ized to enable 3 million people to be on treatment by 2005. Indeed, one critique of a published evaluation of Haiti’s success (Gilks et al., 2001) contends that the human resources and capacities employed (i.e., the logistics with regard to clinical input and staff time) were not clearly listed in the evaluation and that the criteria for selecting individuals for treatment were not stringent. These observations highlight the need for scientific evaluation criteria, without which important lessons that might be applied in other settings cannot be derived.
Pilot initiatives are by design relatively limited in size, often directed toward a small fraction of those in need and supported by unusually committed and resourced staff. Scale-up activities will introduce factors that may not be especially relevant in a pilot implementation. If a particular health resource is scarce in a country—for example, trained people who may already be staffing other essential health services—a large program scale-up effort may have collateral negative effects on already marginal health care systems by diverting a critical fraction of the limited human resource pool. These threats would not necessarily be evident in a pilot program whose relatively modest requirements were, in the grander scheme, absorbable or even resourced external to the underlying health care infrastructure. Various avenues might be pursued to mitigate this risk of collateral negative effects, although scientific evidence would likely be needed to assess the advisability of different options.
A common discussion in international health revolves around the relative advantages of “stovepiped” or “vertical” condition-specific implementation programs, versus the integration of condition-specific programs into a health sector–wide approach whereby the resources of various programs are effectively integrated into one system, usually under government leadership (see Box 6-1). The care of HIV/AIDS patients is not as easily “stovepiped” as that provided in, for example, an immunization program or even a tuberculosis (TB) program. Rather, HIV/AIDS patients require care that has longitudinal continuity, care that is often complex considering the myriad of opportunistic infections that can occur. The care they receive must also encompass reliable access to quality pharmaceuticals, attention to nutritional and psychosocial needs, counseling to prevent the spread of infection, and the requirements of palliative care. Thus the establishment of programs for HIV/AIDS care requires access to a wide range of services that need to exist for many other purposes as well.
On the other hand, while it would appear that the ideal HIV/AIDS care program would be integrated so as to strengthen the entire health care system, to accomplish such integration in a nondisruptive way takes time.