Theme 1: Interactions between TB IPC training, risk perception and experiences of workplace safety.

Health workers flagged insufficient TB IPC training as an important contributor to poor TB IPC implementation [21, 26, 27, 30, 32–34, 41, 47–49, 52–55]. They described existing IPC training as targeted towards health workers working in TB wards, Directly Observed Therapy Short-course (DOTS) programmes, or at facility managers [32, 55]. In some instances, having TB training led to TB work being delegated to or reserved for these health workers [21, 48]. Gaps in training included uncertainty about which IPC measures are important, a lack of refresher training after graduation, and difficulties in training new staff when there is high employee turnover [32, 37, 50, 53, 56].

Self-perceived risk of developing TB contributed to the importance health workers attached to implementing TB IPC [25, 26, 29, 30, 32, 34, 38–42]. (Table 2, Key finding 1) They described being influenced by the experiences of colleagues who had developed TB [27]; for example, they recalled a medical student who developed drug-induced hepatitis due to TB medication and died [33]. Health workers who had witnessed the suffering of drug-resistant TB patients felt incapable of going through it themselves. One nurse at a drug-resistant TB treatment centre in South Africa mentioned, ‘All that medications and injections—I’d rather kill myself than go through what those poor patients go through’ [51]. A contributory fear of health workers was that they would infect family members [27, 28, 35, 45].

Health workers’ response to occupational risk fell into three groups. In the first, they perceived themselves not to be at risk of TB and did not value TB IPC training [29]. Misconceptions contributing to this included: the belief that latent TB infection would be protective against developing TB disease [29], and that BCG vaccination [53], ‘eating well’ before examining patients [35] or taking multivitamins [27] prevented TB infection. A second group believed they were at risk of TB, valued TB IPC training and felt equipped to care for TB patients [34]. A third group were aware of their occupational TB risk and received TB IPC training, but were faced with inadequate resources to implement the preventative practices that they knew were important [29, 36]. This is conceptualised as ‘powerlessness’ by Chapman and colleagues, when discussing their findings:

“If HCWs perceive that their occupational risk is high and beyond their control, irrespective of their use of infection control measures, such as in understaffed clinical areas, excess workloads, or absence of an isolation ward, then they may feel powerless in their consistent use of M. tuberculosis infection control measures” [29].

Some health workers felt their TB exposure was part of broader concerns about feeling unvalued and unsafe at work, and perceived themselves to have limited power over their working conditions [25, 26, 29, 30, 32, 34, 38–42] (Table 2, Key Finding 2). Health workers phrased this as working in a system that did not care about their health. One primary care nurse in South Africa said “….nobody cares, even if I die, it doesn’t matter” [32]. As an example of feeling uncared for by the health system, health workers mentioned that they did not trust the respirators that they are provided as PPE, because they often fitted poorly, involved counterfeit manufacture or unknown brands, or they received a surgical or cloth mask instead of an N95 respirator [27, 30, 35, 40, 41, 51, 54].

Working in conditions that feel unsafe, and feeling poorly equipped to implement TB IPC, may influence patient care. One nurse working in a TB ward in South Africa described this as: “I feel like I am putting myself in danger to be honest. I am scared to come into contact with those patients because most of them are rude and dangerous” [41]. In some instances, health workers may stigmatise patients because of their concerns about being exposed to TB, and therefore compromise clinical care. Examples included health workers who avoided contact with TB patients, shortened consultations, and restricted patient movement through placing fencing around wards and locking patient rooms [31, 33, 34, 41, 43–48] (Table 2, Key finding 3). Health workers described how fear of being infected with TB leads them to avoid TB work, with some health workers viewed caring for TB patients punishment [43] or work for junior staff [45]. They also described examples where patients were declined intensive care, such as dialysis, or care at a referral hospital because of infection risk [46].

For health workers, working in conditions where they felt unsafe blunted their occupational TB risk perception [36]. One professional nurse decribed this transition as: ‘I used to be scared when I started but I have gone past that now… yesterday we were so busy here, there is no time to get paranoid’ [37]. This included both a high clinical load [29] and the emotional load association with caring for patients who are very ill. A nurse reflected on this: “The most painful and traumatizing thing in the TB wards is that patients die in large numbers and it is so stressing to see people dying just before you every day. … We do not get even debriefing sessions. As a result, we feel that we are losing it, at times we just can’t cope” [41].

Kallon and collegues group many of these workplace considerations under organisational culture as framework, and argue that the influence of organisational culture on TB IPC use is underresearched and underappreciated. The researchers conclude that the responsibility of implementing TB IPC is incorrectly shifted to individual workers:

‘The lack of role modelling, the under-supply of masks, and the vague and poorly enforced protocols around mask-wearing, for example, all communicated that wearing a mask was less a collectively held and enforced responsibility and more a matter of personal choice and agency’ [36].

Theme 2: Technology and organisational implementation challenges of TB IPC in health facilities.

Health workers viewed unavailability of sufficient stock of particulate filter respirators as the main barrier to using respiratory protective equipment [21, 26–31, 33, 35, 36, 40, 41, 45–52]. Doctors or health workers in TB wards or DOTS centres had preferential access over nurses, housekeeping staff [49], or staff in other non-TB sections [28] (Table 2, Key finding 4). Other factors that contributed to poor access to respirators included high unit cost, poor supply chain management, lack of prioritisation by managers as reflected, for example in poor supply chain management and health workers being expected to purchase their own protective equipment [31, 49].

Fluctuations in the availability of PPE prevented health workers from developing a habit of wearing a respirator, as well as making them feel that sporadic use wass futile. A TB staff member in Mozambique said: ‘It should be there all the time. Because if the material is there today, but finished tomorrow, I am not interested, I am not protected at all […] do I not contract the bacteria when I don’t have the material?” [28] Health workers described respirators as difficult to use, “suffocating… and feels hot”[37] and interfering with make-up. Respirators caused communication difficulties with patients who spoke a different language or who were impaired by ototoxic drug-resistant TB medication and depended on lip reading [31]. Health workers paradoxically described using respirators less in poorly ventilated settings, where they described greater discomfort with use [36].

Health workers often worked in facilities where the infrastructure was not designed for airborne infection control measures, with waiting areas, consultation rooms and wards that lacked natural or mechanical ventilation and insufficient isolation rooms [21, 27–30, 32, 33, 36, 38, 40, 41, 42, 44, 46–56]. (Table 2, Key finding 5) One health worker working in Uganda described the struggle they faced to establish a safer waiting room:

“When we were conducting the TB infection assessment, the ventilation was found to be at 0%. … The worst thing is that these HIV patients are seated together with TB patients and some of them are … not yet on treatment. We requested to move the clinic to a … better ventilated place out there on the veranda, but it was rejected. So [name of donor funded project] came in and gave us a tent, which didn’t help us much because it came without seats and immediately [after] we put it up, it broke down and we didn’t use it. So we have remained in the same place up to today” [44].

Some health workers responded to spatial limitations by moving DOT consultations and sputum sample collection outdoors. In some instances this compromised privacy, but was easy to implement [47]. The ventilation measures described by health workers involved optimising natural ventilation—such as opening windows and shifting consultations outdoors. We did not find descriptions of frequent use of mechanical ventilation, air filtration such as High Efficiency Particulate Air (HEPA) filters or upper-room germicidal ultraviolet disinfection [52].

Health workers in primary care and rural health care facilities described receiving fewer resources and worse TB IPC infrastructure than urban and hospital settings [21, 32]. Yet in our broader dataset we noted that health workers in middle income countries (compared to low income countries) and in tertiary hospitals described similar TB IPC implementation barriers to those of low income or rural hospital settings [30, 50], suggesting that TB IPC is difficult to implement across the whole resource spectrum.

Theme 3: Policy and implementation plan gaps relating to TB IPC and occupational health.

Health workers and researchers highlighted the following obstacles to implementing TB IPC: the lack of a national IPC policy [45], conflicting national TB IPC guidelines [32, 40] and the absence of a local TB IPC implementation plan for facilities [24, 37] and uncertainty about how TB IPC policies translate to the work of different health workers (for example dentists) [36]. They viewed national TB IPC guidelines as instructions that also provided a mechanism to motivate health workers to comply, for example, through a central government audit [32].

Yet health workers mistrusted whether TB IPC policies adequately protected them. For example, they were uncertain about whether such guidelines described the correct duration of infectiousness of TB patients [32] or how TB IPC policy should translate into daily practice. A health worker from South Africa explained the uncertainty: “Always wear the N95 mask, or conserve them due to expense? Separate XDR and MDR TB patients, or allow them to mix at times?” [40] Health workers described trying to make individual management decisions but felt there was a need for TB IPC implementation plans to address practical concerns, including around respirator re-use [36].

At one facility, a hospital manager said policy implementation, in contrast to policy availability, was the biggest challenge [40]. Involving health workers in discussing and developing the guidelines of a facility was suggested as way to ‘create a sense of ownership’ [28]. One facility identified a facility-based champion as a way of making TB IPC a priority, yet cautioned it should be balanced with maintaining TB IPC as a collective responsibility [37].

Health workers expressed uncertainty about whether TB was legally defined as an occupational illness. Underreporting obscured the extent to which health workers are affected by occupational TB [32, 37, 38, 40] (Table 2, Key finding 6) and influenced risk perceptions of health workers and managers [38]. Several health workers viewed poor TB IPC implementation and occupational TB to be an individual’s problem, instead of a health system’s problem [36, 37]. A health worker in South Africa said: ‘The thing is with TB and being a health worker, should I get it, I know it’s going to be my problem. I won’t be able to prove that I got it here. So I guess if you work here, it is at your own risk. That’s how I feel’ [37]. Even among health workers who perceived TB to be an occupational illness, and who were aware that they could qualify for compensation, did not trust the bureaucratic progresses of accessing compensation, with one example provided from South Africa [37].

Health workers also described occupational health service shortcomings: lack of trained staff [38], a shift of the responsibility to implement an occupational health policy to untrained staff [38], lack of confidentiality within staff health clinics or occupational health services [38, 40], and absent occupational TB screening programs [27–29, 30, 40, 52, 54]. At times, TB IPC and occupational health policy implementation was viewed as a lower priority than clinical work [49].

Theme 4: Health worker perceptions of patient responses to TB IPC.

Many health workers described patients as non-compliant with the TB IPC measures that they initiated [27, 33, 37, 40, 42, 48, 50, 51, 54–56]. Health workers framed this as a knowledge deficit [57] which they tried to correct through providing health education or pamphlets, but that patients did not adhere to nor understand [44]. A South African nurse said: ‘Most of the [TB] patients are stubborn and their behaviour is uncooperative. … They have a bad attitude, don’t want to eat, carelessly cough around, refuse their tablets. They are a risk for everyone. Often I don’t trust them’ [50]. Health workers distrusted TB patients because they felt patients might pretend to have TB [50] or use coughing as way to skip the queue [28]. Some health workers felt patients were deliberately trying to infect them ‘through wicked strategies’ like putting their sputum in the health worker’s food [43]. One counter-example was cited where patients reminded nursing staff to wear their protective equipment [27]. Tudor and collegues described similar narratives of ‘deviant patients’ not ‘complying’ and observed that this may be a way for health workers to shift the blame for TB transmission in healthcare settings from health workers to patients [51].

Yet health workers also considered stigma to be an important impediment to using TB IPC measures [21, 28, 30–32, 36, 39, 56, 57] (Table 2, Key finding 7) Specifically, they viewed patients to as feeling stigmatised by TB IPC measures, such as being singled out to wear a mask for source control [44]. Brouwer and colleagues reflected on this, noting that masks can have ‘an alienating or depersonalizing effect and reduce the HCWs’ ability to provide compassionate care’ [28]. Similarly, health workers struggled to decide between closing their consultation room door to respect the privacy of patients, or leaving an open door to improve ventilation [47]. Reducing the stigma associated with wearing a mask or a respirator was highlighted as an IPC priority [28], with one study suggested universal mask-wearing by health workers and patients as potential solution [36].

Instead of looking at ways to mitigate the stigma associated with wearing a mask, some health workers did not use PPE and avoided prolonged contact with patients. A physician from the Dominican Republic said: “It is more cultural than anything. I put on a mask as if I am going to become infected [by a patient]. It looks ugly, so we do not do this. We do, however, keep our distance from patients and avoid speaking closely face-to-face” [30]. Health workers avoided entering the rooms of patients with TB or spent the shortest possible time there [39]. They also avoided making patients wait during the consultation while they fetched PPE [29], deeming it acceptable to ‘rather endanger [their] lives in order to save lives’ [41]. Chapman and collegues offered an alternative to framing TB IPC to be in opposition to providing compassionate care. Instead, they believe health workers ‘who fail to adhere to recommended M. tuberculosis infection control measures may not be upholding their ethical and moral responsibility to protect their own as well as their patients’ health’ [30].

Dodor and colleagues drew a distinction between using IPC measures rationally (to prevent becoming infected) and using it irrationally (out of fear, when it is not needed, when is stigmatising and not effective in reducing transmission) [43]. From the overall dataset contributing to this theme, we noticed that this distinction is based on the underlying assumptions that health workers and researchers make about what constitutes ‘effective’ IPC measures. While effectiveness was not the focus of the review, there is evidence to suggest that patients are concerned about the spread of TB in health facilities, and that their views towards TB IPC measures might be more nuanced than health workers perceive [57].

Theme 5: Underlying uncertainties that influence how TB IPC tools are used.

We indentified three important uncertainties that health workers described with regards to TB IPC implementation. Firstly, health workers described prioritising TB IPC implementation in clinical areas where patients known with TB or DR TB were seen, in particular, TB wards and DOTS centres [21, 29, 32–34, 36, 41, 42, 45, 47, 48, 50, 52–55] By contrast, housekeeping, administrative staff and community health workers were overlooked, as well as health workers providing undifferentiated patient care. (Table 2, Key finding 8) This targeting of TB IPC resources also demonstrated localisation of TB risk, with the contribution of undiagnosed people with TB to transmission being underappreciate both by health workers and researchers undertaking the studies. Kallon and collegues described what they observed in primary health care facilities in South Africa: ‘These acts of locating risks in specific physical spaces within the facility or associating it with exposure to particular individuals emerged as an important shared, implicit knowledge in many facilities. It was not always clear whether HCWs really believed TB risk could so easily be located and managed, or they believed this inadequate approach was the only practical one available to them (or some mix of the two)’ [36]. They note that with an airborne illness, locating risks only to certain areas in a health care facility (the TB room) or certain patients (those with drug-resistant TB) is not an effective or logical use of TB IPC resources [36].

The second uncertainty that was prominent in our data relates to duration of infectiousness. Health workers described a wide variation in the duration over which that they perceived people with TB to be infectious [32, 43, 50] (Table 2, Key finding 9). This ranged from deeming patients less infectious immediately after starting treatment, to continuing infection control measures for the duration of TB treatment [50]. This has important implications for TB IPC resource allocation in facilities.

Thirdly, health workers described uncertainty about the modes of TB transmission.

They used a variety of IPC measures aimed at preventing the spread of TB, including contact, droplet and airborne precautions [23, 28, 34, 40, 42, 44, 46, 47, 49, 54, 56]. (Table 2, Key finding 10). While WHO guidelines describe TB as exclusively transmitted by the airborne route [10], some of the researchers and health workers focussed on droplet and contact precautions [45, 56] including ‘showering and changing clothes so they did not carry the bacillus home’ [35]. Health workers placed detailed focus on containing and disposing expectorated sputum—for example closing the sputum bottle to prevent it from being aerosolised as it dries [43], asking patients to bring a tin with sand to cough in [28], prefilling the sputum cup with water [49], putting a disinfectant in the sputum cup [46], or requiring separate bins for Tuberculosis waste [56].

We found other examples of TB IPC measures being implemented outside of what is recommended in international guidelines, this is described in Table 3.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Contextualising and evaluating TB IPC implementation considerations identified through this qualitative evidence synthesis.

https://doi.org/10.1371/journal.pgph.0000292.t003