Theme 1: evidence generation and use

Most interviewees described the evidence on ACF as being relatively limited and emphasised the need to generate different types of evidence to inform ACF policy development and implementation. They stressed the importance of disseminating and exchanging evidence, of the demand for evidence by decision-makers and stakeholder engagement to enable evidence use. Apart from highlighting specific types of evidence, interviewees across the different settings had similar views with regard to this theme.

Interviewees highlighted that a variety of evidence is needed and demanded by decision-makers working on ACF; from effectiveness and health economic evaluations to implementation and operational research. One interviewee from a university in a high-income country stressed that to demonstrate effectiveness, there is a need ‘to do ACF in the context of randomized controlled trials’ (I-32). Another interviewee from a non-governmental organisation (NGO) in a low-income country highlighted the importance of distinguishing clearly where the decisions are being made; be it at the community, district or national level:

I think this is very important, ie what types of evidence you would need to make decisions at various levels (…). What evidence is enough evidence at what level to take the decision. (I-7)

Local evidence was said by many to play a significant role in, for instance, available health and diagnostic facilities, and health workers’ capacity and experience in communicating with communities. In particular, evidence from national TB prevalence surveys was described as significant for TB policy development more broadly. Two interviewees from funding organisations in high-income countries concluded that countries should be encouraged ‘to adopt [ACF] policies based on the local evidence and then move forward, rather than waiting for systematic reviews’ (I-30) and ‘you should implement enough to figure out what’s practicable and what works, and then that should become policy’ (I-15). According to the interviewees, evidence use in ACF policy development and implementation necessitates evidence dissemination and exchange, especially to share unpublished findings. One interviewee from an international organisation highlighted:

Unfortunately, we are [from a low-income country] and we are not very good at publishing. We’ve got a wealth of experience that is unpublished (…) but it has been presented at several conferences. (I-35)

Depending on the country context, gaps may exist between evidence and policy and/or between policy and practice. As one interviewee from an international organisation in a high-income country pointed out:

Countries are different. As I said, in [that country] (…) from evidence to policy was difficult. But once it [ACF] was inside the policy or even without the policy, they used to easily convert it to practice. But here [in our country] (…) evidence to policy is easier, but policy to practice is more difficult. (I-28)

Interviewees emphasised that researchers should engage with key stakeholders from the beginning of the research process to foster research use in ACF policy development and implementation; stakeholders may include the WHO, the Ministry of Health and the National TB Programme.

‘Make sure that you have the right partners from the beginning; partners who are going to take your results and actually do something with them. Because otherwise you are kind of doing it [research on ACF] in a vacuum,’ said one interviewee from a university in a high-income country. (I-37)

Moreover, to spark dialogue through stakeholder engagement, an interviewee from an NGO in a low-income country stressed that one must ‘create platforms, or you need to use the platforms which are already there’ (I-7). Regular review meetings at subnational and national levels to discuss challenges and successes related to ACF offer one such platform. Overall, evidence use was said to be influenced by who is being engaged and by personal contacts which may be ‘more important than they should be’, as another interviewee from a university in a low-income country described (I-36).

Theme 2: factors influencing ACF policy development

According to the interviewees, many different stakeholders influence ACF policy development, specifically governments, donors and NGOs. Interviewees underlined stakeholder involvement as being necessary for policy development and the contextualisation of global policy into local realities. Interviewees did not have any contradicting views with regard to this theme, but rather highlighted the specific roles of certain stakeholders they thought were most influential in ACF policy development. The leadership, buy-in and commitment of governments and National TB Programmes were described as being vital for ACF policy development and implementation. India was mentioned as a prime example where political push ‘forced the technocrats to develop policies and implement them’ (I-27, an independent consultant in a lower middle-income country). Governments make decisions for political reasons or donor incentives, even if these contradict the evidence. One representative from an international organisation in a high-income country highlighted an example of action perceived to be contradicting their view of the evidence:

Women and children of reproductive age (…) should only be included as part of the passive system not as a priority for ACF ever. But when you talk to NTP [National TB Programme] managers, there is strong political pressure and a perception that donors want them to focus on women and children. (I-24)

Donor organisations such as the Global Fund to Fight AIDS, Tuberculosis and Malaria and the case-finding initiative TB REACH (the latter is coordinated by the Stop TB Partnership and funded largely by Global Affairs Canada) were described as being influential in ACF policy development, for example, TB REACH was said to have ‘brought this concept of ACF to the country’ (I-2, interviewee from an international organisation in a low-income country), while the Global Fund ‘hold[s] every power to change things and not to change things’ regarding ACF policy development (I-7, representative from an NGO in a low-income country). Likewise, interviewees pointed out that donors’ influence was linked to WHO’s influence, as donors request countries to adopt WHO guidelines to be eligible for funding:

‘Why national policymakers are looking mainly at things like WHO documents: because a lot of them get Global Fund money and Global Fund money is often aligned with countries implementing WHO policies,’ described an interviewee who is based at a research institution in a high-income country. (I-9)

This observation was shared by another interviewee, from an NGO in a low-income country (I-8). Linking back to the preceding theme on evidence generation and use, TB REACH projects have the potential to generate useful evidence for future policy and practice, as an independent consultant in a lower middle-income country pointed out (I-27). Interviewees said that NGOs are often the ‘implementers’ of ACF whose years of experience are of great value for ACF policy development and they should therefore be involved in the same, for instance, in policy dialogues with the government and other key stakeholders. One representative of an NGO in a low-income country stated:

We [NGOs] are the one who really deal with the people (…). We have the evidence. We have the good photographs. We have the data. (…) We are the ones who can influence [ACF policy development]. (I-8)

Theme 3: factors influencing ACF policy implementation

Interviewees elaborated on available resources, systems and processes within a given health system, donor and government stakeholders, as well as the motivation and incentives for health workers as major factors influencing ACF policy implementation. Interviewees emphasised the role of particular stakeholders, as well as barriers and facilitators they thought were most influential in ACF policy implementation, while no clearly contradictory views on this theme emerged. The implementation and scale-up of ACF policies depend on the availability of financial resources, as many interviewees stressed.

‘We realized that in a country like [our country], we have great policies. The problem is the implementation. (…) And this is where the support of the development partners, funded through PEPFAR [The President's Emergency Plan For AIDS Relief], have been key to implement these policies, particularly ACF policies,’ one interviewee described. (I-39, interviewee from an international organisation in an upper middle-income country)

ACF implementation may ‘just stop because [there is] no funding’ (I-29, interviewee from an international organisation in a lower middle-income country). An interviewee from a funding organisation in a high-income country provided a different perspective regarding the funding for ACF by highlighting that ‘ACF through government funding can be more difficult than doing it through donor funding’ (I-15). This perspective may inhibit long-term thinking about ACF, as it seems to focus on immediate action to implement rather than sustainability, which is more likely to come with government investment. In addition to limited financial resources, human resource constraints for ACF were highlighted as a major challenge by experts from low-, middle- and high-income countries. These constraints could hinder National TB Programmes in thinking more strategically and ambitiously about how to address TB comprehensively.

The use of existing systems and processes in a given health system was said to be central because ‘if you start from scratch, it [ACF] is much more difficult than if there are already things to which you can link,’ as an interviewee from an international organisation in a high-income country pointed out (I-17). Interviewees mentioned that ACF policy implementation can build on experience from existing screening programmes (eg, cervical cancer screening), activities for vulnerable populations (eg, needle exchange programmes), healthcare infrastructure (eg, chest X-ray buses) and already known locations for screening in high-incidence areas and trained human resources (eg, those involved in prevalence surveys). Yet, pursuing synergies may be challenging due to the fragmentation of activities. In addition, the structure and financing of TB within a health system matters in terms of availability of resources:

‘TB has tended to fall into the preventative [arm of the health system] and that has limited the availability for resources,’ described an interviewee from a university in a high-income country. (I-32)

Processes including supportive supervision, monitoring and the use of standard operating procedures are critical for ACF policy implementation and are necessary to avoid corruption, interviewees discussed. In one country, the ‘whole case-finding system collapsed along with the supervision’ (I-16, interviewee from an international organisation in a high-income country). Moreover, processes that strengthen communication with, engagement of and awareness raising among communities were described as instrumental for ACF policy implementation, for example, to help reduce stigma. One interviewee from a university in a high-income country mentioned how the community ‘has started to advocate loudly for ACF services’ (I-20).

Many interviewees underlined that donors influence the implementation of ACF policies in countries with no or insufficient domestic resources. ‘The piper will determine what music you play’ (I-35, interviewee from an international organisation in a high-income country), which, again, highlights the power which donor organisations are perceived to have in influencing ACF policy implementation, and the possible resulting lack of a sense of policy ownership in some countries. Donors influence ACF policy implementation by setting targets for their funding recipients and pushing them towards reaching them. One interviewee from an international organisation in a high-income country described:

These targets that countries have set, that donors have set; people are very anxious (…) and that often means the easiest short cut is to do ACF, even if it is a little bit unethical or little bit using low specificity tools, so you have a little bit of over diagnosis. Donors are very comfortable with that. (I-24)

The consequences of implementing ACF under donor pressure are unclear and should be balanced against the unethical nature of inaction on the TB epidemic, but scale-up of inaccurate diagnostic strategies might lead to heightening the potential risks of ACF such as increasing false-positive diagnoses, as the interviewee mentioned.

ACF policy development and implementation depend on ‘power plays plus push’, for instance, in a country with no written ACF policy, ACF was still being implemented because the National TB Programme manager was respected and able to push for it (I-29, interviewee from an international organisation in a lower middle-income country). The aforementioned pressure by politicians, donors and WHO may be seen as additional examples of ‘power plays plus push’. Many interviewees highlighted the important role of power dynamics in ACF policy implementation. It seems crucial to be aware of such dynamics, while the use of evidence may help mitigate them. ACF policy implementation is in itself a balancing act, which power imbalances might negatively impact.

The motivation of health workers and volunteers is an important enabler for ACF policy implementation. These ‘implementers’ can be strongly motivated by their desire to help people, by understanding the benefit of ACF for communities, by receiving feedback on the outcomes of their work (eg, using performance indicators) and/or by feeling ownership of the ACF process, according to the interviewees. Financial and non-financial incentives (eg, salaries, transportation allowances, provision of motorbikes or mobile airtime) have a significant role in motivating health workers and volunteers to implement ACF as an outreach activity, interviewees discussed. Nevertheless, incentives can raise expectations and distort ACF policy implementation in the long term, for example, if government health workers are paid extra as part of an ACF project, they will also expect an extra pay for such activities in the future and for other work; another balancing act. While incentives should be in line with what a country could adopt later, they are often difficult or impossible for governments to sustain, an interviewee said.

Theme 4: WHO guidelines on systematic screening

This theme focuses on stakeholders’ perceptions of the WHO guidelines on systematic screening, the need for their contextualisation and suggestions for improving them. This theme elicited different views among stakeholders, which are described in the following.

The WHO guidelines on systematic screening are perceived positively by many, for instance, as a reference document when planning ACF activities as well as to put ACF on the agenda. Positive perceptions of the guidelines were described by interviewees from different countries, while negative perceptions were only voiced by interviewees in high-income countries. Such negative perceptions included the guidelines being vague, lacking information about the how-to of ACF and being unduly negative in terms of mentioning the risk of increasing false-positive diagnoses through ACF. Low-income countries may be more receptive to and reliant on WHO guidelines, while in a middle-income country ‘you’ve got really serious domestic universities providing the formal policy evidence. And the country kind of says “Thanks but no thanks” to outside opinions. They are really driving their own decisions. WHO is really not consulted very much, if at all,’ a representative of a funding organisation in a high-income country described (I-15). An interviewee from an international organisation in a high-income country said:

When you have something that is so broad—and you’re talking about ACF which can be so many different things—it’s just very hard to have something that works the same way in different countries (…). I think that’s the main shortcoming around the guidance. (I-28)

Interviewees emphasised the necessity of contextualising the WHO guidelines on systematic screening, for example, depending on a country’s income level, epidemiology and availability of diagnostic tools. One interviewee from a funding organisation in a high-income country pointed out that ‘you just can’t be as prescriptive and exact as you are in the more clinical guidelines’ (I-15), which seems like an important observation and reminder about the limitations that ACF policies will always have. According to the interviewees, contextualisation of guidelines can happen in a stepwise approach, for instance, a country pilots the use of a guideline before adopting and adapting it.

Review meetings with WHO and other partners can provide a platform for discussions around guideline adaptation, interviewees said. Yet, countries have faced challenges in contextualisation, for example, WHO recommends using chest X-ray which was too expensive in a country and could thus not be used. In another instance, WHO describes how contacts of an index patient with TB should provide their address, while individuals were hesitant to do so due to the stigma surrounding TB in the country. More support for the contextualisation of guidelines may be needed.

The interviewees suggested that the WHO guidelines for systematic screening9 must be updated based on new evidence, for instance, evidence from prevalence surveys, gender analyses, studies about specific risk groups (eg, drug users and indigenous populations) and what works and how, with regard to ACF. In this process, WHO should be aware of and avoid conflicts of interest, for example, by ensuring potential conflicts of interests are adequately declared and managed. This comment is in line with what an interviewee previously highlighted about the role of personal contacts to bridge the research–policy gap. These types of biases may undermine the integrity of the process and the resulting quality of guidelines and policies.

Some interviewees lamented that WHO can be paralysed by the need to use the strongest evidence available and suggested that the organisation should consider more programmatic, less scientifically rigorous data. One interviewee from a university in a high-income country described:

Usually we’re relying very heavily on WHO for global policy using the GRADE approach [Grading of Recommendations Assessment, Development and Evaluation was developed for creating summaries of research evidence to help guide health decision-making. It is currently the most widely used tool for evaluating the quality of science, with more than 110 organisations endorsing the method.23] with the PICO [P—Patient, Problem or Population; I—Intervention; C—Comparison, Control or Comparator; O—Outcome(s). The PICO process or framework is a mnemonic used in evidence-based practice to frame and answer a clinical or healthcare-related question. The PICO framework is also used to develop literature search strategies, eg, in systematic reviews.24] and all that stuff. I think that’s laudable, but sometimes I find that weird, subjected to the tyranny of the great process, and you don’t make progress in smaller areas with a paucity of evidence. (I-21)

In addition, interviewees pointed out that WHO recommendations should be based on what should be done, not on what can be done. For example, countries (not WHO) have to be the ones to decide about their ability to pay for Xpert MTB/RIF as a diagnostic tool. This point of view illustrates a stark contrast to the contextualisation challenges mentioned above, for instance, where the use of X-ray was recommended, but, frustratingly, was unable to be applied in a country as it was not feasible to implement. Moreover, the WHO guidelines could be improved by not only describing the what, but the how of systematic screening including ACF, many interviewees said. One interviewee from an NGO in a low-income country suggested:

You can come up with different scenarios: ‘If the context is this, then…’, ‘If the context is that, then…’. (…) Unless guidelines presents [the] ‘how’ better, (…) it’s meaningless. (I-7)

Theme 5: sustainability of ACF

The sustainability of ACF was a cross-cutting theme in this analysis. Interviewees elaborated on opportunities and challenges related to sustainability. ‘TB is not a like smallpox or polio. It’s a long-term sustainable (…) matter,’ an independent consultant in a lower middle-income country described. (I-27)

That is, even more perseverance and long-term thinking may be required to end TB. Interviewees expressed similar views and concerns regarding this theme. Interviewees highlighted that the interest in and leadership for ACF through the government and the National TB Programme are important for the sustainability of ACF. Additionally, the sustainability of ACF requires its integration in and funding through the given health system. An interviewee from an international organisation in a low-income country described:

If this [ACF] were to be sustainable, it should start with the initiation of the NTP [National TB Programme]. (…) It has to be supported, facilitated, monitored. Because it is actually the NTP which later needs to uptake that. (I-2)

Many interviewees highlighted the important role of National TB Programmes. The sustainability of ACF may be restricted in places with frequent government and staff turnover, which makes it difficult to get long-term commitment for ACF from decision-makers, interviewees stressed. Of course, such turnover will affect areas beyond ACF. Also, ACF cannot be sustainable, if it depends on donor funding. One interviewee from an international organisation in a high-income country summarised the situation as follows:

It [ACF] is difficult to sustain. Most of the activities that have been done for ACF have been project-based. (…) So, the Global Fund comes and says: ‘Here is a pot of money for ACF for the next three years.’ (…) And then USAID [United States Agency for International Development] comes (…). Or TB REACH (…). And people do it. But that’s not a sustainable way of doing this and this should be part and parcel of routine programming. (I-35)