Experience with alternative approaches to public drug supply is limited. Good analyses of their long-term performance and sustainability are lacking. However, expanding on questions raised with respect to contracting-out health services in general [69,74], the following questions must be considered in assessing the feasibility of using market mechanisms in public drug supply:

• Will real competition take place? Both the direct delivery system and the prime vendor system rely on competition to stimulate efficiency. If the private sector is poorly developed then contracting-out may simply replace a government service monopoly with a private one - with no visible cost saving or service improvement.

• Can drug quality and service quality be maintained? Contractors may try to cut corners in order to reduce costs. This may adversely affect both the quality of the service (as measured by the rate of shortages and stockouts) and the quality of the drugs delivered.

• Will efficiency actually improve? If lower costs are achieved at the price of lower quality then the result in terms of improved efficiency is ambiguous.

• Can government effectively negotiate and monitor contracts? The benefits of competitive contracting will not be reaped unless government has adequate negotiating and monitoring capacity. Contract specification must cover the quality of the service in adequate detail and must include sanctions against contractors who break quality standards. Contracted services must be monitored to ensure that they are provided as specified in the contract. Government may operate under constraints in capacity that prevent it from carrying out these tasks effectively.

• Will there be sufficient financing? Alternative supply arrangements may result in greater quantities of drugs being provided for a given budget but they will not solve problems of inadequate financing. Late payment of contractors by government is often the main reason given by private sector companies for not wanting to bid for government contracts. So the question of adequate funding may also affect the level of competition.

• Will there be wider unforeseen consequences? Establishing a long-term contract with a private sector company may drive other companies out of the market, resulting in less competition in the future. Tying up a substantial amount of funds in one contract may distort resource allocation in the pharmaceutical sector as a whole. Government officials responsible for contracting need to be aware of the system-wide and long-term effects of the contracts they negotiate.

Just how suitable the models described in Section 5.1 are to a particular country or region depends critically on an analysis of the current problems and institutional conditions there, particularly with respect to the six questions outlined above. Although at this point the evidence is insufficient to reach firm conclusions, some rules of thumb about appropriate policy options can be drawn.

• The efficiency of a direct delivery or prime vendor system depends on a well-developed private sector. In countries without a developed private sector these supply solutions do not make sense, at least in the short term. Instead efforts should focus on improving the efficiency of existing public supply systems and perhaps creating the type of business environment that may attract private firms to enter the market.

• If an existing CMS is beset by problems that can be traced to overly rigid government regulations (such as inability to hire and fire staff, or forced reliance on an ineffective transport pool) then a public autonomous supply agency may provide some advantages. However, without true government commitment to the idea of an autonomous agency and without adequate mission statements and terms of reference, an autonomous supply agency may suffer from problems very similar to those of the CMS.

• Different models place different types of demands on government capacities. Where CMS and autonomous supply agency models require substantial physical infrastructure (to procure, store and deliver drugs), the direct delivery and prime vendor systems require capacity to negotiate, contract and monitor contracts. A government should consider which aspects of its capacity are strongest.

• The success of contracting arrangements and the type of contracts which are appropriate may depend considerably on the nature of the organization contracted to provide the service. Not-for-profit organizations are more likely to share objectives similar to those of the government. If a not-for-profit organization such as MEDS (see Box 2) is the prime vendor, government monitoring procedures may not need to be as rigid as if a for-profit company is contracted.