Comparison of Various Clinical Methods of Birth Weight Estimation in Term Pregnancy

Background: Knowledge of fetal weight in utero is vital for the obstetrician in deciding whether or not to deliver the fetus as well as in fixing the mode of delivery. Both low birth weight and excessive fetal weight at delivery are associated with increased risk of newborn complications during labor and the puerperium. Various clinical formulae like Johnson’s formula & Dare’s formula are in use for fetal weight estimation. Objectives: The aim of this study was to assess the fetal weight in term pregnancies by various clinical methodsDare’s formula & Johnson’s formula and to compare the methods after knowing the actual weight of the baby after birth. Study Design: It is a prospective observational study of 227 women at term pregnancy at GMERS medical college & Hospital, sola, Ahmadabad from April 2014 to April 2016. The formulas used in this study are: Johnson’s formula and Dare’s formula. The measurements were compared with actual birth weight after the birth of baby. Results: Results vary in terms of accuracy with various methods employed for estimating the fetal weight. This study showed that Dare’s formula was the best method for fetal birth weight estimation. Conclusion: Dare’s formula is an inexpensive method for fetal birth weight estimation. It continues to be used in many countries on large scale because of its low cost, ease of use, and need for little training as the setup for ultra sonographic evaluation is not readily available in rural setups.

This was a prospective observational type of study. The prevalence of full term antenatal patient in our institute was 30%. So with 90% confidence interval and 5% allowable error, the desired sample size was as follows N=p(1-p)×z÷(me) 2 Where, N=sample size p=prevalence 30% z =confidential limit which is 90%=1.645 Methodology 227 antenatal patients were included in our study as per inclusion criteria. All the eligible patients were given patient information sheet and were included in study after obtaining written informed consent. The patient was then asked to empty her bladder and her symphysiofundal height (SFH) and abdominal girth (AG) were measured using a flexible, non-elastic, standard measure tape. Both measurements were performed with the patient lying flat on her back, with her legs extended. The SFH was measured from the midpoint of the upper border of the pubic symphysis to the highest point of the uterine fundus. For the AG measurement, the tape was repositioned to encircle the woman's waist, at the level of the umbilicus, without applying pressure to tighten the tape around the abdomen. Then pelvic examination was performed to evaluate degree of descent (station) of the fetal head into the pelvis. Both measurements (SFH and AG) and information on the fetal station were recorded on the individual data sheet and later used to calculate the fetal weight according following formula; A.
Dare's formula: Weight in grams=AGxSFH B. Johnson's formula: Weight in grams=155x(SFH-X) X=13 when presenting part at minus station X=12 when presenting part is at 0 station X=11 when presenting part is at plus station Immediately after delivery, Actual Birth Weight (ABW) was measured using a digital balance. If delivery did not occur within a week of estimations, the estimations were repeated and repeat estimations were taken in to consideration. Estimated Birth Weight (EBW) calculated by using various methods was compared with actual birth weight after the birth of baby.
The average error, average percentage error, correlation of coefficient, standard deviation of prediction error and P value were calculated with the help of statistician and analysis was done. A P value <0.05 was considered significant. Student paired t test was used for comparison of various method of birth weight estimation with actual birth weight. Analysis of variance (ANOVA) test was used for comparison of different methods of birth weight estimations with each other.

Results
The demographic characteristics of the study population are shown in Table 1. In our study the mean maternal age was 25.63±3.68 years and the median was 25 years (range 18-35 years). The mean parity was 0.89±0.82 and the median was 1 (range 0-3). The mean maternal BMI was 21.57±2.49kg/m 2 and the median was 21.51kg/m 2 (range 15.4kg/m 2 -31.39kg/m 2 ). The mean gestational age at the time of delivery was 38.73±0.83weeks and the median was 38.6 weeks (range 37-40 weeks). Table 2 shows mean error of overestimation and underestimation by various clinical methods. Mean error in over and under estimations was more in Johnson's formula than Dare's formula. Mean error of over estimation were 166.30gm and 324.31gm for Dare's formula and Johnson's formula respectively. In case of under estimation, mean error was least with Dare's formula as compared to other method. (101.33gm and 111.25gm for Dare's formula and Johnson's formula respectively).   Table 3 shows the mean fetal weight measured by clinical and ultrasound methods. Mean actual birth weight at the time of delivery was 2827±365gm in study population. Maximum ABW was 4120gm while minimum was 2125gm in present study. The mean estimated birth weights (EBW) by Dare's formula and Johnson's   Table 4 shows the errors in Birth weight estimation by various methods compared to actual Birth weight. The mean error for Dare's formula and Johnson's formula were -92.92±159.43gm and -284.51±215.43gm respectively. (Difference was statistically significant as P value <0.0001) The mean absolute error for Dare's formula and Johnson's formula were 147.37±110.78gm and 304.11±186.60gm respectively. (Difference was statistically significant as P value <0.0001). The mean percentage error for Dare's formula and Johnson's formula were -3.510±5.688% and -10.529±7.983% respectively. (Difference was statistically significant as P value <0.0001). The mean absolute percentage error for Dare's formula and Johnson's formula were 5.298±4.067% and 11.091±7.178% respectively. (Difference was statistically significant as P value <0.0001).

Main findings and comparison with existing literature
Numbers of articles had been published in various journals regarding comparison of various method of birth weight estimation. They also include USG as one of the method of birth weight estimation. Raghuvanshi et al. [11] had found that mean of estimated fetal weight by Dare's formula (2696±394.2gm) were almost closer to the mean of actual birth weight (2593±427gm) but comparing it with ABW, difference was found statistically significant. (P value was 0.01) Amritha et al. [12] had conducted a study for comparative analysis of accuracy of various method of birth weight estimation on 200 antenatal women. Standard deviation of prediction error was less with Dare's formula compared to Johnson's formula in their study (272.66 gm compared to 309.98 gm) In present study, the mean error for Dare's formula and Johnson's formula were -92.92±159.43gm and -284.51±215.43gm respectively (statistically significant as P value <0.0001).Maximum error in estimated birth weight was found highest with Johnson's formula between two clinical methods in studies conducted by Amritha et al. [12] and Raghuvanshi et al. [11] (1135gm and 1771gm respectively). In present study, maximum error was found highest with Johnson's formula (980gm). Amritha et al. [12] had found that Dare's formula had least average error (224.37gm) between two clinical methods of birth weight estimations. In present study, least average error (147.37±110.78 gm) was found with Dare's formula estimated EBW. Number of cases with accuracy of fetal estimates within 10% of ABW was higher with Dare's formula in study conducted by Raghuvanshi et al. [11] (65%) and Amritha et al. [12] (67%). In present study we found 89.43% of cases with error of <10% ABW with Dare's formula. Johnson's formula had only 48.46% of cases with fetal birth weight estimates within 10% of ABW. Limitation: Sample size of present study was small and present study was conducted in single center. Multi centric study should be conducted to validate the observation made in our study. Present study included full term babies with cephalic presentation without any complication. Application of this data may not be useful in determining low birth weight babies in complicated pregnancy. It is necessary to conduct study to verify its usefulness in detecting low birth weight babies.

Conclusion and implications
Among the clinical methods, Dare's formula is more accurate than Johnson's formula. Average error in estimated birth weight compared to ABW is least with Dare's formula. Dare's formula is more accurate in estimating birth weight within 10% of ABW as compared to Johnson's formula. Both methods of birth weight estimations in our study have positive linear correlation with ABW. As the actual birth weight increases the estimation of birth weight by clinical methods also increases. In all term pregnant women, birth weight should be estimated before delivery as it helps us to predict maternal and perinatal outcome. In low resource settings, the Dare's formula should be used as against the Johnson's formula for estimating birth weight.