It is now widely accepted that evidence rarely feeds directly into policy and, moreover, that policy is not always implemented in such a way to achieve the impact desired.1 2 What determines the implementation of evidence-informed policy and what prospects are there that recent research pointing to evidence-based interventions will be widely rolled out across Pakistan to curb the HIV epidemic? This paper addresses both of these questions but, first, what is policy analysis and how can it help?

Policy and policy change are thought to result from complex interactions among institutions (the formal and informal norms governing how decisions are made), interest groups that stand to gain or lose from the policy and ideas (discourse, persuasion and advocacy for or against particular policies based on argument and, sometimes, evidence).3 Making sense of this complexity is the field of policy analysis.

Policy analysis can help to explain why some issues, problems or solutions receive the attention of policymakers. An analysis of the political dimensions of an issue, famously described as “who gets what, when, how”,4 can also identify which stakeholders are supportive or resistant and can therefore be used to develop strategies and tactics to change the political landscape. Policy analysis can assist in identifying and addressing obstacles to the implementation of policies and may establish more realistic expectations of pro-poor reform. Policy analysis can also improve the prospects that technical advice is considered during policy formulation, thus increasing the possibility of implementing evidence-informed policy.5

Our approach to assessing political feasibility involved a consideration of the institutions, interests and ideas presently prevailing in Pakistan with respect to HIV and the specific interventions. In particular, we drew from the mainstream policy agenda-setting, formulation and implementation theories to distil the key determinants of political feasibility.

What decision-makers are paying attention to at any one time depends on the convergence of three separate “streams” of ongoing activities—problem, policy solutions and politics.6 “Policy entrepreneurs” affect the coming together of the streams or take advantage of agenda-setting opportunities when the streams converge (ie, all the conditions are met including that a problem is widely recognised, a technically feasible solution exists and there is sufficient political support).

Once decision-makers have decided to act, the processes of policy formulation and decision-making commence. This involves assessing alternative courses of action proposed by interest groups and technical experts, weighing up their pros and cons, and bargaining and negotiating over which policy option to adopt. Some of the important considerations related to the role of evidence in decision-making include the background and training of the decision-maker and exposure to the research process.7

Hill and Hupe identify seven factors that affect policy implementation.8 These include the manner in which the proposed intervention will affect frontline service staff (but also their moral approaches to the problem and their understanding of the policy) and intended beneficiaries. While acknowledging the complexity and serendipity of the policy process, certain features of the policy-making process, the content of policy and context within which it is made play a role in determining the prospects of getting specific policies onto the agenda and implemented.9

CONTEXT OF THE HIV RESPONSE IN PAKISTAN

Pakistan is a low-income country in which two out of three people live on less than $US 2 per day, and there are consequently relatively poor health outcome indicators at the national level. Pakistan’s progress towards the Millennium Development Goals is limited in many areas, particularly in those areas which impact directly on HIV including gender and reproductive health.

The HIV epidemic in Pakistan is “concentrated” among people with specific risk behaviours (especially injecting drug use, IDU), but the possibility of further epidemic spread has now been documented.10 11 A number of environmental and behavioural factors may further predispose to the spread of HIV in Pakistan inter alia, including entrenched gender inequalities, high levels of violence and unrest and a reported rise in injecting behaviours among drug users. Addressing current and future HIV epidemics in the country will probably require action at both the level of modifying individual behaviours and addressing the underlying structural determinants of those risks.12

The HIV response is situated within the context of an authoritarian state punctuated by periods of democracy, governing a fractious population with weak institutions and rule of law, preoccupied with issues of security and stability in the context of threats to its border and lacking a monopoly of violence within those borders. The state apparatus is marked by poor governance and significant corruption (bottom quarter in Transparency International’s 2007 Corruption Perception Index). Notwithstanding a high profile international meeting of parliamentarians in Pakistan in 2006 to address AIDS stigma and discrimination, parliament is not considered a particularly strong policy institution within an executive-dominated political structure. While select non-governmental organisations (NGOs) play a vital role in service provision (especially to reach the most at-risk populations such as sex workers and drug users), there are few strong civil society advocacy organisations involved in policy dialogue on HIV issues.

Legal context of HIV responses

Responses to AIDS/HIV take place in the context of an Islamic republic, but one marked by recurrent conflict between conservative social values and more liberal cosmopolitan views. Social values may be reinforced formally by an enabling legislative framework. Injecting drug use, sex with minors (<18 years for men and <16 years for women), “carnal intercourse … against the order of nature”, sex outside marriage and prostitution all constitute offences under the Pakistan Legal Code (including Section 377) and carry, theoretically, heavy penalties for offenders. Pakistan recently ratified the International Covenant on Economic, Social and Cultural Rights and signed the International Covenant on Civil and Political Rights and the Convention against Torture. However, during a May 2008 Universal Periodic Review (of progress on human rights), Pakistan publicly stated that there would be exceptions to its commitments, including the decriminalisation of non-marital consensual sex and adultery.

METHODS

A study of STI/HIV vulnerability in Pakistan made a large number of short- and medium-term recommendations for the National AIDS Control Programme (NACP) and its partners. Epidemiological analysis followed by modelling exercises identified four priority interventions: needle and syringe exchange programmes (NESP) for IDUs; targeted behaviour change communication (BCC) for people at high risk; comprehensive sexual health care for male and transgender sex workers (MTSWs); and sexual and reproductive health care for female sex workers (FSWs). A fifth intervention area was identified by the research team as important for addressing the context in which high-risk transmission takes place. It focuses on addressing stigma and discrimination and promoting and protecting the human rights of those at greatest risk, through the training of members of the uniformed services who perpetuate many of these risks.

A questionnaire was developed to gather stakeholder perceptions of the political feasibility of the widespread implementation of five recommended interventions. The questionnaire was piloted on two knowledgeable stakeholders, revised and then administered by email to approximately 40 stakeholders at national and provincial levels identified on the basis of their knowledge of, or role within, the HIV programme. Respondents were asked whether they agreed (on a scale of 1–10) with statements related to 15 potential determinants of successful policy implementation. These determinants were derived from theoretical and empirical literature as outlined in the introduction. These findings are discussed and interpreted in the context of some of their wider political, cultural and economic dimensions concerning the issues.

A visit to Pakistan in April 2008 enabled the authors to have meetings with 12 key stakeholders (some individually and some in groups, but all drawn from those who had completed a questionnaire) to probe for clarifications to the responses provided on the questionnaire and to explore the history, context and dynamics of HIV policy in Pakistan. Government documents and the very limited peer-reviewed literature on the national response were reviewed. The analysis and interpretation of the findings were based on a triangulation of the questionnaire, interviews/meetings and document analysis as well as the authors’ extensive knowledge of the national context.

RESULTS

Twenty-two informants returned a completed questionnaire, including current and former officials from NACP, Ministry of Health, Provincial AIDS Control Programmes (PACP), Ministry of Narcotic Control, multilateral and bilateral donor agencies, service delivery NGOs, international NGOs, think tanks, as well as a serving national legislator.

The position of HIV/AIDS is perceived by a range of stakeholders as occupying an unwarranted and exaggerated position on both development and health agendas in terms of its limited prevalence and burden of disease in the context of competing needs. AIDS is perceived as an externally imposed agenda, driven in part by donors. The country is a recipient of significant development assistance. In 2005, aid comprised 10.4% of central government expenditure.13 In contrast, approximately 85% of expenditures on HIV/AIDS are thought to be externally financed. Donors’ expectations that the government would absorb an increased proportion of HIV expenditure in the forthcoming programme is not likely to materialise to any significant degree. Nonetheless, funding for HIV/AIDS is not a limiting factor to the national programme; as much as 45% of the annual funding provided to the provincial AIDS programmes has to be returned unspent to the federal government, reflecting a common trend across the country resulting from capacity constraints. There is considerable variation among provinces in terms of leadership on AIDS.14

HIV policy and resource framework

HIV legislation (the Prevention and Treatment Act) was drafted in 2006 and provides the framework and standards for the national response.15 The legislation has been held up in the Ministry of Law and it may take some time before it reaches the legislature. Like most other legislative initiatives, this one was taken by bureaucracy with strong donor support. A National HIV and AIDS Policy, finalised in 2007, presents the broad outline of a 10-year vision but has yet to be approved by the Ministry of Health (MoH).16 The draft policy asserts that the “key aim of the legal and policy framework is to provide and maintain an enabling environment for HIV and AIDS prevention and care programs and services”. A National HIV and AIDS Strategic Framework 2007–2012 has also been drafted.17

Results from questionnaire

The results of the email questionnaire are presented in table 1. The wide ranges reflect disparate perceptions across stakeholders, while the medians suggest where the broad consensus lies in a relative manner.

Harm reduction for IDUs

Harm reduction for IDUs probably has the best chance of successful implementation. The problem is widely acknowledged and the intervention is considered evidence-based, technically and administratively simple, and easy to monitor. As the intervention is delivered by contracted NGOs, often relying on ex-users, little resistance to its implementation by frontline staff is anticipated. While scaling-up will require significant resources, respondents indicate that at least one donor will support it. The policy apparently has strong supporters and no strong opposition.

The final draft HIV policy indicates that the AIDS response will be targeted at the most at-risk populations and vulnerable groups, and places IDU at the top of the list. The draft legislation provides legal protection for service providers. In practice, successful comprehensive harm reduction interventions are already provided in a number of large cities by a few NGOs, with an informant estimate of 15% of the at-risk population covered.

Scaling-up will not be without its challenges. In the near term it is highly unlikely that the MoH will deliver this intervention as its frontline staff lack understanding, capacity and the required flexibility to work with the target population. The PACP leadership will, however, have to reach out to other ministries, and past success was often based on personal relationships. Similarly, local ex-user champions will need to come forward and NGO staff will have to develop trust with law enforcement agencies and local community leaders in new locales. These obstacles are not insurmountable but require time and money to address.

The history of the IDU programme might hold some lessons for the success of the other priority interventions. The policy initiative came from DFID and UNAIDS with subsequent support from the World Bank and domestically driven by specific individuals in the NACP. An inter-ministerial task force, chaired at the Director General level, was established. The policy and interventions were initially resisted by police, law enforcement agencies and bureaucrats in the MoH, not because of any organised interests but due largely to inertia and lack of understanding—reflected in the harassment of NGO staff delivering the interventions. Provincial PACP managers had to personally convince relevant district officials, relevant community leaders as well as get those responsible for the policy (eg, Inspector Generals) on side. NESP is now a widely accepted programme—mainstreamed in relevant government agencies—and its beneficiaries now face less discrimination and stigma. Nonetheless, it took 10 years to get to even this modest level of coverage.

Provision of comprehensive sexual health care for MTSWs

Although the epidemiological analysis suggests that male—and particularly transgender—sex workers will play a central role in the future of the HIV epidemic in Pakistan, the policy analysis suggests that the provision of effective interventions (such as distribution of condoms and lubricants, treatment of sexually transmitted infections, voluntary testing and counselling for HIV) will be less than straightforward. Respondents noted that the policy requires a significant degree of change from the status quo; indeed, the draft national AIDS policy does not make explicit mention of homosexuality or refer specifically to men who have sex with men (MSMs) or MTSWs—referring instead to “people who engage in sexual behaviour that puts them at risk”.

The proposed intervention is thought to have the lowest level of “resonance” with society of the five interventions considered. The intervention was thought to be very difficult to explain and, although most respondents expressed agreement that the intervention is evidence-based, the level of agreement was lower than for the other interventions. Respondents were less likely to agree that powerful and committed stakeholders are supportive of this intervention. While informants thought there were a number of donors supportive of the proposed intervention, the World Bank was the only specifically named donor likely to be willing to fund services. MSMs, MSWs and TGSWs are not networked or organised politically to demand recognition or services. As yet, no prominent champion has emerged to raise awareness about and frame the threat and solutions to unprotected commercial male and transgender sex.

Respondents thought that this intervention was most likely to witness opposition from powerful and committed stakeholders. Conservative social groups are held to be opposed to a public health response to the needs of these populations. Provision of services is seen to condone and encourage these activities and, consequently, conservatives favour clamping down on the supply of what is perceived as irreligious behaviour. The view was also expressed that bureaucrats would oppose allocation of domestic budget to services for this group.

While the interventions are technically simple, delivering and ensuring their uptake are complex. On the one hand, it was thought that the interventions are difficult to explain. On the other hand, although male to male commercial sex is fairly prevalent, the behaviour is highly stigmatised. Very few MTSWs currently seek care in the public sector due to stigma; it may therefore prove more effective to use public funds to ensure they receive high-quality and effective care in their current sites of health care seeking. Other challenges will also have to be overcome. It has proved difficult to establish dedicated clinics due to local opposition. In establishing the age of consent for HIV testing at 18 years, the draft policy excludes a proportion of the at-risk population who report age at first sex at a mean age of 13.3–14.1 years.10

Provision of comprehensive sexual and reproductive health care in public facilities which are accessible to FSWs

In Pakistan, FSWs report high levels of unmet need for comprehensive sexual and reproductive healthcare services. The draft HIV policy promotes the integration of HIV services into existing programmes, partly to avoid “unnecessary and unsustainable HIV-specific services”. However, the policy does not refer explicitly to FSWs or their clients.

Informants suggested that the proposed intervention represents a significant degree of policy change that will require significant levels of increased funding. Nonetheless, the problems faced by FSWs are relatively widely acknowledged. The proposed intervention is thought to be based on strong evidence and enjoys reasonable levels of support from powerful stakeholders (including from donors), but is not straightforward to explain. Concerns were raised that the intervention does not resonate widely with society and that socially conservative elements may be opposed to it, although it was thought that their level of commitment to such opposition was low. More problematic was the anticipated opposition to the delivery of services by frontline staff; it was suggested that this could be overcome with monetary incentives.

Provision of targeted BCC on reducing HIV risk

The draft national policy aspires to ensure that “all persons will be provided with access to the information and support they need to protect themselves against HIV infection”. Informants suggest that the prospects for implementing an effective BCC intervention are mixed. The use of explicit information remains problematic due to prevailing social norms, and hence the proposed intervention is thought to represent more than incremental change. Of the interventions considered, this one responds least to a widely acknowledged problem. It was also thought to be the most complicated in terms of the number of organisations and administrative layers involved, and there were concerns that the NGOs who deliver it alongside other inventions may lack capacity to deliver it at a sufficiently high quality. The intervention is thought to be highly evidence-based, easy to explain and supported by powerful and committed stakeholders (including one or more donors). Moreover, despite opposition to explicit messages, powerful and committed stakeholders are thought unlikely to oppose the intervention directly as it is targeted at specific at-risk populations out of mainstream view.

Use of public funds to support NGOs to conduct training workshops with police and corrections officers

All groups in the epidemiological surveys reported high levels of abuse of their basic human rights.10 In common with the other interventions, respondents acknowledged that addressing the human rights abuses suffered by those most at risk would require quite major change for the national AIDS response but, in contrast to the other inventions, would require few additional resources. Despite the issues involved, it was reported that the intervention addresses a relatively widely acknowledged problem and that it would resonate more with social values than the other interventions. Respondents did not think that it rests as firmly on sound evidence as do the IDU or FSW interventions, but that the evidence was just as strong as the interventions for MSWs. It was thought that the intervention is relatively easy to explain, has relatively straightforward technical features, slightly lower levels of support from powerful stakeholders (compared with other interventions, except MTSW), but no real opposition.

The national policy on prevention of HIV currently recognises the potential increased risk that uniformed personnel face and proposes that vulnerability can be addressed through targeted programmes. Sensitisation and awareness-raising programmes on HIV are currently being conducted on an ad hoc basis within uniformed service training programmes. The intervention could build upon the successful efforts made to integrate a module on violence against women in training programmes for new recruits into the police force. This took 10 years to develop, was funded by donors and will require ongoing funding to provide the required incentives for participation. It was suggested that success in working with the uniformed services will depend on getting them to understand their personal risk in having unsafe and forced sex with these populations as opposed to protecting and promoting the rights of vulnerable populations, although these are not mutually exclusive approaches.

DISCUSSION

Each of the proposed interventions requires a considerable shift from the status quo in terms of policy content, implementation strategies or scale of delivery—often all three. Many informants suggested that the AIDS response is increasingly driven by evidence (including evidence gained from operational experience), and that the researchers should feel confident that their findings will be reflected in the national programme. The idea that the strength of the research alone will bring about the required changes is debatable, particularly in the short term. One theory of the research policy nexus argues that ideas derived from research filter into the policy networks that shape the policy process in a particular field and have a cumulative indirect effect rather than an immediate direct effect on policy.18 The literature also suggests that it is more likely that evidence will be used by policymakers if researchers engage directly with them, if evidence is turned to arguments and stories which resonate with key decision-makers, and if evidence is used to influence the values and beliefs of decision-makers instead of just being used to contribute to the knowledge base.6 19 The unusually close interaction between the researchers and NACP provides grounds for optimism, as does the fact that the researchers are presently preparing policy briefs which aim to convince wider audiences of the need to think about risk and vulnerability in new ways.

The role of donors in initiating and supporting policy change (and funding the research upon which the case for reform is often based) is pronounced in the AIDS sector in Pakistan.20 Informants agreed relatively strongly that one or more donors is supportive of each of the proposed interventions. This good news is a double-edged sword as external leadership may undermine national ownership, domestic accountability and the sustainability of the interventions. Nonetheless, addressing the HIV epidemic in the most cost-effective manner is of global concern and, if local leadership is not forthcoming, it is not clear what the alternatives are.

A strong case can be made that, in addition to supporting scaled-up delivery of the priority interventions, donors should support the development of constituencies among the highly stigmatised groups for whom the interventions are intended in order to build their capacity to demand government accountability for their delivery. Experience suggests that the effectiveness of nascent coalitions is enhanced through the work of well connected champions who front the cause in appropriate policy circles. Of particular relevance is the successful leadership and work of Nai Zindagi, a domestic NGO, in rolling out interventions for IDUs in Pakistan.21 A number of reinforcing factors are believed to have facilitated this particular champion, including strong political connections, charismatic leadership, technical competence as well as an emic perspective derived from the experience of ex-IDUs.

Surprisingly, the media was not mentioned by any informant in relation to policy change—a channel which advocates might consider in changing public perceptions of potentially stigmatised groups.

The feasibility of the implementation of four (if not all) of the interventions is improved by virtue of the fact that they will be delivered by contracted NGOs. This largely obviates potential problems in mobilising government service providers to meet the needs of these communities, and creates domestic interest groups to demand the financing of these interventions. This approach does raise questions about the sustainability of the interventions, works against the integration agenda and perpetuates the treatment of MSWs, TGSWs, FSWs and IDUs as “other” and is possibly stigmatising. Nonetheless, this process may also forge increasing solidarity and lead to a political identity as it has elsewhere.

Limitations of the study

The results of the study need to be treated with some caution as they are based on a survey of a very limited number of purposely selected informants carried out during a short period of time (a function of the resources available for the study). Most of the informants were drawn from the field of HIV-related work; most were policy elites and we did not have the opportunity to interview service providers or people from most at-risk or vulnerable communities. Moreover, some informants may have not fully understood the implications of the recommendations or the statements that they were asked to rate or, indeed, may have concealed their true positions, particularly given their understanding of the purposes of the exercise. These are common constraints to policy analysis.22

CONCLUSIONS

The political prospects of each the five interventions differ; each presents unique bottle necks—particularly in implementation. The manner in which these constraints are addressed will determine which interventions are successfully scaled up. While the interventions mark a considerable change from the status quo, NACP—which is tasked with coordinating the development and implementation of policy—is largely supportive. Further research is required to gauge the support of other government departments, particularly in relation to the human rights intervention. In contrast, it appears that there is less support from society at large to address what is considered a problem affecting populations engaging in behaviours which are highly stigmatised (particularly the sex-related ones). We conclude that it is likely that the scaling-up of needle and syringe exchange will be least problematic, whereas those addressing the risks associated with MSM and male and transgender sex work will be most challenging. It is anticipated that there will be little outright opposition to the implementation of these interventions as HIV is not perceived as a priority issue by those who are negatively predisposed to the most at-risk groups (eg, conservative religious groups). Despite the lack of outright opposition to the interventions and a good evidence base at the disposal of government, support for the reforms is rather fragile. Demand for the interventions can be deepened and widened by assisting the formation of advocacy coalitions over the longer term.2 More in-depth research to understand the political dimensions of each of the interventions is called for to support the coalitions that generate the political will to sustain them.23 24