[e-drug] Analysis of South Africa's NDP process

["Andy Gray" <Graya1@nu.ac.za>]

E-drug: Analysis of South Africa's NDP process
---------------------------------------------

[Warning: long message]
Dear E-druggers

South Africa's 1996 National Drug Policy has been closely watched for 
a number of reasons. An analysis of the process of developing and 
implementing the policy has recently been completed by the Centre for 
Health Policy, School of Public Health, University of the 
Witwatersrand. The authors of the report - entitled "POLICY CHANGE IN 
A CONTEXT OF
TRANSITION: DRUG POLICY IN SOUTH AFRICA 1989-1999" - are Andy Gray, 
Thulani Matsebula, Duane Blaauw, Helen Schneider and Lucy Gilson. A 
full text version of the report has been placed on the Centre's web 
site at http://www.hst.org.za/chp/pubs/policy.pdf (374kb)

Below is the Executive Summary. We would welcome comments on the 
report from E-druggers. Those interested in the overall financing 
scene could also look at Gilson L, Doherty J, McIntyre D, Thomas S, 
Brijlall V, Bowa C, Mbatsha S. The Dynamics of Policy Change - Health 
Care Financing in South Africa 1994-1999 - also available on the CHP 
site at http://www.hst.org.za/chp/pubs/policy_change.htm

regards
Andy
~~~

EXECUTIVE SUMMARY

Rationale and Methods

The 1996 National Drug Policy (NDP) was one of the most comprehensive 
policy statements to emerge from the South African Department of 
Health after the change of government in 1994. Drawing on earlier 
experience with aspects of drug policy reform, this policy was the 
first attempt to introduce a coherent set of reforms aimed at 
ensuring more equitable access to drugs throughout the country.

Implementation of the 1996 NDP was, however, characterised by a 
mixed, and at times controversial, set of outcomes. Several reviews 
of NDP implementation have highlighted important gains, notably in 
the development of an essential drugs list and standard treatment 
guidelines for various levels of public sector care. An arguably less 
successful aspect of policy implementation has been the programme of 
legislative reform.

What explains the experience and outcomes associated with the process 
of developing and implementing the NDP? This report tackles this 
question by presenting an analysis of the roots and evolution of the 
NDP between 1989 and 1999. In addition, it specifically discusses the 
processes involved in developing and taking forward public sector 
Essential Drugs Lists and legislative reform. The objectives of this 
evaluation were to:
* Document the evolution of the NDP in relation to
  -  The actors and interests affected by the policy, and
  -  The processes by which the various actors had interfaced with the NDP,
from agenda setting to policy evaluation;
* Identify the factors facilitating and constraining the development 
and initial
implementation of selected aspects of the NDP;
* Draw broad conclusions for future drug and other health policy 
processes in South Africa.

It should be emphasised that this report is not an evaluation of the 
extent to which the NDP has led to change being achieved on the 
ground. Rather, its focus is on documenting the chronology of the 
NDP's development and identifying some of the key factors that 
explain the outcomes of this policy process.

Key Issues the Study Looked at and How These Were Addressed

The detailed design of the study drew extensively on the approach of 
a large review of health financing policy reform in South Africa that 
was undertaken by the Centre for Health Policy and the Health 
Economics Unit, University of Cape Town (Gilson et al., 1999). The 
conceptual framework of both studies promotes detailed investigation 
of four key sets of factors that influence the overall process of 
policy formulation and implementation: factors of context, actors, 
content and process (Walt and Gilson, 1994).

The framework begins from the premise that these factors ultimately 
shape the nature and extent of change achieved by the intended 
reform. Given the diversity of policy changes encompassed by the NDP, 
this evaluation focussed on two specific changes: implementation of 
the essential drugs list in the public sector and legislative reform. 
These two aspects were chosen because of the different actors 
implicated as well as the somewhat different outcomes: a degree of 
progress or success in relation to drug selection,
and blockage in the case of key legislation. A focus on both aspects 
of the NDP allows for a fuller appreciation of how different content 
areas, actors, processes combine to produce particular policy 
outcomes.

Comprehensive information was gathered through extensive review of 
all available documentation, 15 in-depth interviews with 
representatives of key stakeholders, and finally, a critical review 
of the draft document by three key informants. This report presents 
an integrated analysis of these different data sets. Such analysis 
requires interpretation of the information collected and judgments 
about impacts and key influences. Every effort has, therefore, been 
made to ensure a rigorous process of analysis leading to valid 
conclusions and recommendations. Four key steps should be noted.
Firstly, the research team benefited from bringing together both 
'insiders' (researchers with detailed knowledge of the policy process 
under investigation) and 'outsiders' (researchers who had previously 
had less or no involvement in the policy process). Secondly, a 
grounded approach to analysis was adopted to allow the views and 
perspectives of the different interviewees and data sources to be 
fully investigated. Thirdly, data were triangulated within and 
between sources. Finally, a deliberate and thorough review process of 
the draft report was carried-out. There were, nonetheless, two major 
limitations that may have influenced the analysis. Biases in analysis 
may have been introduced by the presence in the research team of 
'insiders' who continue to play various roles in policy change. In 
addition, the research team had some concerns about interviewee 
balance and availability.

Directions Through the Report

The key questions of the evaluation as well as the conceptual 
framework, and a more detailed description of the methods, are 
presented in Chapters 1 and 2.

Chapters 3 and 4 provide a description of the local and international 
context of drug policy development in South Africa, including the 
major drug policy developments in the apartheid era. Chapter 5 then 
provides a detailed account of the events between 1989 and 1999 that 
have shaped the South Africa National Drug Policy. It describes how 
drugs became part of the overall health policy agenda and then 
discusses the processes that led
up to the publication of the NDP in 1996. It also examines in more 
detail the processes underlying the development of Essential Drugs 
Lists (EDLs) and legislative reform.

Building on the detailed description of the NDP's evolution, Chapters 
6-9 present an attempt to understand and analyse these experiences. 
These chapters look specifically at how the context of political 
transition influenced drug policy reform, what positions and stands 
key actors took, and how their actions shaped the process, as well as 
how different types of knowledge were used within the process.

Finally, Chapter 10 highlights the key conclusions derived from this 
analysis of the South African experience of comprehensive 
pharmaceutical policy reform.

An Outline of the Policy Process

Debate about post-apartheid health policy was initiated before 1994. 
On the one hand, the apartheid government continued to formulate 
health policy and produce legislation - such as the 1993 draft 
National Pharmaceutical Policy. On the other hand, a strong 
progressive health movement evolved outside government comprised of 
health activists, academics
and returning exiles. As early as 1993, for example, the African 
National Congress (ANC) established its own Drug Policy Commission 
(DPC) to begin debate on drug policy issues. This Commission's work 
then fed forward into both the ANC's overall economic and social 
policy programme and its 1994 National Health Plan. As a result, by 
the time of the 1994 elections, drug policy had both been clearly 
placed on the new government's agenda,
and significant progress had been made in identifying the key aspects 
of such policy. Indeed, in many respects the two sets of pre-1994 
proposals including some of the ideas developed by the apartheid 
government, provided the basis for the 1996 National Drug Policy 
document.

Nonetheless, the 1996 National Drug Policy was ultimately the result 
of a policy process initiated by the new government. Soon after 
coming into office, the newly appointed Minister of Health convened a 
12-member National Drug Policy Committee (NDPC) that included some 
members from the ANC's DPC. Although not specifically tasked with 
drafting a policy document, its mandate was to develop a plan to 
reduce the cost of drugs
while ensuring that drugs are used effectively, rationally and 
cost-effectively. It was given a short time frame for its work (three 
months) and the final report consisted of a list of options that did 
not provide an adequate basis for rapid implementation decisions. The 
next stage of policy development, therefore, combined a process of 
drafting a formal and comprehensive NDP with direct moves to 
implement specific policies such as an Essential Drugs List.

The process of developing the NDP drew both on the NDPC report and a 
series of consultative meetings with a range of stakeholders, 
including the pharmaceutical industry. It was also discussed within 
the 1995 Committee of Inquiry into Social Health Insurance at its 
provincial hearings. At this stage, the most vocal opposition came 
from the South African Pharmacy Council and the Medical Association 
of South Africa, with the pharmaceutical manufacturers remaining 
relatively quiet in the debates. Although attempts
were made by the policy drafting team to accommodate some of the 
expressed concerns, the government made it clear that it would keep 
its options open on contentious issues such as generic substitution 
and parallel importation within the NDP.

Ultimately the final version of the NDP reflects the thinking and 
work of a small policy network. This network consisted of new cadres 
within government (some who had been involved in pre-1994 debates), 
supportive technical experts, external consultants from the World 
Health Organisation's (WHO) Drug Action Programme and elsewhere, and 
some officials from the former government. They not only ensured that 
momentum was
maintained during the process of drafting, but also made certain 
important strategic decisions over policy content. One important 
influence over their thinking was exposure to international 
experiences. During the 1994 NDPC, for example, three committee 
members visited Zimbabwe to review the implementation of its NDP. 
Sponsored by WHO, the trip made a strong impression on the South 
African delegates. Perhaps not surprisingly,
therefore, the final version of the NDP closely resembled WHO's 
prototype drug policy statement.

Following the NDP's finalisation, the South Africa Drug Action 
Programme (SADAP) was established to facilitate its implementation. 
SADAP proposed a three phase implementation process that would allow 
the staggered introduction of policy changes, beginning with those 
that were less complex (such as the EDL for primary care) and ending 
with those that required major changes (the legislative reforms). In 
practice, however, the proposed reforms were implemented more or less 
simultaneously.

A first draft public sector EDL was distributed for comment in 1995, 
drawing on work undertaken before the 1994 elections as well as 
subsequent work by the 1994 NDPC and the 1995 SHI Committee of 
Inquiry. Developed by a National EDL Committee, the first EDL was 
received favourably by public sector stakeholders. The process of EDL 
development had involved extensive consultations with health 
professionals and provincial policy makers conducted by two EDL 
Committee members who worked on the issue full time for some months. 
The ultimate approach chosen, to develop a drugs list based on a set 
of standard treatment guidelines for PHC rather than a generic WHO 
drug list, was the product of the process of buy-in and compromise 
achieved through the consultation. Although, as with the draft NDP, 
professional associations with a predominantly private sector base 
voiced strong objections to the draft EDL, these concerns were not 
seen as
relevant to a public sector EDL. Eventually 160 drugs were selected 
and the PHC EDL and STGs were distributed widely in April 1996. An 
evaluation as early as 1997 found high levels of availability of STGs 
in health facilities. Immediately after the launch of the primary 
care EDL/STGs a review process was initiated to develop a hospital 
level EDL. Again the EDL development was rooted in the development of 
STGs, and involved wideranging consultation with public sector 
professionals. The full set of EDL/STGS were finally completed in 
December 1998 and included: the revised PHC EDL, a hospital EDL for 
adult patients, and a paediatric hospital EDL. At this point 
implementation slowed down. The outstanding EDL implementation tasks 
included strengthening provincial structures and ensuring 
coordination between provinces.

In contrast to the relatively smooth development and implementation 
of the EDLs, legislative reform was a much more confrontational and 
slow process. It was subject to a series of critical challenges that 
delayed and undermined the reforms proposed in the NDP. As a result, 
by February 2001, five years after the launch of the NDP, none of the 
legislative reforms had been implemented.

Unlike the EDL development process, the proposed legislative changes 
were initially drafted by a small working group of seven senior 
government officials and the Minister's special adviser's office, 
which was established in 1996. The pharmaceutical regulatory 
authority, the Medicines Control Council, was not represented on this 
working group. The Medicines and Related Substances Control Act 
(MRSCA) of 1965 was identified as the key vehicle for legislative 
reform, and, as a first step, a series of regulations to this Act 
were published for comment in July 1996. These regulations elicited 
strong reactions from various stakeholders, including manufacturers 
opposed to bearing the costs of a proposed new drug labelling system 
and the medical profession opposed to proposals on the regulation of 
dispensing. Other concerns focused on the technical weaknesses in the 
legislation. The Minister was forced to withdraw the regulations 
after public hearings within parliament.

The legislative working group then began drafting amendments to the 
MRSCA, and tabled the MRCSA Amendment Bill in the National Assembly 
in May 1997 without having released a draft for public comment. Some 
of the most controversial of the previously proposed regulations were 
again included, and in addition the Bill gave the national Minister 
the power to grant compulsory licenses and allow parallel 
importation. Almost
immediately the Bill was met by vocal opposition from manufacturers 
and professional organizations. Soon after the Minister withdrew the 
Bill and a four person working group, including the Minister herself, 
was established to re-draft it. Although some changes were suggested 
as a result of the technical input of the MCC, in practice the 
decision was taken
to table a new draft Amendment Bill little different from the earlier 
version. Public hearings were again held in parliament and again 
there was vocal opposition, as well as public criticism from the MCC. 
In a meeting with the Minister the Pharmaceutical Manufacturers 
Association (PMA) specifically requested her to withdraw the 
components of the Bill that allowed compulsory licensing and parallel 
imports. However, the Bill was
passed by parliament and assented to by the President in December 1997.

On February 18th 1998, the PMA and 41 co-applicants sought an interim 
interdict from the High Court preventing the President from bringing 
the Act into effect, largely on the grounds that it was 
unconstitutional in giving the Minister unrestricted powers to revoke 
patent rights. The court case took another three years to resolve. 
Finally, in the face of a national and international alliance between 
the government and AIDS activists, the
industry made an unconditional withdrawal from the case in April 
2001, allowing the government to promulgate the Medicines and Related 
Substances Control Amendment Act (Act 90 of 1997).

Meanwhile, however, the government's parallel attempt to amend the 
legal basis of the national pharmaceutical regulatory authority was 
also blocked. The South African Medicines and Medical Devices 
Regulatory Authority Act (Act 132 of 1998) was criticized during 
public hearings as being well intentioned but poorly drafted and was 
eventually promulgated before any of the regulations or drug 
schedules it required had been prepared. Technical problems 
surrounding the Act led both the DOH and PMA to take legal action to 
request that the promulgation notice be set aside. Eventually, on her 
appointment in 1999, the second Minister of Health announced her 
intention to seek its repeal.

Factors Influencing The Policy Process

Since 1994 South Africa has embarked on a programme of unprecedented 
political, social and economic transformation that aims to redress 
decades of colonialism, racism and repression, and to develop a 
society based on democracy and social justice. The health sector has 
been at the forefront of the new government's ambitious reform 
programme. Its visibility after 1994 is generally attributed to the 
role of the former Minister of Health, Dr.
Nkosazana Dlamini-Zuma, who is frequently characterised as a 
powerful, dynamic leader with strongly held values. She certainly 
provided strong political leadership to the transformation process, 
including for pharmaceutical policy reform.

A period of radical transition, such as that faced in South Africa 
after 1994, creates significant opportunities for policy change but 
may also contain elements that constrain the policy process. Eight 
key features of the South African context thus clearly marked the 
subsequent drug policy reform process. These features were:
* the strong political mandate given to the new government to redress 
the inequities that were inherited from the apartheid era, which was 
translated into a specific mandate for the drug reforms proposed in 
the 1994 ANC Health Plan;
* a need for urgent action by the new government to deliver on its 
election promises, and effect dramatic and speedy improvements in 
people's social and economic status;
* a significant but short window of opportunity for wide-ranging policy reform;
* marked societal polarisation characterised by the ANC's legitimate 
mistrust of former opponents, generating negative attitudes towards 
the private sector, in general, and the pharmaceutical industry in 
particular, as well as suspicion of the mainstream medical profession 
given its poor record during the injustice of the apartheid years;
* perpetual change in government structures, policies and personnel 
after 1994 that provided an unstable environment in which to bring 
about policy change;
* the weak inherited institutional capacity of the civil service 
which was characterised generally by centralised and rigid 
procedures, poor levels of competence, and resistance from 'old' 
bureaucrats;
* limited experience of pharmaceutical regulation including no 
previous attempts to manage drug costs, availability, or 
accessibility and a general lack of information with which to 
regulate;
* the search to prioritise a new set of values based on democracy and 
social justice throughout the transformation programme.

Yet despite the uncertainty and confusion inevitable in a time of 
transition, the initial steps of NDP development, at least until the 
launch of the NDP document, can be characterised as primarily a 
rational and linear approach to policy change. Following an approach 
broadly similar to that proposed by the WHO for the development of 
national drug policies, an initial working group undertook a 
situation analysis which fed into the formulation of a discussion 
document. This document was made available for comment
and discussed with a wide range of stakeholders, before the NDP was 
finalised and launched in 1996. At this point, however, rationality 
began to break down. On the one hand, although SADAP was formed to 
manage implementation, there was less explicit planning of the 
implementation process than proposed in WHO documentation. An action 
plan was not, for example, developed to guide implementation; neither 
were monitoring
and evaluation systems established. On the other hand, the attempts 
to implement legislative reforms were characterised by complexity and 
conflict. Although the potential for conflict was clearly recognised 
by policy-makers, the rational frameworks of the WHO did not appear 
to have assisted them in thinking through how to prepare for and 
manage that conflict. Instead, it appears that these frameworks may 
have encouraged the view that
implementation is an automatic consequence of having a policy 
document, rather than a step requiring explicit management.

In addition, the implementation experience of the two strands of 
reform examined in detail in this analysis had significant 
differences. Where the development of the EDLs was consultative and 
deliberate, attempts to implement the legislative programme were 
characterised by haste and conflict. Perhaps the most important 
difference between these
two reform strands was the configuration of actors caught up in or 
brought into the processes. A much wider array of more powerful 
actors, towards many of whom the new ANC government felt distrustful, 
were affected by the legislative process than by the EDL process, 
which was primarily an internal government affair. Where the new 
policy-makers felt comfortable with an open process of consultation 
with government colleagues and
supportive academics, they were perhaps suspicious of the 
pharmaceutical industry and the professional bodies given their 
actions and inactions during the injustices of the apartheid era.

These important differences in the actor configurations of the two 
reform strands fed forward into different responses to the same 
context. Recognising the potential for conflict, the new 
policy-makers sought to use the short window of opportunity to push 
through speedily the legislative reforms that would provide the 
foundation for tighter regulation of
the pharmaceutical market and actors. They appeared to work from the 
understanding that at such a moment political will was the main tool 
needed to combat those opposed to such reforms. In some respects this 
understanding proved valid. Yet despite the highest political
support, the drive for change was undermined by key capacity 
weaknesses in the bureaucracy inherited from the apartheid era. Basic 
mistakes and errors were introduced into the legal drafting 
processes, and in some cases became the basis for successful 
resistance to the proposed legislative changes.

The final, and linked, factor that seems to have shaped the overall 
NDP policy process relates to the particular types of knowledge that 
were used to guide it. All policy development requires both technical 
knowledge, specific to the policy area, as well as knowledge related 
to the values and context of public policy in the country, how the 
bureaucracy functions and which actors hold what positions in 
relation to the policy topic. In principle this broad range of 
knowledge can be drawn from various sources, including
academic and international organisations. In addition, the civil 
servants at the various levels of government who are generally tasked 
with bringing about policy change, are always likely to be important. 
The political transition in South Africa, however, had an important 
influence on which sources of knowledge were privileged in the 
post-1994 policy processes. Although some of those involved in 
thinking about drug policy reforms
were brought into the new, post-1994 reform teams, many of the old 
civil servants, those who had most knowledge of how the bureaucracy 
functioned, were effectively excluded from these teams. Given the 
inevitable antipathies between former political opponents, this
was not surprising. It did, however, deprive the new policy-makers of 
knowledge important to the process of managing change.

Conclusions

Transition periods are classically windows of opportunity. This was 
clearly demonstrated in South Africa by the setting of a new agenda 
and strong political commitment to challenging vested interests in 
the drug policy arena. However, it is important to recognise that the 
destabilised environment of a transition also contains a set of 
constraints. Changes in leadership, loss of institutional memory and 
a rupture of old mechanisms of policy implementation create their own 
difficulties for the policy process The drug policy process in South 
Africa was driven by actors who had clear intentions and the will to 
implement them, but who lacked direct experience of government. They 
not only had to institute bold new policies through bureaucratic 
structures with old values and forms of governance, but
this structure had a quasi-federal political arrangement imposed on 
top of it, which created a disjuncture between the spheres of policy 
making and implementation. In addition, the NDP was confronted with a 
wide variety of sophisticated extra governmental players, including a 
well-established domestic pharmaceutical industry able to mobilise 
technical and legal resources to oppose policy. Developing and 
implementing policy in such a
context was bound to be a highly complex process in which there are 
few guides to strategy and little prior experience to draw upon.

Nevertheless, this study found that successes are possible within 
such a context and that implementation obstacles in drug policy can 
be traced to weaknesses in the process of policy-making. On the one 
hand, it is not sufficient to simply acknowledge powerful and varying 
interests - their involvement in policy processes needs to be managed 
deliberately.
On the other hand, political will cannot compensate for inadequate 
technical knowledge and experience in a highly complex arena such as 
pharmaceuticals.

As Brinkerhoff (1996) has noted, transitional periods require special 
attention to processes through which to manage and coordinate actors. 
In the absence of such processes, change is unlikely to happen. 
Engagement with actors is not only about co-opting them and 
legitimising new policies, but also about capturing and incorporating 
knowledge required for implementation. In other words, the aims of 
engagement are not only to give 'voice' to actors, but also to test 
reactions to policy, anticipate unintended consequences, uncover 
technical weaknesses and find ways to modify policy in the light of 
these. However, successful implementation does not necessarily 
require universal assent or consensus - for example, the public 
sector essential drugs policy process correctly focused on getting 
public sector buy-in and could, to some extent ignore opposition from 
private professional groupings.

The challenge of implementation is thus less a matter of following 
blue-prints and recipes than of 'learning by doing'. This involves a 
high degree of organisational reflexivity - the ability to learn from 
experience. In the instances where the South African drug policy 
established processes that met these requirements, such as during the 
writing of the policy
and the essential drugs list, outcomes were successful. Where these 
processes were not sustained, gains were not maintained or followed 
through. On the whole, the drug policy process in South Africa showed 
a high awareness of the actor environment, but insufficient 
recognition of the fact that policy implementation is inherently a 
process of constant
negotiation and renegotiation. Over time, the opportunities for 
negotiation have tended to diminish rather than expand. If future 
policy implementation is to be successful, mechanisms for 
organisational learning need to become far more institutionalised. 
Within government, this implies ensuring ongoing mechanisms to 
combine top-down, national with bottom-up, district and provincial 
planning. Also important are regular review and evaluation, creating 
opportunities for external actors to regularly interface with policy 
and legislative processes, and opening up public debate on measures 
that are likely to be controversial and elicit reaction from players 
with strong interests.

~~~~~~~~~~~~~~~~~~~
Andy Gray MSc(Pharm) FPS
Senior Lecturer
Dept of Experimental and Clinical Pharmacology
Nelson R Mandela School of Medicine
PBag 7 Congella 4013
South Africa
Tel: +27-31-2604334/4298 Fax: +27-31-2604338
email: graya1@nu.ac.za or andy@healthlink.org.za
--
To send a message to E-Drug, write to: e-drug@usa.healthnet.org
To subscribe or unsubscribe, write to: majordomo@usa.healthnet.org
in the body of the message type: subscribe e-drug OR unsubscribe e-drug
To contact a person, send a message to: e-drug-help@usa.healthnet.org
Information and archives: http://www.essentialdrugs.org/edrug