Fetal size monitoring and birth-weight prediction: a new population-based approach.

OBJECTIVE: To develop a complete, population-based system for ultrasound-based fetal size monitoring and birth-weight prediction for use in the second and third trimesters of pregnancy. METHODS: Using 31 516 ultrasound examinations from a population-based Norwegian clinical database, we constructed fetal size charts for biparietal diameter, femur length and abdominal circumference from 24 to 42 weeks' gestation. A reference curve of median birth weight for gestational age was estimated using 45 037 birth weights. We determined how individual deviations from the expected ultrasound measures predicted individual percentage deviations from expected birth weight. The predictive quality was assessed by explained variance of birth weight and receiver-operating characteristics curves for prediction of small-for-gestational age. A curve for intrauterine estimated fetal weight was constructed. Charts were smoothed using the gamlss non-linear regression method. RESULTS: The population-based approach, using bias-free ultrasound gestational age, produces stable estimates of size-for-age and weight-for-age curves in the range 24-42 weeks' gestation. There is a close correspondence between percentage deviations and percentiles of birth weight by gestational age, making it easy to convert between the two. The variance of birth weight that can be 'explained' by ultrasound increases from 8% at 20 weeks up to 67% around term. Intrauterine estimated fetal weight is 0-106 g higher than median birth weight in the preterm period. CONCLUSIONS: The new population-based birth-weight prediction model provides a simple summary measure, the 'percentage birth-weight deviation', to be used for fetal size monitoring throughout the third trimester. Predictive quality of the model can be measured directly from the population data. The model computes both median observed birth weight and intrauterine estimated fetal weight. Copyright © 2016 ISUOG. Published by John Wiley & Sons Ltd.

Correlation between prenatal ultrasound and postmortem findings in 1029 fetuses following termination of pregnancy.

OBJECTIVE: A prenatal ultrasound examination and a postmortem examination provide the basis for correct diagnosis in fetuses terminated due to congenital anomalies. The aim of this study was to correlate fetal anomalies detected by ultrasound examination with those identified at autopsy following termination of pregnancy (TOP) over a 30-year period, and to evaluate the correlation between findings at different gestational ages and assess these trends over time. METHODS: The study group consisted of 1029 TOPs performed over a 30-year period, from 1985 to 2014. The gestational age ranged between 11 and 33 weeks. Prenatal ultrasound examinations were performed at the National Center for Fetal Medicine, St Olavs Hospital, Trondheim, Norway. Autopsies were performed at the Department of Pathology and Medical Genetics at the same hospital or a collaborating hospital. RESULTS: There was full agreement between ultrasound and autopsy findings in 88.1% (907/1029) of TOPs, and the main diagnosis was correct in 97.9% (1007/1029). When comparing the 15-year period of 2000-2014 with that of 1985-1999, the difference in the rates of full agreement and agreement in the main diagnosis was statistically significant. In 1.3% (13/1029) of cases, ultrasound findings were not confirmed at autopsy. There were no false-positive diagnoses leading to TOP. Throughout the 30-year period, there was an increase in early TOPs, whereas late TOPs declined. CONCLUSIONS: Our study demonstrates that there is a clear correlation between ultrasound and autopsy findings, which is continuously improving. Despite this high correlation, there is reason to continue the practice of validation to ensure the safety of the diagnostic process leading to TOP. The trend towards an earlier termination emphasizes the necessity of such a practice. Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd.

Maternal obesity affects fetal myocardial function as early as in the first trimester.

OBJECTIVE: To investigate cardiac function from 14 weeks' gestation in fetuses of obese pregnant women (FOW). Animal studies have shown that maternal obesity induces fibrosis in fetal myocardium. We hypothesized that fetal cardiac function would be impaired among FOW. METHODS: A case-control study with longitudinal follow-up was performed at Trondheim University Hospital, Norway. In total, 80 pregnant women were included and the final population comprised 52 obese and 24 of normal weight (mean body mass index before pregnancy, 34.8 ± 4.1 vs 21.0 ± 2.2 kg/m(2) ; P < 0.001). The main outcome measures were global strain rate (GSR) and strain by tissue Doppler imaging, tissue Doppler velocities (TDVs) and interventricular septal thickness assessed by fetal echocardiography at gestational ages of 14, 20 and 32 weeks. RESULTS: In FOW, fetal left ventricle (LV) and right ventricle (RV) GSR and strain were significantly lower than in fetuses of normal-weight pregnant women: LV GSR was 33.3% lower at 14 weeks, 22.4% lower at 20 weeks and 22.8% lower at 32 weeks of gestation (P < 0.001) with no difference in fetal heart rate. Systolic and late diastolic TDVs for LV were significantly lower from 20 weeks' gestation and remained lower throughout pregnancy. Fetal interventricular septum was 26.6% (P < 0.001) thicker in late pregnancy in FOW compared with normal-weight pregnancies. CONCLUSIONS: At 14 weeks of gestation, we detected fetal myocardial dysfunction with reduced LV and RV GSR and strain in FOW compared with fetuses of women with normal weight. Our finding is alarming considering the high prevalence of obesity and may partly explain the predisposition of offspring to cardiovascular disease later in life.

Abnormal gyration of the temporal lobe and megalencephaly are typical features of thanatophoric dysplasia and can be visualized prenatally by ultrasound.

Autopsies of fetuses with thanatophoric dysplasia (TD) have shown abnormal gyration of the temporal lobes. In addition, the head is relatively large compared with the abdomen. We evaluated by ultrasound six consecutive cases of TD at 19 + 0 to 19 + 6 gestational weeks based on last menstrual period. We observed abnormal and deep transverse sulci in the temporal lobes in all cases; these features were confirmed at autopsy. We performed biometric assessment, including biparietal diameter (BPD) and mean abdominal diameter (MAD). For each MAD value in the TD fetuses, we computed mean and SD of the corresponding BPD values from a population-based registry in the relevant age range, and used them to calculate Z-scores for each BPD/MAD ratio. In the general population, the average BPD/MAD ratio was 1.05. In the TD fetuses, the mean BPD was 51.5 (range, 49-54) mm, the MAD was 45 (range, 41-47) mm and the BPD/MAD ratio was 1.15 (range, 1.09-1.20). The average Z-score of the ratios for TD fetuses was 2.44 (range, 1.05-3.39). The ratios for the TD fetuses were significantly higher than were the population ratios (P = 0.016). At autopsy, the mean brain-to-body weight ratio was 20.6% (range, 15.4-24.1%), which was greater than the corresponding mean ratio of 14.9% in normal fetuses. We conclude that abnormal and deep transverse gyration of the temporal lobes can be visualized by ultrasound in mid-second-trimester fetuses with TD. Due to megalencephaly, fetuses with TD have significantly different body proportions, with a larger BPD compared with normal fetuses.

Comparison between prenatal ultrasound and postmortem findings in fetuses and infants with developmental anomalies.

OBJECTIVE: To determine if postmortem examinations of fetuses and infants change the diagnosis obtained at prenatal ultrasound and affect counseling of future pregnancies, and if there has been a change over recent years in the accuracy of prenatal ultrasound diagnosis. METHODS: This was a retrospective review of 455 autopsies of fetuses and infants with developmental anomalies performed at Trondheim University Hospital between 1995 and 2004 and with a prenatal ultrasound examination performed at a tertiary referral center. The routine ultrasound examinations were performed by specially trained midwives and obstetricians, referral scans by fetal medicine experts and autopsies by consultant pathologists with experience in perinatal pathology. The results of this study were also compared with those of a previous similar study performed between 1985 and 1995, with fetuses and infants coming from the same population and diagnosed at the same center. RESULTS: Of all cases analyzed during the study period, there was complete agreement between prenatal ultrasound and postmortem findings in 84% (384/455), i.e. prenatal ultrasound diagnoses were supplemented by postmortem examinations in 16% (71/455). Autopsy findings in four of these cases influenced further counseling. There was agreement regarding the main diagnosis in 98% (445/455) of cases. In the previous 10-year period, there was complete agreement in 75% and the main diagnosis was correct in 90% of cases. These differences between the two time periods were statistically significant (P = 0.0004 and P < 0.0001, respectively). The most frequent defects involved the central nervous system, heart and urinary tract. For these defects, detection rates for the main diagnoses were significantly better in 1995-2004 compared with in the previous 10-year period (P = 0.0125, P = 0.0111 and P = 0.0241, respectively). CONCLUSION: The accuracy of prenatal sonographic detection of developmental anomalies has increased in recent years. However, postmortem examination is still necessary to verify or improve the prenatal diagnosis and may influence future counseling.

Advantages of the population-based approach to pregnancy dating: results from 23,020 ultrasound examinations.

OBJECTIVE: To confirm the results from two previous evaluations of term prediction models, including two sample-based models and one population-based model, in a third population. METHODS: In a study population of 23,020 second-trimester ultrasound examinations, data were prospectively collected and registered over the period 1988-2009. Three different models for ultrasonically estimated date of delivery were applied to the measurements of fetal biparietal diameter (BPD) and two models were applied to the femur length (FL) measurements; the resulting term estimations were compared with the actual time of delivery. The difference between the actual and the predicted dates of delivery (the median bias) was calculated for each of the models, for three BPD/FL-measurement subgroups and for the study population as a whole. RESULTS: For the population-based model, the median bias was + 0.4 days for the BPD-based predictions and - 0.4 days for the FL-based predictions, and the biases were stable over the inclusion ranges. The biases of the two traditional models varied with the size of the fetus at examination; median biases were - 0.87 and + 2.2 days, respectively, with extremes - 4.2 and + 4.8 days for the BPD-based predictions, and the median bias was + 1.72 days with range - 0.8 to + 4.5 days for FL-based predictions. The disagreement between the two sample-based models was never less than 2 days for the BPD-based predictions. CONCLUSION: This study confirms the results from previous studies; median biases were negligible with term predictions from the population-based model, while those from the traditional models varied substantially. The biases, which have clinical implications, seem inevitable with the sample-based models, which, even if overall biases were removed, will perform unsatisfactorily.

Prenatal sonographic assessment and perinatal course of ichthyosis prematurity syndrome.

All cases of ichthyosis prematurity syndrome (IPS), registered at the National Center for Fetal Medicine in Trondheim, Norway between 1987 and 2010 were identified and the findings analyzed. Five fetuses with IPS were identified between 1988 and 2000. All five developed polyhydramnios between 28 and 31 weeks. The fetal stomach appeared to be empty in four cases, and was not described in one case. The fetal skin was described as 'uneven' at ultrasound examination in two cases. Separation of chorionic and amniotic membranes with a peculiar appearance of echo-free fluid in the chorionic cavity and echogenic sediment in the amniotic cavity were observed between 28 + 5 and 32 + 3 weeks in all cases. All fetuses were delivered prematurely between 30 and 34 weeks. All neonates had difficulties in breathing, two developed aspiration pneumonia, and one had bilateral pneumothorax after intubation and died at 6 months because of pulmonary and cardiac sequelae. Prenatal sonographic signs of IPS are separation of the membranes, echogenic amniotic fluid and echo-free chorionic fluid occurring between 28 and 32 weeks' gestation. Delivery occurs at 30-34 weeks and, as there is a high risk of asphyxia, an experienced neonatal intensive care unit team should be present at delivery.

A new population-based term prediction model vs. two traditional sample-based models: validation on 9046 ultrasound examinations.

OBJECTIVES: To compare results of predictions of date of delivery from a new population-based model with those from two traditional regression models. METHODS: We included 9046 fetal biparietal diameter (BPD) measurements and 8776 femur length (FL) measurements from the routine ultrasound examinations at Stavanger University Hospital between 2001 and 2007. The prediction models to be validated were applied to the data, and the resulting predictions were compared with the actual time of the subsequent deliveries. The primary measure was the median bias (the difference between the true and the predicted date of delivery), calculated for each method, for the study population as a whole and for three subgroups of BPD/FL measurements. We also assessed the proportion of births within ± 14 days of the predicted day, and rates of preterm and post-term deliveries, which were regarded as secondary measures. RESULTS: For the population-based model, the median bias was -0.15 days (95% confidence interval (CI), -0.43 to 0.12) for the BPD-based, and -0.48 days (95% CI, -0.86 to -0.46) for the FL-based predictions, and both biases were stable over the inclusion ranges. The biases of the traditional regression models varied, depending on the fetal size at the time of the examination; the extremes were -3.2 and + 4.5 days for the BPD-based, and -1.0 and + 5.0 days for the FL-based predictions. CONCLUSIONS: The overall biases, as well as the biases for the subgroups, were all smaller with the population-based model than with the traditional regression models, which exhibited substantial biases in some BPD and FL subcategories. For the population-based model, the FL-based predictions were in accordance with the BPD-based predictions.

Narrowed beam width in newer ultrasound machines shortens measurements in the lateral direction: fetal measurement charts may be obsolete.

OBJECTIVES: Fetal ultrasound measurements are made in axial, lateral and oblique directions. Lateral resolution is influenced by the beam width of the ultrasound system. To improve lateral resolution and image quality, the beam width has been made narrower; consequently, measurements in the lateral direction are affected and apparently made shorter, approaching the true length. The aims of this study were to explore our database to reveal time-dependent shortening of ultrasound measurements made in the lateral direction, and to assess the extent of beam-width changes by comparing beam-width measurements made on old and new ultrasound machines. METHODS: A total of 41,941 femur length measurements, collected during the time-period 1987-2005, were analyzed, with time as a covariate. Using three ultrasound machines from the 1990s and three newer machines from 2007, we performed 25 series of blinded beam-width measurements on a tissue-mimicking phantom, measuring at depths of 3-8 cm with a 5-MHz transducer. RESULTS: Regression analysis showed time to be a significant covariate. At the same gestational age, femur length measurement was 1.15 (95% CI, 1.08-1.23) mm shorter in the time-period 1999-2005 than in the time-period 1987-1992. Overall, the beam width was 1.08 (95% CI, 0.50-1.65) mm narrower with the new machines than with the old machines. CONCLUSIONS: Technical improvements in modern ultrasound machines that have reduced the beam width affect fetal measurements in the lateral direction. This has clinical implications and new measurement charts are needed.

Biases of traditional term prediction models: results from different sample-based models evaluated on 41 343 ultrasound examinations.

OBJECTIVE: To evaluate two Norwegian traditional, sample-based term prediction models as applied to the data from a large population-based registry. The two models were also compared with an established German model. METHODS: Our database included information from 41 343 non-selected ultrasound scans registered over the years 1987-2005. The prediction models were applied to measurements from the ultrasound examinations, and the resulting term predictions were compared with the actual times of the deliveries. The median bias (the difference between the true and the predicted date of delivery) was calculated for each model, both for the study population as a whole and for subgroups of measurements of biparietal diameter (BPD) and femur length (FL). Secondary measures, i.e. proportion of births within ± 14 days and the rates of preterm and post-term deliveries, were also assessed. RESULTS: The analyses showed that the models had significant biases, predicting delivery date either too late or too early. For each model the size of the bias varied, depending on the fetal size at the time of the examination; the extremes were minus 4 and plus 4 days for the BPD-based predictions. There were similar results with the FL-based predictions. CONCLUSION: Term predictions made with traditional sample-based models had significant biases that varied over each method's measurement range. These models have important shortcomings, probably because of strict selection criteria in the process of constructing the models, and because the methods primarily aim at estimating the last menstrual period-based day of conception, not the day of birth.

Prenatal detection of trisomy 21 by second-trimester ultrasound examination and maternal age in a non-selected population of 49 314 births in Norway.

OBJECTIVES: To assess the contribution of the second-trimester routine ultrasound examination and maternal age (>or= 38 years) to the prenatal detection of trisomy 21 in a large non-selected population in which no other screening methods were carried out. METHODS: A prospective follow-up study was carried out in a non-selected Norwegian population from 1987 to 2004, including a total of 49 314 births. Data from all cases of trisomy 21, detected prenatally by karyotyping following amniocentesis performed for abnormal ultrasound findings or because of maternal age >or= 38 years, or postnatally after delivery, were registered. The study period was divided into three 6-year periods. RESULTS: Eighty-eight cases of trisomy 21 were registered. The prenatal detection rate was 43% (38/88). No significant change in the detection rate was observed over the 18 years. Fourteen percent (12/88) of cases were detected as a result of advanced maternal age and 30% (26/88) by prenatal ultrasound examination. Of all women with a trisomy 21 fetus, 72% (63/88) were under 38 years of age. The percentage of women >or= 38 years opting for karyotyping during the three time periods decreased significantly from 51% to 50% and 36%, respectively. The termination rate of trisomy 21 fetuses was 84%, with no significant change over time. CONCLUSIONS: Our data can be considered as a reference standard for population screening for trisomy 21 based solely on maternal age and second-trimester ultrasound imaging. The prenatal detection rate of trisomy 21 cases was poor and remained unchanged throughout the 18-year study period. If improvement in detection rates is desired, additional programs are necessary.

Prenatal ultrasound detection of facial clefts: a prospective study of 49,314 deliveries in a non-selected population in Norway.

OBJECTIVES: To evaluate prenatal detection of facial clefts by ultrasound examination in a large non-selected population, and to study trends in detection rates over 18 years, as well as the prevalence of isolated cases and those with associated anomalies. METHODS: This prospective follow-up study from January 1987 to December 2004 was divided into two 9-year periods. The study included all registered cases of prenatally or postnatally diagnosed facial clefts in a non-selected population in Norway. RESULTS: A total of 101 fetuses or newborns with facial clefts in a population of 49 314 deliveries were registered. The distribution of clefts was: 25 (25%) cleft lip, 52 (51%) cleft lip and palate, and 24 (24%) cleft palate (CP). No CP was detected prenatally. Cleft lip with or without cleft palate (CL(P)) was detected prenatally in 35/77 (45%) cases, with a significant increase in the detection rate from 34% to 58% between the two 9-year periods (P = 0.03). Over the whole study period CL(P) was detected at a median of 19 + 2 gestational weeks, with no change over time; altogether 24/35 (69%) cases were detected at the routine second-trimester ultrasound examination. Thirty-three of 77 (43%) cases of CL(P) and 14/24 (58%) cases of CP had associated anomalies; 12/101 (12%) had chromosomal aberrations. In 18/101 (18%) the clefts were part of a syndrome or sequence. CONCLUSIONS: The detection rate for CL(P) improved significantly over time. Detection of CL(P) is important because nearly half the cases have associated anomalies.

Lack of evidence of permanent engraftment after in utero fetal stem cell transplantation in congenital hemoglobinopathies.

The use of fetal hematopoietic stem cells for in utero transplantation to create permanent hematochimerism represents a new concept in fetal therapy. In one fetus with alpha-thalassemia, one with sickle cell anemia, and one with beta-thalassemia, we have transplanted fetal liver cells obtained from legal abortions in gestational weeks 6-11. The fetus with alpha-thalassemia was transplanted twice during pregnancy, in the 15th (20.4 x 10(8) cells/kg) and in the 31st weeks of gestation (1.2 x 10(8) cells/kg), and is now two years of age. One fetus with sickle cell anemia received its transplant in the 13th week of gestation (16.7 x 10(8) cells/kg), and is now one year old. The fetus with beta-thalassemia was transplanted in 18th week (8.6 x 10(8) cells/kg), and is now three months old. Engraftment was evaluated by chromosomal analysis (sex chromosomes), red cell phenotyping, HLA class I and II typing, and PCR (polymerase chain reaction) for Y chromosome-specific sequences and DNA polymorphisms in cord and peripheral blood. The children with alpha- and beta-thalassemia underwent bone marrow aspirations at 3 and 7 months of age, respectively. In neither of these cases were we able to detect convincing evidence of stem cell engraftment. Thus, the administration of fetal stem cells to fetal recipients after the 12th week of gestation did not result in permanent hematochimerism. It remains to be determined whether the engraftment process can be promoted by earlier transplantations and/or higher cell doses.

A mathematical model of umbilical venous pulsation.

Pulsations in the fetal heart propagate through the precordial vein and the ductus venosus but are normally not transmitted into the umbilical vein. Pulsations in the umbilical vein do occur, however, in early pregnancy and in pathological conditions. Such transmission into the umbilical vein is poorly understood. In this paper we hypothesize that the mechanical properties and the dimensions of the vessels do influence the umbilical venous pulsations, in addition to the magnitude of the pressure and flow waves generated in the fetal atria. To support this hypothesis we established a mathematical model of the umbilical vein/ductus venosus bifurcation. The umbilical vein was modeled as a compliant reservoir and the umbilical vein pressure was assumed to be equal to the stagnation pressure at the ductus venosus inlet. We calculated the index of pulsation of the umbilical vein pressure ((max-min)/mean), the reflection and transmission factors at the ductus venosus inlet, numerically and with estimates. Typical dimensions in the physiological range for the human fetus were used, while stiffness parameters were taken from fetal sheep. We found that wave transmission and reflection in the umbilical vein ductus venosus bifurcation depend on the impedance ratio between the umbilical vein and the ductus venosus, as well as the ratio of the mean velocity and the pulse wave velocity in the ductus venosus. Accordingly, the pulsations initiated by the fetal heart are transmitted upstream and may arrive in the umbilical vein with amplitudes depending on the impedance ratio and the ratio between the mean velocity and the pulse wave velocity in the ductus venosus.

Methodology and basic problems related to blood flow studies in the human fetus.

A method was developed for non-invasive measurement of human fetal blood flow. The method combines real-time ultrasonography with 2 MHz pulsed Doppler technique. The blood flow is calculated from the blood velocity, estimated from the Doppler spectrum, and the vessel diameter, measured in the real-time image. Time-distance recording was applied for measurements of the pulsatile diameter changes in the fetal aorta. The method proved to possess a good accuracy and reproducibility when tested in vitro experiments and in a comparison with electromagnetic flow measurements in animals. Possible sources of error were analysed and recommendations for minimizing the risk of errors are presented.

Ultrasound during pregnancy and birthweight, childhood malignancies and neurological development.

The present paper summarizes some of the epidemiological studies of in utero ultrasound exposure and subsequent childhood development. Emphasis is placed on birthweight, childhood malignancies and neurological development. A meta-analysis, including neurological outcomes such as handedness, speech development, motor development, hearing and vision, is presented. The epidemiological evidence does not indicate any association between diagnostic ultrasound exposure during pregnancy and reduced birthweight, childhood malignancies or neurological maldevelopment. The possible association between ultrasound and nonrighthandedness among boys needs further evaluation.

Blood flow in the fetal descending aorta; intrinsic factors affecting fetal blood flow, i.e. fetal breathing movements and cardiac arrhythmia.

An ultrasonic method combining real-time ultrasonography and pulsed Doppler technique was used for the examination of blood flow in the fetal descending aorta. The mean aortic blood flow velocity in the last trimester of normal pregnancies was 29.0 cm/s; the peak maximum velocity 97.3 cm/s and the mean blood flow 238.4 ml/min/kg. The blood flow velocity did not change significantly with gestational age, the aorta diameter showed a linear growth. During labour, the aortic blood flow in undistressed fetuses was not different from the flow recorded during late pregnancy. Fetal breathing movements modulate the flow velocity signals in the descending aorta, the inferior vena cava and the umbilical vein of the fetus; therefore, when quantifying fetal blood flow, only periods without fetal breathing movements should be considered. A group of fetuses with various types of cardiac arrhythmias was examined. Postextrasystolic potentiation was found to be present already during intrauterine life. The present method enables quantitative evaluation of the hemodynamic effects of cardiac arrhythmias.

Estimation of the pressure gradient across the fetal ductus venosus based on Doppler velocimetry.

In the fetus, the umbilical vein is directly linked to the inferior vena cava by the narrow ductus venosus. Thus, the ductus venosus blood velocity probably reflects the pressure gradient between the umbilical vein and the central venous system. In a longitudinal study that included 29 normal fetuses, pulsed Doppler velocimetry was carried out in the umbilical vein and the ductus venosus during the last half of the pregnancy. By applying the Bernoulli equation, we estimated the pressure gradient across the ductus venosus to vary between 0-3 mm Hg during the heart cycle; it remained within those ranges during gestational weeks 18-40. During fetal inspiratory movement, pressure gradients up to 22 mm Hg were estimated. The estimated ductus venosus pressure gradient seems to be within ranges compatible with known umbilical venous pressures, and may provide a new opportunity to understand central venous hemodynamics and respiratory force in the fetus once methodological limitations are controlled.

The use of ultrasound in determining the initiation of treatment in instability of the hip in neonates.

We have evaluated the effect of the use of ultrasound in determining the initiation of treatment in neonatal instability of the hip. A total of 99 newborn infants (1.5% of all live births) with neonatal hip instability did not have treatment from birth, but were re-examined at eight to 15 days. In the 31 who had persisting clinical instability and ultrasound abnormality, treatment was then started with a Frejka pillow. The hips in the remaining 68 infants showed spontaneous clinical stabilisation and improvement of the ultrasound findings. Treatment was therefore withheld. There was a marked trend towards normal development in mildly unstable hips, whereas no hips with severe instability did so spontaneously. Further follow-up showed normal development in all the hips which had been treated, and in all except five of the 68 untreated infants. These five infants showed persistent hip dysplasia on both ultrasound and radiological examination at four to five months of age. Treatment with an abduction splint was then started and their hips developed normally. Ultrasound is very useful in deciding on treatment if the examiners have adequate experience with the method. Its use substantially reduces the rate of treatment. Spontaneous resolution occurred in more than half of the unstable hips. Since five of the untreated infants developed hip dysplasia a strict follow-up is essential to identify and treat these cases.