Introduction: Fetal weight estimations are essential to identify and treat high-risk pregnancies. According to the Swedish association for obstetricians and gynecologists, the mean deviation of estimated weight from birth weight should be 0%, and the standard deviation 8%. Aim: To evaluate fetal weight estimations, by comparing the deviation from expected weight at the examination to at birth. Methods: Eight examiners were included from the Department of Obstetrics and Gynecology at Västerås Central Hospital. All singleton pregnancies with live births examined by these examiners from week 36+0, from January 1st to February 10th, 2022, were included, with a few exceptions. The deviations mentioned above were compared using a paired T-test. Mean and standard deviations were calculated for individual examiners. Midwives and doctors were compared using an independent T-test. Results: 86 examinations were included. There was a tendency to overestimate birth weight deviation by 1.0%, with a standard deviation of 7.8%. The differences between the deviation at the weight estimation and at birth, and between the professions, were not statistically significant. Both mean and standard deviation varied between individual examiners. Conclusions: The precision found in this study fell within what is considered acceptable according to the SFOG, and the mean did not differ statistically significantly from the recommended. Differences in fetal growth rate between examination and birth may have induced an error in the results, but the magnitude of that error cannot be determined. The study indicates differences between examiners, but a larger study is needed to draw any definite conclusions.

Conselho Nacional de Desenvolvimento CientÃfico e TecnolÃgico / CoordenaÃÃo de AperfeiÃoamento de Pessoal de NÃvel Superior / Objetivos: testar a validade da relaÃÃo diÃmetro transverso do cerebelo (DTC) /circunferÃncia abdominal (CA) como mÃtodo diagnÃstico ultra-sonogrÃfico da restriÃÃo de crescimento fetal (RCF). Determinar, atravÃs de curva ROC (receiver operator characteristic), o melhor ponto de corte da relaÃÃo DTC/CA. Verificar se a relaÃÃo DTC/CA tem sua acurÃcia modificada na dependÃncia do tipo de RCF (simÃtrica ou assimÃtrica) ou do tempo entre a ultra-sonografia e o parto. Comparar DTC/CA, no ponto de corte obtido, com a relaÃÃo comprimento do fÃmur (CF) /circunferÃncia abdominal (CA). MÃtodo: estudo prospectivo, seccional, envolvendo 250 gestantes com gravidez Ãnica, idade gestacional precisa, feto vivo. Foram realizadas ultra-sonografias obstÃtricas atà a resoluÃÃo da gestaÃÃo, mas somente a Ãltima foi considerada para anÃlise. Os neonatos cujas relaÃÃes DTC/CA estiveram maiores do que o ponto de corte determinado pela curva ROC foram considerados acometidos por RCF. Idem para a relaÃÃo CF/CA. Considerou-se como padrÃo-ouro para o diagnÃstico de RCF os recÃm-nascidos cujos pesos situaram-se abaixo do percentil 10 para a idade gestacional nas curvas de Lubchenco et al. (1963), corrigidas para sexo. Definiu-se RCF simÃtrica neonatos com Ãndice ponderal de Rohrer situado entre 2,2 e 3.0. Aqueles com RCF cujos Ãndices fossem < 2,2 foram classificados como RCF assimÃtrica. Resultados: a prevalÃncia da RCF foi de 12,4%. O ponto de corte da relaÃÃo DTC/CA determinado pela curva ROC foi 16,15. A sensibilidade, especificidade, valores preditivos positivo e negativo, acurÃcia, razÃes de verossimilhanÃa positiva e negativa foram de 77,4%, 82,6%, 38,7%, 96,3%, 82%, 4,5 e 3,7, respectivamente. Na RCF simÃtrica a sensibilidade e especificidade foram de 80,8% e 81,7%, respectivamente. Na assimÃtrica a sensibilidade e especificidade foram 60% e 75%, respectivamente. Resultados menores do que na simÃtrica, porÃm, nÃo estatisticamente significantes (p > 0,05). No intervalo de zero a sete dias entre a Ãltima ultra-sonografia e o parto, a sensibilidade e especificidade foram de 81,5% e 82,1%, respectivamente. No intervalo de oito a 14 dias, a sensibilidade e especificidade foram de 50% e 84,3%, respectivamente, sem diferenÃa estatisticamente significante entre os dois intervalos (p > 0,05). O ponto de corte da relaÃÃo CF/CA foi de 22,65, com sensibilidade, especificidade, valores preditivos positivo e negativo, acurÃcia, razÃes de verossimilhanÃa positiva e negativa de 67,7%, 81,7%, 34,4%, 94,7%, 80%, 3,7 e 2,5, respectivamente. ConclusÃes: a relaÃÃo DTC/CA no ponto de corte 16,15 mostrou-se mÃtodo eficaz no diagnÃstico de RCF, tanto simÃtrica quanto assimÃtrica, nÃo sendo influenciada pelo tempo entre a Ãltima ultra-sonografia e o parto. Sendo mÃtodo independente da idade gestacional, à especialmente Ãtil nos casos em que este dado à ignorado. A relaÃÃo CF/CA mostrou-se menos eficaz do que a DTC/CA no diagnÃstico da RCF. / Objectives: to evaluate the validity of transverse cerebellar diameter (TCD)/abdominal circumference (AC) ratio as an ultrasonographic diagnosis method of fetal growth restriction (FGR). To calculate by receiver operator characteristic (ROC) curve the best cut-off value of TCD/AC ratio. To verify whether TCD/AC has its accuracy modified according to the dependence of type of FGR (symmetric and asymmetric) or according to the time between ultrasonography and deliverance. To compare TCD/AC ratio at its cut-off with the femur length (FL)/ abdominal circumference (AC) ratio. Method: a prospective cross-sectional study, carried out in 250 pregnant women with singleton pregnancies between 20 and 42 weeks of gestation, known accurate gestational age with ultrasound confirmation, living fetuses. Obstetrics sonographic examinations were accomplished until gestation resolution, but only the last one, within 14 days of the deliverance, was used for analysis. Neonates with TCD/AC ratio greater than the cut-off, established by ROC curve were diagnosed as FGR. The same was considered for FL/AC ratio. We classified as gold standard for FGR in new-born infants, who presented birth weight bellow 10th percentile of gestational age according to the growth curves of Lubchenco et al. (1963), corrected according to their sex. Neonates showing FGR and Rohrer ponderal index between 2,2 and 3,0 were labeled as symmetric FGR. Those showing FGR and ponderal index below 2,2 were labeled as asymmetric FGR. Results: prevalence of FGR among the study group was 12,4%. The best cut-off value calculated by ROC curve for TCD/AC ratio was 16,15. The sensitivity, specificity, accuracy, positive predictive values and negative predictive values, likelihood ratio for positive and negative tests were 77,4%, 82,6%, 38,7%, 96,3%, 82%, 4,5 and 3,7, respectively. In the symmetric FGR, sensitivity and specificity were 80,8% and 81,7%, respectively. In the asymmetric FGR, sensitivity and specificity were 60% and 75%, respectively. Results lower than in the symmetric FGR, but not statistically significant (p > 0,05). In the interval zero to seven days between sonographic examination and deliverance, sensitivity and specificity were 81,5% and 82,1%, respectively. In the interval of eight to 14 days, sensitivity and specificity were 50% and 84,3%, respectively, with no statistically significant difference (p > 0,05). The best cut-off value calculated by ROC curve for FL/AC ratio was 22,65, showing sensitivity, specificity, accuracy, positive predictive values and negative predictive values, likelihood ratio for positive and negative tests of 67,7%, 81,7%, 34,4%, 94,7%, 80%, 3,7 and 2,5, respectively. Conclusions: TCD/AC ratio at cut-off 16,15 proved to be an effective method in antenatal diagnosis of FGR, both symmetric as asymmetric, with no influence of interval between ultrasonography examination and deliverance. As a gestational age-independent method, it is useful enough in the occurrence of cases where these data are unknown. FL/AC ratio proved is not so effective as TCD/AC ratio in diagnosis of FGR.