INFECTION WITH
AFTER
SPECIAL
REFERENCE V.
V.
The risk factors
associated
analysed Absence of
using
discharge associated The
mechanisms
regression
analysis
the
controversy
consider of the
associated
whether
late deep infection Aglietti etal. 1974; this
deep
FROM
L.
Hospital,
infection
deep
Franco, whereas discharge
from
The follow-up
infection
a consecutive
total
involved
hip
room or arises some authors source et al.
Salvati and Wilson 1973; etal. 1977; Wilson 1977). In we
were
extensively discharge
unable
to find
the association from the wound
any
between and the
infection.
in the
healing
PATIENTS
of the
AND
the risk factors infection after focused on the
operative
wound.
803 total hip in our clinic
replacements between
31
age
of the Most intervals.
patients patients A final
V. V. Surin, K. Sundholm, Ortopediska Sweden.
MD, Orthopaedic Surgeon MD, Orthopaedic Surgeon Kliniken, Centrallasarettet,
,
1977,
METHODS
December
were
chosen
for this
study.
Box
850,
5-501
97 total the study living to The primary
for reprints British
The
mean
15
Bor#{225}s,
Development,
should
Editorial $2.00
be sent
Society
to Dr V. V. Surin.
of Bone
and
Joint
Surgery
Fack,
5-462
interrelated
risk
together,
in
71
and
other
of the hip or epiphyseolysis in 55, post-traumatic changes
diagnoses
in
eight.
In
on the indications 33 failed hemiarthroplasties procedures.
without
first 1 1 3 total hip antibiotic prophylaxis;
phylaxis Lidgren
according to the method and Lindberg (1977)
defined (Lidwell
Multiple
their effect
answered by 609 patients with We also registered all revision joint for the 86 patients with
dislocation arthritis
previously operated replacement were failed extra-articular The
factors.
potentiate
hip replacements who had died in the course of period. The shortest follow-up for all patients the end of the study was three years. diagnoses for the hip operation were as follows: osteoarthritis in 587 hips, osteoarthritis second-
ary to congenital in 82, rheumatoid
replacements routine described was used
the for
145
hips
total joint and 1 12
were antibiotic
done pro-
by Carlsson, in the 690
operations. infection after total hip replacement was as periprosthetic sepsis which led to reoperation et al. 1982). The most important criteria
considered for the assessment of the presence of sepsis were isolation from the joint of potentially pathogenic micro-organisms (isolated in 28 cases), pus in the joint (observed
in
23
cases),
a
sinus
(17
cases),
raised
erythrocyte sedimentation rate when previously not raised (30 cases) and suggestive histological findings (in 21 out of the 24 cases with stained specimens available). Data analysis. To avoid statistical problems arising from the
Surgeon and System
V#{228}nersborg, Sweden.
030l-620X834071
412
performed on 695 March 1, 1970, and
was 64 years (range 32 to 8 1 years). were being followed up at regular review was made early in 1981 and
L. B#{228}ckman, MBA, Orthopaedic Department of Organisational
1983
hip replacements
total
3 to 10 years.
questionnaires were then 706 total hip replacements. procedures of the implanted
Deep
,
of803
from
when acting
subsequent
Requests
series
ranged
multiple
risk factors,
Baer and Enneking 1977; others did not observe any from the wound and the
(Aglietti, Freeman
controversy,
complications
01
WOUND
Bor#{226}s,Sweden
methods.
that these
after
The purpose of this paper is to study associated with the deep periprosthetic total hip replacement with the interest
The first patients
THE
BACKMAN
infections
wound as an important infection (McLauchlan
report studying more the early superficial late
with
in the operating of cross-infection,
the discharging late periprosthetic
of
actuarial
indicated of the deep infection.
is acquired as a result
1976; Muller 1976; Andrews et al. 1981), correlation between
spite
with 34 deep
standard
REPLACEMENT
DISCHARGE
SUNDHOLM,
the Central
of our data
on the development In
K.
A
HIP
prophylactic antibiotics, complications after operation, discharging wounds, previous and remote infection were related to the development of deep periprosthetic infection. An early from the wound with growth of micro-organisms, though superficial in appearance, proved to be with a high risk of developing late deep infection.
operations
replacement in the ward
TO
SURIN,
From
have been
TOTAL
differences
in the
length
of follow-up
of individual
total hip replacements we used standard actuarial methods for the statistical analysis (Breslow 1975 ; Dobbs 1980). The survival was measured from the date of insertion to the
date
of the THE
first
revision
JOURNAL
procedure. OF
BONE
AND
Life JOINT
tables
SURGERY
INFECTION Table I. Micro-organisms (percentages)
AFTER isolated
TOTAL
from
the
HIP wound
related
No prophylaxis
Micro-organism
were
computed
divided
into
The estimated ratio
of
observed The infection
ratio
presence The
Superficial discharge (n=31)
Deep infection (n=18)
Superficial
Deep
discharge (n=84)
infection (n=l6)
39
61
10
19
Staphylococcus
a/bus
19
22
17
6
Gram-negative
16
11
31
25
Other
23
6
14
31
Nogrowth
35
40
25
whole
deep risk (0)
of the factor significance
prosthetic
version
II
population
infection in a subgroup rate which was simply
to
organisms were cultivated Staphylococcus a/bus grew
to the risk factors.
expected
(E)
was the
number
of having
deep
infection
of
to that in the absence of the risk ratio
in the
of the factor. was tested by
means ofthe log rank test (Peto et a!. 1977). The program used for computations of regression and survival data was the Proportional Hazard Linear Procedure from the Statistical Analysis Inc.,
computer analysis General System,
79.5.
periprosthetic
infection
hip replacements.
In 10 cases
in the first three early deep infections
months started
haematomata, By definition, the fascia, heal during
in seven all nine
diagnosed
the infection
in 34 total was diagnosed
(early deep infection). Nine as spontaneously discharging
cases with haematomata
initially sterile extended
needed extensive surgical the first three weeks.
therapy
culture. beneath
and did not
annual
probability
slightly
of
from
the
during
the
for deep infection (Fig. 1).
seven were
for
of hips years
carried
deep
aureus
found in the discharging
varied
studied.
No
out after
the seventh
without
proven
None
of
periprosthetic
association
with
deep
infection The patient’s age and sex, cortisone treatment, diabetes, operation time, blood loss, perioperative complications, urethral catheterisation for more than 24 hours, temperature rise over 38 degrees day, extreme obesity, the surgeon operation did not show any with
deep
associated
discharge
was
periprosthetic
with
periprosthetic infection Early discharge from observed
infection.
the development
the operative
a
Celsius on the seventh and the indication for statistically significant
of deep
wound.
in 1 15 wounds.
A superficial By definition, all
these wounds were completely healed before first day without active surgical therapy. pathogenic micro-organisms were cultivated
the twentyPotentially from the
draining material in 70 wounds 45 wounds the draining material
(infected discharge). was sterile. According
to the surgeons’ assessment only of superficial inflammation. The hips which developed
32 wounds a superficial
showed
In
signs
discharge
year
was
the
most
deep infections wounds with
common as well positive
microas in the cultures.
time
the
or late) was Bacteriological
were taken during hip replacements 65-B,
factors
recovered.
developed
only
(Table
VOL.
never
18 wounds: 1 1 wounds,
reoperations
of the isolation of Staphylococcus aureus only by the use ofprophylactic antibiotics
I);
was
hip replacements
from from
infection,
at risk,
The frequency was influenced (early
Investigated
in samples in samples
2.0
Staphylococcus organism superficially
reoperation
number
aureus
these 160 total infection.
Factors was
In most of the 24 cases of late deep infection the symptoms developed insidiously and we were unable to find any certain date for the start of symptoms. The calculated
Staphylococcus
association RESULTS
Deep
regime
aureus
according
risk
prophylactic
Flucloxacilin
association between any risk factor and was estimated by the risk ratio, defined
of the
to the
Staphylococcus
the
risk of having as a relative
infections. the deep as the
for subgroups
413
REPLACEMENT
of manifestation
of deep
without influence. samples (opening the operations with antibiotic
No. 4, AUGUST
1983
infection
ILa LJ
cx
La
0
1.0
0 YEARS
and
closing
of 160 consecutive prophylaxis.
Fig.
swabs) total Micro-
5 SINCE OPERATION
The annual versus time
1
probability ofreoperation for deep infection as a percentage since operation. Bars represent standard error (the positive element only being shown).
10
414
V. V. SURIN, Table
II. Observed
numbers
of deep
infections
Factor
K. SUNDHOLM, and
risk
ratios
factor
Prophylactic
antibiotics
Complication Wound
after
discharge
Previous
operations
Remote
infection
Metal-on-plastic
prosthesis
Significance: These
5P