In South Africa, primary health care (PHC) facilities are the entry point for diagnosis, treatment and management of tuberculosis (TB) and multidrug-resistant TB (MDR TB). With a TB incidence of 834 cases per 100 000 population, a 61% HIV co-infection rate¹ and 18.8% HIV prevalence among adults aged 15–49 years,² the health care system is overwhelmed in dealing with the co-epidemic. As TB transmission frequently occurs before an accurate diagnosis is made, appropriate infection control is essential.^3,4 Virtually every country in the world – regardless of their TB incidence – has reported the spread of TB in health care settings to both patients and health care workers (HCWs).⁵ The responsibility rests with HCWs, particularly managers, to ensure the implementation of appropriate TB prevention and infection control measures in health care facilities.^3,4 In support of this, major international and national agencies have issued guidelines for minimising TB exposure and subsequent infection in health care facilities.^6,7

World Health Organization (WHO)^7,8,9 recommends four levels of TB infection control – an overarching managerial and facility-level encompassing administrative and environmental controls, and personal respiratory protection. While WHO guidelines have permeated local policy content,^6,10 South Africa has a record of developing progressive policies and programmes only to see them failing to trickle down to the service delivery level.^11,12,13 Over the past 6 years, several studies have highlighted the poor implementation of TB infection control measures and practices at health facilities in South Africa.^{14,15,16,17,18,19,20,21,22,23,24,25} However, this growing body of research has mainly focused on hospital settings, with a limited number of studies undertaken at PHC facilities. Although limited, available research does suggest a pressing need for evidence and interventions to prevent TB transmission in PHC settings^19,21,25 – including foci on policy content and implementation, TB prevention and infection control strategies, and knowledge and behaviour of HCWs and patients regarding prevention. This article aims to describe the state of implementation of TB infection control measures in a high-burden metro in South Africa. More specifically, the TB incidence, 686 per 100 000 population, and HIV prevalence among 15–49-year-old women attending antenatal care, 30.4%, was higher in the metro than the country as a whole (593 per 100 000 population and 29.7% respectively).^26,27

Prevention is a key element in the strategy to end the global TB epidemic¹ and an important component of prevention is infection control. However, this is largely neglected in South Africa despite the availability of policy and guidelines.³¹ Similar to other studies,^{14,15,16,17,18,19,20,21,22,23,24,25} we found poor adherence to TB infection control measures. Based on self-reports by the nurses, 16.5% of the facilities scored 80% and above on compliance with implementation of TB infection control measures. Actual observations were even more reason for concern, with only 3% of the facilities scoring 80% and above for implementation of TB infection control measures. This implies that HCWs and patients are largely unprotected from the spread of TB in PHC facilities, despite evidence⁵ that well-implemented infection control measures can reduce the spread of TB in health care facilities. This poses a serious risk for the PHC system in general and TB management and control in particular.

More specifically, at facility level, we found a lack of clinic infection control committees^19,21 and plans,^3,19 as well as inadequate support from sub-district, district and provincial managers for TB infection control. This is not surprising given the low rate of PHC supervisory visits in the metro.³² Administrative and environmental controls were also poorly implemented. For example, coughing patients were not separated from non-coughing patients^3,19,21 and coughing patients were not provided with masks.^21,25 Furthermore, windows were generally kept closed in TB consultation rooms and waiting areas.^19,21,25 Poorly implemented administrative and environmental controls force HCWs to be responsible for their own safety by using personal protective equipment. However, the use of disposable respirators was also sporadic,^19,21 and in most instances fit testing for disposable respirators was not done, although it ensures better protection if the facial seal is correctly fitted.³³ As noted by Zelnick et al.,²² the use of personal protection that rests on individual discretion and behaviour is unlikely to be effective when there is stigma associated with the disease and unemployment rates are high – as in the case of South Africa.

The main barriers preventing the implementation of appropriate TB infection control measures were the layout of the clinics which did not allow for the separation of coughing patients, poor ventilation and the lack of basic resources such as masks and disposable respirators. It was recommended that TB infection control could be improved if facilities were renovated to allow for designated TB consultation rooms and separate waiting areas for coughing patients. However, physical changes to facilities are costly, and much of public health care infrastructure is run down and dysfunctional because of funding shortages, mismanagement and neglect.³⁴ In order for PHC HCWs to effectively implement TB infection control, all health system components are required to function well: governance and stewardship, financing, infrastructure, procurement and supply chain management, human resources, health information systems, service delivery and finally supervision.³⁵

Keeping with the hierarchy of TB infection control, simple, cost-effective measures should be prioritised when implementing TB infection control practices in PHC facilities. In this regard, administrative control measures are the first line of defence and should be prioritised, as the aim is to prevent staff and patients from being exposed to TB through rapid diagnostic investigation and treatment of presumptive TB. Environmental control measures form the second line of defence and complement administrative measures. At the lowest level of the hierarchy, it is recommended that personal protective equipment should be used together with administrative and environmental controls in situations where there is an increased risk of TB transmission.^7,8

Therefore, the administrative level, as the first line of defence, should be prioritised. All PHC staff should be trained on cough etiquette and tasked with distributing masks and tissues to coughing patients. Facility managers and TB nurses require guidance on how best to use the space available at clinics to separate coughing and non-coughing patients. With reference to environmental controls, the second line of defence, TB nurses need training on airflow, and how to position themselves in the TB consultation room. All PHC staff and patients must be constantly reminded of the importance of natural ventilation to prevent TB transmission. This could be done through health education, ‘open window’ and ‘open door’ stickers, posters, etc. With regard to the use of personal protective equipment, PHC staff require information and training on disposable respirators, including fit testing of respirators and when to use and discard them. Finally, at the overarching managerial level, active involvement of provincial and district managers in supporting PHC facility managers and TB nurses is recommended. Integral to such support is regular supervisory visits and TB infection control assessments at facilities, individualised feedback and support to facilities regarding the state of their infection control implementation, and where necessary training and mentoring for all PHC facility staff on TB infection control.

Our study adds value to the growing body of evidence on TB infection control in low- and middle-income countries, given that South Africa, along with the other 22 high-burdened HIV and TB countries, does not systematically collect data on TB infection control implementation at a national level.³⁶ Furthermore, while there are many studies investigating the implementation of TB infection control measures in health care facilities, the majority of these studies focus on hospital settings. Our study focused on the PHC setting, as it is the main entry point for screening, diagnosis and care of TB patients in South Africa. In addition, we did not only rely on self-reported implementation of TB infection control methods as respondents tend to provide socially desirable responses³⁷ but also undertook observations at the facilities to verify reports by TB nurses.

However, as with all research, ours has limitations. The data were cross-sectional in nature, which means that we cannot infer strong causal claims. By only interviewing TB nurses, and not all PHC nurses, about infection control in their facilities we may have missed information. We did attempt to counter this by observing certain infection control practices at the facilities. Furthermore, it was beyond the scope of our study to explore in-depth reasons as to why HCWs did not implement appropriate infection control measures: qualitative enquiry and explanatory models are two possibilities for further work in this regard.