- Multi-disease prevention campaigns have the potential to reach large numbers of people with HIV testing and prevention interventions.
- Proof-of-concept studies in Kenya and Uganda have shown that campaign events can achieve coverage of 86 and 74% of the adult population respectively, and HIV testing uptake of 100 and 98%.
- The Kenyan campaign was designed to enhance prevention of HIV, malaria and water-borne infectious diseases. The campaign distributed a CarePack containing a long-lasting insecticide-treated net, a water filter and condoms. Everyone was offered HIV testing and people diagnosed with HIV were linked to care.
- The Kenyan campaign diagnosed people with HIV much earlier in the disease course and linked 80% of diagnosed people with care.
- The campaign had the potential to be cost saving when compared to separate vertical programmes. Additional interventions could be added at a very low marginal cost.
- Receipt of insecticide-treated nets and water filters was associated with a 27% reduction in the risk of disease progression to a CD4 cell count below 350 in people not yet eligible for treatment.
- The Ugandan campaign combined HIV testing with screening for TB symptoms, malaria, diabetes and hypertension. It provided treatment for malaria, de-worming for children aged 1-5 years and linkage to care for people diagnosed with HIV, TB, diabetes or hypertension. Everyone received condoms and insecticide-treated nets.
- The campaign found that almost one-in-four people had hypertension, the majority undiagnosed, and 3.5% of people had diabetes. Approximately one-in-four cases of diabetes was previously undiagnosed.
- People with HIV who had CD4 cell counts below 100 were fast-tracked for care and initiation of antiretroviral therapy. 75% of people in this category had started treatment within 2.5 days.
- The proportion of people with undetectable viral load, and the median viral load in the local population, fell between the first campaign event in 2011 and a follow up in 2012.
- Both campaigns required extensive community consultation to ensure that they were correctly targeted. Both research groups say that this was essential to the success of their campaigns.
- Campaigns require strong logistics, good marketing and training of everyone involved in delivering the interventions.
- Multi-disease prevention campaigns offer a major opportunity for communities to be involved in the planning and delivery of health care. They also offer opportunities for different health programmes to work more closely, and to save money and increase impact by doing so.
- More research is needed to show what can be achieved with different mixes of interventions, and to understand better how TB case-finding can be integrated into multi-disease prevention campaigns.
Reviewed by Gabriel Chamie, Vivek Jain (UCSF), Reuben Granich (UNAIDS), Alexandre Doyen (Vestergaard Frandsen) and Judd Walson (University of Washington, Seattle).
This edition was kindly supported by Vestergaard Frandsen.
The previous edition of HATIP looked at how HIV prevention, testing and linkage to care can be integrated into multi-disease prevention campaigns. It examined possible synergies between prevention of malaria, schistosomiasis and water-borne infectious diseases and HIV prevention and care, and the ways in which non-communicable disease screening might also be integrated. See HATIP edition 206 here.
This edition of HATIP examines in more detail two campaign approaches to multi-disease prevention that included HIV. One campaign was conducted in western Kenya, the other in Uganda.
Western Kenya: multi-disease prevention campaign targeting HIV, malaria and water-borne infections
The campaign in western Kenya was designed to integrate the prevention of malaria and water-borne infections through the distribution of long-lasting insecticide-treated nets and water filters at community events. HIV testing was offered at the same events and people who tested HIV positive received point-of-care CD4 tests and were linked to care.
The multi-disease prevention campaign was designed to promote the prevention of three causes of disease associated with high levels of morbidity and mortality in many regions of sub-Saharan Africa.
A prevention programme that can address all three diseases is likely to have greater reach and achieve substantial economies of scale as a result of the distribution of prevention commodities targeting the three different diseases through one distribution channel.
Both malaria and water-borne infections have been proposed as influencing HIV disease progression, but previous evidence about their effects has been mixed. Unclean water is the major source of infectious causes of diarrhoea both in people with HIV and in the general population. Malaria is one of the major causes of morbidity and mortality in endemic areas, and in people with HIV infection it can also cause temporary increases in viral load that are associated with an increased risk of disease progression. (See HATIP 206 for a discussion of the evidence.)
The proof-of-concept project examined the impact of receiving the insecticide-treated net and water filters on the health of people living with HIV and not yet eligible for antiretroviral therapy, during a multi-disease prevention campaign in communities in western Kenya.
In the Kenyan pilot study, the Ministry of Health partnered with local community groups and an international NGO to run a multi-disease prevention campaign, which advertised free HIV counselling and testing at central locations in the largely rural district, during a one-week campaign that used community mobilisation activities to raise awareness of the opportunity for HIV testing.
Everyone who attended the campaign events was offered voluntary counselling and testing. Previous reports from the campaign showed very high uptake of HIV testing; around 99% among those who received the care package. Of these, 80% had never tested for HIV before.1
Everyone who attended also received a CarePack – a canvas bag containing a long-lasting insecticide-treated net, and a water filter. In this campaign, the water filter distributed was the LifeStraw, a filter which comes in two forms: an individual filter or a family filter. The individual LifeStraw filters approximately 1000 litres of water (enough to last around one year) and the LifeStraw Family filters around 18,000 litres of water (enough to last up to three years for a family of five). The filter removes 99.999% of bacteria and 99.9% of protozoa.
The CarePack also provided condoms, and at the campaign events everyone received education on the use of the insecticide-treated nets, the water filters and condoms, together with general information about the prevention of malaria, water-borne diseases and HIV infection.
What the campaign delivered
During a seven-day period in 2008, over 47,000 people in a rural district of Kenya consisting of 157 villages attended the campaign events, which were well publicised in advance. There was 86% uptake of the CarePack multi-disease prevention commodity package. Of the people who registered for the campaign, 99.7% consented to HIV testing; 80% had never tested before and in this group 4% tested positive.2
A second campaign carried out in a peri-urban community in Kenya reached 5198 people in three days and achieved 100% uptake of the prevention package, HIV counselling and testing and CD4 cell testing.3
Very high acceptability of HIV testing
Up to 80% of the population have never tested for HIV in many high-burden settings. Stigma remains a barrier to the uptake of HIV counselling and testing. Qualitative research on the Kenyan rural campaign showed that the primary motivation for attending the campaign events was to learn one’s HIV status.4
“What motivated me was the idea of a free test for HIV.” (man, HIV negative, mid-40s)
Acceptability was reinforced by the commodities that were distributed to people who attended the campaign:
‘‘I have benefited from what they gave me because they are useful to me. They help me because I no longer drink dirty water, I also have a bag to carry my things inside. The net prevents me and my children from being bitten by mosquitoes.’’ (woman, late-40s)
The campaign reduced stigma and educated people about the new hope provided by HIV treatment:
“They are now seeing that there is hope in the lives of HIV-infected people. They are taken by others in the village as part of them, and they offer support and encourage them, they now feel they are part of the society.” (HIV-positive man, 40s)
“Although stigma has lessened, I think more importantly, many people do not perceive themselves to be at risk of HIV so there wasn’t much aversion to getting tested,” said Judd Walson, Associate Professor of Global Health at the University of Washington, Seattle.
“Many different styles of outreach have been attempted for HIV testing but never so successfully,” he went on.
He attributes part of the success of the campaign to the strong branding of the CarePack and the campaign.
“I think there’s something really powerful going on with the brand that communities associated with good health. The campaign used the same colour packaging for the bednet and the water filter.”
It was possible to achieve strong visual branding for the campaign because the campaign focused on the distribution of commodities and because the main commodities distributed at campaign events were manufactured by Vestergaard Frandsen (also a sponsor of the project).
He identifies reluctance of funders to be associated with corporate branding as a potential obstacle to development of campaign approaches.
“A lot of players are intentionally trying not to associate with a particular brand, but if the branding is part of how the campaign works, how do funders who don’t want to be associated with one particular brand support these sorts of interventions in the future? There are five or more endemic diseases that could be combined in multi-disease prevention campaigns. It’s difficult to come up with a campaign that combines five different brands.”
However, Reuben Granich of UNAIDS disagrees. “Delivering quality products as part of a quality campaign is what worked,” he said. Indeed, as the SEARCH Collaboration campaign in Uganda showed, it is possible to achieve a high level of engagement with a campaign without branding and without the same emphasis on distribution of commodities, if the campaign meets the needs of the community (see below).
Linking people with HIV to care
Everyone diagnosed with HIV was linked to care by a ‘navigator’ – another person living with HIV – to whom they were introduced on site during the campaign, and subsequently received cotrimoxazole (Septrin) prophylaxis. Linkage through a navigator was systematically implemented during the second campaign.
The average CD4 count of people diagnosed with HIV during the campaign was 536. This compares with a median count of 348 at the time of diagnosis in the local HIV clinic cohort.5
Over 80% of those who tested positive were linked to care within 10 months of the campaign.6 Enrolment was higher among people who were linked during the campaign to another person living with HIV who acted as a ‘navigator’.
“Point-of-care CD4 testing was really important. Telling people that they have HIV is very different from being able to document that they are immunosuppressed and at risk and need to be linked to care and starting treatment now,” said Judd Walson.
Providing prevention commodities together has the potential to reduce distribution costs and maximise impact. The estimated cost of the commodities and services delivered at large scale in Kenya was $31.98 per person ($6.27 for malaria, $15.80 for diarrhoea, $9.91 for HIV).7 Adding further interventions to a campaign would allow delivery at a very low marginal cost, provided that the intervention is compatible with the campaign model of focused activity, intensive service delivery and commodity distribution.
The multi-disease campaign approach also has the potential to be cost saving compared with separate vertical programmes, particularly in comparison with door-to-door household counselling and testing, when cost savings from disease episodes averted are taken into account.8
Health economic evaluation of the Kenyan rural campaign showed that the cost per disability-adjusted life-year averted was around $20. Reaching 1000 people resulted in an estimated $85,000 in averted healthcare costs due to the prevention of HIV infections (37% of the cost), malaria (57%) and water-borne diseases (12%).9
The impact of a multi-disease prevention campaign on HIV disease progression
The impact of the multi-disease prevention campaign on disease progression in adults living with HIV and not yet eligible for antiretroviral treatment was presented at the 19th International AIDS Conference in Washington DC in July 2012 and subsequently published in the journal AIDS.10
The distribution of long-lasting insecticide-treated nets and water filters to prevent malaria and water-borne diseases was found to significantly reduce the rate of HIV disease progression among Kenyan adults not yet eligible for treatment. Receipt of these items was associated with a 27% reduction in the risk of reaching the CD4 cell threshold for starting treatment among people living with HIV, but not yet eligible for treatment, during two years of follow-up of a prospective cohort.
A related study estimated that, if the same effect were seen throughout sub-Saharan Africa after the distribution of these products to people living with HIV, the intervention would have the potential to save just over US$400 million a year in antiretroviral treatment costs, if all adults were able to start treatment promptly when their CD4 cell counts fell below 350 cells/mm3.11
Investigators recruited a prospective cohort of 589 HIV-positive people with CD4 cell counts above 350 not yet eligible for treatment and without symptomatic HIV disease (WHO stages 3 or 4), of which 361 received the CarePack intervention. Individuals received HIV care through Kisii Provincial Hospital and Kisumu District Hospitals in Western Kenya.
Quite a high proportion of those in the control group used both water purification methods and insecticide-treated nets, but the intervention group had a significantly higher usage of insecticide-treated nets (97.7 vs 83.1%, p<0.001) and were significantly more likely to report that they slept under it (mosquitoes that spread malaria are most active after dark and in low light, so using the net consistently at night is a very important means of protecting against malaria).
People who received the intervention were significantly more likely to drink purified water (99.5 vs 76%, p<0.001) and reported significantly different patterns of water safety practices. In the control group, the predominant practice was to use chlorine for water purification (45.4%), or else to boil water (29.9%), but in the intervention group 93% used the LifeStraw water filter provided. Only 5% used chlorine for water purification. (These practices were verified during a home visit.)
The intervention resulted in a 27% reduction in the risk of disease progression to a CD4 cell count of 350 or below during the two-year follow-up period (HR 0.73, 95% confidence interval 0.57-0.95), and those in the intervention group were also at lower risk of reaching a composite endpoint of either CD4 cell count below 350 or death from non-traumatic cause (HR 0.75, 95% CI 0.58-0.79). There was no significant difference in risk after controlling for cotrimoxazole prophylaxis use or toilet type.
The study also found that the extent of CD4 cell decline was significantly less among those who received the long-lasting insecticide-treated nets and water filters (-54 vs -70 cells/mm3 per year).
People in the intervention group were also significantly less likely to report diarrhoea (HR 0.65), malaria symptoms (HR 0.75) or to be diagnosed with clinical malaria (HR 0.66).
The SEARCH Collaboration: a community-wide campaign in rural Uganda
Another community-wide campaign was conducted in rural Uganda. It sought to use an HIV testing campaign as the platform to carry out point-of-care screening for HIV, malaria, tuberculosis, hypertension (high blood pressure) and diabetes. This demonstration project shows how an even larger range of interventions can be delivered through a time-limited campaign.
The campaign was intended to demonstrate how HIV resources might be leveraged in order to address the large unmet need for diagnosis and treatment of non-communicable diseases. Its scope therefore extended beyond the 'big three' communicable diseases, but the interventions carried out during the campaign targeting communicable diseases also went beyond those delivered in the Kenyan campaign.
It also serves as the community HIV testing strategy for the SEARCH study, a broader approach to 'test and treat' that is designed to evaluate not only the effects of 'test and treat' on HIV transmission, but also its impact on health, educational and economic outcomes. These are key indicators for development, so it is important to gather evidence that can support the case for donor and national government funding of particular approaches to disease prevention and management.
The study is also evaluating the impact of comprehensive HIV testing and referral for HIV therapy on population-wide HIV RNA (viral load) levels, which are overall measures of the amount of the virus in a community. Population viral load is considered to be a useful measure of how well a system is delivering the full spectrum of HIV services and successful therapy, and so the impact of 'test and treat' activities on this outcome is a key question.
The SEARCH community health campaigns were designed by a collaboration between the Makerere University-University of California-San Francisco Research Collaboration and Mulago-Mbarara Joint AIDS Program – the SEARCH Collaboration (Sustainable East Africa Research of Community Health). It took place in the Kakyerere parish of Mbarara district, a rural area in south-western Uganda comprising nine villages with 6300 inhabitants.12
Residents of Kakyerere parish have access to health care and diagnostic services for HIV and TB at one health facility on the eastern boundary of the parish. The nearest facility offering treatment for diabetes and hypertension is more than 20km away.
“We chose the area for the intervention study because it already had an HIV clinic functioning as an ART delivery site, but there had been no previous campaign activities in the region,” Dr Gabriel Chamie, Assistant Adjunct Professor in the School of Medicine, University of California, San Francisco, told HATIP. “We specifically chose a rural area without any history of major intervention except for government-run testing and education campaigns.”
How can we test large numbers of people?
The SEARCH Collaboration study is part of a larger initiative designed to investigate how 'test and treat' approaches to HIV treatment might be implemented. The first questions relevant to any attempt to increase the proportion of people on treatment are: “How do we test large numbers of people in an economical way?” and “How can we motivate people to learn their HIV status?”.
Previous efforts to increase the proportion of people who know their HIV status have focused on taking HIV testing into the community, for example through door-to-door and household testing. These interventions have achieved high uptake, but they may not be the most cost-effective ways of testing people or linking them to care.
Like the Kenyan campaign, the Ugandan campaign relied on large-scale community mobilisation activities, developed by community leaders, in order to get people to attend campaign events. Announcements on radio, in schools, mosques and churches were accompanied by a leaflet and poster campaign which encouraged people to come and get tested for blood pressure, diabetes, malaria, HIV and TB. HIV testing was just one of the tests on offer, because “we didn’t want this to be perceived as just an HIV testing campaign,” said Chamie.
Ways of promoting the uptake of HIV testing were discussed with local community leaders. “One of the things they identified was the need to address diabetes and hypertension too,” said Chamie. “Particularly among men there is an interest in knowing their diabetes and hypertension status and an awareness that these conditions exist, but a lack of access to testing.”
Although the research group did not study the question, there was a perception among the researchers that interest in knowing about diabetes and hypertension was driven by what one member of the group called “positive stigma” – a view among the local population that suffering from diabetes and hypertension denoted a higher economic and social status, due to the ability to consume a diet that increases the risk of these conditions.
The campaign events took place over five days in May 2011 at three locations in the parish, chosen to ensure that no one would need to walk more than 3km to attend. The core of the events, the diagnostic operation, was provided by an on-site laboratory in a large tent staffed by 18 lab technicians. All diagnostic testing was done with fingerprick blood samples.
What the campaign delivered
The campaign delivered a broad array of interventions (see Table 2):
The campaign had originally planned to distribute long-lasting insecticide-treated nets too, but logistic problems meant that the nets were not in the right place at the right time (the nets were eventually distributed to villagers several months later), illustrating a key condition for successful campaign activities – strong logistics and supply chain management.
The campaign events were attended by around three-quarters of adult inhabitants, the researchers estimated, based on Ugandan 2011 census data. A total of 4343 people attended (2523 adults and 2020 children). There was a very substantial bias towards women among adult attendees: the events were estimated to have reached 95% of adult women but only 52% of adult men. There was no gender difference among children; around two-thirds attended for testing.
Uganda is a country with a high awareness of HIV, and HIV testing and counselling have been available for over 20 years in rural areas. The proportion of people who had not previously tested for HIV was therefore much lower than in the previously described Kenyan campaign (34 vs 80%), but the campaign nevertheless reached a substantial population of previously untested adults. The study found that among all those who tested, the prevalence of HIV was similar to the level expected for south-western Uganda: 8% in 15-49 year-olds, and higher in women than in men (9.4 vs 5.3%). Prevalence in children was low, at 0.5%, as would be expected in a rural area of Uganda. (The low prevalence among children may also indicate the success of activities to prevent mother-to-child transmission over a long period in Uganda, but it may also indicate the extent to which lack of access to antiretroviral therapy has resulted in a high mortality rate among children, so reducing prevalence.)
Among the 179 adults diagnosed with HIV during the campaign just under half (46%) had been unaware of their HIV status prior to attending the campaign event. Twelve per cent were in need of immediate treatment (CD4 cell count below 200). Of the newly diagnosed with CD4 counts above 100, only 58% had linked to care within three months of the campaign event, despite being introduced to clinic staff after diagnosis at the campaign event. In people with CD4 counts below 100, an enhanced referral strategy was employed to ensure that people started treatment as quickly as possible. In this group of patients 75% (6 patients) linked to care and started treatment within a median of 2.5 days.13
Linkage to care in those who received routine referrals (CD4>100) was less good: 57% had linked to care within three months, and home visit interviews subsequently found that those who did not link to care in this group tended to cite transport costs, or linkage to care outside the district, as their reasons for not attending the local clinic. Newly diagnosed people were less likely to link to care than people who had received a previous HIV diagnosis.
The majority of those diagnosed (64%) had CD4 cell counts above 350 and were not eligible for treatment, but Gabriel Chamie said “this is the population we need to reach for a test and treat strategy”. As in Kenya, the campaign approach identified people with HIV infection much earlier than a clinic-based approach to testing.
Impact on non-communicable diseases
Hypertension was found to be a common problem among the Kakyerere population: 23% had a blood pressure above 140/90, and even using a conservative cut-off point of 150/100, 12% were found to have high blood pressure. The majority of people with blood pressure above 150/100 had never been diagnosed with hypertension and were not receiving anti-hypertensive treatment. Unsurprisingly, prevalence rose sharply in those aged 45 and above, and among those aged 65 and over prevalence of hypertension reached 35%. (In contrast the US National Health and Nutrition Examination Survey reported a prevalence of around 65% in its study population among over-60s surveyed between 1999 and 2008.)
The study also found that 3.5% of adults had diabetes; one-quarter had not been diagnosed previously. “This is broadly in line with what we expected to find in a rural area,” said Chamie. Sixty-one per cent were already receiving treatment.
Campaign impact on population HIV RNA
At the 2012 International AIDS Conference, Dr Vivek Jain of the University of California, San Francisco presented results from a follow-up community health campaign conducted in Kakyerere in May 2012, which included measurement of HIV RNA levels in people who tested HIV positive. Viral loads from 2012 were compared to those measured during the May 2011 campaign.14,15 The community had experienced three separate changes relevant to the 'test and treat' model: first, the 2011 campaign had diagnosed many individuals and allowed for referral to clinical care. Second, the Uganda national guidelines on HIV therapy shifted to recommend initiation at CD4 cell counts below 350 rather than below 250, allowing more people to begin therapy earlier. And third, the MU-UCSF research group began a separate study offering HIV therapy to people in Kakyerere Parish who had CD4 counts above 350 and who were thus ineligible for government-based therapy. With these three changes in the community spanning the spectrum of ‘test and treat’, Jain and the MU-UCSF team wanted to assess whether viral loads in the region were lower in 2012.
“The assessment of population-level HIV viral load is an important part of evaluating real-world effectiveness of test-and-treat strategies. Viral load suppression reflects effectiveness of the entire cascade of care from HIV diagnosis to linkage to treatment to suppression. Viral load distribution of a population can also provide insight into transmission,” said Dr Jain.
The 2012 follow-up campaign screened 4872 people (of whom 2271 were adults, 72% of the adult population).
Approximately eight per cent of adults in 2011 and 9.4% in 2012 were found to be HIV positive, a difference that was not statistically significant. In 2011, 35% of the participants were male, compared to 45% in 2012. The median age of participants in both years was approximately 19-20 years old.
The main finding of the study was that HIV viral loads were substantially lower in 2012 compared to 2011. First, there was a significant increase in the percentage of people with HIV who had an undetectable viral load, rising from from 37% in the 2011 campaign, to 55% in the 2012 campaign. Second, the proportion of people with a viral load greater than 100,000 copies/ml – who represent people most likely to transmit HIV due to high viral load – decreased from 13% in 2011 to 3% in 2012. And third, the median viral load also decreased substantially from 2011 to 2012, dropping from 2185 copies/ml (IQR:<486-33,045) to less than 486 copies/ml (IQR: <486-7903). Overall, the results indicate that population viral load measurements were lower in 2012 than in 2011.
Integrating TB case finding into campaigns
The western Kenya campaign did not incorporate any TB case-finding activities.
TB case-finding may support a number of potential activities within a campaign:
- Ruling out TB by the use of the WHO 4-symptom screen in people living with HIV, so that isoniazid preventive (IPT) therapy may be initiated.
- Diagnosing active TB in people living with HIV by the use of the WHO 4-symptom screen, sputum smear, sputum culture and Xpert MTB/RIF where available, so that these patients may be fast-tracked for ART initiation.
Diagnosis of active TB is most urgent in people living with HIV, as is exclusion of active TB. Case finding requires diagnostic capacity, which might be augmented by temporary availability of fluorescent microscopy at the campaign laboratory, together with extra staff to handle the high volume of smear reading.
Modelling the potential impact of a campaign approach on TB prevention, Amitabh Suthar and colleagues at the World Health Organization concluded that the Kenyan campaign would have prevented 342 TB cases in people living with HIV through provision of isoniazid preventive therapy if it had incorporated the WHO 4-symptom screen for TB in people with HIV. The purpose of the 4-symptom screen is to rule out the possible presence of TB, so that individuals may be safely initiated on IPT.16 A further 80 people would have been diagnosed with active TB immediately, and 834 people would have been referred for further investigations on the basis of their symptoms, they estimated, using TB frequencies obtained from a WHO/CDC meta-analysis for a setting with a 5% prevalence of HIV.
Suthar and colleagues argue that TB prevention can be integrated into multi-disease prevention campaigns, especially where high HIV prevalence raises the index of suspicion.
The SEARCH Collaboration campaign carried out TB screening for people diagnosed with HIV.
“We were curious to find out whether you could do an intensified case finding screen in a campaign setting. We were also interested to see how prevalent the symptoms were in the community. Just over 87% of HIV-positive people had one of the symptoms in the symptom screen, and only 13% were symptom-free and eligible for isoniazid preventive therapy. Everyone else was referred [for further testing],” said Gabriel Chamie.
“I’d say the jury’s still out on how best to perform case finding in a community campaign setting,” he commented.
Judd Walson is concerned about the practicality of integrating TB case-finding without much greater thought regarding the diagnostic algorithm. His scepticism reflects the difficulties inherent in developing large-scale models of TB case finding in settings where both TB and HIV are endemic. The very high prevalence of symptoms that triggered a referral for further screening, together with the need to use another diagnostic in addition to smear microscopy in order to address the high prevalence of smear-negative TB in people living with HIV, indicate that further research is needed on how to increase the scale of intensified case finding within the economic and logistic context of a multi-disease prevention campaign.
As the two case studies in this edition show, community-based multi-disease prevention campaigns that incorporate HIV prevention, testing and linkage to care have huge potential for increasing rates of HIV diagnosis and diagnosing people with HIV before they become severely immunocompromised. The SEARCH Collaboration’s work in Uganda also shows that a large-scale campaign can have a subsequent impact on community viral load, potentially reducing HIV incidence over time.
Studying the preventive impact of the various HIV-related components of such campaigns is the next big research challenge, and will hopefully spur researchers, public health officials and community advocates in other settings to develop their own models of integrated prevention campaigns, based on local health priorities.
Above all, what is clear from both case studies is the extent to which campaigns require the involvement of the community at the design stage, in order to identify local health priorities and motivations for health-seeking.
Multi-disease prevention campaigns offer a significant opportunity for the empowerment of communities to demand the reprogramming of health resources to meet local needs, and for donors to demand greater integration and economy of scale in disease prevention.
 Lugada E et al. Rapid implementation of an integrated large-scale HIV counseling and testing, malaria and diarrhea prevention campaign in rural Kenya. PLOS ONE 5(8): e12435, 2010.
 Granich R et al. Achieving universal access for human immunodeficiency virus (HIV) and tuberculosis: potential prevention impact of an integrated multi-disease prevention campaign in Kenya. AIDS Research Treat: 412643, 2012.
 De Ver Dye T et al. A qualitative assessment of participation in a rapid scale-up, diagonally integrated MDG-related disease prevention campaign in rural Kenya. PLOS ONE 6(1): e14551, 2011.
 Granich 2012. op cit.
 Hatcher A et al. Predictors of linkage to care following community-based HIV counseling and testing in rural Kenya. AIDS Behav 16 (5): 1295-1307, 2011.
 Kahn J et al. Cost of community integrated prevention campaign for malaria, HIV and diarrhea in rural Kenya. BMC Health Services Research 11: 346, 2011.
 Kahn J et al. Integrated HIV testing, malaria and diarrhea prevention campaign in Kenya: modeled health impact and cost-effectiveness. PLOS ONE 7(2): e31316, 2011.
 Walson J et al. Evaluation of impact of long-lasting insecticide-treated bed nets and point-of-use water filters on HIV-1 disease progression in Kenya. AIDS 27: 1493-1501, 2013.
 Verguet S et al. Cost savings associated with the provision of a simple care package including long lasting insecticide treated bednets and a point-of-use water filtration device to delay HIV-1 disease progression in Africa. Nineteenth International AIDS Conference, abstract MOPE768, Washington DC, 2012.
 Chamie G et al. Leveraging rapid community-based HIV testing campaigns for non-communicable diseases in rural Uganda. PLOS ONE 7 (8): e43400, 2012.
 Chamie G et al. Outcomes in a routine linkage to care strategy and an enhanced strategy with accelerated ART start; community-based HIV testing and point of care CD4: rural Uganda. Nineteenth Conference on Retroviruses and Opportunistic Infections, Seattle, abstract 1134, 2012.
 Jain V et al. Changes in population-level HIV RNA distribution one year after implementation of key components of an HIV ‘test and treat’ strategy in rural Uganda. Nineteenth International Conference on AIDS, abstract TULBE04, Washington DC, July 2012.
 Jain V et al. Assessment of population-based HIV RNA levels in a rural east African setting using a fingerprick-based blood collection method. Clin Infect Dis 56 (4): 598-605, 2013.
 Suthar A et al. Community-based multi-disease prevention campaigns for controlling human immunodeficiency virus-associated tuberculosis. Int J Tuberc Lung Dis 16: 430-6, 2012.