J. Hyg., Carnb.

(1977), 79, 121 121

Printed in Great Britain

Bacterial contamination in a modern

operating suite. 1. Effect of ventilation on airborne bacteria
and transfer of airborne particles*
BY ANNA HAMBRAEUS, STELLAN BENGTSSON AND
GUNNAR LAURELL
Institute of Clinical Bacteriology, University of Uppsala, Upp8ala, Sweden
(Received 4 January 1977)
SUTLMARY
The effect of ventilation on airborne contamination was studied in a new oper-
ating suite containing operating rooms with conventional ventilation (17-20 turn-
overs/h) and operating rooms with zonal ventilation, where the turnover in the
central part of the room was 80/h. The efficacy of the ventilation was first
examined with gas tracer experiments and found satisfactory. Experiments using
potassium iodide particles showed the transfer between adjacent rooms in the
suite to be less than 10-3 % with closed doors and from 1 % to 2*5 x 10-2 % when
the doors were opened once a minute. The transfer between two adjacent operating
rooms was calculated to be ' 10-4 %. There is thus little risk of spread of airborne
infection between operating rooms.
Experiments with potassium iodide particles showed that in operating rooms
with zonal ventilation the particle concentration in the centre of the room was
about one-tenth that in the periphery; in conventionally ventilated operating
rooms the concentration was about one-half. With bacteria-carrying particles
generated by human activity the concentration in the centre of operating rooms
with zonal ventilation was about half that in the periphery both during experi-
mental activity and operations; in conventionally ventilated operating rooms it
was about equal in both cases. Bacterial counts at the periphery were found to be
lower in rooms with zonal ventilation (- 50 c.f.u./m3) than in conventionally
ventilated ( 70 c.f.u./m3).
INTRODUCTION
There is general agreement that air supplied to operating rooms should be free
from dust and bacteria. For many years operating rooms have therefore been pro-
vided with some form of ventilation equipment. The purpose has been to provide
comfortable working conditions and to maintain dilution of airborne contaminants
derived from human sources inside these areas. The introduction of positive
pressure ventilation for operating rooms led to a substantial reduction of the
number of airborne micro-organisms and some workers claim that this was accom-
* This study was supported by grants from the Swedish Medical Research Council (Project
No. K75-16X-3808-05) and from the Swedish Planning and Rationalization Institute of the
Health and Social Services (Project No. 7116).
122 A. HAMBRAEJuS, S. BENGTSSON AND G. LAURELL
panied by a reduction in the frequency of wound sepsis (Williams, Blowers, Garrod
& Shooter, 1966). Methods for a further reduction of the air contamination by a
factor of ten or more (ultra clean air) have been introduced during the last years.
The claim is made that the use of these would lead to a further reduction of wound
sepsis at least in some kinds of surgery (Charnley, 1964; Charnley & Eftekhar,
1969; Charnley, 1972; Eftekhar, 1973).
In many of these studies, however, the comparisons made have been between
the rates of sepsis before the introduction of the ultra clean air systems and those
experienced subsequent to the change. Other changes were also made during the
period of study and there are those who believe that a conventional plenum
system would have given comparable results (Leading article, Brit. med. J., 1975).
In order to evaluate the gain derived from the introduction of clean air systems it
is, in our opinion, necessary to have a direct simultaneous comparison of the new
and old systems.
The new operating department at the University Hospital in Uppsala is equipped
with positive pressure ventilation systems of two different kinds. In order to
investigate the effectiveness of the ventilation in this new unit we have studied
the airborne transfer of particles as well as the amount of airborne contamination
during activity. When studying transfer of airborne particles we have used a
method described by Foord & Lidwell (1972). We have also tried to discover if the
layout of this operating unit has contributed to maintaining good working routines
within the unit. The results of the clinical study of sepsis rates in operations per-
formed in the suite will be presented in a later paper.

MATERIALS AND METHODS

Operating unit
The operating suite consists of twelve operating rooms. The layout of the unit,
i.e. the operating rooms and ancillary rooms, can be seen from Fig. 1. In principle
it is a double-corridor system. One corridor is used by clean staff and for clean or
sterile supplies only. In this the scrub sink for the operating team is situated. The
other corridor is used for transport of patients and other material and for the
removal of used and soiled articles. Each operating room has three doors. One
leads to the corridor used by the staff, one leads to the anaesthetic room and one
to the exit area.
Ventilation
The unit is equipped with a positive pressure ventilation system and the ven-
tilation is normally balanced between all rooms in the operating suite. The extrac-
tion of air from the operating room can be increased causing an inflow of air from
the corridors. This latter type of ventilation is used during operations classified
as septic, contaminated, or infected according to local rules.
The ventilation rate in the operating rooms is about 17-20 air changes per hour.
In some rooms there is a special zonal ventilation (Fig. 2) where a clean air inlet
is installed in the ceiling over the operating table. This inlet consists of a per-
forated area 2-4 x 3-4 m enclosed by two parallel 2 mm wide slits. By this system
 Ventilation and airborne transfer 123

Fig. 1. Layout of the operating suite. OP, Operating room. The broken line indi-
cates the position of the zonal ventilation; A, anaesthetic room; E, exit area.

Perforated ceiling
Slits r ---
Filtered clean air

Fig. 2. Design of operating room with zonal ventilation.

Broken arrows symbolize evacuation grilles.

the operation area is provided with a ventilation corresponding to 80 air changes

per hour. According to the designers (Abel & Allander, 1966) the clean air is
almost completely prevented from being mixed with the peripheral room air by
the air jets derived from the parallel slits.
Gas tracer experiments
Before the particle tracer experiments were performed the ventilation plant
was examined by using nitrous oxide as a tracer gas (Lidwell, 1960). The gas was
liberated in measured amounts and samples were taken continuously in the oper-
ating rooms. The gas concentration present in an air sample was determined by
infra-red absorption. All experiments were made with the doors closed. This part
of the investigation was done in collaboration with Dr Olander and Dr Faxvall
at the Institute of Technology in Stockholm.
Tracer particle experiments
A spinning disk particle generator was used to produce an aerosol of particles
(Foord & Lidwell, 1975). With the help of an external fan the particles were mixed
124 A. HAMBRAEUS, S. BENGTSSON AND G. LAURELL
with the air in the room. Potassium iodide dissolved in ethyl alcohol was used at
a concentration giving an aerosol with particles with a sedimentation rate close
to 0 3 m/min. In all experiments the generator was run for 30 min and, if possible,
adjusted to reach a steady state of about 500 particles/l at the source. Air sampling
and determination of particle size was carried out as described earlier (Hambraeus
& Sanderson, 1972).

Bacteriological experiments
Airborne bacteria-carrying particles were sampled using a Casella slit sampler
sampling 700 I/min. Blood agar plates were used. The plates were incubated at
37 0C for 48 h before being examined. The total count of c.f.u. of bacteria was
estimated and presumptive Staph. aureus c.f.u. were tested for deoxyribonuclease
activity.
RESULTS
Gas tracer experiments
In more than three-quarters of the experiments with balanced ventilation there
was no detectable transfer of air from adjacent rooms into the operating theatre
and only once did the concentration in this room exceed 3 % of that in the
(adjacent) room in which the gas was being liberated.
When there was a forced exhaust of the air in the operating theatre transfer of
air into the operating room was detected in half these experiments and the con-
centration in the operating room was between 2 and 20% of that in the room
where the gas was liberated.
When tracer gas was liberated in the operating room transfer from the operating
room to adjacent rooms was detected in half the experiments with the ventilation
in the balanced conditions and in as many as one-third of the experiments when
there was increased extraction of air from the operating room. On no occasion
did the concentration found in any room exceed 10 % of that in the operating room.

Particle transfer between the operating room and adjacent areas

(with balanced ventilation)
In this series of experiments the following sequences of experiments were used.
(1) Generation of particles in the operating room measuring the transfer to:
(a) the anaesthetic room; (b) the exit area and adjacent parts of the patient
corridor; (c) the staff corridor.
(2) Generation of particles in the anaesthetic room and staff corridor, measuring
the transfer to the operating room.
(3) Generation of particles in the anaesthetic room, measuring the transfer to
the corridor outside.
In all experiments the doors to the operating room were closed during the first
10 min of particle generation and no activity was carried out. During the following
20 min one person walked from the site of the source to the receiving room and
back again every other minute, i.e. the activity corresponded to 60 door openings
 Ventilation and airborne transfer 125
Table 1. Particle transfer within the operating suite*
Ratio sourcelreceiving room
No activity Activity
(median (log mean
Source/receiving room value) value)
Opera.ting room 1 1 x 105 9-6 x 10
Anaesthetic room
Operating room > 2 x 105 4*8x 102
Exit area
Operating room Just outside exit area > 2x 105 2-3 x 104
Patient corridor : 4 m along the corridor > 2 x 106 5.5 x 104
Operating room Just outside the door > 2 x 106 7.9 x 102
Staff corridor t 4 m along the corridor > 2 x 106 2-4 x 103
Anaesthetic room > 2x 105 2-1 x 103
Operating room
Staff corridor > 2 x 105 1*3 x 103
Operating room
Anaesthetic room Just outside the door > 2 x 105 1-5 x 103
Patient corridor :t 4 m along the corridor > 2 x 105 4 0 x 103
* On an average 11 experiments were made of each kind.

per hour. Air samples were taken at the source for 1 min periods at 2 min intervals.
In the receiving rooms samples were taken continuously for 10 min during the
period of no activity and for 4 min every 5 min during the period of activity. The
transfer of airborne particles from one place to another will be presented as the
ratio of the particle concentration at the source site to that at the receiving site.
The results of the experiments are shown in Table 1. When there is no activity
there is a detectable transfer of particles only from the operating room to the
anaesthetic room. During activity the lowest ratio 9-6 x 10 is found for transfer
from the operating theatre to the anaesthetic room. Transfers to the exit area and
the staff corridor were of the same order with ratios of 4-8 x 102 and 7*9 x 102
respectively. The transfer into the operating room is much the same for the two
doors tested, ratios being 2-1 x 103 and 1-3 x 103, mean value 1-7 x 103. As the third
door was of the same size, particle transfer through this door was not measured.
In some experiments the concentration of particles was measured simultaneously
just outside the source room and about 4 m along the corridor. As can be seen the
concentration fell to between i and J between these two points.
The risk of transfer from one operating room to the adjacent one of a pair, under
the experimental conditions, can be calculated as follows. For transfer via the
exit area:
S/R = S/E x E/R = 4-8 x 102 X 1*7 x 103 = 8-2 x 105,
where S represents the particle concentration in the operating room, acting as the
source room, R that in the receiving operating room and E that in the exit area.
For this calculation the mean of the two experimental values for transfer into an
126 A. HAMBRAEUS, S. BENGTSSON AND G. LAURELL
Table 2. Transfer of tracer particles within an operating room; see Fig. 3
(Ratio of concentration to that close by source, 1.)
Conventional Zonal
Expt. ventilation ventilation
At periphery near source, 2 1 0-30 0-83
2 0-53 0-45
3 2-50 0-11
Mean* 0-74 0-34
At centre, 3 1 0-27 0-18
2 0-31 0-14
3 1-25 0-05
Mean 0-47 0-11
At periphery opposite source, 4 1 0-24 1-43
2 0-32 0.91
3 1-67 0-21
Mean 0-51 0-65
* Means are geometric.

operating room through a door has been taken as the value of EIR. For transfer
via the staff corridor:
SIR = SICR x CR/IR = 2*4 x 103 x 1-3 x 103 = 3-1 x 106,
where CR represents the particle concentration in the staff corridor outside the
receiving operating room door, i.e. 4 m from the source room door.
The overall transfer ratio between the two operating theatres is then
1/(1/(8-2 x 105) + 1/(3-1 x 106)) = 6-5 x 106.
Particle transfer within the operating room
Transfer of potassium iodide particles
Experiments were made in rooms with and without zonal ventilation. The
arrangement of the tests can be seen from Fig. 3 and the results in Table 2. In
order to create a normal turbulence in the operating rooms one person was moving
around in the periphery during the experiments.
As might be expected there is considerable variation in the distribution of par-
ticles from experiment to experiment. In rooms with conventional ventilation the
ratios between the particle concentration at the different sampling sites to that
at the source site varied between 2-5 and 0-24. In rooms with zonal ventilation
the corresponding range was from 1-43 to 0-06. However, the mean peripheral
concentration ratios were similar for both types of ventilation, 0-74 and 0'51 with
conventional ventilation compared with 0-34 and 0-65 with zonal ventilation. The
mean concentration ratio at the centre was, however, much less with zonal ven-
tilation, 0.11 compared with 0-47.
 Ventilation and airborne transfer 127
I

N\\O\ 2

'O 4
N

I I I I
0 1 2 3
m

Fig. 3. OP 72. Transfer of potassium iodide particles in an operating theatre.

Position of generator and samplers. A, Generator. Sampling points: 1, at the source;
2, periphery near source; 3, in the centre; 4, periphery opposite source.

Table 3. Transfer of bacteria-carrying particles during standard

activity in operating rooms
(Ratio of concentration to that in the source area.)
Conventional ventilation Zonal ventilation
Source area Receiving area (means of 4 experiments) (means of 9 experiments)
Periphery Centre 0-83 059
Centre Periphery 0-83 200

Transfer of bacteria-carrying particles generated by standard activity

The distribution of airborne bacteria-carrying particles generated by human
activity was also measured in the two operating rooms with different types of
ventilation. Five technicians performed a standardized activity for 30 min (rapid
walking) either in the centre or in the periphery of the operating room. The bac-
teriological samples were taken in the centre and in the periphery. The mean value
of the ratios between bacterial concentration at the sampling points to that in the
source area are given in Table 3. The experimental activity was sufficiently high
to create airborne contamination of between 31-7 c.f.u./m3 and 74-4 c.f.u./m3 with
an average of 41 c.f.u./m3. Whether standard activity was taking place in the
centre or in the periphery of operating rooms with zonal ventilation the concen-
tration of bacteria within the zone was about half that in the periphery. In rooms
with conventional ventilation the particles were almost evenly distributed in the
room, mean ratios being 0-83 for both kinds of experiments.

Tests on airborne contamination during operations

The same type of sampling in the centre and the periphery was also done during
actual operations on ten occasions in rooms with zonal ventilation and five times
128 A. HAMBRAEUS, S. BENGTSSON AND G. LAURELL
Table 4. Comparison of activity and airborne contamination in operating rooms
with zonal ventilation and operating rooms with conventional ventilation
Zonal Conventional
ventilation ventilation
Mean length of operation 80 min 90 mm
Minutes between door 1-4 1.2
openings (mean value)
No. of people present 13 15
(mean value)
Total number of bacteria
Mean of all samples 46-3 c.f.u.1m3 74*4 c.f.u./m3
Mean of means/operation 48-9 c.f.u./m3 70 9 c.f.u./m3
No. of Staph. aureus 0 03 c.f.u./m3 0*24 c.f.u./m3
mean per operation

in rooms with conventional ventilation. When sampling in the centre the slit
sampler was placed as close to the patient as possible. The level of airborne con-
tamination in rooms with zonal ventilation was about half that found in rooms
with conventional ventilation. In the centre of the rooms the mean values per
operation were 31-5 c.f.u./m3 and 73-2 c.f.u./m3 for rooms with zonal and con-
ventional ventilation respectively. The corresponding values for samples from
the periphery were 53'9 and 89-0 c.f.u./m3. The mean ratio between bacterial
concentration in the centre and the periphery for rooms with zonal ventilation
was 0-6 (range 0.5-0.8), for rooms with conventional ventilation it was 0*8 (range
0.7-1). The difference, although small, is significant (P < 0-01 > 0.005).

Observations concerning working routines and airborne

contamination during operations
During a period of 1.5 months a study was made of ten operations in operating
rooms with each type of ventilation. Two hundred and sixty-eight air samples
were taken in operating rooms with zonal ventilation and 141 in operating rooms
with conventional ventilation. All the samples were taken at the periphery of the
operating rooms. Notes were made concerning length of operation, number of
people present, frequency of door openings, etc. As can be seen from Table 4 there
was no significant difference between length of operation, which usually lasted
between 1 and 2 h, or number of door openings; at least one door was opened
almost every minute. The average number of people present during an operation
was also about the same being 13 and 15, respectively, in the two operating rooms.
However, in this investigation it was not possible to randomize the staff involved.
In operating rooms with zonal ventilation 122 different persons were present at
one time or other during the ten operations studied. The corresponding figure for
conventional operating rooms was 125. Of these people only 46 were the same for
the two series of observations. The mean number of bacteria of the total samples
taken in rooms with zonal ventilation was 46-3 c.f.u./m3 and for rooms with con-
ventional ventilation it was 74-4 c.f.u./m3. This difference is highly significant
(P < 0-001). The mean of the mean number for each operation was almost the
 Ventilation and airbome transfer 129
same: 48'9 and 70*9 c.f.u./m3, respectively. The mean number of Staph. aureus
found per operation was less in rooms with zonal ventilation 0*03 c.f.u./ms as
compared to 0-24 c.f.u./m3 in rooms with conventional ventilation.

DISCUSSION
The aim of the investigation was mainly to study transfer of airborne particles
and airborne contamination in a modern operating unit with positive pressure
ventilation. The initial experiments with tracer gas showed that the ventilation
system functioned in agreement with the plans drawn up for this unit. The transfer
of inert particles from one space to another was very low when there was no
activity. When particles were generated in the operating room the highest con-
centration outside it was found in the anaesthetic room. During activity the ratio
of particle concentration between the two rooms was about 100. This was due to
the fact that the anaesthetic room is a small closed area from which particles
disappear more slowly than from the large open corridors.
The transfer from the outside into the operating room was assumed to be the
same for all three doors and therefore only transfer through two of the doors was
tested. The mean value of the two ratios found was 1-7 x 103 and this was used in
calculations including the third door, i.e. transport from exit area to operating
room. As two operating rooms share the same exit area, the transfer of particles
from one operating room to the other via the exit area was calculated. Under
experimental conditions the ratio of particle concentration between the source and
the receiving operating rooms was 8-2 x 105 for transfer via the exit area only, and
1/(l/(8.2 x 105) + 1/(3.1 x 106)) = 6-5 x 105 if transfer via the staff corridor with
the same rate of door opening, 60/h, is included.
The ratio between source room and receiving room is inversely proportionate
to the frequency of movement through the door (Lidwell, 1972). It is therefore
possible to recalculate the experimentally found value into one that would cor-
respond to particle transfer during actual activity according to the following
simplified formulae:
60
am = a0 x q7g

= aD 60
m
am= X
'IO
a' denoteconcentration in source room
concentration in exit area
a' denotes concentration in exit area
concentration in receiving room
m number of movements through the door in each direction
=
amI and c4 are the ratios at m movements
a40 and aL4 are the ratios at 60 door openings per hour (= experimental
activity).
9 HYG 79
130 A. HAMBRAEUS, S. BENGTSSON AND G. LAuIRELL
The frequency of door openings between the operating rooms and the exit area
was 12 times per hour. For this activity the ratio of particle concentration between
source and receiving operating rooms would be 8'2 x 105 (60/12)2 = 2-1 x 107. It
would seem that this should provide good protection against airborne transfer of
bacteria from one room to the other, since a value of 1 particle/l in the source
room would give rise to only 5 x 1o-8 particles per litre in the receiving room. The
risk of airborne transfer from one operating room to the other through the staff
corridor was more than ten times less.
The potassium iodide particle method has earlier proved to be a useful method
for estimating transfer of airborne particles in different kinds of patient wards
(Hambraeus & Sanderson, 1972; Foord & Lidwell, 1975). It has proved possible
to use this technique in a highly ventilated area such as an operating room. A
comparison between room to room transfer found in an isolation ward (Hambraeus
& Sanderson, 1972) with plenum ventilated rooms with air-locks and that found
in the operating suite showed that the airborne isolation provided for the patient
in an operating room was more than 1000 times better than that provided for the
patient in the isolation ward. Since airborne transfer of bacteria in the isolation
ward was found to be far less important than transfer of bacteria by other means
such as clothes (Hambraeus, 1973), the airborne room-to-room transfer of bacteria
in the operating suite is probably negligible compared with other routes of trans-
mission. The distribution of particles within the operating room itself is therefore
more interesting. In this study operating rooms with zonal ventilation were com-
pared with operating rooms with conventional ventilation.
In operating rooms with zonal ventilation experiments with tracer particles as
well as bacteria-carrying particles showed that the contamination in the centre
was lower than in the periphery. In operating rooms with conventional ventilation
both tracer particles and bacteria-carrying particles were almost evenly distri-
buted within the room.
In operating rooms with zonal ventilation only about one-tenth of tracer par-
ticles were transferred from the periphery into the centre compared with one half
when conventional ventilation was employed. The air currents from the ceiling
are too slow to create a completely effective air curtain over the operating table
and in experiments with bacteria-carrying particles the difference between centre
and periphery and vice versa was substantially less being only about one half.
This may be due to the fact that in both these sets of investigations there were
people in the centre during the measurements. In the series of experimental
activity one person had to attend to the slit sampler in the centre and during
operations the entire operating team was in the centre. Since they generate
bacteria-carrying particles themselves it is natural that the differences will be
smaller than when tracer particles are used.
Airborne contamination was, however, low especially in operating rooms with
zonal ventilation. The contamination in these operating rooms was not con-
siderably higher than has been reported in investigations where ultra clean air
ventilation systems have been used (McDade, Whitcomb, Whitfield & Franklin,
1968; Scott, 1970; Cook & Boyd, 1971).
 Ventilation and airborne transfer 131
This operating unit is planned with two corridors, one for the patients and the
other for the staff. The operating and service rooms are situated in the centre.
One of the reasons for this arrangement is to minimize the traffic to and from the
operating rooms during operations. It was therefore surprising to find that the
activity was high during operations, a door being opened almost every minute.
According to several workers an intensive traffic in operating rooms is considered
to increase the risk of acquiring infections during operations. From observations
during operations it was not possible to calculate how much the entrance of one
or two more people contributed to air contamination. The variations in activity
during the operations were too rapid and frequent. From the experiments in which
bacteria-carrying particles were generated by standard activity it is possible to
calculate the influence of people and ventilation rate on the amount of airborne
contamination. With five people present in a room with a ventilation rate of 20
turnovers per hour the average airborne contamination was 41 c.f.u./m3.
The die-away rate due to sedimentation can be taken as approximately 5.5/h,
assuming a particle settling-rate of 0 3 m/min in a room 3'3 m high. The equili-
brium concentration in the room is given by Ne = (nB)/(R x S) (Bourdillon,
Lidwell & Lovelock, 1948), with n the number of persons, B the rate of dispersal
per person, B the ventilation rate and S the rate of loss due to sedimentation. For
the above conditions B = 41(20 + 5.5)/5 = 209. It is then possible to calculate
the equilibrium concentration for any number of persons present and for any
ventilation rate assuming the same rates of sedimentation and dispersal per person.
Although the mean number present at the operations was 13-15 probably not
more than 10 were active. According to this calculation the number of airborne
bacteria would then be 82/m3 in the conventionally ventilated room, the observed
mean value was 71.
In a following paper the frequency of infections during a three-year period
among patients operated on in this unit will be presented. Attempts will be made
to explain the importance of airborne infection and whether routines involving as
much movement as those described above may contribute to an increased rate
of infection.
REFERENCES
ABEL, E. & ALLANDER, C. (1966). Undersokning av nytt inbiAsnings system for rena rum.
V VS No. 8.
BOuEDILON, R. B., LIDwELL, O. M. & LOVELOCK, J. E. (1948). Studies in air hygiene.
Medical Re8earch Council Special Report Serie8 No. 262. London: H.M.S.O.
CHARNLEY, J. (1964). A clean air operating enclosure. Brii8h Journal of Surgery 51, 202.
CiHNLY, J. (1972). Postoperative infection after total hip replacement with special refer-
ence to air contamination in the operating room. Clinical Orthopaedic8 87, 167.
CHARNLEY, J. & EFTEKHAR, N. S. (1969). Postoperative infection in total prosthetic replace-
ment arthroplasty of the hip joint, with special reference to the bacterial content of the
air of the operating room. Briti8h Journal of Surgery 56, 641.
CooK, R. & BomD, N. A. (1971). Reduction of the microbial contamination of surgical wound
areas by sterile laminar air-flow. British Journal of Surgery 58, 48.
EFTXHAR, N. S. (1973). The surgeon and clean air in the operating room. Clinical Ortho-
paedics 96, 188.
FOORD, N. & LIDWELL, 0. M. (1972). An airbome particle tracer for crossinfection studies.
Journal of Hygiene 70, 279.
9-2
132 A. HAMBRAEus, S. BENGTSSON AND G. LAURELL
FooRD, N. & LIDWELL, 0. M. (1975). Airborne infection in a fully air-conditioned hospital.
II. Transfer of airborne particles between rooms resulting from the movement of air from
one room to another. Journal of Hygiene 75, 31.
HAMBRAEuS, A. (1973). Transfer of Staphylococcue aureus via nurses' uniforms. Journal of
Hygiene 71, 799.
HBAMRAEuS, A. & SANDEmON, H. F. (t972). Studies with an airborne particle tracer in an
isolation ward for burned patients. Journal of Hygiene 70, 299.
LEADING ARTICLE (1975). Risks and uses of total hip replacement. Briti8h Medical Journal i,
296.
LIDWELL, 0. M. (1960). The evaluation of ventilation. Journal of Hygiene 58, 297.
LDWELL, 0. M. (1972). Ventilation in subdivlided isolation units. Joural of Hygiene 67, 649.
M1DADE, J. J., WmTco1m, J. G., WHmTFmI, W. J. & FRANXLI, C. R. (1968). Microbio-
logical studies conducted in a vertical lamina airflow surgery. Journal of the American
Medical A88ociation 203, 125.
SCoTT, C. C. (1970). Laminar/linear flow system of ventilation. Lancet i, 989.
WILiAMs, R. E. O., BLOWERS, R., GARROD, L. P. & SHOOTER, R. A. (1966). Ho8pital Infec-
tion8, Cau8e8 and Prevention. 2nd ed. London: Lloyd-Luke Ltd.