1 INTRODUCTION
6.1 CONTAINMENT EFFECTIVENESS OF AIIRS
Ventilation performance within an AIIR and negative pressure between the AIIR and adjacent areas are identified as key factors in the containment effectiveness of an AIIR (e.g., Li et al. 2007a, Hyttinen et al. 2011), yet the minimum requirements are unclear. Moreover, consistent ventilation and negative pressure recommendations are not agreed for Finland. This study considers effective control methods for limiting the spread of infectious agent outside an AIIR, and the protection of a healthcare worker is also discussed.
The results show that infectious agents spread outside an AIIR when a person moves between the AIIR and corridor despite the high air change rates in the AIIRs (ACH 4−24 1/h) and anterooms (ACH 6−11 1/h), and mean negative pressures (-9.2 ± 7.9 Pa). However, higher negative pressure between the AIIR and corridor was strongly associated with smaller contaminant transmission outside an AIIR.
The airtightness of the enclosure has an important role in the control of pressure differences, which is reflected in the results as there is a higher negative pressure among AIIRs with better tightness. The typical leaking points in the enclosure structures were penetrations, doorways and window structures. Similarly, Saravia et al. (2007) recognized the importance of airtightness among 678 AIIRs assessed in USA: the AIIRs with solid ceilings had a mean pressure difference of -4.4 Pa, whereas in AIIRs with drop ceilings the mean pressure difference of -2.0 Pa was significantly lower. Overall, 32% of AIIRs measured had the recommended pressure difference of -2.5 Pa by CDC (2005) (Saravia et al. 2007). Higher negative pressures were observed in this research, as the pressure difference between an AIIR and corridor was ≤-5 Pa in 79% and ≤-10 Pa in half of the AIIRs studied. This can be explained by the fact the Finnish design values for the negative pressure are higher than the guidelines of ∆p in USA.
Similar air change rates for AIIRs have been published by Rydock et al. (2004) and NIST (2006), who reported 5−21 1/hand 9−22 1/h, respectively. Li et al. (2007b) observed higher ACHs with a mean of 20 ± 7.3 1/h. The improved ventilation performance compared to the present and other studies is reasonable since Li et al.
(2007b) tested AIIRs that were newly built which presented state-of-the-art technologies. Nevertheless, the requirement for an air change rate of 12 1/h set in Nordic countries (NIPH 2004, SSI 2010, SFVH 2016) was followed only in 21% of the studied AIIRs. Previous international studies support the results of this thesis in that not all AIIRs follow the recommendation of 12 1/h: 10−62% were found to violate that (Sutton et al. 1998, Rydock et al. 2004, NIST 2006, Li et al. 2007b, Saravia et al. 2007).
The lower requirement of 6 1/h for older AIIRs (CDC 2005, Fusco et al. 2009) was
72
complied with in 79% of the AIIRs tested in this thesis. Similarly, Sutton et al. (1998) found that the minimum ACH criteria of 6 1/h was met in 83% of the AIIRs. In the present study, the lowest ACHs were measured in the oldest AIIRs (those constructed in 2002 or earlier).
The findings show that a high air change rate of an AIIR does not ensure an efficient contaminant removal from the healthcare worker’s working area around patient bed. This finding is in agreement with the studies by Pantelic & Tham (2013), and Mousavi & Grosskopf (2015) that HCW’s exposure to particles might increase by increasing the air change rate. The higher ACH can generate local airflow patterns that increase the particle concentrations at the HCW’s breathing zone (Pantelic &
Tham 2013, and Mousavi & Grosskopf 2015). In addition, Sutton et al. (1998) indicated that despite most of the AIIRs meeting the ACH recommendation, contaminants moved towards the HCW rather than away from them. Although a higher ACH is more effective in diluting the contaminants, the present study supports findings by Melikov et al. (2010b), Memarzadeh & Xu (2011), Pantelic &
Tham (2013) and Mousavi & Grosskopf (2015) that air distribution and local air patterns within the AIIR are more important than the ACH alone in particle containment.
Results show an ineffective local removal of contaminants in the HCW’s working area in AIIRs where exhaust vents are located in the ceiling opposite the patient’s head site. The highest exposure of the HCW to contaminants presented in a situation where the exhaust was located above the HCW’s working area. This resulted in tracer gas moving from the patient site across the HCW into the exhaust. It is acknowledged that the thermal plume of an actual patient may be different from the simulated plume in the present study. Here, a heated pillow was used, thus different airflow patterns could have been generated compared to a patient. Nevertheless, contaminant removal efficiency at the HCW’s working area was the most efficient when the exhaust vents were positioned above the patient head in the ceiling.
Efficient tracer gas removal near its generation point (i.e., the patient) minimized the contaminant movement towards the HCW’s working area. This finding is supported by the study of Thatiparti et al. (2017), where the particle flow path from patient to ceiling exhaust controlled the particles. In addition, contaminant removal near the patient’s head is in agreement with the studies by Cheong & Phua (2006) and Kao &
Yang (2006), although they stated that extraction behind the patient head on the wall achieves more effective contaminant removal compared to ceiling level since it produces a uniform flow path with minimal air mixing within an AIIR. Nonetheless, as supported by the present thesis, the ventilation design should be based on the introducing clean air to breathing zone of the healthcare worker.
The control of airflows in doorways is difficult during door openings. Tracer gas measurements indicate that particles are able to transit outside an AIIR via airflows during door traffic events. However, leakage into the corridor was generally below 5% and below 1% in 38% of the AIIRs studied. Leung et al. (2013) made somewhat lower estimations, evaluating that 0.01% of the airborne infectious particles transfer
73 from AIIR to corridor through human movement. When they released benign bacteria into an anteroom, 2.7% of the particles were transferred to the corridor by door traffic (Leung et al. 2013). However, the particle migration results by Leung et al. (2013) are most likely underestimated, since the background particles could not be distinguished from the tracer particles. In addition, different room configurations, pressure differences, air change rates, air distribution and other such variables may explain the disparity between the results.
In the present study, the location of supply air vents most likely explain the higher tracer leakages into corridor in some AIIRs. If supply air vents are located near a door, air jets may influence airflows in doorways by enhancing the migration of the contaminated air outside the AIIR. In addition, the fluctuation in pressure differences during door openings can contribute to particle transmission outside an AIIR (Hayden et al. 1998, Tang et al. 2005). According to CDC (2005) and Adams et al.
(2011), negative pressure ≤ -2.5 Pa enhances the particle containment against the fluctuation in pressure differences by e.g., door openings. International limit values for negative pressure vary between -2.5 and -15 Pa (Table 4). The results show that the tracer leakage was below 3% with negative pressure ≤ -10 Pa. On the other hand, particle escape outside an AIIR during door traffic have been observed even with limit values of -15 Pa (Rydock & Eian 2004). Nevertheless, tracer leakage results indicate that higher negative pressure provides better containment since it is associated with a significantly lower contaminant escape outside an AIIR.
Furthermore, according to findings of tracer transfer, an anteroom reduces the transmission of contaminants into corridor.
Although the ACH alone does not predict the containment effectiveness within an AIIR, the negative association between the air change rate and tracer leakage into a corridor suggests that an increased air change rate will limit the transmission of infectious agents through dilution. In particular, to enhance enclosure containment, more attention should be focused on the minimum ACH requirement of an anteroom when entering the anteroom instead of the ACH of the patient room. In theory, the air change rate of an anteroom should be higher than those observed in this thesis (mean 21 min for 90% removal, 41 min for 99% removal) to obtain rapid dilution and enough safety time (CDC 2003) before leaving the AIIR after an aerosol-generating procedure. Generally, a HCW does not stay in the anteroom more than 2–3 min, thus to achieve at least 90% contaminant removal, ≥40 ACH would be needed after entering the anteroom. It is acknowledged that this calculation is based on an assumption of perfect mixing within the space. However, perfect mixing may not occur, and removal times may be longer than those calculated. Overall, air change rates measured in anterooms (5.7−11 1/h) were somewhat lower than those reported in other studies: e.g. 10−47 1/h by Rydock et al. (2004) and 9.6−16 1/h by NIST (2006).
The important role of an anteroom in particle containment must be emphasized, considering the fact that a patient may move around the patient room, releasing, for example, cough particles around the AIIR. In this situation, even well-designed ventilation that provides a protective air distribution for the healthcare worker
74
around patient bed and efficiently removes contaminants from the (assumed) lying patient may not prevent the contaminant transmission outside the patient room.
Thus, an anteroom has a key role in enclosure containment effectiveness.
Finally, despite the observed spread of contaminants outside an AIIR, the infection risk in the surrounding areas is very low, since estimations of an infection risk within an AIIR have varied from 3−5% with 6 ACH (WHO 2009a), 1−3% with 12 ACH (WHO 2009a) and 0.3% with 24 ACH (Knibbs et al. 2011). Mean air volume migration through a doorway during person traffic may vary between 0.5-1.8 m3 for hinged doors and sliding doors (Hayden et al. 1998, Kalliomäki et al. 2016). In the present study, less than 10% of the tracer gas released in AIIRs transmitted into the corridor during door traffic. However, tracer gas concentrations measured outside AIIRs may be higher than the actual spread of infectious agents since tracer gas may not behave in the same way that the largest infectious particles tend to. Particles larger than 2.5 µm (such as cough particles) may not remain airborne as long as tracer gas due to gravitational settling and surface deposition (Gao & Niu 2007) as tracer gas is less influenced by deposition mechanisms. However, tracer gas simulates well behavior of the fine particles, in the size range typical of infectious particles (Gao et al. 2008, Noakes et al. 2009). Therefore, this study gave valuable information about particle dispersion from enclosed spaces. Tracer gas technique may be disturbed by contaminants present in hand disinfection agents used in wards. Nevertheless, efforts were made to take background concentrations into account in the results.
Importantly, to prevent the transmission of infectious agents in health-care settings, standard precautions have a key role (CDC 2007), including performance of good hand hygiene and use of PPE during nursing tasks.