BRITISH MEDICAL JOURNAL
6 DECEMBER 1975
obtained without using such treatment.6 Moreover, it has not
always been possible to show any effect of treatment from
kinetic studies of platelet and fibrinogin turnover.7 Recently
the importance of fibrin deposition in the kidney in the patho-
genesis of the disease has been questioned.8 9 The undoubted
risk of fatal haemorrhage in association with anticoagulation
or thrombolytic treatment makes the need for adequate con-
trolled trials imperative, but the complexity of the trial design
and the problem of unnecessary treatment of mildly affected
patients have so far proved insurmountable.3 A further consid-
eration pointing to the need for proper trials is the apparent
low incidence of residual hypertension in patients treated with
streptokiilase.' Local infusion, as used in our cases and in those
of others,4 has an advantage of providing the drug in high
concentration in the kidney without affecting haemostasis in
the rest of the body, and, moreover, if only one kidney is
perfused the other can serve as a control. A similar approach
in cortical necrosis
other than the haemolytic uraemic syndrome is suggested by
the second case. The use of the gamma-camera for frequent
assessment of renal perfusion and function in impending cortical
necrosis would enable only the more seriously affected cases
to be treated in this way, whereas those with presumed tubular
necrosis alone might be treated conservatively. We cannot
yet conclude that the fall in renal perfusion as estimated by
dynamic renal scintillography will necessarily lead to cortical
necrosis but it is possible to identify these patients with tubular
necrosis who have normal or only slightly diminished renal
Close collaboration between various disciplines within the
hospital service is needed, for without this essential teamwork
the treatment of acute renal failure is unlikely to improve.
ADDENDUM-Since this report was prepared we have treated two
further children, one aged 14 months and the other 5 months.
Both were referred in established renal failure from haemolytic
uraemic syndrome and in both cases renal perfusion as judged
by the appearance on renal scintillography was extremely poor.
It was not known at what time renal perfusion had begun to
deteriorate. One child remained alive and well but with chronic
renal insufficiency, and there was no difference in individual
renal function. The other child died of chronic renal failure
after some recovery in renal function, but at necropsy the
perfused kidney was smaller than the non-perfused kidney.
1Lieberman, E J, Pediatrics, 1972, 80, 1.
2Monnens, L, Acta Helvetica Paediatrica, 1972, 27, 45.
3Stuart, J, et al, BritishMedicallJournal,1974, 3, 217.
4Rosen, S M, et al, British MedicalJournal, 1970, 3, 465.
5McCintyre, W J, et al, in International Symposium: Radionuclides in
Nephrology, Berlin, 1974. ed K Zum Winkle, in press.
6 Tune, B M, Leavitt, T J, and Gribble, T J, Pediatrics, 1973, 82, 304.
7Berberich, F R, et al,J'ournal ofPediatrics, 1974, 84, 503.
8 Katz, J, et al,Journal ofPediatrics, 1973, 83, 739.
9Riella, M C, et al, Kidney International, 1974, 6, 89A.
Plastic isolators for treatment of acute leukaemia patients
under "germ-free" conditions
P C TREXLER,
A S D SPIERS,
British Medical journal, 1975, 4, 549-552
A gnotobiotic isolation system based on those developed
in veterinary research has been constructed for hospital
use. Fifteen patients with leukaemia and neutropenia
spent a total of 110 weeks in plastic isolators, and none
acquired any infection. Endogenous flora was effectively
suppressed by topical antiseptics and gastrointestinal
decontamination effected with nonabsorbable antibiotics.
The isolator system was acceptable to patients and staff
and much cheaper than the use of sterile rooms. Other
advantages of the system are portability, easy storage,
and use on ordinary open wards without prejudice to the
microbiological protection afforded. It is as yet uncertain
whether protective environments of this type will sub-
stantially improve the outcome oftreatment for the acute
Royal Veterinary College, London NW1 OTU
P C TREXLER, MS, reader in veterinary gnotobiotics
MRC Leukaemia Unit, Hammersmith Hospital, London W12 OHS
A S D SPIERS, MD, FRACP, consultant physician
Department of Bacteriology, St Mary's Hospital Medical School,
London W2 IPG
H GAYA, MRCPATH, reader
For the past 60 years microbial isolators have been used to
maintain a sterile environment in which germ-free animals
may be produced for laboratory investigations.' These animals
are free from bacteria, moulds, and other contaminants in the
environment, but some carry vertically transmitted viruses, and
the term "gnotobiotic," signifying that their flora is known, is
more accurate than the term "germ-free." Colonies of gnoto-
biotic rats and mice have been maintained continuously since
1954 and a great many species, including calves and foals,
have been reared in sterile isolators." Experience gained from
veterinary gnotobiotic research enabled the development of
closed isolator systems using barriers made of flexible plastic
film in which many tasks may be performed without violating
the microbiological integrity of the system.
Closed-system isolators that are used for the care of patients
who are at special risk of infection4-8 impose restrictions on
medical and nursing care because of the interposition of a
mechanical barrier between patient and attendants.9 Hence
isolation systems in which a linear flow of sterile air is combined
with partial mechanical barriers and reverse barrier nursing
techniques have been preferred, even though such systems are
microbiologically less secure than closed-system isolators. 9-12
We have designed a patient isolator in which the limitations
imposed by the flexible film barrier have been reduced and a
linear flow of sterile air has been used solely to protect the port
through which sterile supplies are introduced.1 The security
of this type of isolator has been shown in the rearing of gnoto-
biotic animals.2 3 14 15 The isolators can be placed over standard
hospital beds in an open ward; no special facilities are necessary
for their use and they can be taken down and stored.
The isolator is a rectangular envelope of transparent polyvinyl-
chloride film (0-2-0-3 mm) 1 9 m long, 1 8 m wide, and 2 0 m high
supported by a light metal frame attached to the bed. There are two
half suits attached to the walls on both sides of the bed. These consist
of a clear plastic face piece supported by an adjustable head harness
and attached to a conical invagination of the plastic wall which bears
the sleeves and gloves. An attendant can enter this half suit and move
freely within it, wearing a ventilating vest attached to an air line that
provides air for breathing and cooling. The attendant's feet remain
on the floor outside the isolator (see fig), but with four suits available
any point within the isolator is readily accessible. Supplementary
sleeves, terminating in rubber gloves, are attached to the isolator walls
and many procedures-for example, adjusting the rate of an intra-
venous infusion-can be carried out without entering the half suits.
The oval entry port (76 x 62 cm) at the foot of the bed is normally held
closed by a mating port on a sterile supply isolator (see fig). For entry
of the patient the opening between the supply and bed isolators is
closed with plastic film, the supply isolator removed, the patient helped
in, and the supply unit replaced.
Bed isolator showing patient in bed and two half suits in use. A nurse is
working in supply isolator on left. A=Control mechanism for regulating
pressure in isolator. B=Patient entry port. C=Air filter for pressurising
supply isolator. D =Downflow supply port. E =Air supply to half suit.
F- Unoccupied half suit.
The supply isolator is 1-3 m long, 0 75 m wide, and 1 0 m high
and provides space for storing sterile supplies,
and a toaster, and for preparation. Supplies are introduced upwards
through a horizontal oval port (32 x 28 cm) in the floor of the supply
isolator against a stream of sterile air flowing downwards. Smoke
tests have confirmed the absence of backflow into the supply isolator.
When supplies are introduced a plastic curtain is drawn over the
opening into the bed isolator to prevent overinflation, and a supplemen-
tary air supply is turned on to provide a sterile air flow of about 100
cubic feet per minute through the supply port. The outer wrap of the
double-wrapped sterile supplies is removed in the air stream, the
package is introduced, and the inner wrap is removed and discarded
through the downflow port. When not in use the downflow port is
covered with a plastic bag which also serves a a waste bag.
Small items unsuitable for autoclaving, such as ampoules of drugs
and plastic bags containing platelet concentrate, can be rapidly intro-
duced through a flexible tube sealed into the wall at the foot of the
bed isolator and immersed in
At the foot of the bed on the patient's right a small oval port
(32X 28 cm) is for the removal of soiled materials such as linen
a bath of detergent-hypochlorite.
BRITISH MEDICAL JOURNAL
6 DECEMBER 1975
ethylene bag introduced from inside the bed isolator. The bag can
be removed and bulky items passed directly to the outside without
danger of deflating the isolator since this port will remain open and
unobstructed for at least 15 seconds before the pressure within the
isolator falls to that of the room outside.
On the left-hand side of the bed the isolator rests on the floor,
forming a space 0-6 m wide along the length of the bed, which permits
the patient to stand upright and walk and allows room for a chair to be
placed underneath the isolator at the foot of the walkway. A bedside
table is inserted into a pocket sealed to the wall of the isolator. Pockets
in the isolator wall over the head of the bed accommodate shelves
that can be reached by the patient and attendants. The mattress is
fitted into a pocket in the bottom of the isolator and remains outside
the protected space, although the bedclothes can be tucked under the
mattress as usual. Since the bed, mattress, and furniture are outside
the isolator, sterilisation of the interior ofthe isolator is readily accomp-
lished with 750 ml of a 2^, aqueous solution of peracetic acid, dis-
persed from an atomizer.'4
is covered with
Filters for sterilising entering and exhaust air are placed at the
top of the isolator envelope. Those originally used were 1-8 m long
and 7-5 cm in diameter and incorporated three layers of glass-wool
filter mat FM004, as in animal isolator units.'4 More recently compact
HEPA filters have been used. The blower unit which pressurises the
filters is in a sound-absorbing box below the bed and has an alarm
system in case of failure (which is extremely rare). The flow of air
entering the supply filter is controlled by a ball-valve at the foot of the
bed and the flow is indicated on a gauge. The extract filter is connected
to the blower through a ball-check valve which is adjusted by a
mechanical linkage to the roof of the isolator (see fig) and controls
the interior pressure. Slight deflation of the envelope closes the valve
and increases pressure within the isolator: conversely, when pressure
increases 7nd the isolator roof rises the valve opens and pressure is
An isolated patient may be moved to other parts of the hospital
without breaking microbiological isolation by using a transit isolator.
This is a tube of thin plastic film, 85 cm in diameter and 3-5 m long,
with the ends closed by diaphragms of plastic film the same size as the
bed isolator entry port. The transit isolator is placed on a stretcher
trolley, the diaphragm and the film closing the entry port are wiped
with tincture of iodine and attached together, and an opening is cut
for the patient. The patient entry port is closed with fresh film, the
end of the transit isolator is tied off, and the two isolators are separated.
The patient can re-enter the bed isolator through the diaphragm on
the other end of the transit isolator. There are four pairs of sleeves
and gloves on the transit isolator, two on each side. Intravenous
apparatus is accommodated inside the isolator. Air is sterilised by a
small pump and blower attached to the isolator and exhaust air leaves
through a flexible tube with an adjustable clamp which controls the
interior pressure. As with the bed isolator it is possible to converse
with the patient in normal tones while he is in the transit isolator. The
transit isolator is sterilised with peracetic acid or by gamma-radiation
TREATMENT IN ISOLATOR
The impervious plastic film between the patient and his attendants
hampers his treatment surprisingly little. Physical examination and the
insertion of intravenous lines are carried out through gloves; instru-
ments may be sterilised and passed into the isolator-for example,
tongue depressors-or used from outside via invaginations of the
plastic-for example, ophthalmoscopes,
machine can be putin a half suit and theplates placedbeneath the
plastic film, the entire unitremainingoutside the isolator.Sphygmo-
manometer cuffs and electrocardiograph leads areplacedwithin the
isolator: their connections pass out throughtheplasticwall and the
instruments remain outside. The patient cantelephone,the receiver
beinginserted into a sleeve of thinplasticthat transmits soundreadily,
so that sterilisation of the telephone is unnecessary. Sterilebagsor
bottles of parenteral solutions are aseptically introduced into the
BRITISH MEDICAL JOURNAL
6 DECEMBER 1975
Details ofpatients with leukaemia treated under gnotobiotic conditions in plastic isolators
Age and sex
CGL in acute transformation
AML, failed induction therapy
AMML in late relapse, failed reinduction
AML in late relapse, failed reinduction
AEL failed induction therapy
AMML, failed induction therapy
AMML, failed induction therapy
CGL in chronic phase, intensive chemotherapy
AMML secondary to chemotherapy for Hodgkin's disease
CGL in chronic phase, intensive chemotherapy
CGL in acute transformation
AML in late relapse, failed reinduction
Further failed reinduction
CGL=Chronic granulocytic leukaemia. AML=Acute myeloblastic leukaemia. AMML=Acute myelomonocytic leukaemia. AEL= Acute erythroleukaemia. ALL= Acute
*The term "remission" is applicable only to patients with acute leukaemia.
isolator and attached to hooks in sleeves welded to the roof. The height
of these bottles can be adjusted to 1 m above or below the roof level
or to any intermediate position.
Preparation of patients
The patient is first decontaminated by reverse barrier nursing
combined with sterile food and the use of chlorhexidine preparations
-surgical scrub for the skin and hair, mouthwash, and atomized
solutions for the ears, nose, and throat. Chlorhexidine douches and
obstetric cream are applied to the vulva and oral framycetin, colistin,
and nystatin are given to decontaminate gut contents. These measures
are continued after entering the isolator and their efficacy is monitored
by examining bacteriological specimcns from many anatomical sites;
swabs of the isolator itself and slit samples of the air inside are also
examined for contamination. Chloramphenicol and saline enemas are
given 24 and 4 hours before entry to the isolator. Before entry two
gloved and gowned nurses place the patient on a sterile sheet on a
trolley and spray his entire body, excluding the face, with chlor-
hexidine aerosol. Another sterile sheet is folded over the patient, who
is wheeled to the isolator and helped through the entry port, leaving
the sheets behind. Sterile night attire is in the isolator. Once inside,
although the patient is in complete physical isolation from the con-
taminated environment, he may be freely visited by attendants and
relatives without need of barrier nursing precautions. There is less
psychological isolation within the isolator than under strict reverse
barrier nursing conditions, while microbiological isolation is complete.
Almost any object can be sterilised by heat, gamma-radiation, or
chemicals, and items ranging from soft toys to slide rules have been
supplied for the use of patients.
In the past four years we have treated 15 patients in plastic isolators
for a total of 110 weeks (see table): all had leukaemia, neutropenia,
and an enhanced risk of infection.
Patient acceptance of isolator-We could not predict whether certain
patients might be psychologically unsuitable for isolation,
decided to isolate a patient on the basis of his susceptibility to infec-
tion, the status of his leukaemia, and the availability of an isolator.
Patients who already had serious infections were not isolated. No
patient refused to undergo isolation but most had misgivings. Never-
theless, all 15 patients tolerated isolation well. Most found it much
pleasanter than the period of strict reverse barrier nursing that pre-
ceded it, and said that they felt less, rather than more, cut off from
human contact. Most expressed their willingness to undergo another
term of isolation if necessary and one patient was isolated three times
over 13 months. Patients who became seriously ill during isolation
also became depressed, but their distress appeared no different to that
of similarly ill patients nursed under conventional conditions.
Reaction ofrelatives-The patients' relatives thought that the isolator
represented further protection that might improve the chances for
successful treatment. Some relatives of dying patients requested
that they be removed from isolation, whereas others wanted isolation
to be continued to show the patient that hope had not been abandoned.
Our policy was to accede to the wishes of relatives.
Acceptance by nursing staff-The isolation equipment was initially
viewed with misgivings by our excellent nurses. Most became pro-
ficient in the operation of the isolator in a week, and many became
most enthusiastic about its use. The nurses agreed that generally
the care of an isolator patient was less onerous than the care of a
similar patient under conditions of strict reverse barrier nursing. The
avoidance of repeated masking, scrubbing, and gowning more than
compensated for the inconvenience of working in a half suit. Since
nurses carry the main burden of caring for a patient in an isolator
their favourable assessment is important.
Attitude of medical staff-Though procedures such as venepuncture
were marginally more difficult in an isolator, the ease with which the
patient could be seen several times a day more than compensated for
logical examinations were made of the urine, stool, and saliva and of
swabs from the ears, nose, throat, axilla, umbilicus, groin, and vagina.
Faeces were cultured quantitatively by aerobic and anaeorbic tech-
niques and the
chromatographic analysis for fatty acids.'6 During 110 weeks of
isolation in 15 patients there was no evidence of acquisition of any
exogenous micro-organisms, and the
remained satisfactorily suppressed and free from potential pathogens
such as Pseudomonas aeruginosa, the enterobacteriaceae, and Staphylo-
bacteriologically and in their faeces this finding was confirmed by the
gas-chromatographic analyses. Episodes of fever with negative blood
cultures were observed and often resolved without explanation of their
origin. One case of Ps aeruginosa septicaemia was documented; the
apparent source ofthe organism was an old osteomyelitic lesion present
Results of leukaemia treatment-Initially we selected patients who
had a poor prognosis or were likely to require particularly intensive
example, those with chronic granulocytic leukaemia in acute trans-
formation or acute myeloid leukaemia in a late relapse. Later some
new cases of acute leukaemia and two patients with chronic granulo-
cytic leukaemia in its chronic phase who were undergoing intensive
chemotherapy were put in the isolator.'8 These patients were main-
tained in satisfactory condition, despite extreme neutropenia and
immunodepression, for many weeks, and cytotoxic treatment was
continued without the intervention of fatal sepsis. Obviously the
favourable outcome in some patients could not have occurred had
they succombed to infection. On the other hand, in some patients
remission of the leukaemia was not secured, even though freedom from
infection enabled repeated and aggressive cytotoxic therapy. Thus
effective protection from infection is but one facet in the control of
presence of bacteria
also sought by
aureus. In some patients no residual flora was detectable
The simultaneous application of decontamination of the
tract with antibiotics to reduce endogenous
BRITISH MEDICAL JOURNAL
6 DECEMBER 1975
floral9 and isolation to prevent the acquisition of exogenous
organisms20 should substantially reduce the risk of infection
in neutropenic patients with leukaemia who are receiving
cytotoxic drugs. Uncontrolled2' and controlled22 studies of
protected environments and prophylactic antibiotics in patients
with acute leukaemia have shown fewer infective episodes and
fewer deaths from infection. Although it was considered that
patients so protected could tolerate unusually intensive chemo-
therapy2l an increased remission rate was not observed, despite
the reduction in deaths from infection.22 In the controlled study
the antileukaemic chemotherapy administered to the isolated
patients was identical to that received by the control patients-
that is, no advantage was taken of their protected situation by
using exceptionally aggressive therapy. Had this been done a
higher remission rate might have been observed, although our
own experience suggests that this is by no means certain.
In our small and deliberately heterogeneous group of patients,
most of whom were selected because of their poor prognosis,
the results of the antileukaemic treatment cannot be evaluated.
The decontamination and isolation procedures, however, were
highly effective, suppressing endogenous pathogens and com-
pletely preventing the acquisition of contaminants from the
environment. Although we have not shown that such protection
can substantially improve the outcome of antileukaemic chemo-
therapy it is extremely unlikely to make it worse. Extensive
studies of intensive chemotherapy in protected environments
are required to establish whether improved remission rates can
The isolator system proved acceptable to patients, nurses,
and medical staff. It is much cheaper than systems that depend
on the construction of special rooms, and it can be rapidly
dismantled and stored and is easily portable. Plastic isolators
may be used in open wards without prejudicing their bacterio-
logical security. Their use on open wards is desirable, since most
patients become depressed when nursed for prolonged periods
in single rooms, with or without a plastic isolator. Provision of
sterile nursing suites based on the use of special rooms is not
feasible below a certain size-four such rooms is probably the
minimum-but it is easy to operate a single plastic isolator,
securing better microbiological protection at much smaller
financial outlay. And plastic isolators can be made available in
small hospitals as well as in major centres.
1Gordon, H A, and Pesti, L, Bacteriological Reviews, 1971, 35, 390.
2Trexler, P C, Veterinary Record, 1971, 88, 15.
3Drummond, A J, et al, Veterinary Record, 1973, 92, 555.
4Levenson, S M, et al, American Journal of Surgery, 1964, 107, 710.
5Haynes, B W, and Hench, M E, Annals of Surgery, 1965, 162, 641.
6 Robertson, A C, et al, Lancet, 1968, 2, 1376.
7 Dietrich, M, et al, Revue Europeenne d'Etudes Cliniques et Biologiques,
1972, 17, 488.
8 Wilson, R, et al, Germfree Research, p 37. New York, Academic Press,
9 Shadomy, S, et al, Archives of Environmental Health, 1965, 11, 183.
10Penland, W Z, and Perry, S. Lancet, 1970, 1, 174.
11Hummel, R P, et al, Annals of Surgery, 1972, 176, 742.
12 van de Waaij, D, et al, Germfree Research, p 31. New York, Academic
13Spiers, A S D, and Trexler, P C,Journal of Physiology, 1973, 231, 66P.
14Trexler, P C, and Reynolds, L I, Applied Microbiology, 1957, 5, 406.
15Sacquet, E, The Germ-Free Animal in Research, p. 1. New York, Academic
16 Gompertz, D, et al, Gut, 1973, 14, 183.
17 Spiers, A S D, British Medical3Journal, 1973, 3, 528.
"ISpiers, A S D, et al, Lancet, 1975, 1, 829.
19Preisler, H D, Goldstein, I M, and Henderson, E S, Cancer, 1970, 26, 1076.
20Jameson, B, et al, Lancet, 1971, 1, 1034.
21Bodey, G P, et al, Cancer, 1968, 22, 1018.
22Levine, A S, et al, New EnglandJournal of Medicine, 1973, 288, 477.
Value ofDoppler ultrasound in diagnosis of clinically
suspected deep vein thrombosis
A N NICOLAIDES,
C J WALKER,
J D O'CONNELL
BritishMedical_Journal,1975, 4, 552-554
Doppler ultrasound was used to study 120 legs of 106
patients with suspected deep vein thrombosis (DVT)
or pulmonary embolism. Venography was subsequently
performed in all. DVT was confirmed by venography in
44 legs and was confined to the calf in 10 of these. Ultra-
sound detected three calfthromboses and 29outof34more
extensive thromboses. Of five undetected thrombi that
were proximal to the calfone was associated with partial
Waller Cardiopulmonary Unit, St Mary's Hospital, London W2
JEANETTE MEADWAY, MB, MRCP, senior research registrar
Blood Flow Laboratory and Surgical Unit, St Mary's Hospital
Medical School, London W2
A N NICOLAIDES, MB, FRCS, lecturer in surgery
C J WALKER,MB, FRCS, senior registrar (locum) inorthopaedic surgery (now
senior orthopaedic registrar, St Mary's Hospital, London W9)
Department of Radiology, St Mary's Hospital, London W2
J D O'CONNELL, MB, DMRD, FRCR, senior registrar (now consultant radiolo-
gist, St Finbarr's Hospital, Cork, Eire)
occlusion and four with extensive collateral circulation.
Of the 76 limbs without venographic evidence of throm-
bosis 21 were thought to have DVT by ultrasound; 18 of
these false-positive results could be attributed to external
compression of veins, two to excessive tenderness pre-
cluding adequate examination; and in one no explanation
was found. This test gives more accurate results than
judging by clinical signs alone, but users must be aware
of its limitations and, particularly, the causes of false-
positive and false-negative results.
The clinical diagnosis of deep vein thrombosis (DVT)
inaccurate. Half the thrombi in the calf do not produce signs,'-4
and 40% of patients with signs have normal veins on venog-
raphy.5 6 Most clinical pulmonary emboli arise from veins
proximal to the knee,7-9 and only small silent emboli arise from
the calf.'0Early detection of thrombi in deep veins, particularly
those extending into the popliteal and more proximal veins, is
therefore essential for.the prevention of pulmonary embolism.
Some workers have claimed that Doppler ultrasound techniques
may detect thrombi early," 12 both in routine screening'314
and in patients with clinical signs suggestive of venous throm-
bosis,15 but others have questioned its value.'6