Anaesth Crit Care Pain Med 34 (2015) 277–280

Original Article

Influence of needle diameter on spinal anaesthesia puncture failures

for caesarean section: A prospective, randomised, experimental study
Fausto Fama’ b,1,*, Cecile Linard b,1, Damien Bierlaire a, Maria Gioffre’-Florio b,
Jacques Fusciardi a, Marc Laffon a
a
University Hospital of Tours, Department of Anaesthesiology and Intensive Care, Hôpital Bretonneau, 2, boulevard Tonnellé, 37044 Tours cedex 9, France
b
University Hospital of Messina, Department of Human Pathology, Via Consolare Valeria, 1, 98125 Messina, Italy

A R T I C L E I N F O A B S T R A C T

Article history: Objectives: Spinal anaesthesia represents the technique of choice for elective caesarean section. The
Available online 6 October 2015 purpose of this study was to compare the puncture failure rates with 25, 26 or 27 gauge (G) pencil-point,
Whitacre type (with introducer) needles during spinal anaesthesia for caesarean section.
Keywords: Study design: Prospective, randomised, experimental study in healthy subjects.
Spinal anaesthesia Patients and methods: We recruited 330 adults, consecutively scheduled parturients, randomised into
Caesarean section three groups. The subarachnoid puncture procedure was standardised. The flexibility of the three needle
Diameter
types was assessed in vitro, and a force was applied using a dynamometer. The occurrence of postdural
Needle
Failure
puncture headache was also evaluated.
Postdural puncture headache Results: The number of spinal puncture failures was significantly higher in the 27 G group, than in the
25 G (P = 0.006) group and the 26 G (P < 0.001) group, but did not differ between the 25 G and 26 G
groups (P = 0.606). Ten postdural puncture headaches were observed without significant differences
among the groups.
Conclusions: This prospective study showed that puncture failures occur less frequently with the use of
25 G or 26 G pencil-point needles as compared to 27 G needles, probably due to the higher flexibility of
the latter. This characteristic was demonstrated in vitro, in a reproducible model. This experiment
suggests that a 26 G pencil-point needle is the optimal gauge for performing spinal anaesthesia for
scheduled caesarean sections.
ß 2015 Société française d’anesthésie et de réanimation (Sfar). Published by Elsevier Masson SAS. All
rights reserved.

1. Introduction needles, respectively [7]. The probable mechanism is an increase in

Spinal anaesthesia (SA) is used more and more frequently and
has become a technique of choice in elective surgery as well as in
obstetrics [1–3]. One of its complications is the occurrence of
postdural puncture headache (PDPH), especially in parturients
who are more vulnerable for their age and gender [4–6]. The use of
fine, pencil-point needles can reduce such effects [1,4,6]. A meta-
analysis estimated the incidence of PDPH at 2.2% when using
25 gauge (G) needles compared to an estimation of 1.7% for 27 G
needles [4]. Yet, the finer the needle, the greater the risk of
puncture failure. In a prospective study on more than 200 partu-
rients undergoing a caesarean section, the impact of puncture
failures ranges from 0 to 7.4% with 25 G and 27 G fine, pencil-point
* Corresponding author at: Complesso MITO - Residenza Ginestre F/2,
98151 Messina, Italy. Tel.: +39 090 2402767; fax: +39 090 2402767.
E-mail address: famafausto@yahoo.it (F. Fama’).
1
Equal contributors.
needle flexibility inversely proportional to size [8–10]. Therefore,
the more a needle is flexible, the less its path is rectilinear when it
passes through tissues [9,11]. To our knowledge, there are no data
presenting a standardised model describing needle flexibility as a
function of calibre. In addition, the risk for puncture failure and
PDPH has not been evaluated for 26 G pencil-point needles.
The purpose of this prospective, randomised study was to
compare the puncture failure rates with 25 G, 26 G or 27 G pencil-
point needles during SA for caesarean section, as well as to study
needle flexibility in vitro.
2. Patients and methods
2.1. In vivo study
We prospectively included 330 adults parturients consecutive-
ly scheduled for a caesarean section under SA. All subjects gave

http://dx.doi.org/10.1016/j.accpm.2015.05.005
2352-5568/ß 2015 Société française d’anesthésie et de réanimation (Sfar). Published by Elsevier Masson SAS. All rights reserved.
278 F. Fama’ et al. / Anaesth Crit Care Pain Med 34 (2015) 277–280
their written informed consent, and the local ethics committee
(University Hospital of Tours, France) approved the study.
Exclusion criteria were the woman’s refusal of SA, body mass
index superior to 30, acquired or congenital coagulation disorders,
systemic or localised infection at the puncture site, pregnancy
complicated with preeclampsia and eclampsia, neurological
diseases, major kyphoscoliosis, and a history of migraine heada-
ches. Caesarean sections performed under emergency conditions
were not included to avoid statistical interferences, i.e., failures
due to technical difficulties related to the context: procedurally
incorrect assessments, difficulty to maintain the required position,
and increased anaesthetic stress of the parturients. The SA was
performed by 5 trained anaesthetists, 2 with 10 years and 3 with
6 years of experience in obstetrics and gynaecology anaesthesia.
The calibre of the pencil-point needle, 25, 26 or 27 G, commonly
utilised in clinical practice or for obstetrics SA, was assigned on the
base of a computer-generated randomisation table. The subarach-
noid puncture procedure was standardised. All parturients were in
a sitting position, leaning forward, feet on a stool and shoulders
maintained by an assistant. The puncture was done on the lowest
palpable space according to the operators’ experience, most
frequently at levels L4–L5, notwithstanding the lack of ultra-
sounds, which are mainly suggested by French guidelines. In all
cases, after a cutaneous local anaesthesia, each lumbar puncture
was performed using a 20 G diameter introducer. Two high sensory
levels were observed, once in the 25 G (C4) group and once in the
26 G (C6) group. In the present study, a single brand of needles was
tested to ensure uniformity as concerns technical characteristics
and manufacturing qualifications. The pencil-point spinal needle
(WhitacreTM type with introducer, length 90 mm, Laboratories
Vygon1, France) was driven through the introducer sheath and
then advanced until a change in resistance was felt, corresponding
to the passage of the dura mater. The stylet was withdrawn until
the appearance of cerebrospinal fluid (CSF). Once this flow was
clear, the anaesthetic solution was injected: hyperbaric bupiva-
caine 0.5% and sufentanil, associated or not with morphine
according to the practitioner’s preferences. Drugs were selectively
adjusted for each woman and dosages were administered as
indicated in Table 1. At the end of the injection, the persistence of a
CSF flow was checked to detect anaesthesia failure due to a
possible accidental needle displacement during the injection.
The study-data, anaesthetic and demographic features, were
collected in the operating room. The occurrence of PDPH (diffuse
and/or localised in the fronto-occipital area, and enhanced in an
orthostatic position) was assessed via a questionnaire directly
addressed to women in bed by an outside observer, up until five
days after the surgical procedure.
For an alpha risk at 0.05 and a beta risk at 0.20, the number of
women to be included in each group was 108. It was therefore
Table 1
Anaesthetic and demographic characteristics of the three parturient groups studied.
The results are expressed as means  standard deviations (SD), as medians
(interquartile range) and as median [extreme] values.
Group 25 G 26 G 27 G
(n = 109) (n = 121) (n = 98)
Age (years) 31  5 31  5 31  6
BMI (kg.m 2) 29  7 28  5 29  6
Injected volume (mL) 2.9  0.3 3.0  0.3 3.2  0.3
Bupivacaine (mg) 10.9  1.3 10.7  1.2 10.9  1.2
Sufentanil (mg) 31 31 31
Morphine (mg)a 100 100 100
Dermatomeb T3 [C4–T8] T3 [C6–T10] T4 [T1–T11]
G: Gauge; BMI: body mass index.
a
Median (interquartile range [IQR]).
b
Median [extremes].
decided to randomise a total of 330 parturients into three groups of
110 each.
2.2. In vitro study
The flexibility of the three needle types was assessed in vitro
(n = 60, 20 for each needle type) after the positioning of a 20 G
introducer. Needles were driven into the introducer. A perpendic-
ular force starting at 0.1 and then moving up through 0.2 - 0.3 -
0.4 Newton (N) was applied to each type of needle using a
dynamometer, either at a centimetre after the distal end of the
introducer sheath (n = 10 needles per group; proximal force) or at
the distal end of the needle (n = 10 needles per group; distal force).
Flexibility was assessed by measuring the distance between the
ends of the needle before and after the dynamometer application.
The applied force was limited to a maximum of 0.4 N because a
preliminary experiment in this study demonstrated a 908
deformation for a 27 G needle with 0.5 N. For each force applied
in the proximal and distal extremities of the needles, flexibilities
were compared calculating the quotient between the deviation of
the 27 G needle in relation to the 25 G and 26 G needles.
2.3. Statistical analysis
In vivo, qualitative variables were compared with a Chi2 test
with a Yates correction, and the most relevant statistical events
were also investigated by means of a Fisher’s exact test. The
quantitative variables were analysed using ANOVA or Kruskall-
Wallis tests, and then integrated by a Fisher’s test, if statistically
significant. In vitro, a Kruskall-Wallis test was utilised, and
integrated by a Student-Newman-Keuls test, if statistically
significant.
The level for statistical significance was set at P < 0.05. The
results are expressed as means  standard deviations (SD) for
normally-distributed, continuous variables or as median values (with
indication of extreme values) for non-normally-distributed continu-
ous variables and as medians (interquartile range) for discontinuous
variables. The statistical software utilised was Primer of Biostatistics,
version 4.02 by Stanton A. Glantz.
3. Results
3.1. In vivo
Anaesthetic and demographic characteristics were not signifi-
cantly different between groups (Table 1). The study lasted
18 months. The American Society of Anaesthesiologists (ASA)
physical status classification system was assessed for all patients;
223 were ASA 1, and 107 ASA 2. Puncture failure was defined as such
only after three consecutive failures to obtain CSF flow with the same
needle. When a first failure occurred, our protocol indicated a new
attempt with the upper diameter needle (i.e. passing from 27 G to
26 G or from 26 G to 25 G), which was reinserted at the same spot,
and in case of failure, general anaesthesia (GA) was then performed.
In the 25 G group, two cases with failed punctures required general
anaesthesia (GA). In the 26 G group, there was only one puncture
failure, and the SA was easily performed by means of a 25 G needle.
In the 27 G group, 12 failures were followed by SA with a 26 G
needle. In summary, the 27 G group included 98 women, the 26 G
group 121, and finally, the 25 G group 109 because 2 were submitted
to GA (Table 2). The percentages of anaesthetic punctures failures
were 1.8%, 0.9% and 10.9%, in groups 25 G, 26 G and 27 G,
respectively, and their comparison was statistically significant
(P < 0.001). The number of spinal puncture failures was significantly
higher in the 27 G group as compared to the 25 G (P = 0.006) and
26 G (P < 0.001) groups, but was not different between the 25 G and
 F. Fama’ et al. / Anaesth Crit Care Pain Med 34 (2015) 277–280
Table 2
Spinal puncture failures and incidence of postdural puncture headache. The number
of puncture failures was statistically significant in the 27 G group (P = 0.006 versus
the 25 G group, P < 0.001 versus the 26 G group). No statistically significant
difference was found between the 25 G and 26 G groups (P = 0.606). Only 2 general
anaesthesia procedures were carried out after 25 G attempt failures.
Group 25 G 26 G 27 G
Number of patients: n 109 121 98
Failure: n (%) 2 (1.8)a 1 (0.9)a 12 (10.9)
Headache: n (%) 5 (4.6) 3 (2.5) 2 (2.0)
Blood patch: n 1 1 0
a
P < 0.05, 27 G vs. 25G and 26 G.
26 G groups (P = 0.606). Ten PDPHs were observed without
significant differences among groups (P = 0.53) (Table 2). All
headaches appeared before or during the 5th postoperative day.
Only two blood-patches were performed since most of the PDPH
were well-tolerated or responded adequately to simple analgesics.
3.2. In vitro
Figs. 1 and 2 present results concerning the study on the
flexibility of 30 pencil-point WhitacreTM type (Vygon1) needles.
The deviations of needle extremities varied from 4  1 mm (25 G)
to 21  4 mm (27 G) when force was applied to the proximal end, and
from 6  1 mm (25 G) to 34  2 mm (27 G) when applied to the distal
end. Regardless of the applied force, 27 G needles were more inflected
than 25 G ones (P  0.02). The mean flexibility ratios for the 27 G
needle was 1.4  0.1 compared to the 26 G and 2.0  0.1 compared to
the 25 G (supplementary data, Appendix A).
4. Discussion
In our prospective study, during SA for caesarean section,
puncture failures occurred more frequently when 27 G needles
were utilised, with respect to 25 G and 26 G ones, probably due to
the greater flexibility of the former, whose characteristics were all
demonstrated in vitro.
Spinal anaesthesia represents the gold standard for obstetric
procedures, and its use has increased in recent years [1,2].
The most frequent potential complications are failure of the
procedure (with its conversion into GA), and PDPH. Puncture
Fig. 1. Needle deviation obtained in a laboratory model showing the forces on the
abscissa-axis (indicated as Newton [N]), and the deviations on the ordinate-axis
(indicated as millimetres [mm]) when a standardised force is applied at one
centimetre from the distal end of the introducer sheath (n = 10 needles per group;
proximal force A). * P < 0.05, 27 G vs. 25 G and 26 G; # P < 0.05, 25 G vs. 26 G; §
P < 0.05, 27 G vs. 25 G; no other significant differences.
279
Fig. 2. Needle deviation obtained when a standardised force is being applied to the
distal end of the needle (n = 10 needles per group; distal force B). * P < 0.05, 27 G vs.
25 G and 26 G; # P < 0.05, 25 G vs. 26 G; § P < 0.05, 27 G vs. 25 G; no other
significant differences.
failures are more common with 27 G needles than with 25 G and
26 G. Previous studies have found similar results comparing 27 G
needles with 25 G for caesarean section [7,12]. The literature
concerning non-obstetric and obstetric surgery confirms these
data, indicating that the subarachnoid space is more difficult to
detect when the calibre of the needle decreases [12–14]. The uses
of needles as this as 29 G or 30 G makes a subarachnoid puncture
very difficult, mainly due to a too high flexibility [15]. In obstetrics,
one study has evaluated 27 G pencil-point WhitacreTM needles and
has reported an incidence of spinal puncture failure of 7.4%. A
difference in the flexibility of the needles used could explain this
difference, but the study does not specify this feature [7]. Recently,
another study reports a large Chinese experiment using only 27 G
Whitacre spinal needles for elective caesarean section with a much
lower incidence (0.5%) of failure requiring supplementary GA
[15]. Failure percentages concerning 25 G needles ranging
between 0 and 1.3% have also been reported [7,16]. To our
knowledge, there are no data in the obstetric literature about 26 G
pencil-point needles. The only available data concerning 26 G
needles comes from a correspondence, and relates a percentage
failure of 4.3% compared to 6.5% using 29 G needles in spinal
anaesthesia for caesarean section [17]. Nevertheless, the lack of
precision concerning the type of needle utilised and the number of
parturient studied complicates comparisons with our findings.
Previous studies, using a 26 G Quincke bevel needle and excluding
obstetrics, reported a consistent percentage of headaches between
3.7 and 7.5% [18]. The use of pencil-point needles may explain this
difference. Several hypotheses have been proposed to explain the
more frequent puncture failures when using 27 G needles:
 the lack of sensation during the passage through the subcutane-
ous tissues, which are infiltrated in pregnant women [14];
 the slowness of the observable CSF flow through the needle
[14,16];
 especially a greater flexibility of the needle [7,11].
Furthermore, the denser the material pierced by the needle (e.g.
tomato, potato, pig muscle, packing foam) and the thinner the
needle, the more its path is deflected. In summary, we can say that
the presence of deflection increases as a function of the density of
the material crossed [8,11]. Similarly, on two muscle tissue models
(beef meat or paraspinal pork muscle), needle deflection was more
evident when the needle was fine [9,19]. The limit of these models
is their lack of reproducibility.
280 F. Fama’ et al. / Anaesth Crit Care Pain Med 34 (2015) 277–280
The incidence of PDPH varies from 0 to 4% for 25 G needles in
obstetrics, and from 0 to 1.7% for 27 G ones. A prospective study
demonstrated that the median of occurrence happened on the 2nd
postoperative day (48%); however, all headaches occurred before
the 6th day. In a recent meta-analysis performed on 10 studies, it
was found that headache onset usually occurred before the 6th
(82%), and before the 8th (100%) postoperative day, respectively
[4,5,20–22]. PDPH incidence decreasing in relation to finer needle
diameters is largely reported in the literature [1,2,20,23–26] and
the high number of dural punctures increases the incidence of side
effects, in particular PDPH [27].
In this study, we have considered only failures concerning the
subarachnoid space identification related to needle features,
disregarding anaesthesia failures pertaining to other causes. The
existence of these known anaesthesia failures have been explained
as follows:
 an intrathecal high volume of CSF that dilutes the local
anaesthetic;
 a misdistribution of local anaesthetic for a dural ectasia, presence
of extradural cysts;
 a mutation of the local anaesthetic receptors (sodium channel);
 a decrease of the ionised form in situ or a defect in drug
manufacturing or preparation [28,29].
Evidence of a normal local anaesthetic concentration in the CSF
simultaneous with SA failure suggests a local cause [30].
5. Conclusion
The primary endpoint of the present study was to compare
puncture failure rates with 25 G, 26 G or 27 G pencil-point needles
during SA for caesarean section, and to analyse needle flexibility by
applying a standardised force. In our experimental and reproduc-
ible model, the flexibility of the three types of needles was assessed
in vitro using a dynamometer. A limit in the study was the use of
only one brand of needle, WhitacreTM type (Vygon1), yet in our
opinion this also resulted in a guarantee of uniformity. Failure rates
were comparable between the 25 G and 26 G groups. The incidence
of PDPH reached 3%, a percentage comparable to the data found in
the literature concerning 25 G and 27 G needles, even if the
number of women is insufficient for drawing conclusions.
This study shows, assuming that the potential risk of headache
after subarachnoid puncture is directly proportional to needle
diameter, that a 26 G pencil-point needle may be the optimal
choice when performing SA for a caesarean section. The flexibility
characteristics in our reproducible model can be used to compare
different diameter needles proposed by manufacturers and
develop new needles thinner than 26 G but with an equivalent
flexibility.
Disclosure of interest
The authors declare that they have no conflicts of interest
concerning this article.
Acknowledgements
For the accurate revision of the manuscript, we wish to thank
Mr Sam Palella, a native speaker, with an extensive experience in
the correction of scientific papers.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.accpm.2015.05.005.
References
[1] Kestin IG. Spinal anaesthesia in obstetrics. Br J Anaesth 1991;66:596–607.
[2] Auroy Y, Laxenaire MC, Clergue F, Péquignot F, Jougla E, Lienhart A. Anesthésies
en obstétrique. Ann Fr Anesth Reanim 1998;17:1342–6.
[3] Caton D. A century of spinals for childbirth. Int J Obstet Anesth 2000;9:149–50.
[4] Choi PT, Galinski SE, Takeuchi L, Lucas S, Tamayo C, Jadad AR. PDPH is a
common complication of neuraxial blockade in parturients: a meta-analysis of
obstetrical studies. Can J Anaesth 2003;50:460–9.
[5] Douglas M, Ward ME, Campbell DC, Bright SB, Merrick PM. Factors involved in
the incidence of postdural puncture headache with the 25 gauge Whitacre
needle for obstetric anesthesia. Int J Obstet Anesth 1997;6:220–3.
[6] Morewood GH. A rational approach to the cause, prevention and treatment of
postdural puncture headache. CMAJ 1993;149:1087–93.
[7] Smith EA, Thorburn J, Duckworth RA, Reid JA. A comparison of 25 G and 27 G
Whitacre needles for caesarean section. Anaesthesia 1994;49:859–62.
[8] Hatfalvi BI. Postulated mechanisms for postdural puncture headache and
review of laboratory models. Clinical experience. Reg Anesth 1995;20:329–36.
[9] Sitzman BT, Uncles DR. The effects of needle type, gauge, and tip bend on spinal
needle deflection. Anesth Analg 1996;82:297–301.
[10] Ahn WS, Bahk JH, Lim YJ, Kim YC. The effect of introducer gauge, design and
bevel direction on the deflection of spinal needles. Anaesthesia 2002;57:
1007–11.
[11] Kopacz DJ, Allen HW. Comparison of needle deviation during regional anes-
thetic techniques in a laboratory model. Anesth Analg 1995;81:630–3.
[12] Tarkkila P, Huhtala J, Salminen U. Difficulties in spinal needle use. Insertion
characteristics and failure rates associated with 25-, 27- and 29-gauge
Quincke-type spinal needles. Anaesthesia 1994;49:723–5.
[13] Lesser P, Bembridge M, Lyons G, Macdonald R. An evaluation of a 30-gauge
needle for spinal anaesthesia for caesarean section. Anaesthesia 1990;45:767–8.
[14] Carrie LE, Collins PD. 29-gauge spinal needles. Br J Anaesth 1991;66:145–6.
[15] Sng BL, Lim Y, Sia AT. An observational prospective cohort study of incidence
and characteristics of failed spinal anaesthesia for caesarean section. Int J
Obstet Anesth 2009;18:237–41.
[16] Campbell DC, Douglas MJ, Pavy TJ, Merrick P, Flanagan ML, McMorland GH.
Comparison of the 25-gauge Whitacre with the 24-gauge Sprotte spinal needle
for elective caesarean section: cost implications. Can J Anaesth 1993;40:1131–5.
[17] Lyons G, Macdonald R. 29-gauge spinal needles. Br J Anaesth 1991;67:222.
[18] Quaynor H, Corbey M, Berg P. Spinal anaesthesia in day-care surgery with a 26-
gauge needle. Br J Anaesth 1990;65:766–9.
[19] Drummond GB, Scott DH. Deflection of spinal needles by the bevel. Anaesthe-
sia 1980;35:854–7.
[20] Corbey MP, Bach AB, Lech K, Frørup AM. Grading of severity of postdural
puncture headache after 27-gauge Quincke and Whitacre needles. Acta Anaes-
thesiol Scand 1997;41:779–84.
[21] Landau R, Ciliberto CF, Goodman SR, Kim-Lo SH, Smiley RM. Complications
with 25-gauge and 27-gauge Whitacre needles during combined spinal-epi-
dural analgesia in labor. Int J Obstet Anesth 2001;10:168–71.
[22] Shutt LE, Valentine SJ, Wee MY, Page RJ, Prosser A, Thomas TA. Spinal anaesthesia
for caesarean section: comparison of 22-gauge and 25-gauge Whitacre needles
with 26-gauge Quincke needles. Br J Anaesth 1992;69:589–94.
[23] Halpern S, Preston R. Postdural puncture headache and spinal needle design.
Meta-analyses. Anesthesiology 1994;81:1376–83.
[24] Vallejo MC, Mandell GL, Sabo DP, Ramanathan S. Postdural puncture head-
ache: a randomized comparison of five spinal needles in obstetric patients.
Anesth Analg 2000;91:916–20.
[25] Tourtellotte WW, Henderson WG, Tucker RP, Gilland O, Walker JE, Kokman E.
A randomized, double-blind clinical trial comparing the 22 versus 26 gauge
needle in the production of the post-lumbar puncture syndrome in normal
individuals. Headache 1972;12:73–8.
[26] Flaatten H, Felthaus J, Kuwelker M, Wisborg T. Postural postdural puncture
headache. A prospective randomised study and a meta-analysis comparing
two different 0.40 mm O.D. (27 g) spinal needles. Acta Anaesthesiol Scand
2000;44:643–7.
[27] Seeberger M, Kaufmann M, Staender S, Schneider M, Scheidegger D. Repeated
dural punctures increase the incidence of postdural puncture headache.
Anesth Analg 1996;82:302–5.
[28] Bevacqua BK, Cleary WF. Relative resistance to intrathecal local anesthetics.
Anesth Analg 1994;78:1024–6.
[29] Fettes PD, Jansson JR, Wildsmith JA. Failed spinal anaesthesia: mechanisms,
management, and prevention. Br J Anaesth 2009;102:739–48.
[30] Steiner LA, Hauenstein L, Ruppen W, Hampl KF, Seeberger MD. Bupivacaine
concentrations in lumbar cerebrospinal fluid in patients with failed spinal
anesthesia. Br J Anaesth 2009;102:839–44.