X-ray pelvimetry in clinical obstetrics.

X-ray pelvimetry has been used in obstetrics for over 80 years, but its clinical usefulness remains uncertain. Review of 101 consecutive x-ray pelvimetries obtained by Brown's modification of the Colcher-Sussman technique demonstrated the anteroposterior measurements to be of no value in predicting the eventual method of delivery. However, a transverse of the inlet of less than 13.0 cm and a transverse of the midpelvis of less than 10.0 cm were each associated with an increased frequency of operative delivery. In 94 cases no new information was obtained by x-ray pelvimetry; in the other 7 patients the information obtained was equally accessible by either ultrasonography or abdominal scout film. The authors conclude that x-ray pelvimetry by the technique used in their institution has very limited usefulness in clinical obstetrics.

X-ray Pelvimetry in Clinical Obstetrics MICHAEL W. VARNER, MD, DWIGHT P. CRUIKSHANK, MD, AND DOUGLAS W. LAUBE, MD X-ray pelvimetry has been used in obstetrics for over 80 years, but its clinical usefulness remains uncertain. Review of 101 consecutive x-ray pelvimetries obtained by Brown's modification of the Colcher-Sussman technique demon­strated the anteroposterior measurements to be of no value in predicting the eventual method of delivery. However, a transverse of the inlet of less than 13.0 cm and a transverse of the mid pelvis of less than 10.0 cm were each associated with an increased frequency of operative delivery. In 94 cases no new information was obtained by x-ray pelvi­metry; in the other 7 patients the information obtained was equally accessible by either ultrasonography or abdominal scout film. The authors conclude that x-ray pelvimetry by the technique used in their institution has very limited use­fulness in clinical obstetrics. Informed clinical decision-making requires synthesiz­ing information obtained from appropriate diagnostic procedures with information obtained from a careful history and physical examination. With increasing concern for errors of commission as well as those of omISSIOn, the risks and benefits of diagnostic proce­dures--- including established procedures---must be carefully evaluated. A useful diagnostic procedure must provide valuable information for diagnosis and management; this information must be unobtainable by other means. This paper appraises the present value of x-ray pelvimetry in clinical obstetrics. Materials and Methods A retrospective investigation was conducted at the University of Iowa Hospitals of 101 consecutive pa­tients who underwent x-ray pelvimetry between July 1, 1977, and September 30, 1977. X-ray pelvimetry was performed on 12.6% of laboring patients during the study period. All patients were white. The radio­graphic technique used was Brown's modification of From th~ D~artme~t. of Obstetrics and Gynecology, University of Iowa HospItals and ClImes, Iowa City, Iowa. Submitted for publication March 14, 1980. 296 0029-7844/80/090296-05$01.75 the Colcher-Sussman technique. ' All studies were in­terpreted by resident and faculty physicians of the De­partment of Obstetrics and Gynecology. In analyzing each case attention was paid to the in­dications for x-ray pelvimetry, the course of labor be­fore pelvimetry, the presentation as diagnosed before pelvimetry, the plan of management before and after pelvimetry, and the specific radiographic measure­ments. The course of labor after pelvimetry was re­viewed; data included duration of labor, use of oxyto­cin, additional information gained from pelvimetry, and type of delivery. Results Eighty-five of the 101 patients were primigravidas and 10 were secundigravidas. five of the remaining 6 mul­tigravid women had pelvimetry performed because of breech presentation, and in the sixth patient it was performed because of uncertain fetal presentation. In­fant weights were similar to established birth weight distributions for the University of Iowa Hospitals.2 In­dications for x-ray pelvimetry are noted in Table 1. In 3 cases the clinical diagnosis of presentation was in­correct. Except in 1 case in which a hyperextended fetal head was encountered in a known breech presen­tation, pelvimetry did not prevent a trial of labor. The mean pelvic measurements for both the total population and subgroups are presented in Table 2 ac­cording to method of delivery. The only subgroup with pelvic measurements significantly different from the total group was that of patients who underwent midforceps delivery. In that group the transverse mea­surements of both inlet and midpelvis were signifi­cantly smaller than those of the total population. Pa­tients undergoing cesarean section with a vertex presentation had smaller transverse inlet measure­ments, but the difference was not significant. Analysis of the specific pelvimetry parameters (fig­ures 1 and 2) showed that anteroposterior (AP) mea­surements had no prognostic significance. No change VOl. 56, NO.3, SEPTEMBER 1980 Obstetrics & Gynecology Table 1. Indications for and Outcome of X-ray Pelvimetry Method of delivery Spon- low Mid Cesarean Vaginal Indications N taneous forceps forceps section breech Secondary arrest of labor 59 23 13 6 16 l' Breech presentation 21 2" rt 12 Induction of labor 14 7 2 3 2 Clinically small pelvis 4 3 1 Unknown presentation 3 1 2 " Clinical diagnosis of presentation incorrect. t One patient with hyperextended fetal head on abdominal scout film. Table 2. Pelvic Measurements (cm) (Mean ± SO) Delivery Tinlet APinlet Tmidpelvis AP midpelvis Total group 13.1 ± 1.0 12.2 ± 1.2 10.2 ± 0.8 12.2 ± 1.3 Spontaneous, low 13.2 ± 1.0 12.1 ± 1.1 10.5 ± 0.8 12.2 ± 1.1 forceps Midforceps 12.5 ± 0.9" 12.8 ± 1.2 9.8 ±0.8* 12.7 ±0.9 Cesarean section Vertex 12.8 ± 0.7 12.0 ± 0.8 10.2 ± 0.7 12.0 ± 1.0 Breech 13.5 ±0.8 12.6 ± 1.2 10.8± 0.6 12.9± 1.2 Vaginal breech 13.2 ± 0.8 12.2 ± 1.2 10.4 ± 0.7 12.2 ± 1.4 T = transverse; AP = anteroposterior. " Significantly different from total group, P < .05. All others not significant. AP Inlet 100 DSVD-ELF .Midforceps • Vertex C-Section 63 63 57 DBreech C-Section 50 01 ~ Assisted Breech 0 .1 0 ~I II 0 .... < 11.0 em 11.1-12.0 em )12.1 em c:: CIl t...l. CIl Cl AP Midpelvis 100 50 .1 0 o < 11.0 em 11.1-12.0 em )12.1 em Figure 1. AP pelvimetry measurements compared by method of delivery. Numbers are the percentage of vertex or breech presenta­tions by method of delivery in each measurement category. 56, NO.3, SEPTEMBER 1980 Varner et al X -ray Pelvimetry 297 .1 , r I): ! ~ i ' Table 3. Mengert Indices (Mean ± 50) * Delivery Inlet Midpelvis Total group 159 ± 18 128 ± 16 Spontaneous, low 159 ± 17 128 ± 14 forceps Midforceps 157 ± 16 121 ± 13 Cesarean section Vertex 154 ± 15 124 ± 16 Breech 172 ± 31 143 ± 26 Vaginal breech 158 ± 20 128 ± 12 • No statistically significant differences. in the percentage of operative deliveries was observed among vertex presentations with decreasing AP mea­surements in either the inlet or midpelvis. However, there was a progressive increase in spontaneous vagi­nal delivery as transverse inlet and midpelvis mea­surements increased. When the total population was compared against both the Mengert indices (Table 3) and the CoJcher­Sussman indices (Table 4), no statistically significant differences between groups were observed. Even though the studies by Mengert and CoJcher and Suss­man involved radiographic examination of normal ob­stetric patients,'-' the current data revealed larger mean values in all diameters than did those earlier studies. The use of oxytocin for induction or stimulation of labor after x-ray pelvimetry was also reviewed (Table 5). No patient with breech presentation received oxytocin, but 54 of 81 patients with vertex presenta­tion were stimulated with oxytocin. Twenty of 21 pa­tients with vertex presentation who were eventually delivered by cesarean section received oxytocin stimu­lation after pelvimetry; the remaining patient had re­ceived oxytocin stimulation before pelvimetry, and the cesarean section was performed shortly after com­pletion of the radiographic study. Of the patients who received oxytocin stimulation after secondary arrest of labor, those who underwent an operative delivery gen­erally had longer labors (X = 9.2 hours) than those who subsequently had spontaneous or outlet forceps delivery (X = 7.4 hours). In addition, the average weight of infants delivered operatively was 3630 g, as compared with an average of 3110 g for those infants who subsequently were delivered spontaneously. The cesarean section rate in the pelvimetry group was 28o/cr-the overall cesarean section rate at this in­stitution is 10.9%. With the exception of the case with a hyperextended head found in a breech presentation, all cesarean sections were performed after a trial of la­bor failed. No cesarean sections were performed for fetal distress in this study group. In 94% of patients undergoing x-ray pelvimetry no 298 Varner et al X-ray Pelvimetry Table 4. Colcher-Sussman Indices (Mean ± 50)* Delivery Inlet .- Mldpelvis Total group 25.0 ± 1.7 - 22.8 ± 1.4 Spontaneous, low 25.2 ± 1.4 22.8 ± 1.3 forceps Midforceps 25.1 ± 1.3 22.3 ± 1.2 Cesarean section Vertex 24.8 ± 1.2 22.3 ± 1.4 Breech 26.2 ± 2.0 24.0 ± 2.2 Vaginal breech 25.5 ± 1.4 22.7 ± 1.3 • No statistically Significant differences. additional information was obtained for formulating a prospective labor management plan. In 7 patients the clinical diagnosis of presentation was found to be in­correct, an unknown presentation was diagnosed, Or a known breech presentation was found to have a hy­perextended head. In none of these 7 patients was this information related to maternal pelvic architecture. Discussion Although x-ray pelvimetry has been performed for over 80 years, the risks and benefits of the procedure have been questioned in recent years. The minimum amount of radiation to which the fetus is exposed is approximately 1.1 rads,' but repeating films may in­crease this total to as much as 4 rads. The pioneering work of Stewart et al'·6 showed an increased incidence of childhood malignancies in those children exposed to radiation in utero. Subsequent studies have con­firmed these findings/ although criticism has justifia­bly been raised regarding the high degree of selectivity in these studies.8 Certainly, this questionable long­term risk might be acceptable if x-ray pelvimetry is ef­fective in preventing serious and permanent metabolic or mechanical birth trauma. However, in none of the patients in this study was management altered by ra­diographically determined measurements of pelvic ar­chitecture. The 28% cesarean section incidence is similar to that of previously reported pelvimetry studies!·lo In those patients who were delivered abdominally on the basis of fetopelvic disproportion, x-ray pelvimetry did not Table 5. Use of Oxytocin After Pelvimetry in Vertex Presentation Delivery No. Percent Overall 54/81 67 Spontaneous, low forceps 27/51 53 Mid forceps 7/9 78 Cesarean section 20/21 95 Obstetrics & Gynecology y Transverse Inlet 100 DSVD-ELF 50 o .... c: Cb ~ 46 ~ 50 ~ nI%~ .' 131 I 1132 Up i 0 ~ 12.0 em 12.1-13.0 em 77 ;> 13.1 em • Midforceps • Vertex C-Section o Breech C-Section ~ Assisted Breech Cb 0... Transverse Miqpelvis 100 73· 77 7 0.1 ~I 50 o .1 I .(: 10.0 em 10.1-11.0 em ;> 11.1 em Figure 2. Transverse pelvimetry measurements compared by method of delivery. Numbers are the percentage of vertex or breech presentations by method of delivery in each measurement category. shorten the duration of labor; this finding was also re­ported by Kelly et al.l! In addition, even though those patients with a vertex presentation who eventually un­derwent operative delivery had a smaller pelvis in the transverse diameter, this information was not usable in the prospective management of labor as evidenced by the almost universal administration of oxytocin af­ter arrest of labor, as well as by the increased duration of oxytocin administration before eventual operative delivery. There are no significant differences in the pelvic measurements between those patients with breech and those with vertex presentations. In addition, the pelvis was largest, both by individual and combination in­dices, in those patients with breech presentations who were eventually delivered abdominally. This con­firmed that patients with breech presentations who fail to progress in labor are delivered abdominally without augmentation regardless of pelvic measurements. Al­though the only prospectively significant finding in this breech population was not related to actual mater­nal bone architecture, the efficacy of x-ray pelvimetry for breech presentation is still controversial and the authors are not prepared to abandon the procedure in the presence of this indication. However, in vertex presentations little useful infor­mation for the prospective management of labor was , ,VOL. 56, NO.3, SEPTEMBER 1980 obtained from pelvic measurements. Although the fre­quency of operative delivery increased in vertex pre­sentations with smaller transverse measurements, this information was not used prospectively. AP measure­ments of the maternal pelvis were of no prognostic value in any circumstance. In this study the clinically useful information that was obtained could have been derived from either ultrasound examination or ab­dominal scout films with less expense and less radia­tion exposure. Certainly, whereas every patient de­serves appropriate individualization of diagnostic procedures, in the authors' experience x-ray pel­vimetry with vertex presentation provides little clini­cally useful information. References 1. Brown RC: A modification of the Colcher-Sussman technique of x-ray pelvimetry. Am J Roentgenol Radium Ther Nucl Med 115:623, 1972 2. Zlatnik F, Varner MW, Hauser K: University of Iowa Hospitals Growth Curve (unpublished data) 3. Colcher AE, Sussman W: A practical technique for roentgen pelvimetry with a new positioning. Am J Roentgenol Radium Ther Nucl Med 51:207, 1944 4. Mengert WF: Estimation of pelvic capacity. JAM A 138:169, 1948 5. Stewart A, Webb J, Giles D, et al: Malignant disease in child­hood and diagnostic irradiation in utero. Lancet 2:447, 1956 Varner et al X -ray Pelvimetry 299 6. Stewart A, Webb J, Hewitt D, et al: A survey of childhood ma­lignancies. Br Med J 1:1495, 1958 7. McMahon B: Prenatal x-ray exposure and childhood cancer. J Natl Cancer lnst 28:1173, 1962 8. Oppenheim B, Griem M, Meier P: The effects of diagnostic x­. ray expos4re on the human fetus: An examination of the evi­dence. Radiology 114:529, 1975 9. Russell JGB, Richards B: A review of pelvimetry data. Br J Ra­diol 44:780, 1971 10. Hannah WS: X-ray pelvimetry-A criticai appraisal. Am J Ob­stet Gynecol 91:333, 1965 11. Kelly KM, Madden DA, Arcavese J5, et al: The utilization and efficacy of pelvimetry. Am J Roentgenol Radium Ther Nucl Med 125:66, 1975 Address reprint requests to: Michael W. Varner, MD Department of Obstetrics & Gynecology University of Iowa Hospitals and Clinics Iowa City, IA 52242 Accepted faT publication April 17, 1980. . Copyright © 1980 by The American College of Obstetricians and Gynecologists. 14th ANNUAL POSTGRADUATE COURSE: GYNECOLOGIC PATHOLOGY, CYTOGENETICS, AND ENDOCRINOLOGY January 15-21, 1981 The Department of Gynecology and Obstetrics at the Medical College of Wiscon­sin will present the 14th Annual Postgraduate Course in Gynecologic Pathology, Cytogenetics, and Endocnriology at the Pfister Hotel and Tower, Milwaukee, Wis­consin, January 15-21,1981. The 6-day course has been extended to include a complete up-to-date review of endocrinology and cytogenetics in addition to a thorough resume of gynecologic pathology. The course, which is limited to 150 registrants, is designed primarily as a postgraduate refresher course for residents, practitioners, and specialists desiring a current review of the pathology of the re­productive tract. Microscopes and a select group of microscopic slides will be available by special request along with special consultation and microscopic in­structions by faculty. Space limitation requires registration for the entire course. This course is approved for 42 cognates, Formal Learning, by the ACOG and meets the criteria for 42 hours of credit in Category I, Physicians Recognition Award of the American Medical Association. A $475.00 enrollment fee will in­clude 68 selected 35-mm slides for all participants. The registration fee is non­refundable. Among the guest faculty will be H. W. Jones, Jr, MO, Georgeanna S. Jones, MO, J. O. Woodruff, MO, and Robert Young, MD. (NCI) For further details and registration write to Richard F. Mattingly, MO, The Med­ical College of Wisconsin, 8700 West Wisconsin Avenue, Milwaukee, WI 53226. 300 Varner et al X-ray Pelvimetry Obstetrics & Gynecolo~