Background
Tuberculosis (TB) control remains a major global public health challenge, with most of the 9 million people affected each year living in resource-limited countries [
1]. Despite the widespread availability of effective therapies for TB, the global average decline in incidence has been only 2% annually. It is estimated that, each year, three million new cases of TB remain undiagnosed [
1]. Actively screening high-risk groups for TB could increase the proportion of infectious cases that are diagnosed and treated and, thereby, reduce ongoing transmission of TB in the community [
2,
3]. Household contacts of patients with TB have a higher risk of developing TB than the general population. A recent meta-analysis of published contact investigation studies showed that, overall, 3.1% (95% CI 2.2-4.4%) of household contacts had TB disease at the time their infectious household member was diagnosed [
4]. The study also showed high rates of incident TB over the four years following enrolment, suggesting that serial screening of contacts is a high yield strategy. Evidence from community-based studies also points towards a possible impact upon the burden of TB in high-prevalence settings [
5].
Over the past decade, WHO recommended screening for TB in high-risk groups including household contacts [
6], child contacts under 5 years [
7] and contacts who are HIV infected [
8]. National Tuberculosis Programs in high-prevalence settings have sought to apply these policies in their local contexts [
9]. However, they have faced substantial challenges in bridging the gap between screening policies and practice [
10]. In particular, convincing asymptomatic contacts to attend screening appointments has been difficult. Understanding the barriers to screening from the perspective of patients and their contacts is critical to the development of effective contact investigation strategies [
11-
13].
This study aimed to characterise knowledge about TB and perceived barriers to participation in contact investigation among household contacts of adult patients with pulmonary TB managed within the Vietnam National Tuberculosis Program.
Methods
Setting
Vietnam is a south-east Asian country with a persistently high prevalence tuberculosis (209 prevalent cases of TB per 100 000 population), despite having achieved good treatment outcomes for over a decade [
1]. Consequently, enhanced case finding strategies, such as contact investigation, are a priority for Vietnam. The National TB Program has recently adopted a policy of routine symptom screening and preventive therapy for child contacts. However screening of adult contacts is not performed routinely [
14].
Study design
We conducted a nested case–control study between September 2010 and July 2012 at District TB clinics participating in a larger randomised controlled trial (RCT) of contact investigation that is being undertaken in 8 Provinces throughout Vietnam [
15]. District TB units are the primary location of TB diagnosis and treatment for the National TB Program, with chest radiography available at most sites. In the main RCT, which is described elsewhere [
15], 70 District TB units were randomised to be either the intervention or control group. Over 10,000 contacts of patients with smear positive pulmonary TB were recruited in each group. In the intervention Districts, ‘index’ TB patients were asked to bring all household contacts to their local District TB unit for screening four times in two years: at the time of the patient’s initial diagnosis, then after six, 12 and 24 months.
District TB units that were randomised to the intervention group were eligible for inclusion in the case–control study if they had demonstrated their capacity to recruit contacts during the early phase of the RCT. This was defined as recruiting at least 50% of eligible index patients, based upon case notification data. This approach was chosen to ensure that health system-related factors alone did not drive the non-attendance of contacts. Based on this criterion, 12 of 36 intervention Districts from the six of the eight provinces in the main study were selected. The study population for the present case–control study comprised all contacts within these 12 Districts that had been enrolled in the main study before 30th September 2011. Within this study population, cases were those who did not attend their scheduled six-month follow-up visit (that is, the first scheduled review after enrolment into the main study) and controls were selected at random from among those who did attend this appointment.
Questionnaire development and data collection
Contacts were interviewed by District TB clinic staff, using a structured questionnaire (Additional file
1), to ask about their knowledge and attitudes relating to contact investigation. The ‘index’ patients of enrolled contacts also completed a questionnaire. All questionnaires were manually checked for completeness and consistency by study staff, and 10% of contacts were randomly re-interviewed by Provincial staff to verify the accuracy of responses.
Knowledge, attitudes and beliefs about TB were measured using a questionnaire that was administered to all selected cases and controls (and their index patients).
The questionnaire was developed based upon a review of the literature, and three focus group discussions including 26 household contacts and their index patients and health care workers involved in a pilot contact investigation study [
16]. Discussion themes were identified by a search of relevant literature and consultation with local physicians. Recordings from the discussions were transcribed, coded and analyzed according to the grounded theory method [
17]. The final questionnaire was based on themes arising from the discussions, as well as themes identified in published literature relating to knowledge about TB, attitudes towards contact investigation, and compliance with TB screening.
Data analysis
Qualitative interviews were recorded digitally, transcribed and translated into English before coding and analysis.
Questionnaire responses were independently entered twice using EpiData 3.1 [
18]. Groups of contacts were compared using generalized estimating equations (GEE), with clustering at the level of the index patient. Effects were estimated as odds ratios and 95% confidence intervals. Model covariate selection in multivariate analysis was based upon quasi-likelihood under the independence model criteria (QIC) using reverse covariate selection. Data were analyzed using SAS (v9.3, SAS Institute, Cary, NC). Variables with <10% of missing values were imputed using multiple imputation [
19], and those with a higher proportion of missing values were excluded.
We estimated that 136 cases and an equal number of controls would yield 80% power to detect a 40% difference in the prevalence of positive responses between cases and controls as significant (at the 5% level).
Ethics approval and consent
This study was approved by the Human Research Ethics Committee of the University of Sydney, and the Institutional Review Board of the National Lung Hospital, Hanoi, and was conducted in accordance with the principles of the Helsinki Declaration. Study participants provided written informed consent. The main RCT within which this study was nested (the ACT2 study) has been registered in the Australian New Zealand Clinical Trials Registry ACTRN12610000600044.
Discussion
This case–control study, nested within a RCT of serial contact investigation in 70 Districts throughout Vietnam, characterised the knowledge and attitudes of participants towards TB and contact investigation. A substantial proportion of contacts and index patients identified discrimination against people with TB as an important issue, while most believed that screening contacts for TB was beneficial. Cases were less likely to understand TB as an infectious disease, and misunderstandings about the biology of disease and transmission were common.
For the long-term ambition of global TB elimination to be realized [
21], new strategies are needed to substantially enhance case-finding and thereby reduce transmission in the community. Consequently, recent WHO policies have provided a framework for the expansion of TB screening in a range of high-risk populations, including household contacts [
6,
22]. However, if enhanced case-finding strategies are to be successful, international recommendations must be adapted successfully to local clinical contexts [
23]. Consequently, it is important that before ‘scaling up’ new screening programs, National TB programs and local health services should consider how local attitudes towards TB screening are likely to influence their success.
This study has identified a number of important misconceptions about the natural history of TB that are likely to affect attendance. One third of cases believed that they were no longer at risk once the TB patient in their household had completed treatment, and one quarter believed an initial negative test was sufficient to exclude subsequent disease. This reflects a lack of understanding about the nature of latent tuberculous infection, which can frequently lead to tuberculosis months or years after the initial exposure [
4]. This is likely to reduce motivation to attend serial screening for disease. For contact investigation programs to achieve high rates of re-screening, participants must be educated about their ongoing risk of disease. This finding emphasizes the importance of National TB Programs delivering clear and accurate information to contacts about their ongoing risk of reactivation in the several years after exposure, at the time of initial screening.
Many contacts were also confused about the way in which TB is transmitted. Over one third of contacts who did not attend their appointment did not believe TB was transmissible. Over 70% of contacts incorrectly believed that sharing utensils or other common property could transmit TB. While these beliefs could heighten their perception of risk and motivate their desire to participate in screening, fear of contagiousness may also lead to discrimination against people with TB and increase their sense of alienation from society. Simple educational interventions to explain the airborne nature of TB transmission can be incorporated into screening programs to combat this misperception.
Importantly, over half of all contacts hold the traditional Vietnamese belief that an ‘exhausted’ form of TB is non-infectious, as previously described [
20]. In our focus groups some participants described this form of disease as indicating an intrinsic weakness within a family. This idea, although not biologically accurate, is understandable given the observed household clustering and known genetic risk factors for TB. These attitudes and beliefs may further exacerbate a sense of alienation among affected individuals. It is not surprising that this notion was more common among those who did not attend contact screening than among those who did attend.
Importantly, we found that 45% of Districts required contacts to travel more than 500 m from the District clinic to access external radiology facilities. This is likely to prolong each screening encounter. Similar difficulties accessing clinical services have also been shown to reduce treatment compliance in patients treated for active TB in other settings [
24-
26]. As over one quarter of contacts in our study lived over 10 km from the screening clinic, consideration should to be given to a more decentralised approach to screening, particularly in rural areas. Another common barrier to attendance was difficulty remembering appointment dates and times. Innovative strategies, such as scheduled mobile phone based prompts using SMS, or written follow-up reminders could provide a simple and effective solution to address this difficulty. Hence, both practical and psychosocial barriers to attendance need to be addressed by the health service, as local contact investigation strategies are developed and scaled-up in Vietnam.
Observations from this study are highly relevant to other comparable high-burden settings where contact investigation is to be implemented. National TB Programs considering implementing or scaling up contact investigation programs may benefit from conducting similar research to identify locally-relevant barriers to implementation. Surveying participants during the early stages of a newly introduced contact investigation program can provide helpful information for health managers, in order to modify screening processes, improve accessibility, refine health promotion materials and focus National TB Program staff training.
A potential limitation of this study was that contacts selected for the study had already enrolled in the main contact investigation study. For both ethical and practical reasons, it is difficult to evaluate the opinions of subjects who are unwilling to enroll in a study. While selection bias may have underestimated the prevalence of perceived barriers to accessing screening among contacts, it is likely to bias effect estimates towards the null. Nonetheless, we identified barriers using both qualitative methods (focus groups) and quantitative approaches in 12 Districts, in 6 provinces, nested within our main study. Consequently, it is likely that these represent important factors common to many potential participants in contact investigation. Hence, our results have been useful in developing resources to improve participation in the main study, and will enhance the expansion of contact investigation in Vietnam.
This study has demonstrated that contacts and patients in Vietnam commonly perceive prejudice against TB, although most participants did not feel this directly affected them. By identifying some important misunderstandings about TB transmission and pathogenesis, we have identified factors that, if addressed, may improve participation and retention in contact investigation programs.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
GJF, LPL and GBM conceived of the study, and participated in its design and coordination and helped to draft the manuscript. NVN DNS and WJB participated in its design and coordination and contributed to the manuscript. NTL participated in its coordination and implementation and contributed to the manuscript. All authors read and approved the final manuscript.