15 April 1967
Contamination of Hospital Disinfectants with Pseudomonas Species
D. W. BURDON,* M.B., B.S.;
J. L. WMITBYt M.B., M.R.C.P., M.C.PATH.
Brit. med.J., 1967, 2, 153-155
During the investigation of an outbreak of infection caused by
Proteus mirabilis in the infants' nursery it was found that solu-
tions of disinfectants in use there were contaminated with viable
This finding prompted us to examine stocks of disinfectants
in the hospital for the presence of living bacteria.
Subsequent investigation showed that aqueous solutions of
Hibitane (chlorhexidine) and Savlon (cetrimide-chlorhexidine
mixture in the ratio 10: 1) issued from the pharmacy and in
use throughout the hospital were frequently contaminated with
organisms identified as Pseudomonas sp. The insidious nature
of this contamination and its extensiveness seemed important
enough to report, as such contamination may be more wide-
spread than is generally supposed.
sp., and were quite unconnected with the outbreak.
Description of Investigation
In the pharmacy batches of chlorhexidine and Savlon are
made up in 100- and 90-litre quantities respectively by dilution
of stock concentrate and then bottled in litre bottles.
sampled unopened and unused bottles of chlorhexidine 0.05%
from one batch, and Savlon
and cetrimide 0.5%) from three batches.
residues in bottles returned from the wards from five batches
of chlorhexidine 0.05% and five batches of Savlon 1 in 30.
All samples were found to be infected with Pseudomonas sp.
The organisms found were not identical, and three distinct
biochemical types were recognized, which we designated in rela-
tion to initial positive sample numbers as Ps. 10, Ps. 60, and
Ps. 10 was found in all the Savlon batches, and Ps. 60
and Ps. 65, but not Ps. 10, were found in chlorhexidine.
A search was made for the source of these organisms in the
pharmacy, as the finding of positive cultures from unopened
bottles suggested that contamination was most likely to be
Samples were taken from bulk stock concen-
trates of disinfectant,
distilled water, measuring-jars, mixing
the measuring-stick, polyethylene tubing,
filling apparatus, shelves, benches, sinks, and bottle-washing
Though many different organisms were found in
some of these sites, none having the characteristics of Ps. 10,
Ps. 60 or Ps. 65 was recovered.
chlorhexidine and Savlon were also tested immediately before
bottling, and were found to be sterile.
apparatus was sampled before a batch of empty Savlon bottles
was washed and was free of pseudomonads, but after con-
taminated bottles had been washed it was found to have become
Bottles contaminated with Ps. 10 were found to be
still contaminated with that organism after washing, and an
uncontaminated bottle washed immediately after a contaminated
one was found to have become contaminated.
concluded that infection was being maintained by the presence
of a residual inoculum of bacteria in the bottles before they were
refilled with fresh (sterile) disinfectant.
1 in 30 (chlorhexidine 0.05%
We also sampled
Freshly prepared batches of
In addition, washing
of the washing
t Consultant Bacteriologist, Queen Elizabeth Hospital, Birmingham 15.
in Clinical Pathology, United Birmingham Hospitals.
apparatus, so that infection would subsequently be transferred
to clean bottles.
The manufacturers of chlorhexidine and Savlon recommend
that bottles should be thoroughly rinsed in tap-water, preferably
hot, and then drained before refilling, and that isopropyl alcohol
4% v/v or ethyl alcohol 6% v/v should be added to the stock
solutions once or twice a year to eliminate possible contami-
apparatus is connected to the cold-water supply, and before
recommending that a change should be made to hot-water
washing the survival of the Pseudomonas strains isolated was
tested at higher temperatures.
were tested and proved to be remarkably resistant to high
temperature, though it was noted that cultures exposed to
higher temperatures took much longer to grow, and that a high
percentage of the initial inoculum was killed.
exposure to a temperature of 700
sterilize the culture, and this implied that hot water could not
be relied on to sterilize bottles before refilling.
All three strains were found to survive in chlorhexidine 0.05%
freshly prepared in the laboratory, and Ps. 60 was also able to
survive in 0.1% chlorhexidine.
into Savlon 1 in 30 dilution prepared in the laboratory it failed
to survive, though it had been initially isolated from Savlon at
This is at present under further investiga-
tion.None of the strains was found to survive in 0.05%
chlorhexidine containing 4% v/v isopropyl alcohol when tested
in the laboratory, but recently a strain of Pseudomonas having
the biochemical characteristics of Ps. 60 has been isolated from
the residues in 3 out of 16 bottles which had been returned
to the pharmacy after use.
These bottles had contained 0.05%
chlorhexidine and 4%
finding indicates that continued use of isopropyl alcohol may
ultimately select for strains resistant to both isopropyl alcohol
and chlorhexidine, and thus that some other method such as
autoclaving the returned bottles should be used to break the
chain of infection, a conclusion reached by Lowbury (1951) in
relation to bottles of cetrimide.
reported contamination of returned ward stock bottles and
They too recommend disinfection of bottles
Laboratory experiments to determine whether growth of the
three strains of Pseudomonas would occur in Savlon and
chlorhexidine were performed by inoculating 10 ml. of Savlon
1 in 30 with Ps. 10, and 10 ml. of 0.05% chlorhexidine with
Ps. 60 and Ps. 65 separately.
stock concentrate with sterile distilled water in laboratory glass-
Inoculation was performed with a straight wire from a
single colony of the respective organism grown on nutrient
agar. Viable colony counts were performed at intervals during
a period of four weeks by spreading 0.1-ml. volumes of an
appropriate dilution on nutrient agar.
were able to demonstrate growth of Ps. 60 in 0.05% chlor-
hexidine, but not of Ps. 65 in 0.05% chlorhexidine or Ps. 10 in
The viable colony count of Ps. 60 in 0.05%
chlorhexidine showed a 100,000-fold increase during a period of
Two strains, Ps. 10 and Ps. 60,
C. for five minutes failed to
10 was inoculated
Kelsey and Maurer (1966)
a white disinfectant
The dilutions were made from
Using this method we
1 in 30.
General Characteristics of the Organisms
The three strains of Pseudomonas, Ps. 10, Ps. 60, and Ps. 65,
were all motile Gram-negative bacilli, oxidase- and catalase-
positive, and produced nonpigmented colonies on nutrient agar.
Glucose was utilized by oxidation.
was characteristically slow, and the initial isolations from dis-
infectant solution required five to six days' incubation at room
temperature before growth became apparent.
and biochemical properties are listed in Table I.
chemical and cultural tests we have performed Ps. 60 has the
characteristics of Ps. multivorans, as described by Stanier,
Palleroni, and Doudoroff (1966) except that, like three of their
19 strains, it failed to hydrolyse gelatin, and, like one of their
strains, it failed to give an egg-yolk reaction.
that their strain 398, identified as Ps. multivorans, also lacked
these two characteristics, and was one of the strains causing
urinary infection in children described by Mitchell and Hay-
Growth on artificial media
In the bio-
It is interesting
TABLE I.-Cultural and Biochemical Characteristics
Gram stain .
Growth on MacConkey agar
Growth on Cetavlon agar
Growth at 420 C.
Growth at 5'C.
Citrate utilization .. .
Gelatin hydrolysis .. .
Starch hydrolysis.. .
Nitrate reduction.. .
Arginine dihydrolase. ..
Egg-yolk reaction .. .+
The isolation of bacteria from chlorhexidine stock bottles has
been recently recognized, and in 1965 the manufacturers (I.C.I.)
reducing the incidence of contamination.
of contamination of chlorhexidine and of clinical infection
caused by these contaminants appeared in 1966.
Hayward (1966) described
urinary tract in children after cystoscopy.
infection was traced to the chlorhexidine solution used for
disinfecting the bladder-irrigation reservoir.
traced back further to the stock bottles of chlorhexidine 1/5,000
in use in the hospital, and they concluded that the chlorhexidine
solution had deteriorated on storage.
Alcaligenes faecalis from the urine of 30 gynaecological patients
undergoing bladder drainage by indwelling catheter.
organism was isolated from the jar used for storage of spigots
after these had been sterilized by boiling.
In other reports
been implicated. Lowbury (1951) concluded that contamination
in bottles was being maintained by showers of bacteria from the
organisms did not actually survive in the disinfectant they were
investigating (cetrimide) but that organisms were released from
the cork when it was removed and thus passed out of the bottle
with the disinfectant.
Linton and George (1966) described a
substance, probably a tannin present in cork (and tea), which
is a very potent inactivator of chlorhexidine.
that the neutralizing action of this substance was responsible
for the contamination of chlorhexidine solution at Bristol. Cork
closures have not been used in our pharmacy for several years,
and all bottles tested in the present series had bakelite closures
a circular on the subject with recommendations for
cases of infection of the
The source of the
The infection was
Dulake and Kidd (1966)
The jar contained
species cork closures of bottles have at times
with rubber seals.
contamination here described we have so far not been able to
Clinical infections with these organisms have been described
In our hospital we have looked for infection caused
by Pseudomonas sp. other than Ps. aeruginosa in all urine and
wound swabs received over a period of two months.
purpose material was inoculated on to cetrimide agar (Lowbury
and Collins, 1955) and after overnight incubation
keeping the plates at room temperature for a further four days.
On no occasion were any of these strains isolated from urine,
though Ps. 60 and Ps. 65 had both been isolated from jars of
chlorhexidine used for disinfecting cystoscopes.
swabs several other Pseudononas
were all of biochemical types that were distinct from our strains
Ps. 10, Ps. 60, and Ps. 65.
However, disinfectants may well
have been the source of these wound infections, as the disinfec-
tant solution may have become contaminated after issue to the
Some support for this view is supplied by the isolation
of Ps. putida and Aeromonas sp. in addition to Ps. 10 from a
mop and a thermometer standing in Savlon
Another article better sterilized by other methods but com-
monly sterilized by chemicals
where the catheter used had been sterilized in Savlon 1 in 30.
Ps. 10 was isolated from the Savlon solution employed for
Blood cultures were taken of venous blood from
the right atrium via the catheter and of arterial blood from the
brachial artery by means of a stainless-steel cannula that had
been sterilized by autoclaving.
and Ps. 10 was isolated from each of them.
isolated Ps. 10 from blood cultures from suspected cases of
bacterial erndocarditis, but a serious potential threat obviously
It is perhaps not generally appreciated how widespread con-
tamination of chlorhexidine and Savlon solutions can become,
so that ultimately all bottles of the aqueous solutions of these
compounds may be contaminated at issue.
extent of contamination Tables II and III record the isolations
issued by the pharmacy.
Until the introduction of 4% v/v
isopropyl alcohol all samples were contaminated.
very few isolations have so far occurred, but it seems possible
that continued use of 4% v/v isopropyl alcohol may ultimately
result in selection for strains resistant to chlorhexidine and 4%
v/v isopropyl alcohol.
Later samples of all batches tested have
been sterile ; thus the possibility that some organisms mayhave
taken a rather long time to be killed by the low concentration of
the alcohol cannot be excluded.
Recently we have had the opportunity of sampling stock
solutions of aqueous chlorhexidine and Savlon in four other
If a neutralizing substance is involved in the
sp. were isolated, but they
1 in 30 in the
is the cardiac catheter.
two patients undergoing cardiac catheterization
Six cultures were taken in all,
We have not
To illustrate the
sp. from chlorhexidine and Savlon bottles
Isolated from Different Batches of 0.05 %
v isopropyl alcob
Ps. 60 (2).
Ps. 65 (1)
Ps. 60 (1).
15 April 1967
Contamination of Disinfectants-Burdon and Whitby
15 April 1967
Contamination of Disinfectants-Burdon and Whitby
reported were found; 30 out of the 48 samples taken yielded
a growth of Pseudomonas sp.
In all these hospitals organisms similar to those here
In one pharmacy the distilled
TABLE III.-Organisms Isolated from Different Batches of 1 in 30
4% v/v isopropyl alcohol added.
water used for dilution was found to be heavily contaminated
with the same organism found in samples of disinfectant distri-
buted throughout the hospital.
Storage of distilled water presents a special problem, and we
recommend that freshly distilled water, hot from the still, be
But, this apart, it is important to decide what practicable
steps can be taken to prevent contamination assuming significant
If chlorhexidine solutions are sterile at the time of
bottling sterilizing the bottles-will at least ensure that supplies
leave the pharmacy uncontaminated.
isopropyl alcohol is added to aqueous preparations the chance
of subsequent contamination on the wards will be very greatly
Though these organisms are not highly pathogenic
If in addition 4% v/v
they are capable of causing human disease under certain circum-
stances, and unless steps are taken to rid stock solutions of this
contamination further case reports can be confidently predicted.
Contamination of stocks of chlorhexidine and Savlon with
Pseudomonas species resistant to the concentrations of chlor-
hexidine and Savlon in clinical use is reported. The contamina-
tion was maintained by a residual inoculum remaining in bottles
after washing, and was spread to other bottles in the bottle-
It is suggested that some wound infections
may be caused by disinfectants contaminated either at issue
by the pharmacy or during use on the wards.
of 4% v/v isopropyl alcohol to aqueous preparations of both
disinfectants was found greatly to reduce the incidence of
We gratefully acknowledge the help of the group pharmacist, Mr.
A. E. Marston, and other members of his staff throughout this
Anderson, K., and Keynes, R. (1958).
Dulake, C., and Kidd, E. (1966). Lancet, 1, 980.
Kelsey, J. C., and Maurer, I. M. (1966). Mth. Bull. Minist. Huth Lab.
Serv., 25, 180.
Linton, K. B.. and George, E. (1966). Lancet, 1, 1353.
t.owbury, E. J. L. (1951).
--and Collins, A. (1955).
Mitchell, R. G., and Hayward, A. C. (1966). Lancet, 1, 793.
Stanier, R. Y., Palleroni, N.
Microbiol., 43, 159.
Brit. med. Y., 2, 274.
Brit. 7. industr. Med., 8, 22.
7. din, Path., 8, 47.
J., and Doudoroff, M. (1966).
Aorto-jejunal Fistula from Rupture of
Teflon Graft, with Septic Emboli in the
Brit. med.J3., 1967, 2, 155-156
We report the case of a patient who died from a massive
haemorrhage into his gut from an aorto-jejunal fistula four
and a half years after insertion of a plastic prosthesis of the
that ruptured and not the suture line. A week before death
showers of infected emboli appeared in the skin of the lower
part of the body which were thought to be diagnostic of
infective endarteritis of the graft.
It was the middle of the graft material itself
The patient, a tractor driver, was aged 38 when he presented
with a four-year history of increasingly severe pain of the left side
of his back and left thigh.
There was no relevant past or family
history of metabolic, vascular, or Marfan's disease.
radiological evidence of a large abdominal aortic aneurysm that did
not involve the renal arteries was found.
Teflon aorto-ileal bifurcation graft was inserted as an emergency
operation by Mr. D. N. Ross at Guy's Hospital.
started on anticoagulant therapy, which was continued for four and a
half years. He was then admitted to Orpington Hospital on account
of very severe haemorrhage from a radiologically demonstrable non-
malignant gastric ulcer, and anticoagulants were stopped.
weeks later, only two days after discharge from hospital, the symp-
toms of his terminal illness developed: malaise, rigors, and aching
This was resected and a
The patient was
in the body and head.
oxytetracycline, his condition worsened and he was readmitted to
hospital, and on this day septic lesions appeared on the right heel,
great toe, and sole of the foot, and he had signs of left lower-lobe
Septicaemia secondary to his pneumonia was diagnosed
(penicillin-resistant Staphylococcus aureus was cultured from the
blood), and he was immediately started on treatment with cloxacillin,
fusidic acid, and cephaloridine.
" purpura " was noticed on his right thigh, and soon afterwards
he developed bilateral femoral arterial occlusion.
The following week, in spite of systemic
On the ninth day of his illness
He was treated
6St>3B~ ~ ~ ~~~~. .'...........
Septic embolic lesions on lower half of the body.
splinter haemorrhages and finger-clubbing.
Note the absence of