Should Poor Countries Follow WHO's New Advice on When to Start AIDS Treatment?

WHO and UNAIDS have just put poor country AIDS treatment program managers on the spot. Will they choose to follow the new WHO advice to push patients to begin treatment earlier – when many are still healthy? Or will they turn their backs on the new advice and push instead for expanded access to those in need. In conjunction with World AIDS Day on December 1, WHO and UNAIDS have released new treatment guidelines for AIDS patients. Until this week, WHO’s official position has been that AIDS treatment programs should strive to start patients on antiretroviral therapy when the number of CD4 cells in their blood drops from normal levels of between 800 and a 1,000 per cubic millimeter down to 200 per mm3. The fact that poor countries have not been able to achieve this goal accounts for WHO’s current estimate that there are about 5 million people currently needing treatment, twenty percent more than the 4 million receiving it. There’s no question that starting treatment much later than at the 200 CD4 count endangers the patient. Studies show that patients who wait until their CD4 drops to 50 have a 50% greater mortality rate than those who start at 200. In recent years studies [e.g. here and here or here (latter requires subscription)] have shown that mortality risk can be reduced still further by starting even earlier. Now WHO is recommending that patients start ARV therapy at a CD4 count of 350, which might reduce mortality risk for the average patient by about 15%. Should those responsible for AIDS treatment policy in poor countries respond to this advice by designing and implementing programs to identify and recruit patients much earlier, when their CD4 counts drop below the new 350 cell threshold? In the absence of a budget constraint, the evidence of health benefits alone would be enough to argue for adopting the new policy. And for over a decade it has seemed that budgets for AIDS treatment are not constrained. But now the US government, the biggest funder of AIDS treatment, has capped its spending for the next six years to $63 billion, of which perhaps only” about $10 to $20 billion will be on treatment. As the US is expanding its war effort in Afghanistan and recovering from a deep recession, the chances of increasing this amount seem dim. AIDS policy makers need to recognize that there will be tough choices to make in this budget-constrained environment, and carefully consider the tradeoffs. The figure below summarizes calculations that suggest the tradeoff between earlier recruitment and expanded access is stark. The figure shows the total worldwide 15-year cost* of any combination of average starting CD4 counts and average proportion recruited of those who need treatment (uptake rate). The diagonal contour lines represent combinations with equal 15-year costs. (Please ignore the kinks in the lines, which would be smoother if I had run 100 times as many simulations.) The color scheme goes from green, representing lower costs when uptake is small and recruitment is late, to red representing high costs associated with the earliest recruitment and the most universal access. The contours show that for any given budget, earlier recruitment means recruiting a smaller proportion of those in need. The black circle labeled “Historical Policy” captures the constellation of CD4 recruitment and uptake rates that have characterized AIDS policy in resource-limited countries in the last few years. This ellipse straddles the $120-$140 billion contour, which is the estimated cost of continuing past policy. Funding this 15-year commitment is already proving difficult and countries are anticipating reducing their uptake rates in the next few years. But suppose the world decided to spend money as WHO is advising by standardizing earlier recruitment. As captured in the figure, policymakers will need to push national policy to the right as represented by the horizontal arrow, increasing 15-year costs up to $200 billion. But this would imply continuing the current pattern of preferential access in which up to 60 percent of those needing treatment are unable to receive it. With the same $200 billion budget it would be possible EITHER to double access (from its current level up to 70 percent of those in need) OR to push the recruitment threshold up to a CD4 count of 325, NOT BOTH. Given that budget, policy makers must choose one of these options – or another along the “isocost” contour between them. Will they favor better treatment for a smaller number with the new guidelines or greatly expanded access using last-year’s treatment guidelines? And what if the budget is much smaller, perhaps as low as $80 billion over the 15-years? This would constrain AIDS treatment policy to the lowest of the contours in the figure, depressing uptake down to around 15 percent. With so few resources available worldwide, policy makers will find it difficult indeed to heed WHO’s new guidelines – which would lead to even more stringent restrictions in access in the interests of better treatment for the very few. ============================================================== *Technical Note: The cost projections made for the graph above are from 4,000 simulations using the beta of a new version of the AIDSCost software package [previous version available here]. AIDSCost uses data from UNAIDS and other sources to measure how HIV prevalence, deaths, treatment, and need are affected by a variety of parameters selectable by the user. In this case, all but two of the epidemiological and cost parameters available were held constant at their default values. By only varying uptake proportion and starting CD4 count (which itself affects time from infection to treatment need, total numbers needing treatment and various death rates), one can produce a matrix of costs implicitly defined as a function of only uptake rate and starting cd4 count, which is then graphed. The values of the default parameters are as follows: Proportion of HIV+ newly eligible for ART                               .143 ART Death Rate during first year on 1st line                             .095 ART Death Rate during subsequent years on 1st line            .04 ART Death Rate of AIDS patients on 2nd line                            .04 Non-ART Death Rate of AIDS patients                                          .085 Cost computations based on following parameters Lower bound for 1st-line drug costs                                              $88 Upper bound for 1st-line drug costs                                              $261 Lower bound for 2nd-line drug costs                                            $819 Upper bound for 2nd-line drug costs                                             $2,634 Number of bed-days per year per patient                            1.56 Number of out-patient visits per patient                              9.50 Average fixed non-drug cost at ART=1000                                 $750 Elasticity of average fixed cost w.r.t. ART                             -.146 This exercise has highlighted the tradeoff between uptake and starting CD4 count. Of course, AIDS policy makers must consider tradeoffs on many other dimensions, some of which can be studied using this same approach and varying others among the above parameters.