Abstract:
Background: Ultrasound to estimate fetal weight is done often in our country as part of investigations of pregnant women. In most of the developed countries and some developing countries, studies have been done to determine the validity of an ultrasound in estimating fetal weight, but no similar studies which have been done in Tanzania.
Objective: To determine validity of ultrasound in estimating fetal weight in singleton pregnancies at MNH.
Methodology: A diagnostic study was conducted from 1st June 2011 to 31st December 2011 at Muhimbili National Hospital Dar es salaam. All women who met inclusion criteria with indication for elective delivery were identified. Data on demographic, obstetric, social and medical history, the indication for elective delivery was collected. Ultrasound to estimate fetal weight was done 24 hours before delivery. The actual birth weight of the fetus was measured by using a desktop baby scale weighing machine. The estimated fetal weight by ultrasound was correlated with direct birth weight as gold standard. Data was entered in epi info version 6 and analyzed by using SPSS version 16.
Results: A total of 800 pregnant women were recruited for the study. All of women consented and underwent an ultrasound estimation of fetal weight 24 hours before delivery.
The median age was 30 years (range 25-34years). Most were married or cohabiting (97. 8%). about half of the participants (51.2%) had primary school education. The median actual birth weight was 3040 g (range 1200–5000 g). One hundred and twenty two infants (15.2%) weighed less than 2500 g and 31 (3.9%) weighed more than 4000 g.
Good correlation was established between actual birth weight and estimated birth weight(r = 0.892, p<0.001), the sensitivity of ultrasound in detecting birth weight below 2500g was 59.8% and specificity was 99.3%. In the study population in estimating low birth weight, there was positive predictive value of 93.6%, negative predictive value of 93.2% and likelihood ratio of 87. The sensitivity of ultrasound in detecting birth weight more than 4000g was 54.8% and specificity was 97.8%. In this study population there was positive predictive value of
47.2%, negative predictive value of 98.2% and likelihood ratio of 22 in estimating large birth weight babies.
Conclusion; In a clinical setting Ultrasound is useful in diagnosing low birth weight babies while its usefulness in detecting large birth babies is questionable. As the sensitivity of ultrasound estimation of fetal weight to detect larger babies is poor, the use of such an objective measurement in the management of suspected macrosomia in singleton pregnancies should complement other clinical diagnostic methods.
