Incidence of Fistulas After Drainage of
Acute Anorectal Abscesses
Kari-Pekka J. H~l-nfilfiinen, M.D., A. Peter Sainio, M.D.
From the Fourth Department o f Surgery, Helsinki University Central tgospital, tgelsinki, Finland
PURPOSE: The aim of this study was to assess the incidence
of anal fistulas and factors related to this incidence after
incision and drainage of acute cryptoglandular anorectal
abscesses. METHODS: Of 170 patients without previous
anal fistulas, 146 were followed up for an average of 99
(range, 22-187) months after abscess drainage or until a
fistula appeared. RESULTS:Fifty-four (37 percent) patients
developed a fistula, and 15 (10 percent) patients developed
a recmTent abscess. The incidence of fistulas was higher in
females than in males (50 vs. 31 percent; P = 0.0403),
especially regarding anterior abscesses (88 vs. 33 percent).
Abscesses growing Escherichia coil were more prone to
fistula formation than those growing other bacteria (46 vs.
27 percent; P = 0.0368). CONCLUSION:Incision and drain-
age alone of acute anorectal abscesses is recommended,
because an unnecessa,T primary fistulotomy can be avoided
in more than half of the patients by this approach. For
superficial anterior abscesses in females, however, primary
fistulotomy may be considered. [Key words: Anorectal ab-
scess; Anal fistula; Surgical treatment]
H~im~il~iinen K-PJ, Sainio AP. Incidence of fistulas after drain-
age of acute anorectal abscesses. Dis Colon Rectum 1998;
41:1357-1362.
ccording to the cryptoglandular theory, the
A source of anorectal suppurative disease is in-
fected intramuscular anal glands. 1 Traditionally, the
treatment of acute abscesses has been simple incision
and drainage, but incidence rates of up to 66 percent
for recurrent abscesses or subsequent anal fistulas
have been reported, z This has led many surgeons to
look for and to treat underlying anal fistulas already in
the acute abscess phase; but because there are no
reliable clinical factors to predict which abscesses will
develop into fistulas and which will not, some pa-
tients may unnecessarily become subjected to the
adverse effects of definite fistula surgery, such as
disturbance in anal control. 3
Some controversies still exist as to the appropriate
treatment of acute anorectal abscesses, especially sim-
ple drainage vs. primary fistulotomy. We attempted to
assess factors that might predict outcome and that
should be considered in choosing the optimal treat-
ment policy for this disorder. We performed this as-
sessment by studying our long-term results of abscess
drainage.
No reprints are available.
1357
PATIENTS AND METHODS
Between 1980 and 1993, 170 patients (118 males)
with anorectal abscess of nonspecific cause and with
no previous anal fistulas were seen at our hospital. All
patients, except 48 with a subsequent fistula treated at
our hospital, were sent a questionnaire or were inter-
viewed by telephone concerning their anorectat dis-
orders for indications of recurrent abscesses or fistu-
las. Also, relevant case records from other hospitals
were acquired and reviewed. Six patients had died
and 18 could not be reached, either by letter or tele-
phone. Follow-up data were available on 146 (86
percent) patients for a mean of 5.8 years (69 -+ 56;
range, 1-187 months), which ended w h e n a fistula
was detected or the questionnaire was answered. For
those w h o had not developed a fistula, the mean
length of follow-up was 8.3 years (99 -+ 47; range,
22-187 months). There were 102 males. The mean
age was 38.6 +- 10.5 (range, 16-76) years at the time
of presentation, 40.0 -+ 10.7 years for males and
35.4 + 9.2 years for females (P = 0.0295). Fifty (34
percent) of the patients had had anorectal abscesses
previously.
For superficial abscesses simple incision and drain-
age under local anesthesia on an outpatient basis was
done on 85 (58 percent) patients. For deeper ab-
scesses regional or general anesthesia was used and
in these 61 (42 percent) patients the abscess cavity
was usually thoroughly curetted and the loculations
were broken down. An internal fistulous opening was
not sought, nor was primary fistulotomy performed,
according to the treatment policy of our hospital.
Penrose drains were used in 98 (67 percent) cases.
Ninety-five (65 percent) of the patients had received
perioperative antibiotics.
The abscesses were classified as anterior or poste-
rior, but if neither of these circumanal locations were
apparent, they were classified as lateral. Specimens
for microbiologic culture were obtained from 116 of
the 146 (79 percent) abscesses, but in one case there
was no growth.
Data in the text are expressed as means + standard
deviation (SD). For statistical analyses t-test, chi-
1358 H~M)kI~INEN AND SAINIO
Table 1.
Incidence of Anal Fistula After Incision and Drainage of Anorectal Abscesses According to
Their Circumanal Location
Anterior Lateral
Males 13/40 (33) 3/19 (16)
Females 14/16 (88) 2/11 (18)
Total 27/56 (48) 5/30 (17)
Numbers in parentheses are percentages.
* P = 0.0403.
1" P = 0.0129 between distributions by gender.
squared test, or Fisher's exact probability test was
used. The probability level of 0.05 for a Type I error
was considered statistically significant.
RESULTS
Fifty-four of the 146 (37 percent) patients devel-
o p e d an anal fistula and 15 (10 percent) patients
d e v e l o p e d a recurrent abscess but no demonstrable
fistula. The n u m b e r of previous abscesses, the type of
anesthesia, and the use of perioperative antibiotics or
drains had no statistically significant effect on fistula
formation. The incidence of fistulas was greater in
females than in males, particularly after incision and
drainage of anterior abscesses (Table 1). Forty-five
(83 percent) of the fistulas were detected within three
years of abscess d r a i n a g e .
Abscesses growing Escherichia coil developed into
fistulas more often than those growing other bacteria
(Table 2). Furthermore, abscesses with any bowel-
derived organisms were m o r e susceptible to fistula
formation than those with nonbowel-derived organ-
isms, but the difference was not statistically significant
(P = 0.0912). Skin-derived staphylococci or strepto-
cocci solely were cultured from 2 of 44 (5 percent)
abscesses developing into fistulas, c o m p a r e d with 9
of 71 (13 percent) abscesses with no fistula formation
(P = 0.1496). There was no difference in bacterial
growth by gender or location of abscess.
DISCUSSION
Simple incision and drainage undoubtedly is the
most popular w a y to treat anorectal abscesses. 2-6 An
easy technique and the possibility of ambulatory sur-
gery favor this approach. The drawbacks are said to
be a high recurrence rate and a long healing time. A
long healing time is associated with deroofing proce-
dures preferred by some authors7 Our recurrence
rates, t0 percent for abscess and 37 percent for fistu-
las, show that an anal fistula is not an inevitable result
of an anorectal abscess, and in more than 60 percent
Dis Colon Rectum, November 1998
Posterior Total
16/43 (37)1" 32/102 (31)*
6/17 (35)t 22/44 (50)*
22/60 (37) 54/146 (37)
Table 2.
Microbiology of Abscesses in Relation to
Fistula Formation
Organisms Fistula No Fistula
Bowel-derived* 40 (42) 56 (58)
Nonbowel-derived 4 (21) 15 (79)
Escherichia coti 31 (46)1" 36 (54)
No E, coli 13 (27)1" 35 (73)
Numbers in parentheses are percentages.
* Including anaerobes except Bacteroides melanino-
genicus,
1 P = 0,0368.
of the patients, there has b e e n no need for definite
fistula surgery during a long period of follow-up.
Bearing in mind that most recurrences arise early after
abscess drainage, 3' 8-s0 it is very likely that only a few
further fistulas, if any, will ever develop in these
patients, although in one study 11 30 percent of recur-
rences were detected more than two years after ab-
scess drainage.
In only a few studies reporting on the incidence of
recurrence (Table 3) were all acute abscesses man-
aged coherently b y drainage procedures, but those
with a detectable internal opening were subjected to
immediate fistulotomy and were excluded from most
studies, but included in some others. This makes it
difficult to compare recurrence rates a m o n g the stud-
ies and b e t w e e n the studies and our results.
In general, an associated fistula is found in 6 to 43
percent of abscesses at the time of drainage 7, 10, sx, i4-16
and in 76 percent of recurrent abscesses. 17 In primary
abscesses an internal opening was identified in 26 per-
cent of the cases, ¢ but in one article ~8 an internal open-
ing or crypt of origin was found as often as in 88 percent
of the cases. The incidence of recurrence seems to be
slightly higher after drainage of recurrent abscesses than
primaw ones, 4 and the same applies to abscesses with
detectable internal openings compared with those with-
out. 11 In a recent study 5 on primary- abscesses with
 <
o

Table 3.
Incidence of Recurrence After Drainage of Anorectal Abscesses Reported in the Literature
Recurrence
Reference No. of Procedure Follow-Up Note
Patients Fistula Abscess (mean (range))
(no. (%)) (no. (%))
Wilson 9 100 Incision, curettage and 7 (7) 15 (15) 27 (23-36) Months
primary suture
Doberneck 12 50 Drainage 6 (12) 2 (4) -- Primary fistulotomy (n = 15) excluded.
Ramstead 8 139 Incision and drainage 25 (18) -- --
Lai e t aL ~e 292 Incision and drainage 80 (27) 38 (13) -- Primary fistulectomy (n = 32) included. O
Sainio 13 184 Incision and drainage 64 (35) -- (1.5-11.5 Years) Children (n = 30) included (5 with fistula). O
Seow-Choen e t aL e 89 Incision and drainage, or 10 (11) -- 122 (104-136) Weeks* Primary abscesses with no internal opening. >
deroofing or saucerization Primary fistulotomy and fistulectomy (n = 31)
excluded. Type of recurrence not specified. 0
Buchan and Grace 11 25 Incision and drainage 4 (16) 4 (16) @-9 Years)1" Primary fistulotomy within 10 days (n = 13)
excluded.
113 Saucerization 16 (15) 21 (19) >
Scoma e t a/. 2 232 Skin excision and drainage 154 (66) -- (6 Months-13 Years) Primary fistulotomy excluded. Type of recurrence GO

not specified.
Weber and Buchmann 7 48 Skin excision and drainage 15 (31) 4 (8) Primary fistulotomy (n = 18) excluded. c.~

Vasilevsky and Gordon 4 83 Skin excision and drainage 31 (37) 9 (11) (1-106 Months) Intersphincteric abscesses and patients with
additional procedures (n = 20) excluded.
Schouten and van 32 Skin excision and drainage 3 (9) 10 (31) 42.5 (12-57)1" Months* Primary abscesses. Primary fistulectomy (n = 34)
Vroonhoven 3 excluded.
Ramanujam e t a l . 15 668 Skin excision and unroofing 22 (3) 3 (0.4) 36 (12-60)1- Months Abscesses with no internal opening. Primary and
staged fistulotomy (n = 355) excluded.
* Median value.
1" For total material.

%o
1360 H~dVL~LAINENAND SAINIO
proven internal openings, the recurrence rate was as
low as 15 percent within one year.
Primary fistulotomy has not gained widespread
popularity, although g o o d results with regard to re-
currence rates and functional outcome have been
produced b y this method, v' 15, is, 19 However, less sat-
isfactory results have also b e e n reported. 3' z0 The
greatest disadvantage of this approach is the possibil-
ity of an unnecessary fistulotomy, with its adverse
effects on anal function. In addition, an iatrogenic
fistulous track may inadvertently- b e created by
searching an underlying fistula b y careless probing.
We did not study the functional outcome, and be-
cause all our abscesses were m a n a g e d in the same
way, w e could not c o m p a r e the results from various
treatment modalities. In general, w e believe it is jus-
tified to drain only the acute abscess. If a fistula
develops, we deal with it in the second stage wherein
the type of fistula can be assessed better and proper
treatment for cure can be completed safely. Because
most anterior abscesses in females d e v e l o p e d into
fistulas, a finding not previously reported in the liter-
ature, primary fistulotomy m a y be considered in these
cases, if the abscess and the associated fistula have a
low location. On the other hand, laying o p e n of an-
terior fistulas in females has b e e n considered contra-
indicated because of an increased risk of anal incon-
tinence 18, 20 However, in an earlier study from our
hospital, no such an association was noted. 21 We
cannot explain the aforementioned observation of a
high incidence of fistulas associated with anterior ab-
scesses and female gender, but anatomical aspects of
the anorectum m a y play a part.
A frequently presented claim is that high abscesses
are associated with an increased risk of recurrence. 9
We were not able to classify our abscesses as low or
high reliably enough to make comparisons of this
kind. However, abscesses m a n a g e d on an inpatient
basis under regional or general anesthesia very likely
were d e e p ones, and no difference in the frequency
of fistula formation was noted between the abscesses
in this respect. Similar observations were m a d e by
other authors. 1~
Not m u c h attention has b e e n paid to the microbi-
ology of anorectal suppurative disease, probably be-
cause antibiotics do not have an important role in its
management. The available data, however, indicate
that the microbiology of the acute abscess can give
useful information o n the risk of fistula formation. 9' 14
The present study s h o w e d that abscesses growing
bowel-derived organisms, especially E. coli, were
Dis Coton Rectum, November 1998
most susceptible to fistula formation, but fistulas m a y
also develop from abscesses growing skin-derived
cocci only. Thus microbiology does not seem to be a
reliable predictor of fistulas and is of no help in
deciding the optimal treatment of a particular case in
the acute abscess stage.
It is u n k n o w n w h y some anorectal abscesses de-
velop into fistulas and others do not and what part
m a n a g e m e n t plays. The exceptionally low recurrence
rate reported b y Ramanujam el al. 15 suggests that an
aggressive attitude in m a n a g e m e n t is favorable, but
because there are no randomized comparative studies
available on various drainage procedures, no recom-
mendations of this kind can be given. As a general
surgical principle, g o o d drainage without undue de-
lay and breaking d o w n of all loculations probably is
the most important aspect in minimizing the risk of a
persistent fistula.
CONCLUSION
Because a considerable n u m b e r of patients have no
further problems after incision and drainage of acute
anorectal abscesses, this treatment seems to be war-
ranted as the first stage, and definite fistula surgery
can be completed more safely later if necessary. This
policy would avoid an unnecessary primary fistulot-
omy, with its potential adverse effects on anal func-
tion, in more than half of the patients. However,
because anterior abscesses in females are exception-
ally prone to fistula formation, these abscesses, if low
ones, m a y be m a n a g e d by one-stage procedures.
REFERENCES
1. Parks AG. Pathogenesis and treatment of fistula-in-ano.
BMJ 1961;1:463-9.
2. Scoma JA, Salvati EP, Rubin RJ. Incidence of fistulas
subsequent to anal abscesses, Dis Colon Rectum 1974;
17:357-9.
3. Schouten WR, van Vroonhoven TJ. Treatment of ano-
rectal abscess with or without primary fistulectomy:
resuks of a prospective randomized trial. Dis Colon
Rectum 1991;34:60-3.
4. Vasilevsky C-A, Gordon PH. The incidence of recurrent
abscesses or fistula-in-ano following anorectal suppu-
ration. Dis Colon Rectum 1984;27:126-30.
5. Tang C-L, Chew S-P, Seow-Choen F. Prospective ran-
domized trial of drainage alone vs. drainage and fistu-
lotomy for acute perianal abscesses with proven inter-
nal opening. Dis Colon Rectum 1996;39:1415-7.
6. Seow-Choen F, Leong AF, Goh HS. Results of a policy
Vol. 41, No. 11 DRAINAGE OF ANOR£CTALABSCESSES
of selective immediate fistulotomy for primary anal ab-
scess. Aust N Z J Surg 1993;63:485-9.
7. Weber E, Budmlann P. Er6ffnunganorectaler Abscesse mit
oder otme Fistelspaltung? Chimrg 1982;53:270-2.
8. Ramstead KD. Do anal abscesses lead to fistulae-in-ano?
Br J Clin Pract 1983;37:58-60.
9. Wilson DH. The late results of anorectat abscess treated
by incision, curettage, and primary suture under antibi-
otic cover. Br J Surg 1964;51:828-31.
10. Athanasiadis S, Fischbach N, Heumtiller L, Maria B.
Abscessexcision und prim~ire Fistulektomie als Initial-
therapie des periproktitischen Abscesses: eine prospe-
ktive Analyse bei 122 Patienten. Chirurg 1990;61:53--8.
11. Buchan R, Grace RH. Anorectal suppuration: the results
of treatment and the factors influencing the recurrence
rate. Br J Surg 1973;60:537-40.
12. Doberneck RC. Perianal suppuration: results of treat-
ment. Am Surg 1987;53:569-72.
13. Sainio P. Fistula-in-ano in a defined population: inci-
dence and epidemiological aspects. Ann Chir Gynaecol
1984;73:219-24.
14. Henrichsen S, Christiansen J. Incidence of fistula-in-ano
complicating anorectal sepsis: a prospective study. Br J
Surg 1986;73:371-2.
15, Ramanujam PS, Prasad ML, Abcarian H, Tan AB. Peri-
anal abscesses and fistulas: a study of 1,023 patients. Dis
Colon Rectum 1984;27:593-7.
16. Lai CK, WongJ, Ong GB. Anorectal suppuration: a review
of 606 patients. Southeast Asian J Surg 1983;6:22-6.
17. Chrabot CM, Prasad ML, Abcarian H. Recurrent anorec-
taI abscesses. Dis Colon Rectum 1983;26:105-8.
18. Fucini C. One stage treatment of anal abscesses and
fistulas: a clinical appraisal on the basis of two different
classifications. Int J ColorectaI Dis 1991;6:12-6.
19. McElwain JW, MacLean MD, Alexander RM, Hoexter B,
Guthrie JF. Symposiurn: Anorectal problems. Experi-
ence with primary fistulectomy for anorectal abscess, a
report of 1,000 cases. Dis Colon Rectum 1975;18:646-9.
20. Hanley PH. Rubber band seton in the management of
abscess-anal fistula. Ann Surg 1978;187:435-7.
21. Sainio P, Husa A. Fistula-in-ano: clinical features and
long-term results of surgery in 199 adults. Acta Chir
Scand 1985;151:169-76.
Invited Editorial
To the Editor--The authors have attempted to shed
light on the controversy over whether patients with
anorectal abscesses should undergo primary fistulot-
omy. The size and retrospective nature of their study
limits definitive conclusions, but the authors sug-
gested that primary fistulotomy should be avoided
except for anterior abscesses associated w'ith low fis-
tulas in females. Several aspects of the authors' expe-
rience and conclusions merit additional discussion. In
1361
the United States an increasing number of patients are
having their anorectal abscesses treated as outpatients
with the use of local anesthesia and limited or catheter
drainage in an office setting. 1, 2 This method, in addi-
tion to being simple and cost effective, has been
associated with quick healing and a reasonable recur-
rence rate. With the limited anesthesia and exposure
available in an office setting, most surgeons are hes-
itant to perform a fistulotomy during the acute pre-
sentation. This reluctance, which I support, reduces
the importance of the authors' purpose. Anterior fis-
tulotomies in females carry a significant risk of alter-
ing continence. Therefore, even if this subset of pa-
tients has a high incidence of fistula identification
after treatment, most surgeons would rather perform a
fistulotomy, if necessary, after the inflammation has
subsided.
The authors' data on abscess microbiology suggest
that abscesses growing bowel-derived flora have a
higher likelihood of fistula formation. Although it is
not a new finding, the inadequate specificity of this
finding and the cost and time in obtaining cultures has
limited the clinical usefulness of this information. Cul-
tures of the anorectal abscess drainage are not rou-
tinely obtained by most practitioners. 3
Finally, "breaking d o w n loculations" is suggested
as important, but no data to support this is presented.
Very fear abscesses in my experience have been truly
"loculated", and aggressive exploration of the abscess
cavities has been unnecessary. One worries that ag-
gressive exploration might lead to additional tissue
damage.
The authors should be c o m m e n d e d for attempting
to answer questions on the management of a very
c o m m o n condition. Although it does not provide de-
finitive answers, one hopes this article will generate
additional discussion.
REFERENCES
1. Beck DE, Fazio VW, Jagelman DG, Lavery IL, Weakley
FL. Catheter drainage of ischiorectal abscesses. South
Med J 1988;81:444-6.
2. Gregorcyk S, Hicks, T, Opelka F, Timmcke A, Beck D.
Outpatient management of anorectaI abscesses. Dis Co-
lon Rectum 1998;41:A23.
3. Vasilevsky C. Fistula in ano and abscess. In: Beck DE,
Wexner SD eds. Fundamentals of anorectat surgery.
2nd ed. London: WB Saunders, 1998:153-73.
David E. Beck, M.D.
N e w Orleans, Louisiana