into two groups, based on whether or not that blood loss continues after aspirin is discontinued. Although patients who continue to bleed are at high risk for colonic carcinoma, those who cease having any blood loss are at lower risk than are members of the general population. Further diagnostic studies to detect colonic carcinoma should be pursued in the former group, but not in the latter, low-risk group.
Subject(s)
Aspirin/adverse effects , Colonic Neoplasms/diagnosis , Occult Blood , Adult , Aged , Aspirin/therapeutic use , Bayes Theorem , Feces/analysis , Humans , Male , Middle AgedABSTRACT
The lecithin/sphingomyelin ratio (L/S) and the measured value of saturated phosphatidylcholine (SPC), amniotic fluid determinations obtained to assess fetal pulmonary maturity, were evaluated with receiver operating characteristic (ROC) curve analysis. The effects of covariates on the ROC curves were analyzed with a regression methodology that took into account all the available data when constructing an ROC curve for each subgroup. To correct for verification bias the authors used a logistic regression analysis to model the probability of verification, thereby permitting correction for verification bias of a fully stratified data set in spite of small cell frequencies. They examined combination testing with prediction rules using prospective logistic modeling, including as variables test results and clinical features. The L/S was found to be significantly better than SPC for assessing fetal pulmonary maturity. For older gestational age the L/S and SPC performed better than for younger gestational age. Contamination of the specimen degraded the ROC curves. Correcting for verification bias did not influence the ROC curves significantly but changed the cutoff value of the test variable for any particular operating point. Prediction rules to evaluate combination testing showed that obtaining the SPC level in addition to the L/S ratio added no significant information compared with the L/S only. Including gestational age in the prediction rule of either test improved the prediction.
Subject(s)
Lung/embryology , Respiratory Distress Syndrome, Newborn/prevention & control , Bias , Fetal Organ Maturity/physiology , Gestational Age , Humans , Infant, Newborn , Predictive Value of Tests , ROC Curve , Regression Analysis , Respiratory Distress Syndrome, Newborn/diagnosisABSTRACT
The data for medical decision analyses are often unreliable. Traditional sensitivity analysis--varying one or more probability or utility estimates from baseline values to see if the optimal strategy changes--is cumbersome if more than two values are allowed to vary concurrently. This paper describes a practical method for probabilistic sensitivity analysis, in which uncertainties in all values are considered simultaneously. The uncertainty in each probability and utility is assumed to possess a probability distribution. For ease of application we have used a parametric model that permits each distribution to be specified by two values: the baseline estimate and a bound (upper or lower) of the 95 percent confidence interval. Following multiple simulations of the decision tree in which each probability and utility is randomly assigned a value within its distribution, the following results are recorded: (a) the mean and standard deviation of the expected utility of each strategy; (b) the frequency with which each strategy is optimal; (c) the frequency with which each strategy "buys" or "costs" a specified amount of utility relative to the remaining strategies. As illustrated by an application to a previously published decision analysis, this technique is easy to use and can be a valuable addition to the armamentarium of the decision analyst.
Subject(s)
Decision Making , Monte Carlo Method , Operations Research , Humans , Models, Biological , Probability TheoryABSTRACT
The objective of this study was to evaluate to what extent (1) the characteristics of localization, distribution, and size of echodense and echolucent abnormalities enable individuals to be designated as having either periventricular hemorrhagic infarction or periventricular leukomalacia and (2) the characteristics of periventricular hemorrhagic infarction and periventricular leukomalacia are independent occurrences. The population for this study consisted of 1607 infants with birthweights of 500 to 1500 g, born between January 1991 and December 1993, who had at least one cranial ultrasound scan read independently by at least two ultrasonographers. The ultrasound data collection form diagrammed six standard coronal views. The cerebrum was divided into 17 zones in each hemisphere. All abnormalities were described as being echodense or echolucent and were classified on the basis of their size, laterality, location, and evolution. Eight percent (134/1607) of infants had at least one white-matter abnormality. The prevalence of white-matter disease decreased with increasing gestational age. Most abnormalities were small or medium sized and unilateral; only large echodensities tended to be bilateral and asymmetric. Large abnormalities, whether echodense or echolucent, were more likely than smaller abnormalities to be widespread, and the extent of cerebral involvement was independent of whether abnormalities were unilateral or bilateral. Large abnormalities were relatively more likely than small abnormalities to involve anterior planes. Small abnormalities, whether echodense or echolucent, or whether unilateral or bilateral, preferentially occurred near the trigone. Using the characteristics of location, size, and laterality/symmetry, we were able to allocate only 53% of infants with white-matter abnormalities to periventricular hemorrhagic infarction or periventricular leukomalacia. Assuming that periventricular leukomalacia and periventricular hemorrhagic infarction are independent and do not share risk factors, and that each occurs in approximately 5% of infants, we would have expected 0.25%, or about 4 individuals, to have abnormalities with characteristics of both periventricular leukomalacia and periventricular hemorrhagic infarction, whereas we found 63 such infants. Most infants with white-matter disease could not be clearly designated as having periventricular hemorrhagic infarction or periventricular leukomalacia only. Periventricular hemorrhagic infarction contributes to the risk of periventricular leukomalacia occurrence, or the two sorts of abnormalities share common risk antecedent factors. The descriptive term echodense or echolucent and the generic term white-matter disease of prematurity should be used instead of periventricular leukomalacia or periventricular hemorrhagic infarction when referring to sonographically defined white-matter abnormalities.
Subject(s)
Cerebral Hemorrhage/diagnostic imaging , Cerebral Ventricles/diagnostic imaging , Echoencephalography , Infant, Premature, Diseases/diagnostic imaging , Infant, Very Low Birth Weight , Leukomalacia, Periventricular/diagnostic imaging , Brain Mapping , Dominance, Cerebral/physiology , Female , Humans , Infant, Newborn , Male , Prospective StudiesABSTRACT
Duplex ultrasonography was performed on 17 consecutive patients being evaluated for persistent gestational trophoblastic tumor (GTT). All patients had had a prior molar pregnancy evacuated and presented with a rise or plateau in their beta-human chorionic gonadotropin levels. The ultrasonography was considered to be abnormal if the image demonstrated a focal area of altered echogenicity within the uterus or if Doppler scanning revealed a focal area of detectable intrauterine blood flow. The ultrasound findings were compared with the pathologic results from dilation and curettage specimens. Ten of the 17 patients had pathologically proven macroscopic tumor. Of those 10, 7 had an abnormal sonographic image (sensitivity, 70%), and 9 had an abnormal Doppler examination (sensitivity, 90%). In all 10 patients the image and/or Doppler examination was abnormal. Among four patients with microscopic disease, imaging was positive in one case, and the Doppler examination was positive in three. Imaging and Doppler ultrasonography are complementary modalities that can reliably detect persistent uterine GTT, and Doppler ultrasonography appears to be more sensitive than imaging in making this diagnosis.