In my last blog, I suggested that under a specific condition a NICE approach, as outlined in the Gardiner Harris article in the NYT, could help to allocate expenditures between AIDS treatment, HIV prevention and care for affected persons. The condition is that there be no spillover effects, either positive or negative, between these categories of interventions. A further condition is that alternatives being compared are incremental, not radically different public health strategies.But what if there ARE spillover effects between AIDS treatment and HIV prevention? A disheartening possibility is that AIDS treatment leads people to increase their risk behavior. If the extent and magnitude of such a disinhibition effect were known, a modified NICE approach would allocate resources away from treatment towards prevention. Some have argued that the disinhibition effect is small in poor countries, but that AIDS treatment could instead have a strong beneficial spillover on HIV prevention because people who are receiving effective treatment have a reduced viral load and are thus less infectious. This perspective motivates the authors of an article published in the Lancet on November 26which proposes a way to dramatically increase the positive synergies between voluntary counseling and testing (part of the prevention rubric) and treatment. Applying the concept of herd immunity, the authors suggest that testing ALL adults EVERY YEAR and giving newly HIV infected people treatment IMMEDIATELY could turn the epidemic around by preventing a large share of new HIV infections. They calculate that their strategy would cost much more than our current AIDS strategy in the next two decades, but would save money by the year 2032. The NICE approach as outlined in Harris’ New York Times article works best on incremental decisions regarding the treatment of a single individual, where there are no spillover effects of that individual’s treatment on others. The presence of either positive or negative spillover effects requires modifying the NICE approach And it would require considerable extension before the approach could be applied to non-incremental alternatives, such as the choice between completely different public health strategies. Comparing an existing public health program to a radically different one such as that outlined by Granich et al. requires extrapolation beyond the observable data – essentially a leap of faith. When a strategy is failing as badly as current HIV prevention strategies are failing in Africa, such a radically different approach may be worth attempting. The NICE approach is unlikely to be a reliable guide for such a leap.
