132 zyxwvutsrqp BIRTH 20:3 September 1993 zy Accuracy zyxw of Leopold Maneuvers in Screening for Mdpresentation: A Prospective zy Study zy Mona Lydon-Rochelle, CNM, MSN, Leah Albers, CNM, DrPH, Julie Gotwocia, CNM, BSN, Ellen Craig, CNM, MS, and Clifford Qualls, PhD ABSTRACT: This study describes the accuracy of Leopold maneuvers as a screening procedure for fetal malpresentation. The frequency of fetal malpresentation ranges from 15 percent at 32 weeks to 4 percent at term. We prospectively determined fetal presentation by performing Leopold maneuvers on 150 women, followed by a fetal ultrasound examination for comparison. Experienced cert$ied nurse-midwives per- formed the maneuvers with high sensitivity (88%), spec$city (94%), positive predictive value (74%), and negative predictive value (97%) in a population with a I7 percent frequency of fetal malpresentation. We conclude that the maneuvers used by experi- enced clinicians can be effective as a screening tool for fetal malpresentation, particu- larly in settings where ultrasound may not be readily available. (BIRTH zyxw 20:3, Septem- ber 1993) Leopold maneuvers, first described in 1892, have long been used to identify fetal presentation. In the third trimester of pregnancy, the abdomen is sys- tematically palpated in four ways to determine which fetal part occupies the fundus, the location of the fetal back and small parts, which fetal part occu- pies the lower uterine segment, and the position of the presenting part (1-3). As is true for any effec- tive screening test, these maneuvers require high sensitivity to avoid missing problems, and adequate specificity to avoid excess tests that accompany in- correct diagnosis (4). Two studies of the validity of Leopold maneu- vers reported disappointing results (5,6). One re- Mona Lydon-Rochelle, Julie Gorwoda, and Ellen Craig are in the Nurse-Midwifery Division of the Department of Obstetrics and Gynecology; Leah AIbers is in the College of Nursing and Deparlment of Obstetrics and Gynecology; and Clifford Qualls is at the Clinical Research Center, School of Medicine, Univer- sity of New Mexico. This investigation was partially supported by grants from the General Clinical Research Program, DRR, NIH 5 MOI RR00997 1992 and Case Western Reserve University's Frances Payne Bol- ton Alumni Association, Cleveland, Ohio. Address correspondence to Mona Lydon-Rochelle, CNM, MSN, Research Coordinator, Nurse-Midwifery Division, University of New Mexico, 221 1 Lomas Boulevard NE, ACC-4, Albuquerque, N M 87131-5286. ported sensitivity of only 28 percent and specificity of 94 percent (5). Two physicians performed all of the maneuvers. In their third-trimester sample of 154 women, the frequency of malpresentation was only 7 percent. The possibility exists that too few subjects were studied to test the accuracy of the procedure adequately. However, in the other study, which reported sensitivity of 53 percent and speci- ficity of 94 percent, methodologic problems may have biased the results (6). In this study, 20 clini- cians (physicians and nurse-midwives) performed Leopold maneuvers on 176 women, with the num- ber of maneuvers per examiner ranging from 2 to 59. In addition, 10 percent of the evaluations were done on women of less than 28 weeks' gestation. It can be difficult to determine abnormal physical findings clinically among healthy pregnant women because of their low frequency. Fetal malpresenta- tion provides only one example. Breech presenta- tion is an error of polarity that is associated with maternal or fetal abnormalities, and can place the otherwise healthy mother and baby at risk for a complicated vaginal delivery or cesarean section (1,7). Spontaneous version of fetal polarity de- creases as gestation advances, with the reported frequency ranging from 14 to 16 percent at 32 weeks (8,9) to 3 to 4 percent at term (1). Shoulder presentation (transverse lie) occurs