Objective: to Compare Clinical and Sonographic Estimation of Birth Weight Using Hadlock's

Antepartum Detection of Macrosomic Fetus: Clinical Versus Sonographic, Including Humeral Soft Tissue Thickness

Nadia Mudher Sulaiman Al-Hilli

Dept. of Gynaecology and Obstetrics College of Medicine, University of Babylon, Hilla, Iraq.

Abstract Objective: To compare clinical and sonographic estimation of birth weight using Hadlock's (1) equation with new estimation technique that involve measurement of fetal humeral soft tissue thickness to identify newborns with birth weight of at least 4000 g. Patients and Methods: A prospective study conducted in Gynaecology & Obstetrics Department in Al-Yarmook Teaching Hospital, Baghdad, Iraq. Ninety pregnant women were studied between October 2003 and September 2004. They had gestational ages of 37 weeks or more and a suspicion of fetal macrosomia based on the presence of one or more of certain risk factors. Clinical estimation of fetal body weight using Leopold's manoeuvre was done followed by sonographic fetal weight estimation using Hadlock's (1) equation. Fetal humeral soft tissue thickness (the distance from the outer edge of the humerus to the skin surface on transverse views of the upper arm) was measured by ultrasound. Then a comparison of the three methods mentioned above was done regarding their validity in predicting fetal macrosomia. Results: Sonographic fetal humeral soft tissue thickness correlates with birth weight and found to be higher in macrosomic than the non-macrosomic newborns (14.35mm versus 11.6mm) and the difference was statistically significant (P value <0.001). The sonographic fetal humeral soft tissue thickness measurement was more sensitive in predicting fetal macrosomia than the sonographic fetal weight estimation (87.2 versus 75%) but less specific (74.2 versus 86%). The positive predictive value was 84.2 versus 89% and the negative predictive value was 78.7 versus 68% respectively while the clinical estimation had the lowest accuracy in predicting fetal macrosomia compared with sonographic fetal weight estimation and sonographic fetal humeral soft tissue measurement. Conclusion: The sonographic measurement of fetal humeral soft tissue thickness positively correlates with newborn birth weight. It is more accurate than the clinical fetal weight estimation in predicting fetal macrosomia. On the other hand it is more sensitive but less specific than the sonographic fetal weight estimation using Hadlock's (1) equation in predicting fetal macrosomia.

الخلصة الغرض من هذه الدراسة هو مقارنة الفحص السريري والفحص بالمواج فوق الصوتية باستعمال معادلة هادلوك (1) لقياس وزن الجنين مع طريقة جديدة وهي قياس سمك النسجة الرخوة لمنطقة العضد للجنين باستعمال جهاز الفحص بالمواج فوق الصوتية لغرض اكتشاف المواليد الذين يزنون عند الولدة 4000 غراما على القل. اشتملت الدراسة فحص 90 امرأة حامل عمر حملهن 37 اسبوع أو أكثر و لديهن احتمال ضخامة جسم الجنين اعتمادا على وجود عامل خطورة واحد أو أكثر. تم فحص النساء الحوامل سريريا بطريقة (ليوبولد) لتقدير وزن الجنين ثم حساب وزن الجنين باستعمال جهاز الفحص بالمواج فوق الصوتية و باستخدام معادلة هادلوك (1), ثم تم قياس سمك النسجة الرخوة لمنطقة العضد للجنين و الذي يتضمن قياس المسافة من الحافة الخارجية لعظم العضد الى سطح الجلد في المقطع العرضي. أظهرت النتائج أن قياس سمك النسجة الرخوة لمنطقة العضد بجهاز المواج فوق الصوتية يتناسب طرديا مع وزن الجنين كما انه أكثر دقة من تقدير وزن الجنين سريريا و أكثر حساسية من حساب وزن الجنين بالمواج فوق الصوتية. 7- Race and ethnicity: Studies have Introduction demonstrated that Hispanic women acrosomia is a term used have a higher risk of fetal macrosomia rather imprecisely to compared with white, African, Mdescribe a very large fetus- American or Asian women [1]. neonate[1]. Fetal macrosomia has been Consequences of Fetal Macrosomia defined in different ways, including Fetal Consequences birth weight of 4000-4500 g or greater I- Shoulder dystocia: It describes a than 90th percentile for gestational age range of difficulties encountered with after correcting for neonatal sex and delivering the shoulder after delivering ethnicity [2,3]. For example, the 90th the head. It is the most feared percentile at 39 weeks is 4000 g complication of macrosomia and up to whereas the corresponding birth weight one fourth of infants with shoulder at 42 weeks is 4400.[1] dystocia experience brachial plexus or Fetal macrosomia affects between 3% facial nerve injuries, or fractures of the and 15% of all pregnancies, depending humerus or clavicle. It can also lead to upon the racial and socioeconomic asphyxia and hypoxic encephalo- composition of the population under pathy[10] study. [4] II-Birth Trauma: Injuries to the infant Risk Factors for Fetal Macrosomia resulting from mechanical factors 1- Diabetes Mellitus: Delivery of an (compression, traction) during the infant weighing more than 4500 g process of birth are categorized as birth occurs ten times more often in diabetic trauma. women than in a non-diabetic Other consequences include: population [5]. intrapartum asphyxia [4], neonatal 2- Genetic Factors: Large size of the hypoglycemia [11], neonatal parents especially obesity of the hypocalcemia [12] and macrosomic mother [1]. Overweight and obese cardiomyopathy[11]. women are at increased risk for having Maternal Consequences large for gestational age infant and 1. Increased incidence of operative caesarean delivery[6]. vaginal delivery and cephalopelvic 3- Multiparity. disproportion. 4- Prolonged gestation: As many as 2. Increased incidence of cesarean 45% of the fetuses undelivered at their section. expected date of delivery continue to 3. The risk of third or fourth degree grow in utero. With each additional lacerations are increased five folds[13]. week, more babies have a birth weight 4. Increased incidence of postpartum more than 4000 g. hemorrhage[13]. 5- Male Fetus: Male infants typically Strategies to Predict Fetal weigh more than female infants at any Macrosomia gestational age[7, 8]. The three major strategies used to 6- Previous infants weighing more than predict macrosomia are: 4000 gram: Women with a previous  Clinical risk factors. delivery of an infant weighing >4000 g  Clinical estimation by Leopold’s are at increased risk of giving birth to a manoeuvre: Tactile assessment of large for gestational age infant [9]. thickness to identify newborns with fetal dimensions through the excessive weight. maternal abdomen.  Ultrasonography: Patients and Methods The ultrasonographic technique This study was conducted for a represents the newest and most period of one year, from the start of technologically sophisticated method October 2003 to the end of September of obtaining birth weight 2004, in the Obstetrics and estimation[14]. Different models of Gynaecology Department of Al- ultrasound estimation of fetal weight Yarmook Teaching Hospital. The have been proposed by Hadlock 1and study included 90 pregnant women 2, Birnholz, Deter et al , Jordan, with gestational ages of 37 weeks or Shepard, and Warsof et al , using more who had at least one of the biperietal diameter (BPD), following risk factors for fetal occipitofrontal diameter (OFD), macrosomia: anteroposterior and transverse  Gestational diabetes. abdominal diameters (AD1 and AD2)  Pre-existing diabetes. and femur length (FL) in centimeters  Postdate pregnancy. [15].  Prior delivery of a macrosomic Technical limitations of ultrasound infant. estimation Patients either presented in early labor Several technical limitations of the or for induction of labor or for sonographic technique for estimating caesarean section (i.e. prior to fetal weight are well-known. Among delivery). Patients included in the these are maternal obesity, anterior study met the following criteria: placentation, and oligohydramnios. 1. Singleton gestation. Therefore, investigations have been 2. Intact fetal membranes if the attempted to improve the accuracy of patient presented in early labor. ultrasound in predicting fetal weight. 3. The gestational age was Landon et al showed that fetal humeral considered to be reliable based soft tissue thickness measurement (the on the last menstrual period, distance from the outer edge of the regarding that the patient was humerus to the skin surface on sure of her LMP and had transverse view of the upper arm) may regular menstrual cycles, and/or distinguish disproportionately large on early ultrasound estimation fetuses that may be at risk for difficult of the gestational age. delivery. Because macrosomic infants 4. No evidence of congenital tend to have increased subcutaneous abnormality by ultrasound and adipose tissue, we sought to determine after delivery. the usefulness of an objective General examination including assessment of fetal humeral soft tissue maternal body weight and height was thickness in estimating birth weight in performed, clinical estimation of fetal a population at risk for body weight was done using Leopold’s macrosomia[16] maneuver and blood for measuring 2- hours postprandial blood sugar was Aim of study sent (screening of diabetes in To compare clinical and sonographic pregnancy). Patients then referred for estimation of birth weight using ultrasound study. Ultrasound Hadlock's(1) equation with a new examinations were performed using estimation technique that involves “Siemens, Sonoline Versa” ultrasound measurement of humeral soft tissue different categories of women. The scanner. A 3.5 MHz curvilinear mean humeral soft tissue thickness was abdominal transducer was used to statistically higher in infants weighing obtain morphometric measurements at least 4000 g than in those weighing including the fetal abdominal less than 4000g (14.35 mm versus 11.6 circumference and femoral length. mm) P value < 0.001. The difference in Abdominal circumference and femoral HSTT between diabetic and non- length were measured, Estimated fetal diabetic was not statistically weight (EFW) was calculated using significant. Figure (2) shows the Hadlock 1 formula [40]: correlation between sonographic fetal Log10 (BW) =1.3598 + 0.051 (AC) humeral soft tissue thickness +0.1844 (FL) - 0.0037 (AC*FL) measurement and birth weight which The fetal humerus was visualized was statistically significant (p < 0.001). in a longitudinal view, and the Table (3) and figure (3) show the transducer was rotated 90° then moved accuracy of clinical estimation of fetal cephalad until the head of the humerus weight in prediction of fetal was found. The measurement was macrosomia in comparison with taken immediately below the humeral sonographic fetal weight estimation head from the outer edge of the bone to and sonographic measurement of fetal the skin surface. The humeral soft humeral soft tissue thickness. tissue thickness measurement was Sonographic measurement of fetal performed three times at the same humeral soft tissue thickness had the occasion, and the average of the three highest sensitivity (87.2%) and values was compared with the negative predictive value (78.7%), estimated fetal weight for its ability to while the sonographic fetal weight predict macrosomia. After delivery, the estimation had the highest specificity newborn’s body weight was assessed. (85.7%) and positive predictive value Statistical analysis: Data were (89%), while the clinical estimation of analyzed with unpaired t test and fetal weight had the lowest accuracy correlation analysis. Sensitivities and compared with sonographic fetal specificities were calculated from 2x2 weight estimation and sonographic tables. P value <0.05 was considered fetal humeral soft tissue thickness statistically significant. measurement. Results Discussion Table (1) and Figure (1) show a In a study done by Karim et comparison of maternal characteristics al[17], who analysed the prevalence between women who delivered and outcome of macrosomia in macrosomic infants and those who did Pakistan, he found that maternal age not. A statistically significant over 35 years, obesity, difference was found regarding grandmultiparity, postmaturity and maternal weight (87.8 kg versus 84.1 impaired glucose tolerance were kg); height (165.6 cm versus 163.7 associated with fetal macrosomia. cm), parity (5 versus 3) and gestational Similar results were found by Meshari age (39.8 week versus 38.6 week), et al [18] in Saudi Arabia. However, in whereas the maternal age and 2-hours our study maternal age was not found postprandial blood sugar were not to be significantly higher in the statistically significant. macrosomic group probably because Table (2) shows the mean (± the ages of women in our sample were standard deviation) of the sonographic not so diverse and most of their ages fetal humeral soft tissue thickness in tissue thickness was performed for 93 were around 30 years. On the other women with gestational diabetes hand, we did not find 2-hours mellitus during the third trimester. He postprandial blood sugar to be proved that this measurement was the statistically significant as a risk factor most accurate predictor of excessive for fetal macrosomia. We may need to birth weight compared with other do oral glucose tolerance test or even standard ultrasound parameters (i.e. 1-hour glucose screen to detect abdominal circumference, femoral impaired glucose tolerance in women length and others). It had a sensitivity with macrosomic fetuses. Chauhan et and specificity of 82 and 95% al (1998) [19] had studied 661 patients respectively and a positive predictive and compared the accuracy of different value of 90%. In our study, we methods for prediction of term fetal assessed the accuracy of sonographic macrosomia of greater than 4000g. He measurement of fetal humeral soft found that clinical estimation of fetal tissue thickness in predicting fetal weight had a sensitivity and specificity macrosomia in comparison with of 54 and 95% respectively, a positive clinical and sonographic fetal weight predictive value of 60% and a negative estimation. It had the highest predictive value of 93%. In our study sensitivity (87.2%) and negative clinical estimation of fetal weight by predictive value (78.7%), while the Leopold's maneuver found to have a specificity was 74.2% and the positive sensitivity and specificity of 64 and predictive value 84.2% which were 77% respectively, a positive predictive less than that of sonographic fetal value of 81% and a negative predictive weight estimation. value of 57%. The difference may be because we were more cautious and Conclusions our attention was more directed toward 1. The sonographic measurement of the prediction of fetal macrosomia. In fetal humeral soft tissue thickness addition, we have assessed in our study positively correlates with newborn the accuracy of the sonographic fetal birth weight. weight estimation in prediction of fetal 2. The sonographic measurement of macrosomia and found it to have a fetal humeral soft tissue thickness was sensitivity and specificity of 74.5% found to be more accurate than the and 85.7% respectively, a positive clinical estimation of fetal weight in predictive value of 89% and a negative prediction of fetal macrosomia. predictive value of 69%. O'Reilly and 3. The sonographic measurement of Divon (1997) [20] had evaluated areas fetal humeral soft tissue thickness was under receiver operating characteristic found to be more sensitive but less curves for sonographic estimated fetal specific than the sonographic fetal weight as a predictor of fetal weight estimation using Hadlock's (1) macrosomia in prolonged pregnancies. equation in prediction of fetal The sensitivity, specificity, and macrosomia. positive and negative predictive values were 85, 72, 49, and 94% respectively. Recommendations This wide variation in the validity of 1. Because of its sensitivity, we the test may be due to different recommend sonographic measurement sonographic scanner machines used of fetal humeral soft tissue thickness as and different sonographers i.e. additional parameter to predict fetal interobserver bias. In a study done by macrosomia as it is easy to be applied. Landon et al [16], sonographic measurement of fetal humeral soft abdominal circumference and femoral 2. Introduce a new formula for length, so that the validity of sonographic fetal weight estimation sonographic prediction of fetal which includes fetal humeral soft macrosomia can be increased. tissue thickness measurement in addition to other parameters i.e.

Table 1 The maternal characteristics in the group under study.

Significant fetal weight t test p value characteristics 4000g≤ 4000g> Mean ± SD Mean ± SD (Age(year 3.8 ± 30.182 4.04 ± 28.6 0.0687901 NS (Weight(Kg 7.3 ± 87.873 6.3 ± 84.17 0.0132565 S (Height(cm ±1.6 165.6 2.2 ± 163.74 0.0000078 S Parity 2.19 ± 5.1636 2.4 ± 3.74 0.0063531 S (GA(week 1.08 ± 39.836 1.43 ± 38.68 0.0001416 S 2-hrs postprandial 13.9 ± 130.98 8.51 ± 128.45 0.2883651 NS BS(mg/dL) GA: Gestational Age BS: Blood Sugar S: Significant NS: Not Significant

180 160 140 120 100 80 60 40 20 0 ≥ 4000g weight height parity GA postprandial <4000g

Figure 1 The maternal characteristics.

Table 2 The sonographic measurement of fetal humeral soft tissue thickness

Fetal Humeral Soft Tissue Thickness Mean(mm) ±Standard deviation All subjects 13.27 1.79 Birth weight < 4000 g 11.6 1.24 Birth weight ≥ 4000g 14.35 1.20

The difference between macrosomic & non macrosomic groups was statistically significant. P value using unpaired t test < 0.001 The correlation between birth weight & HSTT

5000 4500 4000 3500 t h

g 3000 i e w

2500 h t r

i 2000 B 1500 1000 500 0 0 5 10 15 20 HSTT Figure 2 Birth weight versus sonographic fetal humeral soft tissue thickness

Table 3 Comparison of the validity of clinical estimation, sonographic fetal weight estimation and sonographic fetal HSTT in predicting fetal macrosomia

Sensitivity Specificity Positive Negative Accuracy predictive value predictiv e value Clinical 63.6% 77.1% 81.3% 57.4% 68% estimation sonographic 75% 86% 89% 68% 80% fetal weight estimate sonographic 87.2% 74.2% 84.2% 78.7% 82% fetal HSTT When the actual weight cut off point is 4000 g HSTT: Humeral Soft Tissue Thickness. 100 80 Clinical 60 Estimation 40 20 Ultrasound 0 HSTT

Figure 3 Comparison of the accuracy of clinical estimation, sonographic fetal weight estimation and sonographic fetal HSTT in predicting fetal macrosomia weighing at least 4500 gram. Obstet References Gynecol 1995;85: 558-564. 1. Cunningham, MacDonald, Gant, 9. http://www.tdh.State.tx.us / wichd/ Leveno: Fetal growth disorders- nut/ pdf/ 337pdf. History of Birth of Macrosomia. William’s Obstetrics, Large for Gestational Age Infant. 21st edition 2001. USA. Appleton and 10. Zamorski MA, and Biggs WS: Lange, chapter 29:744-764. Management of Suspected Fetal 2. Jazayari A: emedicine- Macrosomia. American Family Macrosomia. June 26, 2003. (internet) Physician January 2001; 63:302-6. 3. Luis Sanchez-Ramoz, Bernstien S, 11. Milligan DA: Fetal Macrosomia at Kannitz AM: Expectant management Term. Newsletter. January, 1996. versus labor induction for suspected 12. Wollschlaeger K, Nieder J, Koppe macrosomia: A Systematic Review. I: A study of fetal macrosomia. Arch Obstetrics &Gynecology November Gynecol Obstet 1999; 293: 51-55. 2002, vol.100, no.5, part 1, 997-1002. 13. David P, MD, Colleen M. and 4. Nahum G G :Detecting and Stephen K."Intrauterine Growth managing fetal macrosomia. Restriction: Identification & Contemporary OB/Gyn® Archive. management" American Family June 1, 2000. (internet) Physician August 1998. 5. Landon MB, Gabbe SG: Diabetes 14. Nahum G: Estimation of fetal in pregnancy. High Risk Pregnancy weight. emedicine, Obstetrics/ Management Options. 2nd edition Gynecology. July 11, 2002. London, W.B. Saunders 1999; 38: 15. Hebber S: Critical Evaluation of 665-684. Estimating Fetal Weight by 6. Kaufman H, Coleman S, Ganzeier Ultrasound. J Obstet Gynecol 2003; G: Prepregnancy body mass index and 53: 2:131-133. pregnancy outcome. Obstetrics & 16. Landon MB, Sonek J, Foy P: Gynecology April 2001; 97: 4: 71-73. Sonographic measurement of fetal 7. Rodrigues T, Teles TP, Barros H: humeral soft tissue thickness in Risk factors for macrosomia in infants pregnancy complicated by GDM. of non-diabetic women. Arquivos De Diabetes. 1991; 40: (2):66-70. Medicina 1999; 5: 20-23. 17. Karim SA, Mastoor M, Ahmed 8. Lipscom KR, Gregory K, Shaw K. AJ: Macrosomia: maternal and fetal The outcome of macrosomic infants outcome. Asia Oceania J Obstet Gynaecol. March 1994 ; 20(1):73-6. 18. Meshari AA, De Silva S, Rahman I: Fetal macrosomia- maternal risks and fetal outcome. Int J Gynaecol Obstet. July 1990; 32(3): 215-22. 19. Chauhan SP, Hendrix NW, Magann EF: Limitations of clinical and sonographic estimates of birth weight: experience with 1034 parturients. Obstet Gynecol 1998 Jan; 91(1):72-7. 20. O'Reilly-Green CP, Divon MY: Receiver operating characteristic curves of sonographic estimated fetal weight for prediction of macrosomia in prolonged pregnancies. Ultrasound Obstet Gynecol. June 1997; 9 (6): 403-408.