SOME COMPLICATIONS OF TOOTH EXTRACTION

Lectures delivered at the Royal College of Surgeons of England

on
27th April 1961
by

Geoffrey L. Howe, F.D.S.R.C.S., M.R.C.S.Eng.

Professor of Oral Surgery, University of Durham

THE POSSIBLE COMPLICATIONS of tooth extraction are many and varied

and some of them may occur even when the utmost care is exercised.
Others are avoidable if the operator employs a plan of campaign, designed
to deal with the difficulties which he has diagnosed during a careful
pre-operative assessment. Space does not permit the discussion of every
possible complication, and so only a limited number of topics-some of
which are seldom discussed-have been selected for inclusion in this paper.
Standards of tooth extraction
An ideal tooth extraction may be defined as the painless removal of the
whole tooth or tooth root with minimal trauma to the investing tissues, so
that the wound heals uneventfully and no post-operative prosthetic problem is created. Whilst advances in the techniques of local and general
anaesthesia have conquered pain, much progress must be made before
every tooth extraction can be correctly described as " ideal ". Table I
summarizes the findings published in 11 papers describing the results of
pre-prosthetic radiographic surveys of " edentulous " jaws. These
studies reveal that in a period of 36 years, which has been notable for the
advances made in every other branch of dental surgery, progress in exodontia has been disappointing despite the invaluable assistance provided
by improvements in radiography and anaesthesia. Although a strong
case can be made for leaving certain root apices in situ, every endeavour
should be made to avoid fracturing teeth and roots in the first instance.
If a tooth fails to yield when reasonable and controlled force is applied
to it with either an elevator or forceps, the instrument should be discarded
and the cause of the difficulty sought. A pre-extraction radiograph may
prove indispensable and the use of the " transalveolar" technique of
extraction, in which the tooth is dissected from its bony attachments, is
often indicated. The findings of Gardner and Stafne (1929), summarized
in Table I, clearly demonstrate the effectiveness of these measures.
Before embarking upon a dental clearance it is a wise precaution to
radiograph all " edentulous " areas, especially the canine, premolar and
third molar regions, unless the operator has definite proof that no buried
teeth or retained roots are present in these sites.
The elimination of prosthetic difficulties
The removal of the whole tooth or root is not the sole criterion of success
in exodontia. Most extractions are followed by the construction of a
309

GEOFFREY L. HOWE
,;

Co

0
OCZ
-~

CZ
- 0 &YSO
cn~~ :

.CO
o,~C 0

Cq
t

EC) .

Ca

"0

~.

C)

s
O

C)~~~~~C

-.

,~~~~~o
~

C)

C
-t

C
_

'0Ct

04

C0

rs ti oCO

C14

CO
a
C-)

OD

K \

C)

C,)

O Q

ON4N )I
cd

C)

0.
\C

CO~~~~~~~~C
COCxY

-:Eots

U,

N C,;
^

(A

Nt

1oo

l'

0
Cq

0
0

C4ON~

CO

00~
_

N-

_
0

C
C4CS tE o 00a,

0dn

C0,

Us, ,o

0\

0o

a00o

O1

oo-1

mo SMco

tr

0C ,)

0
P
CQ
C4
~
~

~c

CAi

>0

0o0o

00

"0

0COCO

U)

:3C

COCA
r

CO

CCZ

OC

o o v oo0 o o

rO OO

moooCo
Zooo

CO

CA
0

CC
,OeC,

Or.

C
n

CO

Cl~~

~~

cO

lc

(ON

~~

C)

.=

t
_

CO

oZ _
N

C
ON\s,

"0

ON

10

C-4
ON
CO

CO
CZ

bo

0
~1.

ON.

(ON

C-"

C-4

0
0

C)

310

=C)l

SOME COMPLICATIONS OF TOOTH EXTRACTION

denture, and prosthetic difficulties should be eliminated at the time the

teeth are extracted. Many dental prosthetic failures are created by lack
of foresight on the part of the operator at the time the teeth are removed.
Thus, simple forceps extraction of the teeth or roots illustrated in Figures 1,
2 and 3 would leave enlarged tuberosities. Figure 4 shows the maxilla of a
patient in whom an extreme degree of enlargement of the tuberosities was
present. When extractions are contemplated, time spent in assessing the
case from a prosthetic point of view is never wasted.
Blood loss in exodontia
Blood loss is another complication of multiple extractions which is often
either forgotten or ignored. Table IL summarizes the results of investigations into blood loss during multiple extractions and shows that often

Fig I

Roots on enlarged tuberosities

between one-half to one and three-quarters of a pint of blood is lost when

a dental clearance is performed. These workers claim that the average
loss exceeds that sustained in simple mastectomy, inguinal herniorrhaphy
and haemorrhoidectomy, whilst the maximum loss is greater than that
which complicates hysterectomy, nephrectomy, cholecystectomy, and
thyroidectomy and may even equal the blood loss which accompanies
gastric resection. The amount of blood loss is directly correlated to the
operating time and can be reduced by the judicious use of vaso-constrictors.
This degree of blood loss is probably of little significance in fit young
adults, many of whom can donate a pint of blood for transfusion purposes
without suffering any ill-effect. However, most dental cl'earances are
performed upon older patients in whom haemopoiesis is less active and not
infrequently some degree of anaemia exists pre-operatively. Anaemia

311

*!k~ ~ ~.^;h
GEOFFREY L. HOWE

should be diagnosed and treated before extractions are undertaken. Blood

loss can be reduced by performing extractions in easy stages, the use of
vaso-constrictors and attention to post-operative haemostasis.

Post-operative impairment of labial sensation

Post-operative impairment of labial sensation can complicate the
extraction of mandibular cheek teeth or roots. The unpleasant symptoms of labial anaesthesia and paraesthesia may be of varying severity
and may persist for periods lasting from a few hours to many months
depending upon the amount of damage sustained by the inferior dental
nerve and its branches during the extraction of the tooth.

Fig. 2. Fibrous enlargement of tuberosity.

Impairment of labial sensation may complicate the extraction of mandibular third molars due to nerve damage arising in one of two ways:
nerve may be crushed or torn during the dislocation of the
tooth from its socket if the canal and its contents are either grooving
or perforating the roots of the third molar.

(a) The

(b) It may be damaged by instrumentation if it is contained within bone

which has either to be removed or used as a fulcrum for an elevator
during the delivery of the tooth.
In the series reported by Howe and Poyton (1960) the incidence was
equally divided between these two causes. These workers showed that it
was possible after careful interpretation of pre-extraction radiographs to
forecast pre-operatively the presence of grooving, notching and perforation of mandibular third molar roots by the contents of the inferior dental
canal.

312

SOME COMPLICATIONS OF TOOTH EXTRACTION

The inferior dental canal is enclosed within a tube of condensed bone,

the density of which is greater than that of the cancellous bone which
surrounds it. Radiographically this bony wall is seen as two parallel lines
of radiopacity, one roofing the canal whilst the other forms its floor. In
radiographs of some teeth having an intimate relationship to the canal the
continuity of one or both lines is broken, and a band of radiolucency is
present across the tooth root corresponding in position to the inferior
dental canal which may be narrowed as it crosses the root (Fig. 5). This
prior knowledge can be utilized in several ways:
(a) By taking the possibility of post-operative labial anaesthesia into
consideration when deciding whether or not to extract symptomless
third molars.
(b) By warning the patient of the possibility of labial anaesthesia prior
to operation.

..

~~

~~

~~ ~~ ~~

~~

~~

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.

s: _ I X B~~~~~~~~~~~~~~~~ ......... ..

..
.......

..........:M.

Fig. 3. Over-erupted maxillary third molars.

(c) By modifying the operative techniques employed so as to minimize

the risk of nerve damage. To do this the precise relationship
between the tooth and the canal must be known.
It is possible to differentiate grooving of the side of the root, apical
notching and perforation of the root by the canal contents. Grooving
when it occurs is, for practical purposes, always on the lingual side of the
root. By removing much more buccal bone than is usually the case and
moving the tooth buccally away from the inferior dental canal via the
bony defect which has been created, damage to the nerve may be either
prevented or minimized.
Only a minority of the apically notched teeth in the series reported were
vertically impacted. The classical method of removing a mesio-angular or
disto-angular impacted tooth is by the use of an elevator applied to the
mesial or distal surface of the root. As the tooth moves and its long axis
313

GEOFFREY L. HOWE

becomes vertical the apex descends and may damage the canal and its
contents. The chance of damage can be decreased in these cases by
sectioning the tooth with either a bur or an osteotome, thus altering the
" line of withdrawal " of the root or roots so that they are moved away
from the canal.
If possible, " perforated " teeth should be widely exposed by the removal
of buccal bone and then sectioned at the level of the neurovascular bundle.
The root fragments are then removed, leaving the canal contents intact.
If this procedure is not practicable the neurovascular bundle should be
divided with a sharp scalpel and the severed ends placed in apposition after
the tooth has been removed. A cleanly cut nerve regenerates more
quickly than a crushed and avulsed one (Ward, 1955).

Fig 4 Giant tuberosities.

Impairment of labial sensation may also be caused by damage to the
mental nerve unless care is exercised.
Abscesses in the buccal sulcus in the mandibular premolar region
should be opened using Hilton's method (i.e. by plunging a closed pair of
sinus forceps into the abscess cavity via a mucosal incision and then
opening them widely).
Before an incision is made along the crest of the ridge in the edentulous
mandible, the premolar area should be carefully palpated and if a thickened
mental nerve is felt its position should be marked on the mucosa and the
incision curved slightly lingually to avoid it. Elevation of the mucoperiosteal flap outlined by the incisions in such a case should be commenced well in front of and well behind the site of the nerve. Blunt
sub-periosteal dissection will enable the nerve to be identified and preserved. A metal retractor should be used to protect the nerve at all times
during the removal of bone with a bur. When removing bone in this
manner to expose premolar roots the risk of damage to the nerve is
314

SOME COMPLICATIONS OF TOOTH EXTRACTION

minimized if the maximal removal is at the expense of mesial bone when

dealing with first premolar roots and distal bone when second premolar
roots are being removed (Fry, 1958).
TABLE II
SUMMARIZING THE RESULTS OF PUBLISHED STUDIES OF BLOOD LOSS DURING
MULTIPLE EXTRACTIONS
Number of
Largest
Patients
Auithor
Loss
Comments
50
771 ml. 8 lost over 400 ml.
Gores et al. (1955)
(21 with local
(No local
anaesthetic agents)
anaesthetic agents)
2 lost over 400 ml.

Johnson (1956)

175

Connors (1959)

10

(with local
anaesthetic agents)
919 ml.
Average loss
223 ml.
700 ml.
Average loss
316 ml.

Unerupted or partially erupted second premolars may also be grooved,

notched or perforated by the contents of the inferior dental canal, and this
can be diagnosed radiographically using the criteria described in relation
to third molars (Fig. 6).

Reproduced by kind permission of the British Dental Journal

Fig. 5 (a) Radiographic superimposition. (b) Radiograph ot tooth with grooved

root.

Displacement of a tooth or root into the lingual pouch

Displacement of a tooth or root into the lingual pouch is a rarer complication of the removal of mandibular third molars. Predisposing factors
include thinning or perforation of the lingual alveolar plate, the very loose
attachment of the lingual periosteumi in the mandibular molar region and
the presence of fine, easily fractured, lingually situated roots. If the tooth
is lying in "lingual obliquity", not only is the lingual plate thinned, but the
" line of withdrawalR"of the tooth favours its displacement into the lingual
pouch. Errors in technique include inadequate exposure during attempted

elevation of the tooth or root from its socket whilst a blow with an osteotome which is intended to split the tooth may sometimes displace it into
315

GEOFFREY L. HOWE

the lingual pouch. Pre-operative diagnosis of the possibility of this

complication can be made after careful interpretation of a pre-extraction
radiograph. In such cases the tooth should be removed by an operative
technique employing wide exposure and careful elevation of the tooth
whilst a finger of the operator's left hand is kept upon its lingual aspect
(Howe, 1958).
Before the removal of a tooth from the lingual pouch is attempted it
should be accurately localized by bi-manual palpation and radiographs
taken in two planes at right angles to each other. If possible, the tooth
should be " milked " through the tissues to an accessible site for surgery.
An assistant then applies extra-oral pressure to the submaxillary area
whilst the surgeon places one finger of his left hand intra-orally behind the
tooth. This intra-oral and extra-oral pressure should be maintained
throughout the operation. A large lingually based muco-periosteal flap
is then elevated, the defect in the lingual plate enlarged and the tooth

Reproduced by kind permission of Messrs. John Wright & Sons Lid.

Fig. 6. Partially erupted mandi-bular premolar with root grooved by the inferior
dental nerve.

delivered with elevators. The soft tissues are re-apposed with loosely
tied sutures.
Antral involvement during tooth extraction
The roots of all the maxillary cheek teeth may be in close relationship
to the antrum. In some instances as much as one half of their roots may
form part of the wall of the air sinus and be separated from its cavity by
the antral lining alone. During the extraction of such a tooth an oroantral fistula may be created and either the whole tooth or a root be displaced into the antrum. Whilst the basic cause of both these complications is the presence of a large air sinus, fewer roots would be pushed into
antra if the following simple rules were observed:
(a) Forceps should never be applied to a maxillary cheek tooth unless
sufficient of its length is exposed, both palatally and buccally, to
allow the blades to be applied under direct vision.
(b) The apical one-third of the palatal root of a maxillary molar should
be left in situ if it is fractured during forceps extraction unless there
is a positive indication for its removal. Successfula" transalveolar
316

SOME COMPLICATIONS OF TOOTH EXTRACTION

removal of such root fragments necessitates the removal of a large

amount of alveolar bone and in most instances it is better to put the
fragment " on probation ". Palatal root fragments of vital teeth
seldom cause symptoms and can always be removed if and when
they do so. The patient must be informed that a root fragment
has been left and details of the root noted in the case records.
(c) The removal of maxillary roots or teeth by dissection should be
performed under direct vision. After a large muco-periosteal flap
has been raised, sufficient bone should be removed to permit the
insertion of an elevator above the broken surface of the root. Force

(b)

(a)

vI(

A11

Reproduced by kind permission of Messrs. John Wright & Sons

(C)

IMd.

(d)

Fig. 7. Simple repair of a freshly created oro-antral fistula.

can then be applied in such a manner that the root is moved away
from the antrum.
(d) The operator should never atteinpt to deliver fractured maxillary
roots by passing an elevator or Coupland's chisel up the socket.
(e) If a patient presents with a history of previous antral involvement
or an isolated maxillary molar the removal of the tooth should be
undertaken only after a pre-extraction radiograph has been carefully studied. A history of antral involvement complicating
previous extractions probably indicates the presence of a large air
sinus and a thin antral floor. The alveolus supporting an isolated
maxillary molar is often weakened by an extension of the antral
cavity into it.
317

GEOFFREY L. HOWE

..I

(a)

(b)

\N

a...~~~~~~~~~~~LI<.

()a

(c)

*~~~~~~~~~~1 '
Fig. 8. Repair of

an oro-antral fistula with a buccal

318

mucoperiosteal flap.

SOME COMPLICATIONS OF TOOTH EXTRACTION

Oro-antral fistula
Despite every care being exercised, oro-antral fistulae will sometimes
be created during the extraction of teeth from patients with large antra.
If promptly diagnosed and correctly treated they usually heal uneventfully.
If the operator suspects that the antrum has been opened he should
attempt to confirm his suspicions by the " nose blowing " test in which the
patient attempts to blow through his occluded nares with his mouth open.
In the presence of an oro-antral fistula, air will be heard to pass through
the fistula, and will displace a wisp of cotton wool held over the socket
whilst any blood present in the socket will be seen to bubble.

..4~*
*".
..

(a)
(b)
[Fig. 9. Repair of an oro-antral fistula with

(c)
palatal mucoperiosteal flap.

It is seldom, if ever, necessary to pass instruments or silver probes into

the freshly opened antrum and never necessary to squirt fluid into the
antrum to confirm the diagnosis. Unless the suspicion that the antrum
has been opened can be disproved the patient must be treated as if an oroantral fistula is present.

The factors which cause an oro-antral fistula to persist are the size of the
bony defect and infection of the maxillary antrum, and treatment must be
planned with these factors in mind. The fistula heals by organization of a
blood clot, and the surgeon should aim to provide support for this blood
clot and to prevent infection of the air sinus. The patient must not be
allowed to rinse the mouth until the fistula is repaired or the antrum will be
contaminated with oral flora, some of which may become pathogenic in
their new environment and cause infection. An attempt is then made to
cover the defect with a muco-periosteal flap, the actual technique employed
being governed by the surgical experience of the operator and the facilities
and operating time at his disposal when the complication occurs. The
319

GEOFFREY L. HOWE

simplest method is illustrated in Figure 7 and requires the minimum time,

surgical skill and facilities. Muco-periosteal flaps, obtained by reducing
the height of the bony socket, are loosely sutured over the defect. Whenever conditions permit it is better to use an undercut buccal flap lined with
periosteum as advocated by Fickling (1957). This is an excellent method
of repair which requires more operating time and surgical skill and is often
accompanied by some post-operative haematoma formation and loss of
depth in the buccal sulcus (Fig. 8). It is not always possible to cover
fistulae situated high up in the sulcus with a buccal flap raised either
behind or in front of the defect. In these cases it may be preferable to
use a palatal muco-periosteal flap based upon its blood supply (Fig. 9).
Palatal flaps tend to shrink when elevated from the bone, and their thickness and consistency makes them more difficult to reposition than buccal
flaps.
Whichever method is utilized to cover the bony oro-antral fistula,
additional support for the repair should be provided by covering the area
with an acrylic extension to a denture, a shellac base plate, a thin sheet of
black gutta-percha or composition, or the outer covering of a cotton wool
roll impregnated with zinc oxide and oil of cloves paste. The cover may be
either ligatured to adjacent teeth if any are present or sutured to the gum
in edentulous regions. The patient should be told to blow his nose only
when he is forced to and then to blow it without first occluding the nares.
Ephedrine nose drops and friar's balsam inhalations are prescribed and
the method of use detailed.
When a tooth or root has been displaced into the antrum it should be
removed via a modified Caldwell Luc approach if the patient is under
endotracheal anaesthesia. However, usually the patient is an out-patient
being treated under either local or nitrous oxide anaesthesia. In these
circumstances the fistula should be repaired and no attempt is made to
recover the root at this visit either by the passing of instruments or snares
into the antrum via the socket or by enlarging the bony defect.
It is often difficult to tell if a root is in the antral cavity proper or outside
the antral lining. The " head-shaking test " though helpful is by no means
infallible. In my experience, whilst the buccal roots of molars are not
uncommonly pushed up between the antral lining and the bony wall of the
air sinus it is rare for palatal roots to be so displaced. The roots of vital
teeth may for practical purposes be regarded as sterile, and if such a root
is outside the lining it may be put " on probation ". All other roots
should be removed via an extra-nasal antrostomy. The exhibition of antibiotics is indicated if the root fragment is infected, there is pre-existing
disease of the antrum, the antrum was either washed out from below or
otherwise infected during the operation, or acute maxillary sinusitis complicates the healing period. Many surgeons use antibiotics routinely in
these cases.
320

(a)

(b)

(d)

(c)

Reproduced by kind permission of Messrs. John Wright & Sons Ltd.

Fig.

10.

(e)
The treatment of a fractured maxillary tuberosity.
321

GEOFFREY L. HOWE

In the vast majority of patients the measures described above lead to

spontaneous healing of the oro-antral fistula. If the fistula persists the
antrum should be washed out with normal saline twice weekly through
the fistula and the use of nose drops and inhalations continued. A wellfitting acrylic base plate is constructed to cover the defect and to prevent
or minimize contamination of the antrum. The fistula gradually shrinks
and often heals without further surgical intervention if the epithelium
lining the narrowed track is removed periodically with either a caustic
solution or a barbed broach. Persistent oro-antral fistulae can be repaired
by the use of either a buccal or palatal periosteal lined flap once the antrum
is free of infection.
Fracture of the maxillary tuberosity
Occasionally during the extraction of an upper molar, the supporting
bone and maxillary tuberosity are felt to move with the tooth. This
mishap is due to the invasion of the tuberosity by the antrum, which is
common when an isolated maxillary molar is present especially if the tooth
is over-erupted. Pathological gemination between an upper second molar
and an unerupted or semi-erupted third molar is a rare predisposing cause.
When fracture occurs the forceps should be discarded and a large buccal
muco-periosteal flap raised. The fractured tuberosity and the tooth
should then be freed from the palatal soft tissues by blunt dissection and
lifted from the wound (Fig. 10). The soft tissue flaps are then apposed
with mattress sutures which evert the margins and are left in situ for at
least ten days (Howe, 1961).
Fickling (1957) advocates splinting the fragment until union occurs and
the tooth can be removed by dissection. Whilst this method works well
in the treatment of other types of alveolar fractures, I do not favour it for
the treatment of fracture of the maxillary tuberosity complicating tooth
extraction for the following reasons:
(a) The tooth is usually being extracted for the relief of pain due to
either caries or a periodontal condition.
(b) Firm bony union is seldom obtained and the tuberosity may
become detached again when extraction is re-attempted.
(c) The method is laborious and the construction of silver cap splints is

required.
(d) The results are no better. Following removal of the tuberosity
new bone is rapidly laid down in the area and provides a firm base
for the denture (Fig. 10e).
If a pre-extraction radiograph reveals the presence of a large antrum,
an attempt to prevent fracture of the tuberosity should be made by dissecting the tooth from its attachments and sectioning it if necessary.
Should either antral involvement or tuberosity fracture complicate the
322

SOME COMPLICATIONS OF TOOTH EXTRACTION

extraction of teeth from one maxilla it is sound practice to radiograph the

teeth on the opposite side as the other antrum may also be large.
Although the only means by which the dental surgeon can avoid experiencing any complication of tooth extraction is to cease to remove teeth,
the incidence of these mishaps could be reduced considerably by careful
surgery based upon a thorough pre-operative assessment of difficulties
and possible complications in each case. Ill-effects resulting from complications can be either minimized or entirely avoided by prompt treatment.
REFERENCES
CHEPPE, E. (1936) North West U. Bull. 36, 12.
CONNORS, J. J. (1959) Ann. Dentist, 18, 74.
COOK, J. T. (1927) Dent. Cosmos, 69, 349.
ENNIS, L. M., and BERRY, H. M. (1949) J. oral Siurg. 7, 3.
EUSTERMAN, M. F. (1921) Dent. Cosmos, 63, 901.
FICKLING, B. W. (1957) Ann. Roy. Coll. Surg. Engl. 20, 13.
FRY, Sir W. Kelsey (1958) Personal communication.
GARDNER, B. S., and STAFNE, E. C. (1929) Amer. dent. Surg. 49, 321.
GORES, R. J., ROYER, R. Q., and MANN, F. D. (1955) J. oral Surg., 13, 299.
HOWE, G. L. (1958) Brit. dent. J. 104, 283.
(1961) The Extraction of Teeth. Bristol, Wright.
and POYTON, H. G. (1960) Brit. dent. J. 109, 355.
JOHNSON, R. L. (1956) J. dent. Res. 15, 175.
LOGAN, W. H. G. (1921) J. Nat. dent. Ass. 8, 126.
MOLT, F. F. (1925) J. Amer. dent. Ass. 12, 788.
SmITH, E. S. (1946) J. Amer. dent. Ass. 33, 584.
STORER, R. (1957) Brit. dent. J. 103, 344.
SWENSON, M. G. (1944) J. Amer. dent. Ass. 31, 475.
WAGGENER, D. T., and AUSTIN, L. T. (1941) J. Amer. dent. Ass. 28, 1835.
WARD, T. G. (1955) Dent. Delin. 6, 3.

APPOINTMENT OF FELLOWS AND MEMBERS

TO CONSULTANT POSTS
D. AINSLIE, F.R.C.S.
R. ALHADEFF, F.R.C.S.
E. J. HARGADON, F.R.C.S.
C. D. LINTON, F.F.A.R.C.S.
P. MAJUMDAR, F.R.C.S.

S. R. UMRIGAR, F.F.A.R.C.S.

Consultant Ophthalmologist to Moorfields

Eye Hospital.
Honorary Surgeon to the British Hospital,
Buenos Aires.
Consultant Traumatic and Orthopaedic
Surgeon, West Manchester Group of
Hospitals.
Consultant Anaesthetist, Portsmouth Group
of Hospitals.
Assistant Professor of Surgery, B.S. Medical
College, Bankura, West Bengal.
Consultant Anaesthetist, Stavanger Sykehus,
Norway.

323