We explored the potential health impact, net cost, and cost-effectiveness of an integrated mass campaign to distribute commodities and services intended to decrease malaria, diarrhea, and HIV. We found, for each 1000 campaign participants, an estimated health benefit of 442 disability-adjusted life years averted, with a net savings of approximately $16,000. The prevention component yielded 81% of the DALYs averted and large net savings ($85,113). Earlier HIV care yielded additional DALYs and also substantial net costs, due to the high cost of ART. Multivariate sensitivity analyses suggest that overall health benefits reside between 327 and 583 DALYs, the campaign is cost-saving for more than four-fifths of simulation trials, and the cost per DALY averted is less than $20 for 93% of trials.
Compared with the cost-effectiveness of individual interventions, these results are generally more favorable. Malaria interventions cost in the range of $2–15 per DALY averted even for the least expensive strategies 
. Diarrhea prevention has ten- to 100-fold higher cost-effectiveness ratios; filters alone are estimated at $142 per DALY averted 
. HIV prevention is often cost saving, due to the high cost of care, with savings exceeding costs by 25- to 30-fold 
. HIV care with ART costs $500–$800 per DALY averted in Africa 
, and CD4 cell and viral load monitoring of ART $174 and over $5000 per DALY averted, respectively 
Our analysis had several limitations. As with many cost-effectiveness analyses, health impacts and averted care costs are modeled rather than measured directly for the campaign. However, empirical studies of similar interventions have shown evidence of effectiveness in reducing morbidity and mortality over specified time periods, which we adopted and use as the basis for the modeled prevention benefits from the IPC. By including the best available input values and a diversity of inputs (e.g., protective effects for three diseases) we have mitigated this limitation. Further, robust sensitivity analyses allowed us to assess uncertainty in effectiveness, with favorable findings over the range of values explored.
Second, the campaign cost is based on an economic model for scaling up, and is 25% lower than the cost of the initial campaign implementation. We think that uncertainty in this cost estimate is low, based on confirmatory data from subsequent campaign implementation and planning, and thus has little effect on our findings. However, it will be important to observe actual costs in a scaled up implementation. Repeat campaigns in the same geographical location would yield a lower number of new HIV diagnoses, and depending on timing in relation to commodity life spans could yield lower participation and/or health benefits. Finally, we did not explore savings by linking this campaign to other community campaigns, such as annual mass vitamin A administration 
or regular indoor residual spraying (IRS) against malaria.
We did not include the effect of earlier ART on tuberculosis (TB). Co-infection of HIV and TB approaches 50% in Kenya 
, and ART may reduce TB acquisition by greater than 60%, depending on when treatment is initiated 
. This might suggest substantial health gains and economic savings via TB control due to expanded use of ART. In addition, earlier TB diagnosis might lead to easier treatment. However, other factors complicate this picture. Some TB infection that would be detected and treated while under care would never have become clinically significant. Further, ART may induce clinical worsening of TB, due to immune reconstitution (Robin Wood, personal communication). We believe that in the long run, expanded ART will reduce TB transmission and thus prevalence, but in the short run the effects are difficult to anticipate.
This analysis supports a substantial role for integrated multi-disease mass campaigns. Such campaigns are potentially very practical, quickly achieving high coverage of key interventions to reduce the burden of three major diseases, with substantial health benefits, and attractive economics. The campaigns would need to be repeated over time in order to offer ongoing benefits. The optimal timing is unclear, due to the differing duration of campaign interventions: up to ten years for LLIN, three years for water filters, and one year for VCT and condoms. In addition, newly detected HIV cases will drop sharply after the initial implementation, since HIV incidence is much lower than undetected HIV prevalence. Optimal timing would also reflect the local availability of these services through other mechanisms. On balance, we suspect that a three-year cycle would be desirable in most settings. We plan to formally assess this issue in an upcoming analysis.
In conclusion, we propose expanded field implementation of integrated multi-disease mass campaigns, coupled with rigorous evaluation and refinement.
The findings and conclusions in this paper are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.