Content uploaded by Zhang Lin
Author content
All content in this area was uploaded by Zhang Lin on Mar 22, 2020
Content may be subject to copyright.
Conversion of operating theatre from positive
to negative pressure environment
T.T. Chow
a,
*, A. Kwan
b
, Z. Lin
a
, W. Bai
a
a
Division of Building Science & Technology, City University of Hong Kong, Hong Kong SAR, China
b
Department of Anaesthesiology, United Christian Hospital, Hong Kong SAR, China
Received 24 November 2005; accepted 7 July 2006
Available online 14 October 2006
KEYWORDS
Operating theatre;
Airflow performance;
Airborne infection
Summary The severe acute respiratory syndrome (SARS) crisis led to the
construction of a negative pressure operating theatre at a hospital in Hong
Kong. It is currently used for treatment of suspected or confirmed airborne
infection cases, and was built in anticipation of a return of SARS, an out-
break of avian influenza or other respiratory epidemics. This article de-
scribes the physical conversion of a standard positive pressure operating
theatre into a negative pressure environment, problems encountered, air-
flow design, and evaluation of performance. Since entering regular service,
routine measurements and observations have indicated that the airflow
performance has been satisfactory. This has also been confirmed by regular
air sampling checks. Computational fluid dynamics, a computer modelling
technique, was used to compare the distribution of room air before and
after the design changes from positive to negative pressure. The simulation
results show that the physical environment and the dispersion pattern of
bacteria in the negative pressure theatre were as good as, if not better
than, those in the original positive pressure design.
ª2006 The Hospital Infection Society. Published by Elsevier Ltd. All rights
reserved.
Introduction
The severe acute respiratory syndrome (SARS)
crisis in Hong Kong from March to June 2003
resulted in extreme stresses and strains on the
general running of hospitals. Generally, SARS
patients were accommodated in negative pressure
isolation rooms on the ward. When these pa-
tients required operative procedures, a negative
* Corresponding author. Address: Division of Building Science
& Technology, City University of Hong Kong, Tat Chee Avenue,
Kowloon, Hong Kong, China. Tel.: þ852 2788 7622; fax: þ852
2788 9716.
E-mail address: bsttchow@cityu.edu.hk
0195-6701/$ - see front matter ª2006 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhin.2006.07.020
Journal of Hospital Infection (2006) 64, 371e378
www.elsevierhealth.com/journals/jhin
pressure theatre was considered to be more
suitable than a positive pressure environment. In
principle, a positive pressure operating theatre
with adequate air changes could quickly eliminate
the virus from the environment, and it has been
shown that the risk of cross-contamination from
airborne infection is low if staff are adequately
protected with appropriate personal protective
equipment (PPE).
1
However, a negative pressure
operating theatre offers optimal protection to
personnel working in adjacent areas.
Most SARS patients admitted to the United
Christian Hospital during the crisis were residents
of Amoy Gardens, where the largest transmission of
community-acquired SARS occurred. Over a short
period, one of the 11 operating theatres in the main
operating suite of the hospital was converted
temporarily into a negative pressure theatre. This
was achieved by incorporating two strong exhaust
fans next to the original exhaust system. The
pressure differential was maintained by sealing
the entrance doors with disposable sticky tape
after the patient was transported into the room.
This temporary negative pressure theatre worked
well during the SARS crisis.
2
Afterwards, the hospi-
tal management decided that a permanent nega-
tive pressure operating theatre was necessary, in
anticipation of SARS returning and for those pa-
tients contracting similar infectious airborne dis-
eases such as tuberculosis and severe influenza.
The construction of this permanent negative pres-
sure operating theatre was completed in June 2004.
The original operating theatre suite in this
hospital was built in 1994. Similar to all other
operating theatres in Hong Kong around that
period and up to the present day, the operating
theatres were maintained at a positive pressure.
Specifications such as the floor area, airflow
quantity and pressure gradient were designed to
meet the requirements of the British operating
theatre standards, which have been commonly
used as reference in Hong Kong.
3,4
The room pres-
sure was maintained at þ25 Pa. Areas around the
operating theatre were also under positive pres-
sure. By controlling the supply and extract airflow
rates of each room in accordance with the design
data given in Figure 1, a pressure gradient was de-
veloped in continuous progression through zones
with increasing sterility. Outside air was intro-
duced to the operating theatre through a perfo-
rated diffuser at ceiling level, directly above the
surgical area. This led to a downward displace-
ment of air above the operating table. Room air
was extracted through a low-level exhaust grill
located next to the back door and an embedded
exhaust duct in the sidewall.
This paper presents the process of converting
a positive pressure theatre into a negative pres-
sure theatre, and the subsequent performance
evaluation.
Methods
Routine monitoring of airflow performance
The airflow system performance of an operating
theatre in the hospital was monitored regularly
through observations and/or measurements of
pressure gradient, flow pattern, temperature and
humidity levels. To ensure sterility in the operating
theatre, routine bacterial sampling using two
types of plates [tryptone soy agar (TSA) and
Sabouraud agar] was performed after each air
duct cleansing and filter replacement. The plates
were placed in three positions (i.e. high, low and
at the air exhaust). The high position was located
by the anaesthesia apparatus near the operating
table, 2 m above floor level; the low position was
located near the operating table, 1 m above floor
level; and the air exhaust position was in front of
either of the exhaust grilles. The SAS Super 100
Air Sampler (International Pbi, Milano, Italy) was
used to obtain volumes of 500 L into two separate
55-mm culture plates of TSA (Oxoid, Basingstoke,
UK) and Sabouraud agar (bioMe
´rieux, Marcy
I’ Etoile, France). The plates were incubated at
37 Cand30C for two days and five days for bacte-
rial and fungal counts, respectively. A colony count
of less than 30 colony-forming units (CFU)/m
3
for
the TSA agar and 3 CFU/m
3
for the Sabouraud agar
was adopted as the acceptable standard in all oper-
ating theatres, but a more stringent standard was
applied to the theatre using laminar airflow.
5
Theatre selection for negative
pressure conversion
Theatre 1 (OT-1) of the main operating suite was
chosen for pressure conversion for two main rea-
sons. Firstly, OT-1 was the furthest away from the
other operating theatres, making isolation easier to
accomplish. It minimized the risk of air from the
corridor being contaminated as a result of traffic
flow and then being drawn into the negative
pressure operating theatre, which could be a risk
for wound infection. Secondly, it had two free
sidewalls that could accommodate the addition of
a separate exhaust system, and had its own scrub
area and a separate induction room that could be
converted into a room for removal of contaminated
clothing. The main feature of the negative pressure
372 T.T. Chow et al.
design, compared with the positive pressure design,
was the incorporation of a much stronger low-level
exhaust system. The exhaust air passed through
a two-stage filtration system (prefilter plus High
Efficiency Particulate Air (HEPA) filter) before its
final disposal via an exhaust air fan. In order to
achieve the designated airflow criteria, an ante-
room was constructed at the front end of the scrub
and induction rooms leading to OT-1. All doors
leading to these negative pressure rooms were
made airtight and interlocking. The physical layout
and the airflow specification of the negative pres-
sure operating theatre suite are shown in Figure 2.
As OT-1 and OT-2 originally shared the same air con-
ditioning system, a separate air conditioning system
had to be built for OT-2 before the necessary
changes were made to OT-1.
Static pressure heads in OT-1 and in the adjacent
rooms were monitored by differential pressure
gauge measurements. Correct airflow velocities at
the supply diffuser and exhaust grilles were
checked by vane anemometer measurements. The
airflow pattern was examined carefully using smoke
tests. In order to gather more technical information
for assessing the effectiveness of the present
airflow system, the room air distribution before
and after the conversion was examined through
computer analysis.
Airflow evaluation by CFD technique
Computational fluid dynamics (CFD) analysis pro-
vides comprehensive data on airflows within
a room. It demonstrates any deficiencies in air
distribution and in contaminant removal. It has
been applied to the study of airflows and contam-
inant distribution patterns in various operating
theatre applications.
6e10
In this study, the computation models of Cases
A and B, i.e. before and after the pressure
150
Dirty corridor
Scrub Induction
OT-1
Preparation
Sluice
AC plant
room
150
150
150
360
150
360
360
150
850
850
210 210
200
Induction
E.D.
Recovery bay
Clean corridor
Stabilizer
2
Stabilizer
1
Extract
E
SSupply Flow rate in L/s
Figure 1 Floor plan of operating theatre suite before pressure conversion. E.D., exhaust duct; OT-1, Operating
Theatre 1; AC, air conditioning.
Conversion of operating theatre from positive to negative pressure environment 373
conversion, are shown in Figure 3 (a) and (b), re-
spectively. The room dimensions were 6.3 m
(length) 5.9 m (width) 3.1 m (height). In the
computer model, the seven surgical staff standing
upright and the patient lying on the operating ta-
ble were represented as rectangular solid boxes.
In the analysis, it was assumed that each staff
member released infectious particles at a rate of
100 CFU/min from the body surface that faced
the patient. Also, an assumption was made that
an infectious particle release rate of 400 CFU/
min occurred from the surgical incision site at
the waist position and from the patient’s upper
surfaces. The main and satellite medical lamps
were 350 W and 200 W, respectively, and produced
heat fluxes from their downward surfaces. Each of
the eight fluorescent lighting panels surrounding
the perforated supply diffuser released a heat
flux of 70 W. The flow of fresh air was 0.85 m
3
/s.
For Case A (positive pressure), the exhaust
airflow at the exhaust grille was 0.21 m
3
/s and
the balance airflow of 0.64 m
3
/s was a combination
of discharge from the two pressure stabilizers and
the gaps between the doors and the floor. For Case
B (negative pressure), the total air extraction rate
through the two exhaust grilles was 1 m
3
/s. The
balance airflow of 0.15 m
3
/s entered the room
through the two pressure stabilizers. A deflector
plate was positioned 0.15 m in front of Stabilizer
2 to divert the incoming flow upwards. These con-
stituted the only differences between the two
cases, and hence the simulation results can be
readily compared.
Numerical simulations were performed with the
commercial CFD software FLUENT.
11
The standard
empirical model was adopted to simulate the
flow turbulence. Only steady-state conditions
were considered.
Scrub
125
100
150
300
850
Induction
180
150
OT-1
Preparation
Sluice
Induction
Dirty corridor
AC plant
room
Control
panel
E.D.
E.D.
Stabilizer 1Stabilizer 2
Automatic
sliding door
Automatic
sliding door
Automatic
sliding door
OT-2
1000
Extract
E
SSupply Flow rate in L/s
Anteroom
Figure 2 Floor plan of operating theatre suite after pressure conversion. E.D., exhaust duct; OT-1, Operating The-
atre 1; OT-2, Operating Theatre 2; AC, air conditioning.
374 T.T. Chow et al.
Results
Performance tests
After some months of construction, refitting, re-
peated performance checking and testing, most
system requirements were met successfully with
the exception of the differential pressure levels.
Problems in the control of room air pressure mainly
related to the quality of the building construction.
It was not possible to make the room enclosure
totally airtight. The other significant problem
encountered was the backflow of contaminated
air from the negative pressure theatre into the
anteroom, as observed by the smoke test when the
sliding doors of the operating theatre were
opened. Adding a deflector plate in front of
Stabilizer 2 above the entrance door from the
scrub room finally rectified this. Table I compares
some room pressure measurements with the origi-
nal design specifications. Although the differential
pressures measured by repeated checking
remained less than the design specifications, the
actual working conditions were found to be
acceptable. Colony counts of less than 30 CFU/m
3
for the TSA agar and less than 3 CFU/m
3
for the
Sabouraud agar were achieved consistently in the
routine checks.
Stabilizer 1
Stabilizer 2
Exhaust grille
Fluorescent
lighting
Oxygen supply
Anaesthesia apparatus
Equipment
table
Medical lamp
(satellite)
Staff 1
Staff 2 Staff 3
Staff 4
Staff 6
Staff 7
Perforated
supply diffuser
(a)
Medical lamp
(main)
(b)
Stabilizer 1
Stabilizer 2
Fluorescent
lighting
Oxygen supply
Anaesthesia apparatus
Equipment
table
Patient
Exhaust 1
Exhaust 2
Medical lamp
(satellite)
Staff 1
Staff 5
Staff 2
Staff 3
Staff 4
Staff 7
Perforated
supply diffuser
Medical lamp
(main)
Figure 3 Computation models of Operating Theatre 1 before and after the pressure conversion. (a) Positive pres-
sure, (b) negative pressure.
Conversion of operating theatre from positive to negative pressure environment 375
Simulation results
The airflow pattern of Case B (negative pressure)
was found to be generally consistent with the
smoke dispersion pattern as visualized through
the smoke tests during the commissioning period.
The results showed that the airflow systems
performed reasonably well both before and after
conversion from positive pressure to negative
pressure.
Figure 4 shows the simulation results of the dif-
ferential patterns of room air temperature, i.e.
the temperature rise above the supply air temper-
ature at the diffuser outlet for Cases A and B.
Figure 5 shows the distribution of infectious
particles released from the bodies of staff at the
operating level, i.e. 1.1 m from floor level.
Figure 6 shows the concentration of bacteria
released from the patient’s wound site in the ver-
tical plane, i.e. the situation across the incision
site.
In the negative pressure theatre, a deflector
plate was added to prevent backflow of the air
from OT-1 to the anteroom. Figure 7 compares the
velocity profiles before and after the addition of
this deflector plate.
Table I Pressure (in Pascals) recorded in a routine
check of the negative pressure theatre
Location Intended
pressure
Differential
pressure
(all doors
closed)
Differential
pressure
(anteroom with
one door
opened)
Anteroom þ10 þ6.1 0
Scrub room 10 1.9 4.3
Induction
room
10 6.2 8.6
Operating
Theatre 1
15 11.3 14.3
Dirty corridor þ10 þ3.0 1.6
9
810 1
2
3
4
5
5
7
4
6
5
7
7
8
123
4
6
5
7
5
43
2
1
10
98
7
8
7
6
5
4
3
4
43
5
3
3
4
2
2
4
4
5
6
3
6
5
3
5
7
7
8
3
3
33
3
22
22
1
1
12
34
4
5
7
8
9
32
1987
5
4
4
5
Staff 4 Equipment
table
Medical lamp
(main)
(b)
(a)
6
Figure 4 Contours of temperature rise in degrees Cel-
sius (above temperature of air supply) at mid-width
cross-section of theatre. (a) Case A, positive pressure;
(b) Case B, negative pressure.
16
16
14
12
9
9
9
12
14
16
18
Back door
18
20
>20
12
9
7
9
5
1
14
14
16
16
18
20
16
14
14
12
12
9
9
4
4
7
9
9
9
11
11
13
16
11
>20
>20
>20 >20
>20
4
7
7
1
1
3
3
5
5
4
4224
7
9
11
18
16
18
16
13
13
11
9
7
16
18
20
11
2
(b)
(a)
Figure 5 Comparison of concentration distribution of
bacteria (released from staff) at operating plane, 1.1 m
from floor level (unit: colony-forming units/m
3
). (a)
Case A, positive pressure; (b) Case B, negative pressure.
376 T.T. Chow et al.
Discussion
The room temperature conditions shown in Figure 4
were compared on a vertical plane cutting across
the operating table. Steep temperature gradients
were found at the boundaries of the supply air
streams from the diffuser outlet down to the level
of the operating table. This was caused by the
higher flow velocity between the two medical
lamps, with the effect being more obvious in the
negative pressure model. Hence, in the surgical
zone, the patient is in a lower temperature envi-
ronment (closer to the temperature of the supply
air) than the general environment. This amounted
to a difference of 4e5C, being more pronounced
in the negative pressure model. The vertical tem-
perature stratification was not obvious in most
free positions of the room (around 2 C difference
from feet to head level for most room positions).
Hence, thermal comfort was achieved in both
models, consistent with the requirements of the
International Standard Organization’s standard on
thermal comfort.
12
It can be seen from Figure5 that the concentration
levels were low (<10 CFU/m
3
) for all cases at the op-
erating table (centre of the room). Hence, both
models provide good protection for the patient as
the bacterial concentration here is relatively low
compared withthe rest of the room. The vertical uni-
directional airflows were effective in both cases.
The effectiveness of the air stream in removing
the bacteria is obvious in Figure 6. The infectious par-
ticles were shown to flow below the operating table
until being drawn out of the room via the exhaust
grille(s). At the level of the respiratory system, i.e.
1.6 m above floor level, the concentration of infec-
tious particles released from the patient was less
than 10 CFU/m
3
. This means that surgical staff are
at little risk of being infected by the patient during
the operation. At the same time, there was no evi-
dence that adding the deflector plate affected the
airflow streams. The flow pattern in Figure 7(b)
shows a better flow in the surgical zone than in
Figure 7(a). Hence, lower bacteria concentration
levels in the same zone were achieved.
Overall, although the risk of cross-contamination
from airborne infection is low if staff are
adequately protected with appropriate PPE, a neg-
ative pressure operating theatre can offer optimal
protection to personnel working in adjacent areas.
Careful selection of the site is required to avoid
contamination from other sites such as hospital
wards. It has been shown that although the actual
pressure differentials in the negative pressure
rooms after completion may not meet the design
specification, the actual working conditions can
still be adequate if the designated airflow path
Staff 2 Staff 6
1
7
4
>30
>30
30
27
17
14
4
7
11
14
7
7
11
14
4
1
7
>30
4
4
24
>30
4
7
4
1
4
4
(a)
(b)
Figure 6 Comparison of concentration distribution of
bacteria (released from patient) at mid-length cross-
section of the theatre, cutting across the wound position
of the patient (unit: colony-forming units/m
3
). (a) Posi-
tive pressure, (b) negative pressure.
Stabilizer 2
Staff 1 Staff 7
(b)
(a)
Figure 7 Comparison of velocity vector at one-third
length of theatre, vertical plane cutting across the
wall inlet cross with or without the deflection. (a) Nega-
tive pressure design without deflector plate, (b) nega-
tive pressure design after deflector plate added.
Conversion of operating theatre from positive to negative pressure environment 377
and flow rates can be maintained. In the present
case, repeated testing and the results of periodical
air sampling checks confirmed the quality of the
completed work. Moreover, the steady-state sim-
ulation results showed that the level of thermal
comfort as well as the dispersion behaviour of the
bacteria in the negative pressure theatre were as
good as, if not better than, those in the original
positive pressure design.
Acknowledgements
The work described in this article was supported
by a strategic research grant from the City Uni-
versity of Hong Kong (project no. 7001609).
References
1. Seto WH, Tsang D, Yung TY, et al. Effectiveness of precau-
tions against droplets and contact in prevention of nosoco-
mial transmission of severe acute respiratory syndrome
(SARS). Lancet 2003;361:1519e1520.
2. Kwan A, Fok WG, Law KI, Lam SH. Tracheostomy in a patient
with severe acute respiratory syndrome. Br J Anaesth 2004;
92:280e282.
3. Essex-Lopresti M. Operating theatre design. Lancet 1999;
353:1007e1010.
4. National Health Service Estates. Health Technical Memo-
randum 2025. Ventilation in healthcare premises. London:
HMSO; 1994.
5. Friberg B, Friberg S, Burman LG. Inconsistent correction be-
tween aerobic bacterial surface and air counts in operating
rooms with ultra clean laminar air flows: proposal of a new
bacteriological standard for surface contamination. J Hosp
Infect 1999;42:287e293.
6. Chen Q, Zheng J, Moser A. Control of airborne particle
concentration and draught risk in an operating room. Indoor
Air 1992;2:154e167.
7. Colquhoun J, Partridge L. Computational fluid dynamics
applications in hospital ventilation design. Indoor Built
Environ 2003;12:81e88.
8. Chow TT, Yang XY. Performance of ventilation system in
a non-standard operating room. Build Environ 2003;38:
1401e1411.
9. Chow TT, Yang XY. Ventilation performance in the operating
theatre against airborne infection: numerical study on an
ultra-clean system. J Hosp Infect 2005;59:138e147.
10. Woloszyn M, Virgone J, Melen S. Diagonal air-distribution
system for operating rooms: experiment and modeling.
Build Environ 2004;39:1171e1178.
11. Fluent Inc. User’s guide for Fluent 6.1. Lebanon: Fluent Inc;
2003.
12. ISO Standard 7730. Moderate thermal environment ede-
termination of the PMV and PPD indices and specification
of the conditions for thermal comfort. Geneva: Interna-
tional Standard Organization; 1990.
378 T.T. Chow et al.
