Evaluation of compliance with ventilation guidelines for airborne infectious disease control in emergency centres in Cape Town, South Africa



C Groenewald1,2, J Meintjes3, BM Morrow4

1 Kimberley Hospital Complex, South Africa 
2 Department of Emergency Medicine, University of Cape Town, South Africa
3 Unit for Infection Prevention and Control, University of Stellenbosch, South Africa
4 Department of Paediatrics and Child Health, University of Cape Town, South Africa

Correspondence: Prof. Brenda Morrow, 5th Floor ICH Building, Red Cross War Memorial Children’s Hospital, Klipfontein Road, Rondebosch, 7700, South Africa. e-mail: brenda.morrow@uct.ac.za      


Background: Pulmonary tuberculosis (TB) is a major healthcare problem worldwide and is endemic to South Africa. 
Healthcare workers and patients in emergency centres (ECs) in Cape Town may be at risk of acquiring TB.

Objective: To determine whether the isolation rooms used for patients with diagnosed or suspected TB comply with set National Core Standards for airborne infection control.

Methods: A cross-sectional descriptive study was conducted in 19 isolation rooms of the ECs of eight public-sector hospitals in the Cape Town Metropole region, from March to June 2015. The characteristics of the isolation rooms were evaluated, adhering to current engineering standards of measurement. Outcomes with regard to the number of air changes per hour (ACH), negative pressure ventilation, and appropriate discharge of air outdoors, are presented.

Results: None of the 19 isolation rooms complied with the National Core Standards’ ideal requirements. Eleven (57.9%) were designed as isolation facilities; however, only five (26.3%) were under negative pressure, and seven (36.8%) had positive airflow into the adjacent clinical areas. Airflow in naturally-ventilated isolation rooms was significantly lower than airflow in rooms using other ventilation systems (p = 0.004). The number of ACH was 2-3 times higher in rooms ventilated with mixed systems, compared to those with natural ventilation or centrally provision-type ventilation only (p = 0.007).

Conclusion: Adherence to the recommended guidelines of airborne infectious disease control is poor among public-sector hospital ECs in the Cape Town Metropole region, posing a significant risk to staff and patients of acquiring nosocomial airborne infectious diseases. Improving airborne infection control measures in ECs should be a priority disease-prevention measure.


Keywords: tuberculosis, infection control, nosocomial infection 


Tuberculosis (TB) is a major global healthcare problem,1 with a high prevalence, reaching 32 per 1 000 in the Western Cape province of South Africa.2-6 

Airborne infection control strategies to prevent Mycobacterium tuberculosis transmission have long been a neglected component of TB control programmes, with grave consequences, such as failures of programmes to limit the spread of disease, worldwide.7The World Health Organization (WHO) strongly recommends installing and using ventilation systems in healthcare facilities, including emergency centres (ECs).8,9

There is evidence that a large proportion of new TB cases are due to nosocomial contraction of M. tuberculosis; there are valid concerns about the high incidence of TB in healthcare workers in South Africa and other countries.10-15 South African healthcare workers are 5.5-6.5 times more likely than the general population to be admitted for drug-resistant TB, which may be due to occupational exposure.11 Therefore, TB infection control measures should be prioritised in hospitals and emergency departments, especially in high-prevalence settings.9

In South Africa, the Regulations for Hazardous Biological Agents provide detailed guidance for the protection of employees against infective agents.16 The need for occupational health and infection control programmes is also mandated by the National Department of Health for all healthcare facilities, as described in the National Core Standards for Health Establishments of South Africa document.17 Poor or absent infection control in healthcare facilities is contrary to national standards and legislation.

According to South African regulations, a patient with suspected or confirmed TB should be placed in a private room, ideally with: 1) continuously monitored negative air pressure in relation to surrounding areas; 2) a minimum of 6-12 air changes per hour; and 3) appropriate discharge of contaminated air to safe areas, or a monitored high-efficiency filtration of air before circulation. If these conditions cannot be met, then the patient should be placed in a room with a simple extraction fan, allowing at least six air changes per hour (ACH) and/or a room with an open window and adequate ventilation.16

Despite the clear importance and identified need, the adherence to the WHO and national standards for ventilation aspects of airborne infection control has not been studied in ECs in South Africa. This study aimed to evaluate the compliance of EC isolation rooms with all the aforementioned prescribed engineering standards for airborne infection control, in the Cape Town Metropole region of South Africa.



A cross-sectional descriptive study was conducted from March to June 2015. All 10 ECs with specialist emergency physicians, located in two tertiary, three regional or five district hospitals within the Cape Town Metropole, representing all public-sector centres with regular emergency physician cover, were eligible for inclusion.

Standardised data collection forms, developed from WHO recommendations and the National Department of Health’s Core Standards for Health Establishments in South Africa (2011),8,17 were used to evaluate compliance with the stipulated engineering control requirements for airborne infection control. A room needed to have the following characteristics to be categorised as meeting ideal requirements: 1) monitored negative air pressure in relation to surrounding areas; 2) 6-12 ACH; and 3) appropriate discharge of air outdoors or monitored high-efficiency filtration. A room would be deemed as meeting minimum requirements if it had a simple extraction fan, providing at least six ACH, or an open window with adequate ventilation (defined as six ACH). Centres were considered fully compliant if they met the ideal requirements.

Approval for the study was obtained from the Institutional Human Research Ethics Committee (HREC REF 305/2015); hospital superintendents of the participating facilities consented to the evaluation of their ECs.



The technique described in Annexure B of the Draft National Infection Prevention and Control Policy for TB in South Africa18 was used to calculate ACH. The process involved measuring the room diameter with a tape measure, excluding closed fixed furniture in the room, and calculating the volume of the room. A validated and factory-calibrated anemometer (Lutron Anemometer Model AM-4203) was used to calculate the velocity of airflow over each area that contributed to the airflow of the room. The average flow rate over one minute was calculated for each ventilation type. If natural ventilation was used in the room, a measurement was taken over 10 minutes, using averaged values for one-minute interval recordings. Each room was measured once only to evaluate the real-time situation of that room, which reflected any time a patient might be placed in the room.

The number of ACH in each room was calculated using standard methods, as per the example in Box 1.19 Airflow visualisation was done using the dry ice method.18 All measurements were taken by a single operator who received formal training from an expert. The operator’s measurements were checked by staff working in the EC at the time of evaluation.

Box 1. Example of ACH calculation


Statistical analysis

Data collected at each facility were captured on a standardised data capture sheet and transcribed onto a Microsoft Excel spreadsheet by one of the investigators (CG). Accuracy of data entry was confirmed independently by a study supervisor. Continuous data were tested for normality using the Shapiro-Wilk W test. These data were found to be nonparametric and are therefore presented as medians (interquartile range, IQR) unless otherwise specified. Categorical data are presented as numbers and frequencies. Hypothesis testing was performed using the Chi-square or Fisher’s exact tests, as appropriate. Differences between multiple groups (ventilation systems) were determined using Kruskal-Wallis ANOVA by ranks, with post hoc Mann-Whitney U tests in the event of significant between-groups differences. A significance level of 0.05 was used for all hypothesis testing.

Data were analysed using both the PHStat add-in (version 2.5) for Microsoft Excel, for descriptive statistics, and Stata/IC (StataCorp) for Windows, version 12.1, for inferential statistics.



Participating facilities and room characteristics

Two of the eligible ECs were excluded, as superintendent consent was not obtained. In the eight ECs that participated, 19 rooms used for isolating patients were evaluated (1-5 rooms per EC) (Table 1).

Eleven (57.9%) of the 19 rooms were custom-designed to function as isolation rooms with negative pressure ventilation; eight (42.1%) were makeshift isolation rooms, having been built for different purposes (including fracture management and gynaecologicalexamination).


Ventilation systems

Air pressure with relation to surrounding areas

Five (26.3%) of the rooms evaluated were under negative pressure and seven (36.8%) had airflow that moved in both positive and negative directions during the evaluation (defined as erratic airflow). Seven (36.8%) rooms showed continuous positive airflow from the isolation room into the adjacent clinical area where health workers and patients were present (Table 1). Only three (27.3%) of the 11 rooms that were designed to function as isolation rooms had negative pressure airflow movement at time of evaluation, indicating possible problems with regard to maintenance and monitoring.


Types of ventilation systems

Six of the 19 (31.6%) rooms had centrally-controlled air supply (central provision) ventilation systems, with a median (range) flow rate of 1.6 (1-4.6) m/s. Four (21.1%) of the rooms made use of natural ventilation only; one (5.3%) made use solely of centrally-controlled air extraction, exhaust type ventilation system (central extraction); and eight (42.1%) had mixed ventilation systems, i.e. central extraction in addition to local (within-room controlled) extraction, or central extraction (n = 5) as well as natural ventilation (n = 3).

All of the locally-controlled extraction units were fitted to the window or wall of the room, thus directing air outside. Four locally-controlled extraction units had a zero flow rate: two were not switched on, and two were broken. Airflows through the three different ventilation systems were significantly different (p = 0.004), with natural ventilation systems having the lowest flow rates (Figure 1).

Figure 1. Airflow through different ventilation systems (Kruskal-Wallis ANOVA p = 0.004). Brackets and corresponding p values indicate post hoc Mann-Whitney U comparisons between natural ventilation and other systems


Air changes per hour (ACH)

The number of ACH in the isolation rooms ranged from zero (in four rooms) to 112.37 (median 11.9). The highest value was measured in a room with both central extraction and natural ventilation systems (mixed ventilation) and was evaluated on a particularly windy day. It is worth noting that this room was under positive pressure ventilation, with air moving into adjacent clinical areas. 

Table 1. Summary of ventilation standards in each isolation room


The venting of contaminated air to a safe area could be verified only in rooms where locally-controlled ventilation was provided (Table 1). There was an overall significant difference in the number of ACH amongst rooms with different ventilation systems (p = 0.007). The median (IQR) number of ACH was significantly higher in rooms where mixed ventilation systems were used (median 21.7; IQR 13.5-36.2) compared to natural ventilation (median 4.2; IQR 4.0-5.2); p = 0.01, and central provision systems (median 0; IQR 0-9.0); p = 0.007 (Figure 2).

Figure 2. Air changes per hour (ACH) in isolation rooms with different ventilation types (Kruskal-Wallis ANOVA p = 0.007). Brackets and corresponding p-values indicate significant differences between ventilation types, using post hoc Mann-Whitney U tests. The difference between central provision ventilation and natural ventilation was not significant (p > 0.1). The one room with central extraction only was excluded from this analysis



In all cases where air was centrally provided and extracted, staff members were not able to verify if fresh air was provided or if the air was filtered and re-circulated. No EC staff member was able to provide information on the last date on which the ventilation systems were serviced or maintained, if filtration systems were used and, if present, when these were last serviced. None of the rooms was equipped with a negative pressure monitor.


Regulatory requirements

None of the isolation rooms in any of the facilities met the ideal requirements as stipulated in the Regulations for Hazardous Biological Agents16 (Table 1) and, therefore all the rooms were noncompliant, in terms of the definition used in this study. Five (26.3%) rooms met the requirements of negative flow with adequate ACH, without implementing negative pressure ventilation monitoring; the same five met the defined minimum standards (three with mixed ventilation systems, one with central provision-type ventilation, and one with central extraction ventilation). Notably, 50% of the rooms that incorporated central extraction ventilation complied with minimum standards.



This study demonstrates that the ventilation aspects of airborne infectious disease control have been largely neglected in public-sector hospital ECs in Cape Town, despite the high prevalence of such diseases, especially TB. This neglect is in keeping with published literature.7, 9, 19 Airborne infectious diseases remain an occupational risk for healthcare workers and non-clinical workers in these areas. The risk is higher where infection control strategies are inadequate.20

None of the isolation rooms that were evaluated totally fulfilled the legislative and regulatory requirements, and only a handful complied with the minimum ventilatory requirements. It was concerning that no staff member in any EC could indicate when the last ventilation check was done, and did not know whom to contact regarding evaluation thereof, suggesting that the lack of compliance had been long-standing. Due to the lack of negative pressure monitors in all rooms, there was no way for staff to evaluate airflow, which further impacts on adherence to best practice. We did not determine the amount and level of training received in this regard, but we recommend that these factors be evaluated in future studies.

Despite the high prevalence of airborne infectious diseases in the Western Cape province of South Africa,2-6 very few emergency facilities had isolation rooms designed to manage patients with infectious diseases such as TB, which is not in keeping with international guidelines.23,24 In the absence of sufficient isolation facilities, it is likely that many patients with known or suspected TB and/or other infectious diseases, would wait and be managed in general areas within the EC, potentially exposing non-infected individuals. Even where isolation rooms were available, our results suggest that many high-risk patients are placed in rooms that are inadequate for isolation purposes.

It is especially concerning that, amongst facilities with isolation rooms specifically designed to provide negative pressure ventilation, some provided positive pressure ventilation. Rooms with positive pressures place staff, patients and visitors at high risk for infection when they are in the immediately adjacent areas. Importantly, it is unlikely that staff in these adjacent areas are aware of the infection risk, and would therefore be unlikely to use the necessary personal protective and infection control precautions.

Four (21.1%) of the 19 isolation rooms included in the study used natural ventilation only, which was not always associated with negative pressure, and has other drawbacks, such as unpredictable flow direction and the likely closure of windows during tumultuous weather conditions, which would further limit airflow.25,26 In addition, our results show poor airflow and low ACH with natural ventilation compared to other systems. By contrast, rooms with mixed systems, including central extraction ventilation, showed the highest number of ACH. Adequate airflow rates were noted for both centrally- and locally-controlled extraction ventilation systems.

The choice of environmental controls is closely related to building design, construction, renovation and use which, in turn, must be tailored to local climatic and socioeconomic conditions.28 Our results suggest that airborne infection control standards may be best achieved with mixed ventilation systems, and that natural ventilation should not be relied upon. However, owing to the small sample size we cannot make a definite recommendation in this regard.



This study was limited by the fact that measurements and observations were conducted on a single day for each EC, which may not have accurately reflected variation within the facilities over time. The findings cannot be generalised to all ECs in South Africa, which might differ by sector (private versus public) and province. In addition, observations were largely made when the weather was mild and sunny, which could have influenced natural ventilation measurements. The study was, however, strengthened by the use of standardised, objective and repeatable measurement techniques.



There is an urgent need to redress the lack of compliance with the engineering aspects of airborne infection control. A study with a larger sample size is recommended to assess which ventilation system is most effective in achieving the ideal outcomes. New ECs should consider local infectious disease prevalence when planning their buildings, to include appropriate ventilation systems to limit transmission of airborne infectious diseases. Existing facilities should look at ways to improve compliance with ideal airborne infection control measures. Installation of negative pressure monitors for isolation rooms is strongly advocated.

This study focused on ventilation in rooms in ECs where patients with known or suspected airborne infectious diseases are managed. Future studies should also address airborne infection control measures in the entire facility in which the EC is situated, to protect patients and staff from nosocomial transmission of pathogens from asymptomatic individuals. Future research should be conducted to evaluate healthcare workers’ knowledge of infection control for the spread of airborne disease.

We did not measure or record nosocomial infection rates within each centre. This would be useful to determine the compliance with airborne infection control measures and to report associated nosocomial infection rates of airborne pathogens such as TB.



The design, layout, implementation, maintenance and monitoring of airborne infection control is generally poorly implemented in ECs in the Cape Town Metropole region of South Africa. This may promote the nosocomial spread of airborne infectious diseases, including TB, and may be a contributing factor to the documented failure of existing TB infection control programmes.6 Urgent redress is required to optimise compliance to the ventilatory component of infection control practices in public-sector ECs.

• Mixed ventilation systems may offer the greatest number of room air changes per hour, with acceptable flow rates.
• Improving and maintaining ventilation systems in ECs might prevent transmission of infection amongst patients and staff.
• Natural ventilation does not provide adequate ventilation to control airborne spread of infectious particles safely.



The authors acknowledge the co-operation of emergency centre staff  in the eight centres. 



The authors declare no conflicts of interest.



The investigators did not receive funding support for this study.



1. World Health Organization. Global Tuberculosis Report. Geneva: WHO;2015. Available from: http://apps.who.int/iris/bitstream/handle/10665/191102/9789241565059_eng.pdf (accessed 24 Oct 2018).

2. Hermans S, Horsburgh Jr CR, Wood R. A century of tuberculosis epidemiology in the northern and southern hemisphere: the differential impact of control interventions. PLOS One. 2015; 10(8):e0135179.

3. Claassens M, Van Schalkwyk C, Den Haan L, Floyd S, Dunbar R, Van Helden P, et al. High prevalence of tuberculosis and insufficient case detection in two communities in the Western Cape, South Africa. PLOS One. 2013; 8(4):e58689.

4. Mortality and causes of death in South Africa, 2015: Findings from death notification. Pretoria: Stats SA; 2017. Available from: http://www.statssa.gov.za/publications/P03093/P030932015.pdf (accessed 15 Aug 2018).

5. Mayosi BM, Lawn JE, Van Niekerk A, Bradshaw D, Karim SSA, Coovadia HM, et al. Health in South Africa: changes and challenges since 2009. Lancet. 2012; 380(9858):2029-2043.

6. Wood R, Lawn SD, Johnstone-Robertson S, Bekker L-G. Tuberculosis control has failed in South Africa: time to reappraise strategy. S Afr Med J. 2011; 101(2):111-114.

7. Shenoi SV, Escombe AR, Friedland G. Transmission of drug-susceptible and drug-resistant tuberculosis and the critical importance of airborne infection control in the era of HIV infection and highly active antiretroviral therapy rollouts. Clin Infect Dis. 2010; 50 Suppl 3:S231-237.

8. WHO policy on TB infection control in health-care facilities, congregate settings and households. Geneva: WHO; 2009. Available from: http://apps.who.int/iris/bitstream/handle/10665/44148/9789241598323_eng.pdf?sequence=1 (accessed 15 Aug 2018).

9. Escombe AR, Huaroto L, Ticona E, Burgos M, Sanchez I, Carrasco L, et al. Tuberculosis transmission risk and infection control in a hospital emergency department in Lima, Peru. Int J Tuberc Lung Dis. 2010; 14(9):1120-1126.

10. Joshi R, Reingold AL, Menzies D, Pai M. Tuberculosis among health-care workers in low- and middle-income countries: a systematic review. PLOS Medicine. 2006; 3(12):e494.

11. O’Donnell MR, Jarand J, Loveday M, Padayatchi N, Zelnick J, Werner L, et al. High incidence of hospital admissions with multidrug-resistant and extensively drug-resistant tuberculosis among South African healthcare workers. Ann Int Med. 2010; 153(8):516-522.

12. Pai M, Gokhale K, Joshi R, Dogra S, Kalantri S, Mendiratta DK, et al. Mycobacterium tuberculosis infection in healthcare workers in rural India: comparison of a whole-blood interferon γ assay with tuberculin skin testing. J Am Med Assoc. 2005; 293(22):2746-2755.

13. Mirtskhulava V, Kempker R, Shields KL, Leonard MK, Tsertsvadze T, del Rio C, et al. Prevalence and risk factors for latent tuberculosis infection among health-care workers in the country of Georgia. Int J Tuberc Lung Dis. 2008; 12(5):513.

14. Adams S, Ehrlich R, Baatjies R, Van Zyl-Smit RN, Said-Hartley Q, Dawson R, et al. Incidence of occupational latent tuberculosis infection in South African healthcare workers. Eur Respir J. 2015; 45:1364-1373.

15. McCarthy K, Scott L, Gous N, Tellie M, Venter W, Stevens W, et al. High incidence of latent tuberculous infection among South African health workers: an urgent call for action. Int J Tuberc Lung Dis. 2015; 19(6):647-653.

16. South Africa. Department of Labour. 2011. Occupational Health and Safety Act, 1993 (Act No. 85 of 1993). Regulations for Biohazardous Agents, 2011. Government Gazette No. 22956:1390 27 Dec. Available from: http://www.labour.gov.za/DOL/downloads/legislation/regulations/occupational-health-and-safety/Regulation%20-%201390%20-%20OHS%20-%20Hazardous%20Biological%20agents.pdf (accessed 15 Aug 2018).

17. National core standard for health establishments in South Africa; 2015. Available from: http://phisa.co.za/National%20Core%20Standards.pdf (accessed 15 Aug 2018).

18. South African National Department of Health. The draft national infection prevention and control policy for TB, MDRTB and XDRTB; 2007. Available from: http://www.who.int/hiv/pub/guidelines/south_africa.pdf (accessed 24 Oct 2018).

19. American Society of Heating, Refrigerating and Air Conditioning Engineers Ventilation and acceptable indoor air quality in low-rise residential buildings. Atlanta: ASHRAE; 2013. Available from: http://shop.iccsafe.org/media/wysiwyg/material/8950P815-toc.pdf (accessed 24 Oct 2018).

20. Klingenberg S. Smoke studies: clearing the mystery of air flow visualization. Journal of Validation Technology. Summer 2010; 11-17.

21. Nicas M, Sprinson JE, Royce SE, Harrison RJ, Macher JM. Isolation rooms for tuberculosis control. Infect Control. 1993; 14(11):619-622.

22. Menzies D, Joshi R, Pai M. Risk of tuberculosis infection and disease associated with work in healthcare settings. Int J Tuberc Lung Dis. 2007; 11(6):593-605.

23. Reiling J. Safe design of healthcare facilities. Qual Saf Healthcare. 2006; 15(suppl 1):i34-i40.

24. Sehulster L, Chinn RY, Arduino M, Carpenter J, Donlan R, Ashford D, et al. Guidelines for environmental infection control in health-care facilities. Atlanta: Centres for Disease Control (CDC); 2003. Available from: https://www.cdc.gov/infectioncontrol/pdf/guidelines/environmental-guidelines.pdf (accessed 24 Oct 2018).

25. Allard F, editor. Natural ventilation in buildings: a design handbook; London: James and James (Science Publishers) Ltd.; 1998.

26. Atkinson J. Natural ventilation for infection control in health-care settings. Geneva: WHO; 2009. Available from: http://www.who.int/water_sanitation_health/publications/natural_ventilation.pdf (accessed 24 Oct 2018).

27. South Africa. Department of Health. 2013. National Health Amendment Act, 2013 (Act No. 12 of 2013). Available from: https://www.gov.za/sites/default/files/36702_gon529_0.pdf (accessed 2 Nov 2018).
28. World Health Organization. WHO policy on TB infection control in health-care facilities, congregate settings and households. Geneva: WHO; 2009. Available from: http://www.who.int/tb/publications/2009/9789241598323/en/ (accessed 24 Oct 2018).

Download this Article


Email address
Forgot password?