NEWS - July 2002
Tough lessons and rapid reform for vaccine procurement system
THE ALLIANCE must adopt a more efficient and transparent process for buying vaccines, its Board has agreed. The decision follows a review (1) of the first round of procurement and introduction of new vaccines which identified significant shortfalls in the system and recommended immediate changes.
In large part, the problems with the first round of buying vaccines, which took place during 2000 and 2001, resulted from the ambitious timetable that GAVI set itself. Immediately after its launch, the Alliance set out to enable low-income countries to introduce new vaccines against two important diseases hepatitis B and Haemophilus influenzae type b (Hib). In the process, it sought to demonstrate to a sceptical industry that there is a real market for newer vaccines in developing countries.
|[ Photo: Lisa Jacobs ]Piers Whitehead: the world has changed
But while the review found that the Alliance had made remarkable progress in getting new vaccines to more children at reasonable prices, it found a lack of credible and predictable demand for vaccines, mixed messages to suppliers, and muddled aggregate performance of a group of partners with "unclear overlapping roles", none of whom was ultimately accountable for the process. In these respects, the review concluded, the first round was "a missed opportunity".
Most serious was the gap between the awards to manufacturers to supply vaccine and the amounts actually bought. For 2001, UNICEF indicated that it would want 98 million doses of vaccine but bought only 18 million. For hepatitis B monovalent vaccine, only 11% of the volume that was requested was actually bought.
Mercer Management Consulting, a firm that has previously analysed vaccine supply issues for UNICEF, was commissioned by GAVIs Financing Task Force and the Vaccine Fund to produce the review. The goal was to learn lessons for the immediate future, not to assign blame for the past. Piers Whitehead, a co-author of the report and vice-president in the firms San Francisco office, told Immunization Focus that the first procurement round had been a steep learning curve for all involved.
"Its important not to ignore the fact that until now, no new antigens have been introduced into the Expanded Programme on Immunization since it was set up in the early 1970s," said Whitehead. In the past, he added, the public sector had been able to get the vaccines it has needed, and so had not needed to invest as much time and energy in accurate forecasting of demand as it does today. "But the world has changed," said Whitehead.
To ensure the lessons of the first round of procurement are learnt before the second begins, immediate action is needed. The tendering process for the second round is due to start just weeks from now, in September. At its meeting in Paris on 19 June, the Board approved the reviews recommendation that a single project manager will coordinate the procurement of vaccines and the roles of each partner in the process. The project manager will be accountable to a small oversight body to be appointed by the GAVI Board.
As well as assessing the first procurement round, the Mercer review also made an assessment of the global vaccine market and the implications of the market structure for GAVIs procurement strategy. The review confirmed that, although the market has grown sharply, more than doubling since 1992, the growth is primarily in vaccines for the industrialized countries. GAVI is a small player with relatively little leverage in determining the activities of the major pharmaceutical companies, although it is a major player for emerging suppliers in middle-income and low-income countries.
Mercer reviews key recommendations:
- Institute a project management approach to planning and implementing key Alliance initiatives, such as procurement of vaccines
- Pilot the project management approach with the upcoming procurement round for 2004-6
- Create or instruct an oversight body to monitor progress and hold the project manager and relevant individuals and institutions accountable for performance
- Ensure that information on demand, product preference and future needs is shared with industry, unless there is a well-defined reason not to do so
Dr Tore Godal, executive secretary of GAVI, said the review had highlighted the complexity of vaccine procurement and the importance of understanding a market where supplies had once been abundant but are now more closely balanced against demand. "Another lesson is that we need longterm strategic direction," he said. "The industry needs credible multi-year plans from the public sector before it can invest in producing vaccines. Donor commitments need to take account of this."
The review did note some clear positive outcomes of the first round of procurement. First, the Alliance had achieved a "unique accomplishment" in successfully increasing childrens access to under-used vaccines in an unprecedentedly short period of 14 months from start to finish: in total 41 million doses of DTP-HepB and DTP-HepB-Hib are available for delivery during 2002-2003. Second, and equally important, the orders for the vaccines have created a new "pull" on the industry, stimulating manufacturers to increase their capacity to make these vaccines and to compete with each other. In the longer term this will mean more plentiful and affordable vaccines.
Third, the prices negotiated for the vaccines were also reasonable, the review found. For example, the DTP-HepB-Hib combination was bought for almost half the average price at which monovalent Hib vaccine is sold in industrialised countries. These prices benefited the public sector, but also gave manufacturers an appropriate return for their products, breaking the tradition that vaccines for the Expanded Programme on Immunization should be sold at near-cost price. While Mercer acknowledged that this break with tradition is "controversial to some", its review advised that vaccine producers will need appropriate returns if the supply of vaccines for low-income countries is to be sustainable.
The reviews analysis of the problems with the first round showed that several factors were responsible for the severe mismatch in forecasted and actual demand. One key reason was that GAVI had promoted the use of combination vaccines before the supply situation was made clear by industry partners. When it emerged that supplies of these products would fall well short of demand, the Alliance was forced to "ration" them and some countries were offered only the option of using the plentiful supplies of monovalent hepatitis B vaccine instead. As a result, two big countries that had initially indicated that they would want hepatitis B vaccine, but in combination, decided to delay its introduction. This was "a significant source of error" in the forecast, said Whitehead. The review says that, although the gap between the awards and the amount of vaccines bought will narrow as these countries come online, "from the supplier perspective, this delayed volume should probably be considered permanently lost."
Another factor was that awards to manufacturers were based on the most optimistic of three demand scenarios. Further, the award process added an extra 17 million doses to that high figure. Several possible explanations for this decision are suggested by the review. The process of awarding tenders appeared to have been required to serve two contradictory purposes: first, ensuring adequate supplies for countries, which UNICEF has traditionally done effectively, and second, establishing from the industry how much vaccine was available.
The extra 17 million doses may have been added because companies had made bids for certain quantities that UNICEF had to accept or reject in their entirety. "Regardless of explanation, the gap between award and offtake does not support the objective of credible and predictable demand," the review concludes.
Because time was so short, vaccine orders had to be placed before demand requirements were clearly known. "Obviously, given the pressure of time and the new ground being broken, some inaccuracy was unavoidable," says the review.
Three "semi-independent" processes had been used by the Alliance to establish countries vaccine requirements, the review found. These were the countries own proposals for awards, a forecast developed for the Financing Task Force, and UNICEFs own consultations with countries. "In aggregate, from a procurement and forecasting perspective, these yielded the disappointing outcome for 2001," the review concludes. According to Steve Jarrett, Deputy Director of the UNICEF Supply Division, "The start-up was slower than expected due to delays in the availability of vaccines as well as countries decisions to change the timing, quantity, and in some cases, types of vaccine demanded."
From the standpoint of the vaccine industry, the problems were centred around a lack of openness and collaboration. Although the forecasting of immunization demand was shared with the industry for the first time by GAVI, the information was in the form of the number of children to be immunized with given antigens, rather than guidance about the actual products such as combination or monovalent vaccines that countries wanted. "Further, country level demand data was not shared with suppliers as it became available," says the review.
One reason for this, says Whitehead, is because the public sector is traditionally wary of providing the private sector with data that might, in theory, be misused by companies to exert inappropriate influence on countries to buy their own particular products. However, says Whitehead, a well functioning Interagency Coordinating Committee in countries should be able to detect any such inappropriate behaviour.
Of course, the Alliance must ward against conflict of interest such as giving "insider" information to one company in favour of another. But, says Whitehead, the old tradition in which the public sector saw industry almost as an adversary has to end.
The Mercer review also examined the way the various Alliance partners worked together. It praised the fact that "an incredible amount of work" has been done in a short time. But it highlighted the difficulties faced by a loose alliance in implementing policy as distinct from developing it. The roles and responsibilities of GAVIs different players were not clear. "On a self-reported basis, no partner claims lead responsibility for forecast development, or advising countries on vaccine choices," says the review, while two partners claim lead responsibility for procurement strategy and development.
GAVI can add value to the multidisciplinary process of forecasting and procurement by bringing different partners with the relevant disciplines and skills together, says the review. But to deliver efficiently, there must be a clear chain of command and a single coordinator to carry responsibility the model followed by project managers in the private sector.
1. Lessons Learned: new procurement strategies for vaccines: Report to the GAVI Board.
Immunization Focus July 2002 - Contents