Background
With a population of approximately 43 million, South Africa (SA) has one of the highest incidence rates of
Mycobacterium tuberculosis (TB) in the world, with annual notification rates in some areas in the Western Cape exceeding 1000/100000 [
1]. Moreover, SA has one of the world's largest human immunodeficiency virus (HIV) epidemics, an estimated 5.54 million people being infected [
2]. Measures for TB-infection prevention and control (IPC) in SA remain the responsibility of individual healthcare facilities [
3]. There is growing evidence that hospital transmission is a critical factor in epidemic HIV-associated TB [
4,
5]. IPC can reduce the risk of TB transmission even in settings with limited resources [
6]. A study, published elsewhere, on the potential to transmit TB at our hospital confirmed the high incidence of TB and a substantial risk for transmission [
7]. This situation intensifies the need for a comprehensive hospital-based IPC programme to prevent the transmission of TB.
Internationally, TB-IPC is based on a three-level hierarchy of controls, including administrative, environmental, and respiratory protection [
8]. The magnitude of the local TB burden, exacerbated by limited financial and human resources at public healthcare facility level, challenges the applicability and impedes the implementation of international guidelines.
Limitations in effective TB control worldwide have caused a shift in perspective, it is no longer considered a mere technical bio-medical intervention [
9]. This applies to IPC practices at both hospital and community level. Effective TB-IPC requires adherence to measures which should be regarded as a chain of responsibilities, involving healthcare staff and decision-makers, as well as patients, and society [
10]. Several recent studies have looked at non-biological influences on TB control, from the point of view of the patient [
11,
12], community [
13] and health care providers [
14,
15]. Nurses play a central TB-IPC role in detecting the disease, providing and coordinating appropriate treatment, and assuring emotional support [
16], but it seems a neglected area of research in high TB burden countries. There is a lack of information concerning the realities faced by nurses in implementing TB-IPC measures. The absence of nurses' voices constrains the quality and quantity of human resources for TB control and care [
17]. As a result, health systems in a number of countries are weak and ineffective in meeting the growing need for TB control services [
17]. One study in SA reported nurses' lack of awareness of beliefs and attitudes about TB harboured by communities they serve, nor of their behaviour concerning illness [
11].
This paper reports findings from a qualitative study seeking to explore factors influencing TB-IPC practices at hospital level from the experiences of ward nurses in order to identify risks associated with potential nosocomial transmission, and to emphasize the crucial role nurses play in TB control and care.
Methods
Setting
At the time of the study, the tertiary academic hospital in Cape Town, South Africa, contained 1291 beds, of those approximately 26% surgical, 20% intensive care unit (ICU) and emergency (both internal medicine and surgery), 12% obstetric/gynaecological, 16% internal medicine, 20% paediatrics and 6% for minor sub-specialities. There were 1,502 nurses employed, of whom about 75% were of coloured origin, 19% African, and 6% white. Exact figures were not available, but all nurses were conversant in English of whom approximately 80% spoke Afrikaans and some Xhosa as their first language.
Design
The qualitative data was collected during one-to-one in-depth interviews with a semi-structured interview guide (Table
1) to explore ward nurses' experiences of factors influencing TB-IPC practices, using phenomenological approach. The purpose of phenomenological research is to describe what people experience in regard to certain phenomena, as well as how they interpret the experiences or what meaning the experiences hold for them [
18]. Therefore, phenomenology is an approach that concentrates on a subject's experience rather than on the person as a subject or object [
18].
1. Do you usually deal with diagnosed TB patients at your workplace? |
2. What protective device are in place and used on the wards: |
• Surgical masks |
• N95-respirators |
• Gloves |
• Curtains around beds |
• Isolation of patients |
3. If a patient with TB is admitted - what action is taken and when?
|
4. Is action different when there is only suspicion of a TB?
|
5. If the patient is transferred - are there any protective measures for suspected or confirmed cases? |
If yes, what are they? If no, why not?
|
6. Which of these tasks in TB management do you find difficult to do? |
• education on TB for patients and family- If yes, why?
|
• psycho-social counselling - If yes, why?
|
• training of family members - If yes, why?
|
• DOTS treatment - If yes, why?
|
• managing TB and co-infections - If yes, why?
|
• Caring for TB-patients in hospital - If yes, why?
|
• Supervision of health workers - If yes, why?
|
• TB-risk prone procedures - If yes, why?
|
• Other tasks not mentioned - Which? Why?
|
7. Are you afraid of acquiring TB at work? - If yes, why, if no, why?
|
8. Do you have concerns treating or working with TB patients? If yes, why?
|
9. Do you have concerns treating or working with TB patients with AIDS? If yes, why?
|
10. How do you feel when these patients are on treatment for more than a week? Explain.
|
11. Could you imagine referring a patient to a traditional healer for TB treatment, if requested by the patient? Explain.
|
12. Could you imagine referring a patient to a traditional healer as a therapist for DOTS, if requested by the patient? Explain.
|
13. Are there local beliefs that may influence TB-patients not to come to the hospital? |
If yes: what are they?
|
14. Are there local beliefs that may influence TB-patients not to take TB-medication regularly? |
If yes: what are they?
|
15. Did you participate in any TB-training or -workshop in the past two years? If yes, what?
|
16. Do you have sufficient access to information about TB-care, protective measures and other TB-related subjects to confidently manage patients? If yes, from whom, how?
|
17. Do you know the national TB-guidelines for health workers? Explain.
|
18. Do you know the policy for dealing with TB patients in your hospital? Explain.
|
19. Do you feel positive about the CPD-system (continuous professional development) that is already in place for doctors being applied for nursing staff as well in future? Why?
|
The interview guide was tested for content relevance and ease of application prior to data collection. Owing to a chronic staff shortage, compounded by a national health sector strike during the sampling period, a maximum of 20 nurses only were approved by the hospital management to participate in the study. Quota sampling was applied, including hospital wards where TB patients were managed on a routine basis (wards with "TB-routine", internal medicine, paediatrics and internal emergency wards), as well as wards where TB was not a clinical focus (wards "without TB routine", surgery, emergency trauma and obstetric). Nurses (n = 20) from the selected wards appearing on the off-duty plans during the sampling week were randomly selected and grouped into either working on such TB-routine wards (n = 10): emergency internal medicine (n = 2), paediatric (n = 4), internal medicine (n = 4); or not (n = 10): obstetric (n = 3), general surgery (n = 4) and emergency trauma (n = 3). The study included auxiliary nurses (AN) with one year training, staff nurses (SN) with two years training, and professional nurses (PN) with at least three years training. Nurses with different training curricula were included because due to staff shortage, all nursing levels were equally utilised for routine patient care.
Data collection
The interviews, undertaken in private rooms offering strict confidentiality, were conducted in English by the Principal Investigator (PI), lasting approximately 30 minutes each. The participant responses were recorded in writing during the interview. Following the interviews, the PI immediately typed and cross-checked the coded scripts to ensure full and accurate data capture. Interviews were not audio-taped as the nurses felt that recordings could make them identifiable. This concern might have had a negative impact on their participation.
Trustworthiness
The trustworthiness of the data was assured by testing the interview guide to identify and correct any ambiguities and/or errors, via one-to-one discussions with individual participants and peer reviewers, and by prolonged PI engagement in the field prior to and following data collection in order to achieve deeper understanding of the working context of the participants. The PI made every effort to clarify participants' responses and to verify the contextual appropriateness of the coding and emerging themes during data analysis. Two faculty members, and two IPC nurse specialists, served as peer reviewers to verify the thematic analysis, disagreements were debated to reach general agreement.
Throughout the research process the PI adhered to the importance of reflexivity [
19], making explicit from the outset personal experiences, opinions and preconceptions about the field of research [
20]. Throughout the research process, the PI employed "bracketing" [
21] to suspend such personal perspectives and biases in order to reduce the phenomenon under study to its authentic basic components and actively searching for the fundamental concepts. The PI was an experienced clinical researcher who had spent time in the clinical IPC arena of TBH but came from a non-nursing background, thus providing professional working distance to the interviewees.
Data analysis
The data was subjected to thematic analysis [
22]. This method involved identifying, coding, analysing and clustering recurring factors into overarching themes with respective key and sub-themes. The identified themes are presented together with quotes from participants in order to add depth and richness to the findings. Participant quotations are used verbatim and presented in italics, and followed by a unique number indicated in brackets (#No.), to provide some context for the data [
23] (Table
2).
Table 2
Nurses codes and characteristics (education, speciality, "TB routine")
1 | professional nurse | surgical | no |
2 | professional nurse | surgical | no |
3 | staff nurse | obstetric | no |
4 | professional nurse | obstetric | no |
5 | professional nurse | obstetric | no |
6 | auxiliary nurse | emergency trauma | no |
7 | professional nurse | emergency trauma | no |
8 | professional nurse | surgical | no |
9 | professional nurse | internal medicine | yes |
10 | auxiliary nurse | internal medicine | yes |
11 | staff nurse | internal medicine | yes |
12 | auxiliary nurse | internal medicine | yes |
13 | auxiliary nurse | internal medicine | yes |
14 | auxiliary nurse | internal medicine | yes |
15 | auxiliary nurse | pediatric | yes |
16 | staff nurse | pediatric | yes |
17 | professional nurse | pediatric | yes |
18 | staff nurse | pediatric | yes |
19 | professional nurse | surgical | no |
20 | staff nurse | emergency trauma | no |
Ethics
Ethical approval for the study was obtained from Stellenbosch University, and written consent was obtained from all participants. The interview scripts were coded, and personal identifying details were not collected.
Discussion
The study has obtained the experiences of ward nurses concerning multiple factors influencing TB-IPC practices and increasing the potential risk of nosocomial transmission at hospital level. The key perceived contextual influences fall into broad interconnected overarching themes related to the healthcare system, wider contextual conditions and patient behaviour. These broad themes correspond with those elicited in a study exploring barriers to TB care in Russia [
14].
The major influences were inadequacies associated with the healthcare system. Routine IPC audits at different departments of our hospital had showed inconsistently applied transmission-based precautions, even where known TB cases were admitted on a routine basis [
7]. Effective TB-IPC practices by nurses were hampered by the lack of clear TB policy directives, the lack of appropriate isolation facilities and availability of PPE, the lack of TB training for staff and patients, and a persistent work overload.
The findings illuminate the need for a comprehensive TB-IPC policy, with associated standards for provision and practice. The infection control policy must reflect current guidelines on infection control, as well as legislative issues and evidence of annual audits thereof [
26]. It should also be in line with the International Standards for TB Care [
27]. Healthcare workers cannot be blamed for applying inconsistent TB-IPC measures when appropriate, accessible and clear guidelines are not implemented, monitored and evaluated. According to our experiences, and in line with internationally agreed standards, IPC practices should include: rapid clinical evaluation (TB-triage) of all persons (staff, patients and visitors) with symptoms suggestive of TB, segregation of patients with known or suspected pulmonary TB, use of effective local exhaust ventilation in connection with high-risk procedures, employee training, and ongoing risk assessment [
28]. Especially participants from wards without TB-routine complained that patients were managed inadequately with regard to TB. There was no consistent TB-triage by nurses on patient admission, and suspected TB cases were not included in TB-IPC measures such as isolation and usage of PPE. On transfer, only patients with MDR-TB were usually provided with masks. Other PTB patients, confirmed or suspected, were dealt with by chance, dependant on individual staff practices. A large hospital without basic measures for airborne IPC, combined with the high HIV infection prevalence among patients and visitors, could provide a prime setting for the rapid spread of TB.
Work practices and administrative control measures are the first line of defence against TB transmission within facilities caring for people with HIV infection [
5]. A previous study reported that at the hospital 54% (107/199) of TB patients were HIV-positive and one fifth of PTB patients were diagnosed on wards without TB-routine [
7]. HIV infected persons must be protected from exposure to TB as far as possible, with special care needed in healthcare facilities, especially hospitals. TB suspects and diagnosed patients should be managed in separate places and environmental controls including good ventilation must be in place [
26]. The findings of the study suggest that the hospital was not adequately equipped to protect HIV positive patients from acquiring TB transmission, especially on wards without TB - routine. Patients were segregated according to their clinical profile and gender, not to their TB status.
Similar to the study by Dimitrova et al. [
14], the findings reveal staff shortage as a significant problem, leading to work overload, competing clinical priorities and working in a high risk environment. Despite such negative working conditions, the participants expressed a strong sense of duty for care.
All healthcare staff, including support staff, must receive training in infection control [
29]. Through ongoing training and CPD, nurses could develop the skills and enhance their confidence in TB and HIV care in a high burden environment and make informed clinical judgements on appropriate management. The findings of the study identify the pressing need for TB training for all nurses. The participants did not seem fully to comprehend the real risk of TB exposure, and most were afraid of becoming infected with and developing TB. Concerns and stigma related to TB emerged particularly among participants from wards without TB-routine. Irrational work practices and fear might lead to a higher risk of exposure and nosocomial transmission.
IPC training interventions should urgently be offered for nurses and other healthcare workers to improve knowledge, clinical competency and quality of care.
Concerns and stigma related to TB and HIV, and the role of traditional healers, comprised wider contextual influences which impeded TB-IPC at ward level. At the time of the study, approximately 80% of nurses at the hospital spoke Afrikaans as first language, and three quarters were of coloured ethnicity, followed by an equal proportion of African and white nurses. A previous study at the hospital found that about half of the patients with PTB were of African origin, the other mainly coloured [
7]. In the catchment area of this hospital, the African population predominantly spoke Xhosa as first language. Most of the study participants who reported problems concerning cross-cultural understanding of stigma and communication, originated from a different cultural background. The findings suggest a deficit in knowledge of cross-cultural health beliefs. Similar observations were reported in the Limpopo Province of SA [
11].
The impact of traditional healers on patient health was raised by participants. Patients with TB, especially those from an African cultural background, would commonly seek a first opinion from traditional healers, resulting in the late uptake of hospital care. A South African study reported that 74% of patients with TB had visited a traditional healer before attending hospital, and that patients who visited traditional healers took longer to access anti-tuberculosis chemotherapy, were in worse condition by the time they presented, and were more likely to die after they had presented with TB [
11]. The strength and depth of the cultural links between traditional healers and their local communities suggest that such authorities cannot simply be ignored or chastised [
30‐
32]. The involvement of traditional healers in treating TB patients was rejected by the study participants, suggesting a general poor opinion about and understanding of the role of traditional medicine. Only traditional healers trained in TB therapy support were acceptable to some participants.
Although there is strong evidence that the accessibility and acceptability of health services remain the most important factors in patient adherence [
33], patients behaviour is also of critical importance in TB-IPC. Patients' late presentation to hospital was reported in other SA studies [
34,
35]. While the authors saw this phenomenon mainly caused by health system failure, Dimitrova et al. [
14] identified fear of unemployment and stigma towards TB as a major obstacle. In our study, most nurses linked patients' late presentation to hospital mainly to the impact of poverty and traditional medicine. In the study by Dimitrova et al [
14], the 'willingness' to access and adhere to treatment was seen as a critical factor in successful treatment outcomes. In our study, side effects related to TB treatment, patients' concerns and stigma related to HIV and the lack of TB knowledge of patients and their family members/carers were seen as major influences on patient adherence.
In reviewing the role of TB education for patients, Sumartojo [
36] emphasized the need to facilitate behavioural change rather than simply to provide disease information. Edginton et al. [
35] found, that although non-attending patients were visited at their homes by researchers who told them about TB and the need for treatment, more than one third still did not attend. This indicates the complexity of contextual influences on decision making. In a local context of multiple cultures, ethnicities and languages, highly specific and culturally sensitive educational interventions are required and should be offered both to patients and their family members/carers in their mother tongue. Acknowledging the role of traditional medicine, consideration should be given to involving traditional healers in such TB educational interventions to facilitate early medical care, treatment support and continuity of care, and to improve TB-IPC at both community and hospital level.
Strengths and Limitations
TB nurses perform the bulk of work in TB care and control [
37]. This study recognises their crucial role and emphasizes the importance of well-trained staff. A focus group with nursing participants to explore in more depth and/or illuminate further contextual influences on TB-IPC, and to serve as a form of triangulation, was not approved by the hospital management owing to staff shortage. Similar to other qualitative studies, this study was context-specific, hence the findings cannot be extrapolated to other settings. The restricted sampling period resulted in failure to include males and nurses from African ethnicity in the random sample. Their inclusion might have yielded additional insights. It would have been invaluable to explore the cultural context of African patients from the perspectives of African nurses. The staff crisis at the time of sampling also lead to limited transferability, as only motivated nurses and staff being able to deal with work overload were accessible. Nevertheless, this study highlights several issues of critical importance in TB-IPC as experienced by those in the frontline of care provision. The findings might reflect the realities and influences experienced by many nurses in other settings, and pose implications which could be considered for TB-IPC policy and practice in similar settings elsewhere.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
DS conducted the in-depth interviews. All the authors have (1) made substantial contributions to study conception and design, (2) been engaged in data analysis and interpretation of data, (3) been involved in drafting the manuscript or revising it critically for important intellectual content; and (4) have given final approval of the version to be published.