Reducing Nosocomial Infections in ORs
By Farhad Memarzadeh, PhD, PE and Andy Manning, PhD
Figure 1. Source and routes of infection in the operating room5
The National Institutes of Health (NIH) Office of Research Services, Division of Engineering Services, conducted an extensive study on the issue of operating room (OR) ventilation systems and their effect on the protection of the surgical site.
The risk of nosocomial infection is present in all surgical procedures, but can be particularly serious in certain operations, such as joint replacement. The many factors that influence infection are shown in Figure 1.
The standards suggested for air conditioning systems for ORs in different countries vary widely in content and recommendations and are often based on old research. For example, the ASHRAE Handbook1 is based on work originally carried out in the 1950s. As a point of reference, the ASHRAE Handbook suggests that "the delivery of air from the ceiling, with a downward movement to several exhaust inlets located on opposite walls, is probably the most effective air movement pattern for maintaining the concentration at an acceptable level."1
The handbook indicates that the temperature range should be between 62 degrees Fahrenheit (16.67 Celsius) and 80 degrees F (26.67 C), and that positive pressurization should be maintained. It also suggests that the air should be exhausted or returned from at least two locations near the floor. It suggests that supply diffusers should be of the unidirectional type and that high-induction ceiling or sidewall diffusers should be avoided. The recommended ACH is 15 ACH for systems that use all outdoor air and 25 ACH for re-circulating air systems.
Figure 2. Layout of baseline operating room --Mayo stand view
Some studies have been published which consider the relative merits of different systems, such as Lidwell2 and Schmidt.3 However, since they do not include specific system design data, it is difficult to establish definitive recommendations. Further, while laminar flow systems provided lower general concentration levels in the room, they are sometimes blamed for higher infection rates than more conventional systems, for example, Salvati et al.4 The theory put forward by Lewis5 is that laminar flow systems cause impingement on the wound site. However, with Schmidt defining a laminar system as having velocities of at least 90 fpm (0.45m/s), this probably is due to high supply velocities.
The aforementioned studies were experiment based; however, an alternative technique, Computational Fluid Dynamics (CFD), sometimes known as airflow modeling, has been proved to be very powerful and efficient in research projects involving parametric study on room airflow and contaminant dispersion (Ziang, et al.6 and Haghighat et al.7) In the study documented here, airflow modeling is used to consider the dispersion of squame-sized particles (which are recognized as the main transport mechanism of bacteria in operating rooms (Woods, et al.)8 under various ventilation system-design conditions. To establish the relative ranking of the different systems, two target areas of concern are considered: the surgical site and the top surface of the back table where instruments are located.
The CFD and particle tracking routine methodology are described in detail in Memarzadeh and Manning.9 (They are not included here because of space limitations. A copy of the paper is available at: http://des.od.nih.gov/eWeb/research/farhad/index.htm.)
Outline of Base Model
A typical OR layout was considered for the baseline model. The room measured 20 feet (6.1m) by 20 feet (6.1m) by 12 feet (3.66m) high; general features are given in Figure 2 and Table 1. The room's layout was agreed upon by a panel of physicians and engineers. The total heat dissipated in the room was 2166W.
Several ventilation systems were considered in this study and are shown in Table 2. These systems approximately replicate those outlined in Schmidt.3 Models of the various diffuser types considered in this project were all validated prior to the room parametric study against manufacturers data.
Figure 4. Flow Field in Case 1
Squames are cells which are released from exposed regions of the surgery staff and patient, for example, neck, face, etc., and are the primary transport mechanism for bacteria in the OR. They are approximately 25 microns (mm) by 3 to 5 microns thick. Approximately 1.15 by 106 to 0.9 to 108 squames are generated during a typical (two-hour to four-hour) procedure.10 In this study, the squame particles would be tracked to see how many hit the back table, shown in Figure 3, or the surgical site. The surgical site was taken as a 1-foot (0.3m) by 1-foot (0.3m) square with the surface temperature at 100 degrees F (37.78C), and is shown in Figure 3.
A representative number of particles was introduced from three arrays of sources in the CFD models. The locations and sizes of the sources, designated as Main, Nurse and Surgery, are shown in Table 3. As the particles could readily pass to the instruments at this point, the surgery source/top surface of back table target analysis was not performed in this study.
There are three potential particle outcomes:
- The particle exits the room via exhaust grilles
- The particle strikes the surgical site or top surface of back table.
- The particle remains in the room at the time where particle tracking is stopped (one hour)
- Percentage of particles removed by ventilation varying with time
Figure 5. Thermal Plume From Surgical Site in Case 2 (Laminar Design)
Table 4 shows there is a wide range of particle removal effectiveness. Cases that have the same ACH show marked differences in terms of the percentage of particles removed via ventilation. For example, Case 10 demonstrates a more effective removal of particles than Case 1.The reason is that the ventilation system in Case 1 results in the formation of two large re-circulations in the room where particles can become trapped; this is illustrated in Figure 4. In Case 10 the ventilation system works with the thermal plume in the center of the room in driving the particles up to the high level exhausts. Taking Cases 3, 4, 5, 6 and 9 as a group that adopts the same general approach to ventilation, the percentage vented from all locations becomes more uniform on increasing ACH and supply array size. The reason is, for the smaller laminar arrays, the areas outside the direct influence of the supply have very low velocity flow fields, and particles are more likely to drop to the floor via gravity.
Percentage of Particles, Which Hit Surgical Site or Top Surface of Back Table
Figure 6. Flow Field in Case 9
Table 5 shows that the percentages of particles which hit the surgical site from the main or nurse sites are low (less than 1 percent) because of the relative dominance of the thermal plume caused by the surgical site. For example, Figure 5 shows such a plume for Case 2. It is only when the particles are released close to the site, in particular, the surgery source, that the percentage becomes significant. It is also clear that ACH is not as significant in the surgery source/surgical site analysis as design of the ventilation system. In particular, a lower percentage of particles hit the site in Case 4, which has an ACH of 20, than Case 2, which has an ACH of 150.
The results also show, with the exception of Case 11, the percentage of particles that hit the back table from the main or nurse sites is relatively low. In Case 11, the particles are blown onto the table. The results for Cases 4, 5 and 6 indicate that a mixture of exhaust location levels is better than low or high only. Finally, the cases which can be placed together in a laminar flow type group, namely, Cases 2, 3, 4, 5, 6 and 9, show lower strike rates than the other systems. In fact, Cases 4 and 9 represent the lowest strike percentages of all the cases considered.
Conclusions and Discussion
Conclusions of the study are:
- Cases which have the same ACH show marked differences in terms of the percentage of particles removed via ventilation.
- The practice of increasing ACH to high levels results in excellent removal of particles via ventilation, but it does not necessarily mean that the percentage of particles which strike critical surfaces will continue to decrease.
- The percentages of particles which hit the surgical site from the Main or Nurse sites are low (less than 1 percent). This is because of the relative dominance of the thermal plume associated with the warm surgical site. Only when the particles are released close to the site, in particular, the surgery source, does the percentage become significant.
- ACH is not as significant as the design of the ventilation system. In particular, a lower percentage of particles hits the site in a case which has an ACH of 20, than one which has an ACH of 150.
- In a system which provides a laminar flow regime, a mixture of exhaust location levels works better than either low or high level locations only. However, the difference is not significant enough that the low- or high-level location systems are not viable options.
- Systems that provide laminar flow regimes represent the best option for an operating room in terms of contamination control; however, care needs to be taken in the sizing of the laminar flow array. A face velocity of around 30 to 35 fpm (0.15 to 0.18 m/s) is sufficient from the laminar diffuser array provided that the array size itself is set correctly.
To expand on the issue of diffuser array size, it appears that the main factor in the design of the ventilation system is the control of the central region of the OR. In particular, the operating lights and surgical staff represent a large heat density in the middle of the room. Particulates could get caught in buoyant plumes created by these heat-dissipating objects, at which point control of them is lost; however, if a laminar flow type system is employed, the particles are instead driven by the flow to the exhaust. Ideally then, the array size should be large enough to cover the main heat dissipating objects.
This is illustrated in Figure 6, which shows the flow field for Case 9.
Further, another factor is the thermal plume created by the surgical site, shown for Case 2 in Figure 5.A laminar flow regime which provides air at 30-35 fpm (0.15 to 0.18 m/s) ensures that particles are not impinged on the surgical site, a danger highlighted by Lewis5 as the thermal plume should be sufficient to protect the surgical site.
Farhad Memarzadeh, PhD, PE, is chief of technical resources in the Office of Research Services, Division of Engineering Services at the National Institutes of Health in Bethesda, Md. Andy Manning, PhD, is director of engineering for Flomerics, Inc. of Southboro, Mass.