Original Research

Forceps Compared With VacuumRates of Neonatal and Maternal Morbidity

Aaron B. Caughey, MD, MPP, Per L. Sandberg, MD, Marya G. Zlatnik, MD, MMS,Mari-Paule Thiet, MD, Julian T. Parer, MD, PhD, and Russell K. Laros Jr, MD

OBJECTIVE: To compare perinatal outcomes betweenforceps- and vacuum-assisted deliveries. Our hypothesiswas that the force vectors achieved in forceps deliverywill lead to fewer shoulder dystocias, but greater perineallacerations.

METHODS: This was a retrospective cohort study of4,120 term, cephalic, singleton, nonrotational operativevaginal deliveries at a single institution. Outcomes exam-ined included rates of neonatal trauma, shoulder dysto-cia, and perineal lacerations. Potential confounders, in-cluding maternal age, birthweight, ethnicity, parity,station at delivery, episiotomy, attending physician, an-esthesia, and length of labor, were controlled for usingmultivariate logistic regression.

RESULTS: Among the 2,075 (50.4%) forceps- and 2,045(49.6%) vacuum-assisted deliveries, the rate of shoulderdystocia was lower among women undergoing forcepsdelivery (1.5% compared with 3.5%, P < .001), as was therate of cephalohematoma (4.5% compared with 14.8%,P < .001), whereas the rate of third- or fourth-degreeperineal laceration was higher (36.9% compared with26.8%, P < .001). These differences in perinatal compli-cations persisted when controlling for the confounderslisted above. The adjusted odds ratio for shoulder dysto-cia was 0.34 (95% confidence interval [CI] 0.200.57), forcephalohematoma was 0.25 (95% CI 0.190.33), and forthird- or fourth-degree lacerations was 1.79 (95% CI1.522.10) when comparing forceps to vacuum.

CONCLUSION: Vacuum-assisted vaginal birth is moreoften associated with shoulder dystocia and cephalohe-matoma. Forceps delivery is more often associated withthird- and fourth-degree perineal lacerations. These dif-ferences in complications rates should be consideredamong other factors when determining the optimalmode of delivery.(Obstet Gynecol 2005;106:90812)

LEVEL OF EVIDENCE: II-2

The modern-day vacuum extractor was first intro-duced in 1954,1 and its use slowly diffused intoobstetric practice during the ensuing three decades.However, since the late 1980s the use of the vacuumextractor has increased whereas the use of forceps hasdecreased.2,3 These trends of obstetric managementhave occurred despite little evidence regarding whichform of assisted vaginal delivery is optimal in differentsettings.

A recent meta-analysis included all of the pro-spective randomized trials of forceps compared withvacuum.4 There were only 10 such studies, withsample sizes ranging from 365 to 637.6 This collectionof analyses revealed that vacuum extractors have ahigher rate of failure than forceps and lead to greaterrates of cephalohematoma and retinal hemorrhage.They were also associated with a trend toward agreater rate of 5-minute Apgar scores less than 7.Forceps were associated with greater rates of perinealinjury and short-term postdelivery pain. A variety ofother outcomes were examined, and no differenceswere demonstrated. However, most outcomes weretoo underpowered in the analysis for any conclusionsto be drawn.

There have also been several large birth-certifi-catebased analyses that have demonstrated higherrates of cephalohematoma79 and lower rates of peri-neal laceration,8,9 but no difference in intracranialhemorrhage or perinatal mortality.7,8 However, whenintracranial hemorrhage was subdivided further, an

From the Department of Obstetrics, Gynecology and Reproductive Sciences,University of California, San Francisco; and the Department of Obstetrics andGynecology, California Pacific Medical Center, San Francisco, California.

Dr. Caughey is supported by the National Institute of Child Health and HumanDevelopment, Grant # HD01262 as a Womens Reproductive Health ResearchScholar.

Corresponding author: Aaron B. Caughey, MD, MPP, MPH, Department ofObstetrics and Gynecology, University of California, San Francisco, 505Parnassus Avenue, Box 0132, San Francisco, CA 94143; e-mail:abcmd@berkeley.edu.

2005 by The American College of Obstetricians and Gynecologists. Publishedby Lippincott Williams & Wilkins.ISSN: 0029-7844/05

908 VOL. 106, NO. 5, PART 1, NOVEMBER 2005 OBSTETRICS & GYNECOLOGY

nyomanbagusdonnyaryatmamahadewaHighlight

increase in subarachnoid hemorrhage was noted inone analysis.9 These studies are limited with respect tocausality because little is known about a variety ofconfounding variables, including type and experienceof provider and indication for delivery. One of theseanalyses also demonstrated a higher rate of shoulderdystocia for vacuum-assisted vaginal delivery whencompared with either forceps delivery or vacuum andforceps used sequentially.8 The largest prospectiverandomized controlled trial of forceps compared withvacuum also identified a trend toward a higher rate ofshoulder dystocia among their vacuum-assisted vagi-nal deliveries.6

The proposed mechanisms for the differences inthese outcomes include the increased occupation ofspace in the vagina, and thus greater distention of thevagina walls and perineum by the forceps, leading toincreased perineal and vaginal lacerations. Alterna-tively, the placement of the vacuum directly on thescalp has been proposed to increase the rates ofcephalohematoma, retinal hemorrhage, and intracra-nial hemorrhage,7 although only the first two havebeen demonstrated. Finally, the ability to pull harderwith forceps is proposed to explain the increased rateof success. It is of note that this last theory shouldactually support a higher rate of shoulder dystociaamong forceps as opposed to vacuum deliveries.

We propose, however, that it is not the totalmagnitude of the resultant force vector that leads toshoulder dystocia, but rather where it acts and in whatdirection. The forceps are placed cradling the entirefetal head, thus the resultant force vector applied tothe fetal head acts further caudally along the fetalhead, approximately half the distance of the blades(Fig. 1), and closer to the fetal shoulders thorax.

Moreover, because of the placement around the fetalhead, force vectors can be generated ranging fromperpendicular to the ground to horizontal. Becausethe resultant force vector used when delivering byforceps is directed farther down the head and in amore downward direction, it is likely to facilitate indirecting the anterior shoulder down and possiblyunderneath the pubic symphysis. The vacuum, be-cause it is attached directly in front of the posteriorfontanelle and the direction of its resultant forcevector is at most 45 degrees to the floor, actually maylead to more traction force on the fetal head, pullingthe anterior shoulder into the pubic symphysis andlodging it in this location.

Given this background and our proposed mech-anism, we generated the hypothesis that forcepsshould be associated with fewer shoulder dystociasthan vacuum. However, because of the generation ofgreater downward force by the forceps, we hypothe-sized that they would be associated with a higher rateof perineal laceration.

METHODSWe designed a retrospective cohort study of all sin-gleton neonates delivered beyond 37 weeks gesta-tional age in the vertex presentation by nonrotationalforceps or vacuum from January 1, 1981, to Decem-ber 1, 2001, at the Moffitt-Long Hospital at theUniversity of California, San Francisco. Exclusioncriteria included delivery before 37 weeks of gesta-tion, fetal anomalies, noncephalic presentation, occi-put transverse position, and multiple gestation. Thisstudy was approved by the Investigational ReviewBoard at the University of California, San Francisco.Our primary outcomes, shoulder dystocia and peri-neal laceration, were entered into data entry sheets bythe delivering clinician and verified by trained ab-stractors. Shoulder dystocia is defined at our institu-tion as any delivery requiring additional maneuversto deliver the shoulders of the infant, includingMcRoberts maneuver, suprapubic pressure, Woodsor Rubin maneuvers, or delivery of the posterior arm.We also collected information on the following sec-ondary outcomes: vaginal lacerations, cervical lacer-ations, cephalohematoma, intracranial hemorrhage,Erbs palsy, facial nerve palsy, neonatal jaundice,skull, clavicle and humerus fractures, neonatal sei-zures, Apgar scores, umbilical artery pH, umbilicalartery base excess, and neonatal intensive care unit(NICU) admission. The following data were alsocollected and examined as potential confounders:maternal age, ethnicity, education, body mass index,diabetes mellitus (both preconceptional and gesta-

Fig. 1. Component and resultant force vectors exerted byforceps and vacuum extractor. A. Forceps: The force vectoris applied farther down fetal head. The force vector rangesfrom perpendicular to outward. B. Vacuum: The forcevector is applied at fetal vertex. The force vector rangesfrom approximately 45 down to outward.Caughey. Forceps or Vacuum: Obstetric Complications. ObstetGynecol 2005.

VOL. 106, NO. 5, PART 1, NOVEMBER 2005 Caughey et al Forceps or Vacuum: Obstetric Complications 909

tional), length of labor, station at delivery, position atdelivery, obstetric provider(s), year of delivery, parity,anesthesia, and birthweight.

The data were then compiled and analyzed usingSTATA 7 software (StataCorp, College Station, TX).Univariate analyses of the primary predictor, forcepscompared with vacuum, were performed examiningthe primary and secondary outcomes of interest andcompared using the 2 test. Because the rates of manyof the neonatal complications were low, a summaryvariable called severe complications was createdthat included birth trauma (including skull fracture,brachial plexus injury, facial nerve palsy), neonatalseizures, and intracranial hemorrhage. For each ofthese primary and secondary outcomes a multivariatelogistic regression was performed, including the po-tential confounders. The confounders contributionsto the model were tested using the maximum likeli-hood ratio test, and they were only kept in the modelif they were statistically significant. Statistical signifi-cance was defined as P .05.

RESULTSDuring the study period, there were 4,120 womenwho met the inclusion and exclusion criteria outlinedabove. Of these, 2,075 (50.4%) delivered by forcepsand 2,045 (49.6%) delivered by vacuum-assisted vag-inal delivery. The women who underwent forceps-assisted vaginal delivery were more likely to be agedyounger than 35 years, nulliparous, and have Medic-aid insurance, an episiotomy, and epidural analgesia(Table 1). When the primary outcomes were exam-ined, we found that the rate of shoulder dystocia waslower among women undergoing forceps delivery(1.5%) compared with those undergoing vacuum-assisted vaginal delivery (3.5%, P .001). The rate ofthird- or fourth-degree lacerations was higher amongforceps deliveries (36.9%) compared with vacuum-assisted vaginal deliveries (26.8%, P .001). Amongthe secondary outcomes, the rates of 5-minute Apgarscores less than 7, cephalohematoma, admissions toNICU, and neonatal jaundice were higher amongwomen undergoing vacuum-assisted vaginal delivery(Table 2), whereas vaginal lacerations, cervical lacer-ations, and facial nerve palsy were higher amongwomen undergoing forceps deliveries. No differenceswere seen in umbilical artery blood gases, intracranialhemorrhage, or other types of neonatal trauma.

In multivariate analyses, when controlling forpotential confounders, most of the significant univar-iate findings persisted (Table 3), with adjusted oddsratios (AOR) that did not cross unity. Shoulder dys-tocia was lower among women with a forceps delivery

(AOR 0.34, 95% confidence interval [CI] 0.200.57),and both vaginal lacerations and third- and fourth-degree perineal lacerations were higher (AOR 1.79,95% CI 1.522.10). A vacuum-assisted vaginal deliv-ery was still associated with higher rates of 5-minuteApgar scores less than 7, cephalohematoma, admissionto the NICU, and neonatal jaundice. However, thedifferences in the rates of cervical laceration and facialnerve palsy were no longer statistically significant.

To examine the effects of forceps compared withvacuum among women of differing parity, subgroupanalyses were performed. Among nulliparous pa-tients, the prior observed differences in the primaryoutcomes persisted, with women undergoing forcepsdeliveries having a lower rate of shoulder dystocia(1.4% compared with 3.1%, P .001) and higher ratesof third- and fourth-degree perineal laceration (40.9%compared with 30.7%, P .001). These results wereconfirmed by multivariate analyses; when comparingforceps to vacuum, the differences were AOR of 0.39(95% CI 0.230.65) for shoulder dystocia and AOR of1.48 (95% CI 1.281.73) for third- and fourth-degreeperineal laceration. However, among multiparas thedifference persisted only for shoulder dystocia.Women with forceps deliveries had a lower rate ofshoulder dystocia (1.7% compared with 4.5%, P

Table 1. Maternal and Obstetric CharacteristicsBetween Forceps and Vacuum Deliveries

Forceps Vacuum

Variable n 2,075 n 2,045 P

Maternal age 35 15.5 19.1 .002Ethnicity .292

African-American 11.7 10.7Asian 33.6 36.2Hispanic 10.5 9.6Caucasian 44.2 43.5

Medicaid insurance 39.4 24.4 .001Nullipara 77.7 74.3 .009Obese (BMI 29) 6.2 6.5 .665GDM 3.3 3.3 .999OP position 12.4 12.3 .936Birth weight 4,000 g 10.7 11.1 .6792nd stage 3 h 38.8 38.9 .913Episiotomy 68.0 50.9 .001Induction 15.9 15.8 .884Augmentation 44.8 43.5 .437Epidural use 84.5 76.3 .001Station .001

Mid (0, 1) 1.9 2.3Low (2, 3, 4) 89.3 84.9Outlet (5) 8.8 12.8

BMI, body mass index; GDM, gestational diabetes mellitus; OP,occiput posterior.

Values are %.

910 Caughey et al Forceps or Vacuum: Obstetric Complications OBSTETRICS & GYNECOLOGY

.012), but no statistically significant difference in therate of third- and fourth-degree lacerations (19.6%compared with 15.7%, P .107). In multivariateanalyses, comparing forceps with vacuum for shoul-der dystocia, the outcome was different (AOR 0.41,95% CI 0.180.96), but for third- and fourth-degreeperineal laceration it was not (AOR 1.19, 95% CI0.811.69).

DISCUSSIONWe found lower rates of shoulder dystocia amongwomen with forceps deliveries as compared withwomen undergoing vacuum-assisted vaginal delivery,consistent with our proposed hypothesis. We alsofound higher rates of perineal and vaginal lacerationsamong the women who had forceps deliveries. Inaddition to the differences in our primary outcomes,we also found that women undergoing forceps deliv-ery had infants with lower rates of cephalohematoma,5-minute Apgar score less than 7, admission to theNICU, and neonatal jaundice.

When we examined these outcomes by parity, wefound the differences in the primary outcomes ofshoulder dystocia and third- and fourth-degree peri-neal laceration persisted for nulliparas. However,among multiparas, although they still had higher ratesof shoulder dystocia among women undergoing vac-uum-assisted vaginal delivery, there was no statisti-cally significant difference noted in third- and fourth-degree perineal lacerations. This difference couldsimply be attributed to the overall higher rates ofperineal laceration among nulliparas. However, theclinical effect seems to differ, with adjusted odds ratiosof 1.5 among nulliparas and only 1.2 among multip-aras. Thus, it may be that the downward vectorproduced with forceps leads to a greater increase inperineal trauma among nulliparas than it does inmultiparas.

Our findings of greater shoulder dystocia as wellas several measures of short-term neonatal morbidityamong the neonates of women delivered by vacuum-assisted vaginal delivery are of concern. Several ofthese differences have been noted in small random-ized trials of forceps compared with vacuum. Despitethese findings, the rate of vacuum use seems only tobe increasing. Moreover, one recent analysis con-cluded from its analysis of birth certificates thatDelivery by vacuum extraction is at least as safe asdelivery by forceps.8 Such a conclusive statement issupported neither by our study nor the broaderliterature.4,6,9

Our study is not without limitations. A nonran-domized, retrospective, cohort study can be compro-

Table 3. Multivariate Comparison of Maternal andNeonatal Morbidity Between Forcepsand Vacuum Deliveries

Outcomes*

ForcepsCompared

With Vacuum P

Shoulder dystocia 0.34 (0.200.57) .0013rd or 4th degree 1.79 (1.522.10) .001Vaginal laceration 1.77 (1.492.11 .001Cervical laceration 1.36 (0.832.24) .221UA pH 7.0 1.53 (0.643.68) .339UA BE 12 1.16 (0.701.93) .5575-min Apgar 7 0.67 (0.460.98) .042Cephalohematoma 0.25 (0.190.33) .001Serious neonatalcomplications 0.98 (0.611.58) .949

NICU admission 0.69 (0.490.97) .033Neonatal jaundice 0.74 (0.600.92) .006

UA, umbilical artery; BE, base excess; NICU, neonatal intensivecare unit.

Values are adjusted odds ratio (95% confidence interval). Thebaseline group is patients delivered by vacuum.

* Controlling for maternal age, ethnicity, Medicaid insurance,parity, body mass index, diabetes mellitus, fetal position andstation, birthweight, length of labor, episiotomy, year of deliv-ery, attending provider, induction of labor, and epidural use.Maternal age, birthweight, length of labor, and year of deliverywere controlled for as continuous variables.

All patients with Erbs palsy, facial nerve palsy, skull, clavicularfracture, or intracranial hemorrhage.

Table 2. Comparison of Maternal and NeonatalMorbidity Between Forceps and VacuumDeliveries

Forceps Vacuum

Outcomes n 2,075 n 2,045 P

Shoulder dystocia 1.5 3.5 .0013rd or 4th degree 36.9 26.8 .001Vaginal laceration 28.6 22.2 .001Cervical laceration 3.0 1.5 .001UA pH 7.0 0.9 0.8 .669UA BE 12 1.9 1.6 .4405-min Apgar 7 2.8 4.1 .021Cephalohematoma 4.5 14.8 .001Serious neonatalcomplications* 1.7 2.1 .389Skull fracture 0.05 0.1 .314Clavicle fracture 0.6 0.9 .269Erbs palsy 0.5 0.7 .393ICH 0.2 0.4 .272Facial nerve palsy 0.4 0.1 .037NICU admission 3.7 5.7 .002Neonatal jaundice 10.7 13.3 .010

UA, umbilical artery; BE, base excess; ICH, intracranial hemor-rhage; NICU, neonatal intensive care unit.

Values are %.* These were nerve palsies noted by pediatricians at delivery. All patients with Erbs palsy, facial nerve palsy, skull, clavicular

fracture, or intracranial hemorrhage.

VOL. 106, NO. 5, PART 1, NOVEMBER 2005 Caughey et al Forceps or Vacuum: Obstetric Complications 911

mised by having multiple confounding variables. Weattempted to control for these confounders usingmultivariate techniques in our analysis. Not only didwe control for confounders, such as station, birth-weight, episiotomy, and parity, that have been asso-ciated with the outcomes of analysis, but we alsocontrolled for year of delivery and obstetric providerin an attempt to control for the more intangiblefactors surrounding forceps and vacuum delivery. Ofcourse, we could not control for the decision-makingprocess that clinicians undergo when choosing whichinstrument to assist in operative vaginal delivery.Although we were adequately powered to examineour primary outcomes of shoulder dystocia and per-ineal lacerations, we were underpowered to examinesome of our secondary outcomes, such as the morerare neonatal complications. Another possible limita-tion pertains to the generalizability of our studypopulation to that of all pregnant women. The pa-tients in our study were managed at an academiccenter in California where the majority of the provid-ers use both instruments, and deliveries are in con-junction with a resident provider. Other than this, ourpatient population represented a wide spectrum ofethnicities, ages, and educational levels. Furthermore,when we controlled for the various maternal charac-teristics, our findings were robust.

Determination of which instrument to use for anoperative vaginal delivery should entail weighing therisks and benefits of the instrument to both themother and the fetus. It seems that among nulliparas,the substantial benefit of vacuum to the maternalperineum should be weighed against the increasedrisk of shoulder dystocia and cephalohematoma. In-terestingly, among multiparas, there seems to be nobenefit from decreased perineal trauma, whereas theincreased risk to the neonate persists. To examinethese outcomes, a large, multicenter, randomized trial

should be performed. Another important aspect ofthe decision regarding which instrument to use isprovider comfort with these two modes of operativevaginal delivery. If providers are not trained in theuse of forceps, they will not be able to provide thisform of operative vaginal delivery, nor make a choicebetween the two types of operative vaginal delivery.Because we do not know which is better, and eithermay be the instrument of choice in particular clinicalsituations, it is imperative that providers continue tobe trained in the use of forceps as well as vacuum.

REFERENCES1. Malmstrom T. The vacuum extractor, an obstetrical instru-

ment. Acta Obstet Gynecol Scand Suppl 1954;33:131.

2. Kozak LJ, Weeks JD. U.S. trends in obstetric procedures,1990-2000. Birth 2002;29:15761.

3. Roberts CL, Algert CS, Carnegie M, Peat B. Operative deliv-ery during labour: trends and predictive factors. PaediatrPerinat Epidemiol 2002;16:11523.

4. Johanson RB, Menon BK. Vacuum extraction versus forcepsfor assisted vaginal delivery. Cochrane Database Syst Rev2000;(2):CD000224.

5. Fall O, Ryden G, Finnstrom K, Finnstrom O, Leijon I. Forcepsor vacuum extraction? A comparison of effects on the newborninfant. Acta Obstet Gynecol Scand 1986;65:7580.

6. Bofill JA, Rust OA, Schorr SJ, Brown RC, Martin RW, MartinJN Jr, Morrison JC. A randomized prospective trial of theobstetric forceps versus the M-cup vacuum extractor. Am JObstet Gynecol 1996;175:132530.

7. Towner D, Castro MA, Eby-Wilkens E, Gilbert WM. Effect ofmode of delivery in nulliparous women on neonatal intracra-nial injury. N Engl J Med 1999;341:170914.

8. Demissie K, Rhoads GG, Smulian JC, Balasubramanian BA,Gandhi K, Joseph KS, et al. Operative vaginal delivery andneonatal and infant adverse outcomes: population based ret-rospective analysis [published erratum appears in BMJ 2004;329:547]. BMJ 2004;329:249.

9. Wen SW, Liu S, Kramer MS, Marcoux S, Ohlsson A, Sauve R,et al. Comparison of maternal and infant outcomes betweenvacuum extraction and forceps deliveries. Am J Epidemiol2001;153:1037.

912 Caughey et al Forceps or Vacuum: Obstetric Complications OBSTETRICS & GYNECOLOGY