American Journal of Infection Control 40 (2012) 750-5

Contents lists available at ScienceDirect

American Journal of Infection Control American Journal of

Infection Control

journal homepage: www.ajicjournal.org

Major article

Traffic flow in the operating room: An explorative and descriptive study on air
quality during orthopedic trauma implant surgery
Annette Erichsen Andersson RN a, b, *, Ingrid Bergh RN, PhD c, Jón Karlsson MD, PhD d, e,
Bengt I. Eriksson MD, PhD d, e, Kerstin Nilsson RN, PhD a
a
Institute of Health and Care Sciences, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden
b
Department of Anesthesia, Surgery and Intensive Care, Sahlgrenska University Hospital, Gothenburg, Sweden
c
School of Life Sciences, University of Skövde, Skövde, Sweden
d
Department of Orthopedics, Sahlgrenska University Hospital, Gothenburg, Sweden
e
Institute of Clinical Sciences, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden

Key Words: Background: Understanding the protective potential of operating room (OR) ventilation under different
Surgical site infection conditions is crucial to optimizing the surgical environment. This study investigated the air quality,
Door opening expressed as colony-forming units (CFU)/m3, during orthopedic trauma surgery in a displacement-
Air sampling
ventilated OR; explored how traffic flow and the number of persons present in the OR affects the air
Colony-forming units
contamination rate in the vicinity of surgical wounds; and identified reasons for door openings in the OR.
Methods: Data collection, consisting of active air sampling and observations, was performed during 30
orthopedic procedures.
Results: In 52 of the 91 air samples collected (57%), the CFU/m3 values exceeded the recommended level
of <10 CFU/m3. In addition, the data showed a strongly positive correlation between the total CFU/m3 per
operation and total traffic flow per operation (r ¼ 0.74; P ¼ .001; n ¼ 24), after controlling for duration of
surgery. A weaker, yet still positive correlation between CFU/m3 and the number of persons present in
the OR (r ¼ 0.22; P ¼ .04; n ¼ 82) was also found. Traffic flow, number of persons present, and duration of
surgery explained 68% of the variance in total CFU/m3 (P ¼ .001).
Conclusions: Traffic flow has a strong negative impact on the OR environment. The results of this study
support interventions aimed at preventing surgical site infections by reducing traffic flow in the OR.
Copyright Ó 2012 by the Association for Professionals in Infection Control and Epidemiology, Inc.
Published by Elsevier Inc. All rights reserved.

The prevention of surgical site infection (SSI) after orthopedic Strategies to minimize the risk of SSI can be focused on 3 major
implant surgery is a hot topic for politicians, hospital administra- areas: the patient, the surgical technique, and the surgical envi-
tors, and clinicians, given the enormous amount of resources these ronment. Optimizing the patient preoperatively by applying
infections consume in terms of extra costs of medications, reoper- current knowledge about the risks associated with smoking,
ations, and extended length of hospital stays.1-4 Adding the human malnutrition, ongoing infections and wounds, diabetes, and other
perspective, a recent study indicated that afflicted patients suffer underlying diseases and conditions compromising immunologic
deeply, both physically and emotionally, from the consequences of defense systems can improve postoperative outcomes signifi-
a deep SSI for a prolonged period.5 cantly.6-9 Optimizing the surgical technique by not exceeding the
estimated 75th percentile of surgery time based on the type of
surgical procedure reduces the risk of SSI and also minimizes blood
loss, thereby avoiding the need for (allogeneic) blood transfusions
* Address correspondence to Annette Erichsen Andersson, RN, Department of and eliminating postoperative hematomas.10-15
Anesthesiology/Surgery, Sahlgrenska University Hospital/Östra, Smörslottsgatan 1, The present study focused on strategies aimed at optimizing the
SE-416 85 Gothenburg, Sweden. surgical environment, in particular the air quality in the operating
E-mail address: [email protected] (A.E. Andersson). room (OR). Enhancing air quality by reducing airborne contami-
Author contributions: A.E.A., I.B., B.E., J.K., and K.N. designed the study; A.E.
nation has been shown to be of great importance, especially in
A. performed data collection and coordination; A.E.A. and I.B. analyzed data; and
A.E.A., I.B., B.E., J.K., and K.N. wrote the manuscript. relation to implant surgery.16-18 It has been suggested that levels be
Conflict of interest: None to report. maintained at <10 CFU/m3 during implant surgery, and that clinical

0196-6553/$36.00 - Copyright Ó 2012 by the Association for Professionals in Infection Control and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved.
doi:10.1016/j.ajic.2011.09.015
 A.E. Andersson et al. / American Journal of Infection Control 40 (2012) 750-5 751

benefits can be expected by reducing it to 1 CFU/m3,18 given that closed-fracture surgery using plates and screws, intramedullary
very low levels of clinically relevant coagulase-negative staphylo- nails, or hemiarthroplasty. Sampling and data collection were done
cocci can initiate a device-related infection.19 A landmark study during the daytime and in most of the cases once a week, over
found a strong linear relationship between the level of bacterial air a 7-month period from April to November 2010, with the exception
contamination and the prevalence of deep SSI.20 of the holiday month of July.
The most common ventilation systems in use today are turbu-
lent, displacement, and laminar airflow (LAF) systems. Whereas Air sampling method
turbulent and displacement ventilation systems differ primarily in
the methods used to supply clean air, both are incapable of opposing A Sartorius MD-8 air scanner (Sartorius Mechatronics, Göttin-
heat emissions from people and lamps. Both types of systems are gen, Germany) was used to collect airborne microorganisms. Air
sensitive to movement, leading to the formation of local eddies.21 was sampled at a flow rate of 3 m3/hour (0.83 L/second) in
The most important source of airborne contamination is related to 20-minute periods continuously during the operations. The
the dispersal of particles from persons present in the OR and their instrument was placed outside the sterile zone, and a sterilized
movements.22-24 Clothing OR staff in scrubs with lower air perme- flexible hose was extended to reach the wound area, with a filter
ability compared with conventional scrubs can reduce the dispersal holder attached to the end. The filter holder with a gelatin filter (3
of microorganisms by the OR staff, thereby significantly reducing the mm pore size; 80 mm diameter) was placed 20-40 cm from the
airborne contamination.23,25,26 An experimental study has indicated wound. The filters were placed vertically (n ¼ 60), slightly upward
that the protective ability of tightly woven clothing systems can (n ¼ 23), slightly downward (n ¼ 17), or horizontally (n ¼ 3). In
deteriorate after repeated washing and sterilization.27 Another those cases in which the OR nurse had problems attaching the filter
study concluded that unnecessary conversation in the OR can holder close to the wound (n ¼ 13), the holder was placed on the
contribute to an increased risk of airborne contamination,28 and Mayo stand. Data on filter placement was absent in 4 cases. The
a pilot study indicated a possible association between high levels of filter was changed every 20 minutes by the scrub nurse or the
noise during surgery and SSI.29 The impact of OR door openings on assistant and given to the researcher, who immediately placed it on
air quality has been investigated in several studies,30,31 but clinical a nonselective Colombia agar base plate with 5% horse blood. Agar
tests of this have proven difficult. Ritter et al32 found no significant plates were incubated at 30 C for 4 days, after which the total
difference in OR airborne bacterial counts between closed doors aerobic bacterial count was measured. Microbiological results are
(mean, 15.2 CFU) and swinging doors (mean, 14.5 CFU). Stocks et al13 expressed as CFU/m3. A total of 116 samples were analyzed; 4
reached the same conclusion. Only one study to date has reported samples were accidentally contaminated and thus excluded from
a correlation between OR door openings and elevated airborne the analysis. Filters and plates were handled using strict aseptic
bacterial counts33; however, that result was based on 69 passive technique. To evaluate the technique, filters that had not been used
samples (on settle plates) and only 13 active samples (single-stage for air sampling were placed on agar plates and incubated in the
slit impact) placed outside the surgical wound area. The aims of same way as the used filters; no bacterial growth was detected.
the present study were to investigate the air quality, expressed as
CFU/m3, during orthopedic trauma implant surgery in a displace- Observational method
ment ventilated OR; to explore how traffic flow and the number of
people present in the OR affect the air contamination rates in the Data was collected using a pretested, structured observation
vicinity of the surgical wound; and to identify reasons for door form. The following variables were included: date and time, OR,
openings in the OR. room temperature, type of surgery and fixation method. The period
from incision time to wound closure was divided into 20-minute
METHODS intervals corresponding to the ongoing air sampling. During 119
intervals (each interval corresponding to 20 minutes of air
Setting sampling), traffic flow was measured, as well as the reasons for
door openings, and the current step in the surgical procedure was
The study was performed at a Swedish university hospital that recorded. The number of people present in the OR, patient and
performs approximately 9,000 surgical procedures annually. Data researcher excluded, was recorded.
was collected in 3 parallel ORs of equal size (39 m2), each equipped
with an upward air-displacement system supplying cool air (2-3 C Data analysis
below room temperature) above the floor in each of the 4 corners of
the room. By thermal convection, the air is evacuated via 4 exhaust Primary analyses showed that CFU/m3 could not be considered
fans installed in the ceiling. Each OR is supposed to be maintained a variable with a normal distribution. For this reason, the linear
at positive air pressure by adjusting the inflow rate to exceed the relationship between CFU/m3 per 20-minute interval and traffic
outflow rate; however, the desired difference in pressure between flow per 20-minute interval was investigated using Spearman’s rho.
the outer hall and the OR is not specified. Normally the pressure To investigate the strength and direction of the linear relationship
difference is w3 kPa, and an alarm is activated if the pressure falls between the total traffic flow per operation and the total CFU per
so that the difference is neutralized. Each OR has only a single entry operation, partial correlations were conducted, enabling the
point, with the door opening inward, leading directly to the outer removal of duration of surgery as a potentially confounding vari-
hall. The OR teams wore conventional cotton/polyester 50/50 mix able and thereby giving a more accurate description of the rela-
shirts and trousers, long surgical hoods tucked in, and private shoes tionship between the variables. Investigations of correlations
and socks. The scrubbed team also wore reinforced disposable between normally distributed variables (ie, traffic flow, duration of
sterile gowns, facemasks (RII), and double-sterile gloves. Adher- surgery, and number of people present) were performed using
ence to this practice was recorded for every operation. During Pearson’s product-moment correlation coefficient. Significance was
almost half of the operations, at least one of the air inlet supply defined as P < .05. All tests were 2-tailed. In relation to hierarchical
devices was partially blocked by medical equipment. multiple regression analysis, preliminary analyses were conducted
Data were collected during 30 consecutively selected full- to ensure no important violations of the assumptions of normality,
length orthopedic trauma operations involving different types of linearity, and multicollinearity.
752 A.E. Andersson et al. / American Journal of Infection Control 40 (2012) 750-5

One-way between-group analysis of variance with post hoc Table 1

tests found no significant difference in mean CFU counts among the CFU/m3 values and sampling positions

3 ORs. However, applying the same test on sampling device posi- 95% confidence interval for mean
tioning indicated that these variations can lead to differences in Position n Mean SD Lower bound Upper bound
mean CFU/m3 values. The mean difference between vertically
40-20 cm from wound
placed filters and filters placed on the Mayo stand was significant Vertically 60 15.8 13.9 12.2 19.4
(P ¼ .01) (Table 1). In 2 operations involving tibia fractures fixed Downward 17 15.2 10.2 10.0 20.5
with an intramedullary nail, the sampling filters had been placed Slightly upward 23 13.0 13.4 7.1 18.8
Horizontally 3 8.6 3.7 0.74 18.0
vertically on the opposite leg. During surgery, the injured leg was
Mayo stand
flexed at 90 degrees, thereby partially or completely blocking the Vertically 13 6.6 4.4 3.9 9.4
sampling filters with the sterile drape during most of the operation. Total 116 13.9 12.6 11.6 16.3
For this reason, further analysis of air quality in the vicinity of the
n, number of samples.
wound area, samples obtained on the Mayo stand and during the 2
operations for tibia fracture were excluded, leaving 92 samples for
analysis. Four operational phase were defined: 1, incision phase; 2, Table 2
dissection phase; 3, implantation phase; and 4, wound closure Air quality and related variables
phase. Content analysis was used on observational data.34 n 95% CI for Median
Variables (missing) Mean (SD) mean (range)
Ethics CFU/m3 91 (1)* 15.9 (13.4) 13.1-18.7 13 (0-55)
Total CFU/m3 per 24y 60.4 (55.9) 36.8-84 33.5 (7-187)
The study was approved by the University of Gothenburg’s operation
Number of people 111 (9)z 5.4 (1) 5.2-5.6 5 (3-10)
Ethics Committee (157-10). Written and oral information was
Traffic flow rate 119 (1)z 4.3 (2.9) 3.8-4.8 4 (0-14)
provided in line with the 4 principal requirements of the Helsinki Traffic flow rate per 30y 17.4 (13.5) 12.4-22.4 14 (0-67)
Declaration (autonomy, beneficence, nonmalfeasance, and operation
justice).35 Accordingly, informed consent was obtained from all of Duration of surgery, 29 (1)x 83.5 (39.7) 68.4-98.5 60 (20-200)
the OR teams before observations and sampling. minutes

*Number of air samples.

y
RESULTS Number of operations.
z
Measured in 20-minute intervals.
x
From incision time to end of closure in minutes.
Air sampling was performed during 30 orthopedic operations in
a total of 120 air sampling intervals. The distributions of surgical
procedures were 73 plates and screws (60.8%), 26 intramedullary exemplify, this could mean that a staff member would enter the OR,
nails (21.7%), and 21 hemiarthroplasties (17.5%). The variations in take a look around, and then walk out.
CFU/m3 values were found between operations rather than during Traffic flow rates in relation to the previously mentioned 4
operations (P ¼ .001). In 52 of 91 samples, the CFU/m3 values phases of the operation were analyzed by one-way analysis of
exceeded the recommended level of <10 CFU/m3. In 14 of 24 variance with post hoc tests showing no significant difference in
operations, the mean values exceeded 10 CFU/m3; in 5 of these mean traffic flow rate per 20-minute intervals. In addition, no
operations, the mean values exceeded 25 CFU/m3. The highest significant differences in mean CFU/m3 values were found among
mean values were 37.5 and 44.3 CFU/m3. Qualitative analysis the different phases. No correlation was detected between the
revealed high activity levels (ie, movements within the OR as well number of people present and traffic flow rates in the OR.
as traffic flow) during these operations, along with other poten-
tially negative variables, such as hair hanging outside the surgical
Number of people and the effect on air quality
hood, the presence of a sneezing person, and more than 5 people
present in the OR. In 5 operations, mean values were <5 CFU/m3,
A minor correlation was found between CFU/m3 and the
with the lowest values being 1.6 and 2.3 CFU/m3, and notes written
number of people present in the OR (r ¼ 0.22; P ¼ .04; n ¼ 82).
during these operations reveal that there was no traffic flow and
low activity. Basic results on air quality, expressed as CFU/m3, and
related variables are provided in Table 2. Duration of surgery and type of surgical procedure

Traffic flow No correlation was found between CFU/m3 rates measured in

20-minute intervals and duration of surgery measured in minutes.
The relationships between the total traffic flow rate per opera- A positive correlation was found between the total CFU/m3 per
tion and the total CFU/m3 sampled per operation and between operation and duration of surgery (r ¼ 0.62; P ¼ .01; n ¼ 23). No
traffic flow rate per 20-minute interval corresponding to 20 correlation was found between traffic flow rate per 20-minute
minutes of air sampling were investigated. A positive correlation interval and duration of surgery, but a strong correlation was
was found between CFU/m3 and traffic flow rates when measured noted between total traffic flow rate per operation and duration of
in 20-minute intervals (r ¼ 0.309; P ¼ .003). The data show a strong, surgery (r ¼ 0.79; P ¼ .01; n ¼ 23). Differences in mean CFU values
positive correlation between the total CFU/m3 per operation and in relation to type of surgical procedure are presented in Table 4.
total traffic flow rate per operation (r ¼ 0.74; P ¼ .001; n ¼ 24
operations). Because duration of surgery correlates to the total CFU Predictors of CFU
and traffic flow rates, duration of surgery was controlled for in the
analysis. Hierarchical multiple regression was used to assess the ability
A total of 529 door openings were recorded. Reasons for OR of traffic flow and number of people present in the OR to predict
entries and exits were grouped into categories, as shown in Table 3. CFU/m3 levels after controlling for duration of surgery. Duration of
No reason could be identified in relation to 93 entries and exits. To surgery was entered in step 1, explaining 36% (adjusted R2 ¼ 0.359)
 A.E. Andersson et al. / American Journal of Infection Control 40 (2012) 750-5 753

Table 3
Reasons for traffic flow

Necessary door openings* n Semi-necessary door openings n Unnecessary door openings n

Expert consultations (eg, help needed from senior surgeons, 40 Surgical team members entering after incision 76 Logistic reasons planning next 30
expert nurses, or anesthesiologists) or leaving before closure or other operation
Instruments or other material needed 137 Lunch and coffee breaks 108 Social visits 45
No detectable reasons 93
Total 177 184 168
529

*The need assessed in relation to patient safety and ongoing procedure.

Table 4 of door openings; surgical team members entering or leaving the

Relationships among CFU, surgical procedures, and traffic flow, analysis of variance OR when the wound was open, for 14%. Reductions in all of these
n Mean SD 95% CI P value large categories of traffic flow are possible. Door openings for
Mean CFU/m3 value in relation to surgical procedure* .001
logistic reasons could all be avoided by telephone communication.
Plates and screws 69 18.7 13.3 15.5-21.9 Door openings related to social visits and for no detectable reasons
Hemiarthroplasty 11 4.73 9.87 1.1-18.6 together accounted for 27% of the traffic flow, possibly reflecting an
Intramedullary nails 11 4.73 3.1 2.6-6.8 OR culture that accepts door openings for no special reason.
Mean traffic flow rates in relation to .004
Although it is reasonable to think that an individual who enters an
surgical procedurey
Plates and screws 69 4.5 2.7 3.8-5.1 OR always has a good reason for doing so, in those cases we could
Hemiarthroplasty 11 2.3 1.4 1.3-3.3 find no link to the ongoing procedure. Blaming individuals for lack
Intramedullary nails 11 2.2 2.3 0.6-3.77 of discipline is not be a fruitful way to address this problem, given
*Number of air samplings corresponding to type of surgical procedure. that the cause probably extends the individual level. In addition,
y
Number of surgical procedures corresponding to traffic flow rate per 20-minute merely counting exits and entries while failing to analyze the
interval. reasons behind traffic flow behavior could lead to misdirected
interventions.37
of the variance in total CFU/m3 per operation. After entering traffic Directing the focus of change at an organizational level,
flow and number of people present, the total variance explained by including enhanced knowledge, logistics, and perioperative plan-
the model as a whole was 68% [F(3,16) ¼ 14.32; P ¼ .001]. The 2 ning, would give the OR staff the necessary tools to minimize door
control measurements, traffic flow and number of people, openings in the OR. This would not only minimize traffic flow, but
explained an additional 34% (adjusted R2 ¼ 0.336) of the variance in also likely reduce the duration of wound exposure. Lynch et al38
CFU/m3 when controlling for duration of surgery (R2 change ¼ 0.34; reported a mean rate of 40 door openings per hour for orthopedic
F change (2,16) ¼ 9.91; P ¼ .002). In the final model, only traffic flow total joint surgery, and Young et al39 reported a mean rate of 19.2
was statistically significant (standardized b ¼ 0.95; P ¼ .001). per hour for cardiac surgery, compared with the rate of 12.9/hour in
the present study. The traffic flow patterns reported in these 3
DISCUSSION studies must be considered in light of the high correlation between
door opening rate and elevated CFU levels, representing a major
In orthopedic surgery, large-scale efforts and research activities patient safety problem.
have focused on infection control, mainly in relation to elective The large variation in CFU values among operations in the
primary joint replacement surgery. The findings of the present present study is in line with previous reports.13,20,33 This supports
study show that the recommended limit of >10 CFU/m3 was the perception that CFU/m3 level should not be discussed as an
exceeded in 57% of the samples analyzed. Patients with orthopedic independent variable with a presumed normal distribution in the
trauma carry an extra burden of preoperative soft tissue and skel- OR, because it is highly dependent on other variables and can be
etal damage, and have minimal opportunities to be optimized in reduced to almost nondetectable levels under optimal conditions.
relation to comorbidities that are known to be major risk factors in The importance of the duration of surgery in relation to CFU/m3
this group of patients.36 Adding smoking habits and old age (the levels measured at 20-minute intervals was of minor importance.
latter of which is common in patients with osteoporotic hip frac- However, the duration of surgery is of clinical relevance, given that
ture), a picture of a highly vulnerable group of patients emerges. the total CFU level increases with increasing duration of surgery,
Reducing risk factors in the surgical environment clearly would be thereby exposing the wound to an increased total number of CFUs
beneficial for this group of patients. One of the most important and increasing the risk of SSI.10,40 In addition, longer duration of
findings of the present study is the highly negative impact of traffic surgery was associated with in higher total OR traffic flow rates. In
flow in the OR on bacterial contamination of the air close to the this sample, only very small variations in relation to the number of
wound; that is, a high rate of door openings was associated with people present in the OR were observed; as a result, the effect of the
high rates of CFU/m3 values. This correlation is weaker when number of people present in the OR on CFU level could not be
analyzing CFU/m3 per 20-minute interval compared with the total investigated thoroughly. The differences in CFU levels related to
CFU/m3 per operation, which may be related to the unorganized type of surgery, with fixation with plates and screws associated
manner in which bacterial dispersion reaches the wound area after with the highest levels, can be explained by the fact that these
an OR entry or exit because of turbulent air flow patterns as well as procedures were associated with 50% more door openings. The fact
movement of people in the OR. Analysis of the factors affecting that in almost half of cases, at least one of the air inlet supply
traffic flow found that only 7% of the door openings were related to devices was partially blocked by medical equipment might suggest
the need for expert consultation. Supply issues represented the that the staff has poor knowledge of how the ventilation system
largest category (26%); improving preoperative planning and works and how to deal with the reality of underdimensioned
communication between the surgeon and OR nurse in charge could operating rooms. To investigate the consequences of blockage of air
possibly reduce these door openings. Staff breaks accounted for 20% inlets, it would be necessary to control for how close the medical
754 A.E. Andersson et al. / American Journal of Infection Control 40 (2012) 750-5

equipment was placed in relation to the inlet device and also for the observed and the observer. To estimate the effect of the pres-
how large an area of the inlet supply was blocked. These data were ence of an observer, the traffic flow rates at the beginning of the
not registered in the present study, precluding analysis of the study period (May) were compared with rates measured after 6
possible impact on air contamination rates. Given that our regres- months (November); no statistically significant differences in traffic
sion model explains 34% of the variance of total CFU/m3 per oper- flow rates were detected.
ation, future research should aim at developing a clinically relevant
predictive model for estimating bacterial contamination under CONCLUSION
different environmental and behavioral conditions, taking into
account clothing systems and activity levels in the OR. This study has clearly linked elevated airborne bacterial counts
in the surgical area to door openings in conventionally ventilated
Methodological considerations ORs, thereby providing the scientific evidence needed to initiate
interventions aimed at preventing SSI by reducing traffic flow in the
Conducting representative air sampling in the OR in live condi- OR. In addition, analyzing the reasons for door openings seems to
tions proved highly challenging, and many methodological and be of great importance to the success of any intervention
technical issues had to be addressed both before and during the implemented.
present study. The choices of sampling velocity, time, and culture
media were based on recommendations from infection control Acknowledgment
practitioners performing surveillance sampling on a regular basis.
Studies have reported that the viability of microorganisms might be The authors thank the OR staff and orthopedic surgeons for their
affected by prolonged sampling times and high airflow rates.41,42 participation in this study, and L.O. Persson for statistical advice.
Evaluation of the Sartorius air sampler demonstrated no reduction
in the viability of cocci after drawing 2.6 m3 for 20 minutes, but References
negative effects for Escherichia coli.43 Various cocci were the main
relevant species found in the OR,24 and these bacteria also are the 1. Bozic KJ, Ries MD. The impact of infection after total hip arthroplasty on
hospital and surgeon resource utilization. J Bone Joint Surg Am 2005;87:
leading cause of infections related to implanted medical devices.44 1746-51.
Based on this, we believed that the sampling time was an accept- 2. Coello R, Charlett A, Wilson J, Ward V, Pearson A, Borriello P. Adverse impact of
able compromise between the purpose of the study and the need to surgical site infections in English hospitals. J Hosp Infect 2005;60:93-103.
3. de Lissovoy G, Fraeman K, Hutchins V, Murphy D, Song D, Vaughn BB. Surgical
avoid being overly intrusive during the procedures. However, we
site infection: incidence and impact on hospital utilization and treatment costs.
consider the wide variety of sampling positions in this study Am J Infect Control 2009;37:387-97.
a limitation, which might have led to underestimation of CFU values. 4. Monge Jodra V, Sainz de Los Terreros Soler L, Diaz-Agero Perez C, Saa
The literature typically reports on the distance from the wound to Requejo CM, Plana Farras N. Excess length of stay attributable to surgical site
infection following hip replacement: a nested case-control study. Infect Control
the sampling device (striving to be as close as possible), sampling Hosp Epidemiol 2006;27:1299-303.
velocity, and time; unfortunately, methodological issues are rarely 5. Andersson AE, Bergh I, Karlsson J, Nilsson K. Patients’ experiences of acquiring
studied or discussed. Further studies addressing the positioning of a deep surgical site infection: an interview study. Am J Infect Control 2010;38:
711-7.
sampling filters, the angle between filters and air flow, sampling 6. Neumayer L, Hosokawa P, Itani K, El-Tamer M, Henderson WG, Khuri SF.
velocity in relation to air flow patterns produced by different Multivariable predictors of postoperative surgical site infection after general
ventilation systems, and their impact on outcome data are needed. and vascular surgery: results from the Patient Safety in Surgery study. J Am Coll
Surg 2007;204:1178-87.
Standardizing an optimal air sampling method would produce 7. Zhan C, Kaczmarek R, Loyo-Berrios N, Sangl J, Bright RA. Incidence and short-
reliable data and facilitate comparisons between studies to provide term outcomes of primary and revision hip replacement in the United States.
insight into the protective capacity of different ventilation systems J Bone Joint Surg Am 2007;89:526-33.
8. Lindstrom D, Sadr Azodi O, Wladis A, Tonnesen H, Linder S, Nasell H, et al.
during operations. Upward- displacement ventilation systems have
Effects of a perioperative smoking cessation intervention on postoperative
been demonstrated to more effectively remove particles compared complications: a randomized trial. Ann Surg 2008;248:739-45.
with conventional systems.45,46 An experimental study47 comparing 9. Thomsen T, Tonnesen H, Moller AM. Effect of preoperative smoking cessation
interventions on postoperative complications and smoking cessation. Br J Surg
conventional ventilation and upward-displacement ventilation
2009;96:451-61.
confirmed that the upward displacement system was more efficient 10. Pull ter Gunne AF, Cohen DB. Incidence, prevalence, and analysis of risk factors
in removing particles <10 mm, but found no difference between the for surgical site infection following adult spinal surgery. Spine (Phila Pa 1976)
2 ventilation systems for particles >10 mm.47 More importantly, the 2009;34:1422-8.
11. Bierbaum BE, Callaghan JJ, Galante JO, Rubash HE, Tooms RE, Welch RB. An
bacterial air counts were generally higher in the displacement analysis of blood management in patients having a total hip or knee arthro-
systems than in conventional systems. Considering this in the plasty. J Bone Joint Surg Am 1999;81:2-10.
present study, it is possible that the displacement system produced 12. Culver DH, Horan TC, Gaynes RP, Martone WJ, Jarvis WR, Emori TG, et al.
Surgical wound infection rates by wound class, operative procedure, and
higher mean CFU/m3 values than what would have been registered patient risk index: National Nosocomial Infections Surveillance System. Am J
under the same circumstances in a conventionally ventilated OR. Med 1991;91:152S-7S.
However, the present study demonstrates that even in a displace- 13. Stocks GW, Self SD, Thompson B, Adame XA, O’Connor DP. Predicting bacterial
populations based on airborne particulates: a study performed in nonlaminar
ment ventilated OR, very low CFU/m3 levels could be obtained by flow operating rooms during joint arthroplasty surgery. Am J Infect Control
keeping the doors closed and reducing the number of people 2010;38:199-204.
present. 14. Innerhofer P, Klingler A, Klimmer C, Fries D, Nussbaumer W. Risk for post-
operative infection after transfusion of white blood cellefiltered allogeneic or
Even structured observations can be susceptible to bias.48
autologous blood components in orthopedic patients undergoing primary
Human perceptual errors can affect the obtained information, as arthroplasty. Transfusion 2005;45:103-10.
well as behavioral distortions, due to the presence of an observer. 15. Willis-Owen CA, Konyves A, Martin DK. Factors affecting the incidence of
infection in hip and knee replacement: an analysis of 5277 cases. J Bone Joint
Several measures were taken to address potential bias: (1) The
Surg Br 2010;92:1128-33.
observational form was pretested and modified, (2) the observer 16. Gruenberg MF, Campaner GL, Sola CA, Ortolan EG. Ultraclean air for prevention
had no previous connection with the ward under observation, and of postoperative infection after posterior spinal fusion with instrumentation:
(3) the observer underwent self-training sessions to maximize a comparison between surgeries performed with and without a vertical
exponential filtered air-flow system. Spine (Phila Pa 1976) 2004;29:2330-4.
accuracy. Concealed observations to reduce reactivity were not 17. Hansen D, Krabs C, Benner D, Brauksiepe A, Popp W. Laminar air flow provides
feasible and were considered a possible source of distrust between high air quality in the operating field even during real operating conditions,
 A.E. Andersson et al. / American Journal of Infection Control 40 (2012) 750-5 755

but personal protection seems to be necessary in operations with tissue 33. Scaltriti S, Cencetti S, Rovesti S, Marchesi I, Bargellini A, Borella P. Risk factors
combustion. Int J Hygiene Environ Health 2005;208:455-60. for particulate and microbial contamination of air in operating theatres. J Hosp
18. Lidwell OM, Lowbury EJL, Whyte W. Effect of ultraclean air in operating rooms Infect 2007;66:320-6.
on deep sepsis in the joint after total hip or knee replacement: a randomised 34. Silverman D. Interpreting qualitative data: methods for analysing talk, text and
study. BMJ 1982;285:10-4. interaction. London: Sage; 2001.
19. McCann MT, Gilmore BF, Gorman SP. Staphylococcus epidermidis device-related 35. World Medical Association. Declaration of Helsinki: ethical principles for
infections: pathogenesis and clinical management. J Pharm Pharmacol 2008; medical research involving human subjects. J Postgrad Med 2002;48:
60:1551-71. 206-8.
20. Lidwell OM, Lowbury EJL, Whyte W. Airborne contamination of wounds in 36. Bachoura A, Guitton TG, Smith RM, Vrahas MS, Zurakowski D, Ring D. Infirmity
joint replacement operations: the relationship to sepsis rates. J Hosp Infect and injury complexity are risk factors for surgical-site infection after operative
1983;4:111-31. fracture care. Clin Orthop Relat Res 2011;469:2621-30.
21. Chow TT, Yang XY. Ventilation performance in operating theatres against 37. Parikh SN, Grice SS, Schnell BM, Salisbury SR. Operating room traffic: is there
airborne infection: review of research activities and practical guidance. J Hosp any role of monitoring it? J Pediatr Orthop 2010;30:617-23.
Infect 2004;56:85-92. 38. Lynch RJ, Englesbe MJ, Sturm L, Bitar A, Budhiraj K, Kolla S, et al. Measurement
22. Whyte W, Hodgson R, Tinkler J. The importance of airborne bacterial of foot traffic in the operating foom: implications for infection control. Am J
contamination of wounds. J Hosp Infect 1982;3:123-35. Med Qual 2009;24:45-52.
23. Tammelin A, Domicel P, Hambraeus A, Ståhle E. Dispersal of methicillin- 39. Young RS, O’Regan DJ. Cardiac surgical theatre traffic: time for traffic-calming
resistant Staphylococcus epidermidis by staff in an operating suite for thoracic measures? Interact Cardiovasc Thorac Surg 2010;10:526-9.
and cardiovascular surgery: relation to skin carriage and clothing. J Hosp Infect 40. Campbell DA Jr, Henderson WG, Englesbe MJ, Hall BL, O’Reilly M, Bratzler D,
2000;44:119-26. et al. Surgical site infection prevention: the importance of operative duration
24. Edmiston JCE, Seabrook GR, Cambria RA, Brown KR, Lewis BD, Sommers JR, and blood transfusion. Results of the first American College of Surgeons
et al. Molecular epidemiology of microbial contamination in the operating National Surgical Quality Improvement Program Best Practices Initiative. J Am
room environment: is there a risk for infection? Surgery 2005;138:573-82. Coll Surg 2008;207:810-20.
25. Tammelin A, Hambreaus A, Ståhle E. Source and route of methicillin-resistant 41. Whyte W, Niven L. Airborne bacteria sampling: the effect of dehydration and
Staphylococcus epidermidis transmitted to the surgical wound during cardio- sampling time. J Parenter Sci Technol 1986;40:182-8.
thoracic surgery: possibility of preventing wound contamination by use of 42. Godish DR, Godish TJ. Relationship between sampling duration and concen-
special scrub suits. J Hosp Infect 2001;47:266-76. tration of culturable airborne mould and bacteria on selected culture media.
26. Whyte W. The role of clothing and drapes in the operating room. J Hosp Infect J Appl Microbiol 2007;102:1479-84.
1988;11(Suppl C):2-17. 43. Parks SR. An assessment of the Sartorius MD8 microbiological air sampler.
27. Ljungqvist B, Reinmüller B. Aseptic production, gowning systems, and airborne J Appl Bacteriol 1996;80:529-34.
contaminants. Pharm Technol 2005;29(5 Suppl):S30-4. 44. Rupp ME, Archer GL. Coagulase-negative staphylococci: pathogens associated
28. Letts RM, Doermer E. Conversation in the operating theater as a cause of with medical progress. Clin Infect Dis 1994;19:231-45.
airborne bacterial contamination. J Bone Joint Surg Am 1983;65:357-62. 45. Memarzadeh F, Manning AP. Comparison of operating room ventilation
29. Kurmann A, Peter M, Tschan F, Mühlemann K, Candinas D, Beldi G. Adverse effect of systems in the protection of the surgical site. ASHRAE Trans 2002;
noise in the operating theatre on surgical-site infection. Br J Surg 2011;98:1021-5. 108(Part 2):3-15.
30. Shaw BH, Whyte W. Air movement through doorways: the influence of 46. Breum NO. Air-exchange efficiency of displacement ventilation in printing
temperature and its control by forced airflow. Build Serv Eng 1974;42:210-8. plant. Ann Occup Hyg 1988;32:481-8.
31. Wilson DJ, Kiel DE. Gravity-driven counterflow through an open door in 47. Friberg BF, Burman S, Lundholm LG, Östersson R. Inefficiency of
a sealed room. Build Environ 1990;25:379-88. upward-displacement operating theatre ventilation. J Hosp Infect 1996;33:
32. Ritter MA, Eitzen H, French ML, Hart JB. The operating room environment 263-72.
as affected by people and the surgical face mask. Clin Orthop Relat Res 1975;111: 48. Polit DF, Beck TC. Nursing research: principles and methods. 7th ed. Phila-
147-50. delphia [PA]: Lippincott Williams & Wilkins; 2004.