Journal of Obstetrics and Gynaecology (2 0 0 1 ) V ol. 2 1 , No.
6 , 5 7 6 5 7 9
OBSTETRICS
A matched controlled study of Kiellands forceps
for transverse arrest of the fetal vertex
E. SCHIFF, 1 S. A. FRIEDMAN, 2 M. ZOLTI, 1 A. AV RAHAM, 1 Z. KAYAM, 1
S. MASHIACH1 and H. CARP1
1
Department of Obstetrics and Gynecology, Sheba Medical Center, Tel Hashomer, Israel,
2
Department of Obstetrics and Gynecology, Oregon Health Sciences University, Portland,
Oregon, USA
Summary
This study attempted to determine whether delivery with
Kiellands forceps for deep transverse arrest is less favourable
than other instruments. One hundred and forty-six women
who underwent rotation and delivery with Kiellands forceps
between 1994 and 1997 were matched by parity and birth
weight to one of two control groups: delivery by non-rota-
tional forceps or the vacuum extractor. No significant
differences were found in maternal or neonatal outcome (vag-
inal lacerations, 3rd- or 4th-degree perineal tears, postpartum
haemorrhage, fever, blood transfusion, duration of hospitali-
sation, Apgar score, asphyxia, scalp trauma, admission to the
intensive care unit or neonatal hospitalisation). The incidence
of heart rate abnormalitie s prior to instrumental delivery was
similar. The failure to deliver rate (89% after Kiellands
forceps) was not different to the 75% and 68% found in each
control group. These data indicate that the outcome after Kiel-
lands forceps delivery is similar to other instrumental
deliveries if performed by experienced obstetricians.
Introduction
As confirmed by the studies of Friedman (1954), rota-
tion of the fetal head frequently occurs before complete
dilatation of the cervix. Arrest of rotation, whether
before or after complete dilatation, is a matter of con-
cern. Even if the head descends to the mid or low pelvic
level (more than 1 cm below the ischial spines) in an
occipito-transverse position, most women will deliver
spontaneously without deep transverse arrest. It is there-
fore possible to await spontaneous rotation and delivery
as long as the fetal heart tracing is reassuring, the mater-
nal condition is stable and the length of the second stage
is within an acceptable time range. Nevertheless, it is not
uncommon for the fetal or maternal condition to neces-
sitate delivery from an occipito-transverse position, in
the mid or low pelvis. The majority (99%) of obstetric
training programmes in the United States and Canada
use forceps for midpelvic delivery. Kiellands forceps
are used for the majority (76%) of rotational mid-cavity
extractions (Healy and Laufe, 1985). However, there
has been a dramatic decrease in the use of these instru-
ments for transverse arrest in North America, the United
Kingdom and other areas in the last two decades (Jain
et al., 1993). In Israel, too, most obstetricians prefer a
Correspondence to: Dr. H. J. A. Carp, Department of Obstetrics and Gynecology, Sheba Medical Center, Tel Hashomer 52621,
Israel, Tel: 972 3 5302697; Fax: 972 9 9574779; Email: carp@netvision.net.i l
ISSN 0144 3615 print/ISSN 1364 6893 online/01/060576 04
DOI: 10.1080/01443610120085500
trial of vacuum rotation and extraction for transverse
arrest even if the presenting part is low. Some even con-
sider cesarean delivery to be the primary choice for mid-
pelvic transverse arrest, or the secondary choice if the
vacuum fails to effect delivery. Hence, in most pro-
grammes, rotational delivery by Kiellands, Bartons or
similar forceps is not taught, while non-rotating forceps
and vacuum deliveries from mid- to low pelvic stations
are still considered safe.
In this study we assessed whether Kiellands forceps
delivery for a transverse presentation is associated with
a less favourable outcome than other instruments used
for mid-to low pelvic delivery.
Materials and methods
Between January 1994 and June 1997, there were 146
women (06% of all deliveries) with deep transverse
arrest in the Sheba Medical Center, Tel Hashomer.
They were delivered by rotation and extraction with
Kiellands forceps. Other positions, such as occipito-
posterior, were delivered by non-rotational forceps or
vacuum during this period. These 146 women consti-
tuted the study group. The following patients were
excluded: deliveries in which Kiellands forceps were
used for rotation, but not extraction, and multiple gesta-
tions in which only one fetus was delivered with
Kiellands forceps. There were two control groups: (a)
pregnancies delivered by mid- or low non-rotating for-
ceps (n= 146), and (b) pregnancies delivered by mid- to
low vacuum extraction (n= 146). These patients were
selected for inclusion in the control group by a retro-
spective review of the hospital database seeking
patients delivered by non-rotational forceps or vacuum
deliveries. The patients casenotes were then reviewed.
Patients were only included in a control group if there
was a vertex presentation and the head was in an occip-
ito-anterior or occipito-posterior position 1 or 2 cm
below the ischial spines. Patients were matched for
parity and birth weight (3000 g). If a patients notes
were reviewed and she was not matched according to
the above criteria, the next case in the list was examined
until a matched control patient was found. 146 patients
Taylor & Francis Limited, 2001
 Kiellands forceps for transverse arrest of the fetal vertex 577

from each group were allocated randomly to each con-
trol group.
The following features were assessed: duration of the
first stage of labour (from 4 cm to complete cervical
dilation), duration of the second stage of labour (from
complete dilation to delivery of the infant), intrapartum
fever (38C during labour or within 72 hours of deliv-
ery), postpartum fever ( 38C from 2 hours to 3 days
after delivery), fetal heart abnormalities (fetal bradycar-
dia or repeat decelerations necessitating instrumental
intervention), incidence of failed extraction with the
originally selected instrument, vaginal lacerations (in
addition to episiotomy), asphyxia (arterial cord pH
<71) and neonatal renal, cardiac or cerebral injury,
trauma to the fetal head (cephalaematoma, skin lacera-
tions, facial haematoma or facial nerve injury),
neonantal sepsis (clinical signs of infection and a posi-
tive blood or spinal fluid culture), neonatal
hyperbilirubinaemia (hyperbilirubinaemia necessitat-
ing light therapy for more than 24 hours).
Results are expressed as the incidence or meanSD.
Comparisons were made between groups using c2 anal-
ysis, Fishers exact test, unpaired Students t-test or
one-way analysis of variance as appropriate. P values
<005 were considered significant.
Results
During the study period, 24 488 women delivered at the
Sheba Medical Center, Tel Hashomer. One hundred
and forty-six women were delivered by Kiellands for-
ceps with rotation and extraction; they formed the study
group. An additional 2171 instrumental deliveries were
performed, 1221 with the vacuum extractor, 874 with
non-rotating forceps and 76 with both the vacuum and
forceps. As can be seen from Table I, the study group
was similar to both control groups in regard to maternal
age, weight, height and gestational age
Table II presents the details of labour and delivery.
There was a significantly longer second stage of labour
and longer time interval from rupture of the membranes
until delivery in the patients delivered by Kiellands
forceps compared to each control group. Patients in the
study group also required epidural anesthaesia more
often. The mean birth weight was similar in all three
groups of patients.
Table III presents the maternal outcome variables in
the study and control groups. Failure to complete
extraction occurred with a similar frequency in the
study and control groups. There was no significant dif-
ference in any of the features assessed.
Table IV describes the neonatal outcome. Again,
there were no significant differences in any outcome
variables assessed, except for a small but significantly
longer neonatal hospital stay in infants delivered by the
vacuum extractor.
There were no maternal or perinatal deaths in any of
the groups studied.
All instrumental deliveries in this series were per-
formed in the presence of a senior physician.
However, significantly fewer Kiellands forceps deliv-
eries (17%) were performed by junior physicians than
non-rotational deliveries (47%).
Discussion
Although retrospective, these results show that rotation
and extraction with Kiellands forceps is a safe mode
ofdelivery if the fetal vertex is in a transverse position
at 1 or 2 cm below the ischial spines. These results.

Table I. Details of patients in the study

Study group Control group 1 Control group 2

n = 14 6 non-rotating vacuum extractor
forceps n = 14 6 n = 14 6
Age (years) 2 42 44 240 5 1 252 5 3 P = NS
Nulliparous (%) 52 52 52 *
Weight (kg) 7 6 84 759 1 04 749 1 16 P = NS
Height (cm) 164 0 90 1 633 9 2 165 8 116 P = NS
Gestational age (weeks) 40 5 19 403 2 1 401 1 7 P = NS
NS = not significant, * Matched variable.

Table II. Details of labour and delivery

Study group Control group 1 Control group 2

n = 146 non-rotating vacuum extractor
forceps n = 1 46 n = 146
Length of 1 st stage (hours) 92 3 3 86 2 9 89 29 NS
Length of 2 nd stage (hours) 19 0 6 16 0 5 16 06 P < 000 1*
RUM to delivery (hours) 66 2 0 62 2 4 60 17 P = 00 4*
Epidural anesthesia (%) 80 1 685 699 P = 00 5*
Intrapartum fever (%) 48 34 4 8 NS
2nd stage AFM (%) 11 6 191 164 NS
Birth weight(gm) 3 471 5 17 3 415 6 20 35 25 39 5 **
RUM = rupture of membranes. AFM = abnormal fetal heart-rate monitoring. *T hese statistics were cal-
culated by analysis of variance (A NOV A) for both control groups compared to the study group.
**Matched variable ( 30 0 g).
57 8 E. Schiff et al.

Table III. Maternal outcome

Study group Control group 1 Control group 2

n = 146 non-rotating forceps vacuum extractor
n = 14 6 n = 146
Failed extraction (%) 8 9 75 68 NS
Episiotomy (%) 986 973 966 NS
V aginal lacerations (%) 103 96 89 NS
III/IV degree tear (%) 0 7 14 07 NS
Postpartum haemorrhage (%) 6 2 62 68 NS
Postpartum fever (%) 6 8 62 55 NS
Postpartum blood transfusion (%) 4 1 48 34 NS
Maternal hospitalisation (days) 32 11 33 1 2 30 09 NS

Table IV. Neonatal outcome

Study group Control group 1 Control group 2

n = 146 non-rotating forceps vacuum extractor
n = 14 6 n = 146
5 -A pgar < 7 (%) 3 4 48 48 NS
Evidence of asphyxia (%) 27 4 1 21 NS
Scalp/face trauma (%) 14 2 1 27 NS
Admissions to intensive care unit (%) 41 6 2 48 NS
Neonatal sepsis (%) 27 0 7 27 NS
Neonatal hyperbilirubinaemia (%) 55 3 4 89 NS
Neonatal hospitalisation (days) 37 13 37 1 5 4 1 15 P = 00 2

concur with those of several other reports, which dem-
onstrate a relatively favourable and safe outcome for
the mother and fetus when delivered by Kiellands for-
ceps (Cardozo et al., 1983; Traub et al., 1984; Lacreta,
1986; Chow et al., 1987). In this series, rotation and
extraction by Kiellands forceps was not associated
with increased maternal or fetal morbidity compared
with non-rotational forceps or vacuum extraction (when
applied at the same station, or when compared to the
vacuum estractor for a transversely presenting vertex).
Cardozo et al. (1983) have reported 80 cases of Kiel-
lands rotation and extraction procedures in 2708
deliveries, and have shown a similar outcome to that of
spontaneous delivery or non-rotational forceps. Glee-
son et al. (1992) have analysed 98 consecutive
rotational deliveries in primiparae, and have found
Kiellands forceps to be as safe as the vacuum extractor
and more effective. In this present study, strict criteria
were used to define the height of the head, and outlet
procedures were excluded. Careful matching also con-
trolled for possible bias due to parity or fetal size.
Additionally, Kiellands forceps deliveries were per-
formed by a senior physician, or by a junior physician
under supervision. Hence, the results are valid only for
rotation forceps procedures performed by skilled and
experienced personnel.
The purpose of this paper was to compare the safety
and efficacy of Kiellands forceps for rotation and
extraction. They were therefore compared to other
modes of instrumental delivery, non-rotational forceps
and the vacuum extractor. It was considered appropriate
to separate these modes of delivery into two control
groups, rather than one, as complications in one control
group may be obscured by the lack of complications in
the second control group. Both the forceps and vacuum
extractor effected delivery in a relatively high propor-
tion of cases, and was similar for all instruments
regardless of whether rotation of the fetal head was
required (911932%). This efficacy was similar to that
reported by other authors (Chow et al., 1987; Gleeson
et al., 1992; Tan et al., 1992) (9397%) using Kiel-
lands forceps.
Baker and Johnson (1994) have suggested that deliv-
ery by Kiellands forceps may result in disturbance of
fetal acid-base balance. However, Gleeson et al.
(1992), have reported that low Apgar scores and cord
arterial pH values of less than 72 occurred less fre-
quently after Kiellands rotation than rotational vacuum
delivery. In this study, the incidence of Apgar scores of
6 at 5 minutes and the incidence of neonatal depres-
sion were similar whether delivered by rotating or non-
rotating procedures. However, only 12% of Kiellands
deliveries were performed when the fetus showed sig-
nificant heart rate abnormalities, whereas there was a
slightly, but not significantly, higher rate (1619%) in
the control groups. It is possible that a higher proportion
of fetuses in the control groups had abnormal acid-base
status before the procedure was performed. The inci-
dence of fetal trauma was low in both study and control
groups (1427%). This is comparable to the low inci-
dence of trauma reported by Herabutya et al. (1988) in
117 Kiellands and 259 vacuum deliveries, performed
for deep transverse arrest. The higher incidence of
trauma reported for rotational forceps in the 1960s
(Rubin and Coopland, 1970) may be due possibly to
applying forceps to a high or unengaged head.
Catastrophes such as spinal cord transection have
been reported as rare complications of Kiellands rota-
tional deliveries (Gould and Smith, 1984), indicating
the potential risk of this procedure if performed by an
imprudent or inexperienced obstetrician.
Drife (1983) has raised the question of whether cae-
sarean delivery should replace Kiellands forceps for
deep transverse arrest. This study does not compare
these two options. However, neither Traub et al. (1984)
nor Cibils and Ringler (1990) were able to show that
 Kiellands forceps for transverse arrest of the fetal vertex
caesarean delivery was associated with a significant
benefit in the second stage of labour, as opposed to vag-
inal delivery with Kiellands forceps. Maternal
morbidity, however, was higher after caesarean section.
Both studies concluded that mid-forceps rotation, when
properly indicated and executed, offers a safe alterna-
tive to caesarean delivery. It should he noted, however,
that considerable controversy persists with regard to the
use of mid-pelvic instrumental deliveries in general
regardless of the need for rotation (Friedman, 1987;
Robertson et al., 1990; Bashore et al., 1992). As most
studies evaluating the outcome of mid-pelvic proce-
dures have shown no increase in serious neonatal
morbidity, and as the long-term follow-up data have
shown that infants delivered by mid-forceps procedures
have similar outcomes to other infants (Nilsen, 1984;
Dierker et al., 1986), many physicians still perform
non-rotational mid-pelvic forceps and vacuum deliver-
ies. Kiellands forceps rotational deliveries may be the
only way of delivering some women without recourse
to caesarean section in circumstances where manual
rotation and forceps or the vacuum extractor are likely
to fail. Hence, in our opinion, abandoning the prudent
use of Kiellands rotation and extraction is unjustified.
The present study confirms that rotation and extraction
from mid- to low pelvic positions by skilled physicians
using Kiellands forceps is efficacious and safe.
References
Baker P.N. and Johnson I.R. (1994) A study of the effect of
rotational forceps delivery on fetal acid-base balance. Acta
Obstetricia et Gynecologica Scandinavica, 73, 787789.
Bashore R.A., Phillips W.H. and Brinkman C.R. (1992) A
comparison of the morbidity of mid forceps and cesarean
delivery. American Journal of Obstetrics and Gynecology,
162, 14281435.
Cardozo L.D., Gibb D.M., Studd J.W. and Cooper D.J. (1983)
Should we abandon Kiellands forceps? British Medical
Journal (Clinical Research), 287, 315317.
Chow S.L., Johnson C.M., Anderson T.D. and Hughes J.H.
(1987) Rotational delivery with Kiellands forceps. Medi-
cal Journal of Australia, 146, 616619.
579
Cibils L.A. and Ringler G.E. (1990) Evaluation of midforceps
delivery as an alternative. Journal of Perinatal Medicine,
18, 511.
Dierker L.J., Rosen M.G., Thompson K. and Lynn P. (1986)
Midforceps deliveries: long term outcome of infants.
American Journal of Obstetrics and Gynecology, 154,
764768.
Drife J.O. (1983) Kielland or Cesarean? British Medical
Journal (Clinical Research), 287, 309310.
Friedman E.A. (1954) The graphic analysis of labor. Ameri-
can Journal of Obstetrics and Gynecology, 68, 1568
1575.
Friedman E.A. (1987) Midforceps delivery: no? Clinical
Obstetrics and Gynecology, 30, 93105.
Gleeson N.C., Gormally S.M., Morrison J.J. and ORegan M.
(1992) Instrumental rotational delivery in primiparae. Irish
Medical Journal, 85, 139141.
Gould S.J. and Smith J.F. (1984) Spinal cord transection, cer-
ebral ischemia and brain-stem injury in a baby following a
Kiellands forceps rotation. Neuropathology and Applied
Neurobiology , 10, 151158.
Healy D.L. and Laufe L.B. (1985) Survey of obstetric forceps
training in North America in 1981. American Journal of
Obstetrics and Gynecology, 151, 5458.
Herabutya Y., Prasertawat P. and Boonrangsimant P. (1988)
Kiellands forceps or ventousea comparison. British
Journal of Obstetrics and Gynaecology, 95, 483487.
Jain V., Guleria K., Gopalan S. and Narang A. (1993) Mode
of delivery in deep transverse arrest. International Journal
of Gynaecology and Obstetrics, 43, 129135.
Lacreta O. (1986) {More on the Kielland forceps}. Zentral-
blatt fur Gynakologie 108, 104111. {In German}.
Nilsen S.T. (1984) Boys born by forceps and vacuum extrac-
tion examined at 18 years of age. Acta Obstetricia et
Gynecologica Scandinavica, 63, 549554.
Robertson P.A., Laros R.K. and Zhao R.L. (1990) Neonatal
and maternal outcome in low pelvic and mid-pelvic opera-
tive deliveries. American Journal of Obstetrics and
Gynecology, 162, 14361444.
Rubin L. and Coopland A.T. (1970) Kiellands forceps.
Canadian Medical Association Journal, 103, 505506.
Tan K.H., Sim R. and Yam K.L. (1992) Kiellands forceps
delivery: is it a dying art? Singapore Medical Journal, 33,
380382.
Traub A.I., Morrow R.J., Ritchie J.W. and Dornan K.J.
(1984) A continuing use for Kiellands forceps? British
Journal of Obstetrics and Gynaecology, 91, 894898.
Copyright of Journal of Obstetrics & Gynaecology is the property of Taylor & Francis Ltd and its content may
not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.