HIV/AIDS - Getting Results 51901 These reports describe activities, challenges and lessons learned during the World Bank Global HIV/AIDS Program World Bank's HIV/AIDS work with countries and other partners. Could India's business skills improve lagging public health outcomes? Jody Zall Kusek, David Wilson and Austin Thomas Could a marriage of strong business and management skills with sound public health programming boost health outcomes in India and elsewhere? If much is known about how to solve public health problems, why do so many health programs in India and around the world, fail to meet their objectives? Do we have strong examples of such "marriages" improving health outcomes? And if so, what accounts for their success? The Avahan program applies India's outstanding management and business expertise to a difficult public health challenge: preventing new HIV infections among the highest risk groups where most new infections occur. Clearly focused goals, a strong drive for results, carefully chosen extensive partnerships, and intensive use of data to monitor, manage and enhance performance enabled a remarkably fast and extensive scale-up that is Is there an example of successful use of achieving good outcomes. business expertise to solve a public health challenge in India? A management team with extensive business and marketing skills and a strong record of meeting high India's successes in telecommunications, engineering, e- performance targets partnered with groups already business, and building a strong cadre of world class offering the relevant services. Within the Avahan business management professionals are acclaimed partnership, grass-roots outreach workers "market" safe around the world. By contrast, some of India's health sex and other health-protecting behaviors to hard-to- outcomes lag: only two countries have a higher proportion reach clients. Up-to-date detailed data on progress are of children under the age of 5 who are malnourished (low th used to identify good performance, flexibly fix lagging weight for age), and India ranks 137 in maternal mortality performance and bottlenecks, and spread and reinforce rates (UNDP, 2008). In 2002, although India's HIV good practice. prevalence was relatively low, it was estimated to have risen above 1% in 51 districts. Already an estimated 2.45 This evaluation note describes how the Avahan million people were living with HIV, and there was real approach applies key business management concern that an explosive increase might be in prospect. principles to a public health challenge, and considers whether the model is replicable and Could India use its advanced business know-how to adaptable to other health challenges. improve its health care systems, delivery of core health The World Bank has long been a close partner with India services and health outcomes? In particular, could in responding to HIV and AIDS, with three loans to effective HIV prevention be implemented at scale and support India's national AIDS Program, and extensive contain the epidemic? Business expertise is not common analytical work to understand the epidemic and its in public health: only about 6% of HIV/AIDS program impact. The World Bank's direct link with Avahan is managers of leading programs in India surveyed by the through David Wilson (of our Global HIV/AIDS Program) authors have a background in management. Most leaders serving on the Avahan technical advisory board. This of public health programs are doctors or government note is part of the Bank's efforts to understand how to administrators; very few have training or know-how in achieve better results in HIV prevention, and to work modern project and program management. India has not with countries and other partners to "make the money taken advantage yet of relatively new offerings of health work" through more effective HIV/AIDS responses. care management training by a consortium of 30 top-rated Management Schools around the world. Schools, such as efficiency. But delivering speedy, high-quality innovative the Anderson School of Management of UCLA and solutions in treatment and prevention should not be limited Wharton School of the University of Pennsylvania have to the private sector. The development community has joined developing country business schools to create the begun to realize the value of partnering with the private Management Education and Research consortium sector to deliver more effective health services, and also 1 (MERG). the value of harnessing modern management skills for delivering public sector health services. Public-private Applying private sector business practices to improve public partnerships have been used to improve national health health services is not new. Many health services have been care in developed countries for decades, and have started privatized and transformed using bottom-line performance to be used in developing countries too (Box 1). standards to enhance customer service quality and improve Box 1. Healthcare Public-Private-Partnerships in India Private providers play a very significant part in India's curative healthcare, accounting for 80-85% of doctors, 80% of out- patient visits, 93% of hospitals and 64% of hospital beds (World Bank, 2001). NGOs and for-profit businesses have stepped in to fill gaps in the inadequate public health infrastructure. A wide spectrum of Public-Private-Partnership (PPP) models are broadening coverage and providing quality services to underprivileged patients. Contracting is common for simple services -- diagnostics, canteen, janitorial and other non-core services. Contracting-out management services can alleviate health departments' operational pressures. Having started in 1996 with a single Primary Health Center (PHC), the Karuna Trust (www.karunatrust.org) now runs 30 PHCs across Karnataka state and 9 in Arunachal Pradesh. Innovations like 24 hour operation, community-based health insurance, and a Health Management Information System (HMIS) have been introduced in these PHCs. Alliances/Joint Ventures (JV) often leverage government infrastructure with private sector technical and management expertise. To provide advanced medical care in the poor, remote and underserved district of Raichur, the Government of Karnataka entered a JV with a private sector leader, Apollo Hospitals (www.apollohospitals.com/raichur), to run the Rajiv Gandhi Super-speciality Hospital. It has become a 422-bed state-of-the-art medical centre with several super-specialities. Social Marketing uses commercial marketing techniques to promote desired health-seeking behaviour, often in under-served populations. Hindustan Latex Limited ­ one of the largest global condom manufacturing and marketing companies ­ through its foundation (www.hlfppt.org) has worked with various state governments and HIV/AIDS programs to couple their brand promotion and product distribution with private clinic networks and public sector advocacy and reach to promote condom use in thousands of remote villages. Finally, Social Franchising builds a network of certified private sector providers to offer specific subsidised services, allowing vastly expanded reach within existing infrastructure. The Chiranjeevi Program in Gujarat is a good example (gujhealth.gov.in/Chiranjeevi_Yojana). It has used a voucher based system to greatly reduce maternal and infant mortality rates. Limited access to affordable qualified care at birth contributed to high maternal and infant mortality especially in scarcely populated tribal belts. Rather than expensive and slow expansion of public health infrastructure, Gujarat opted for a PPP in which 867 local private gynaecologists participate. For Rs.1795 (about $40) per birth, the results are impressive. It is reckoned that half of all mothers living below the poverty line avail the scheme. Institutional deliveries have gone up in Gujarat from 54 per to 86 per cent, and by early 2009, 270,000 deliveries had taken place under the scheme. "We estimate", says Amarjit Singh, the state's principal secretary, health, "that over 1,500 maternal and 8,000-9,000 infant deaths have been averted by Chiranjeevi". Gujarat's maternal mortality rate has fallen, in three years, from over 300 to 136/100,000. (13February 09, www.deccanchronicle.com/node/15862/print) Mother and newborn in post-natal hospital ward 1 The MERG was incubated by the World Bank to help developing countries improve health care delivery through building health sector capacity in business practices and management training. The aim is to improve health programs in developing countries by training health directors in business management skills. The MERG focuses on national and sub- national government programs that provide treatment and prevention to poor people in developing countries. 2 While privately run hospitals and clinics can be more convinced that to make a difference in India's epidemic efficient than their public counterparts, they are usually there was an urgent need for very high coverage of more expensive and consequently not available to people sharply focused prevention programs, in the highest with lower incomes. Programs in which public-private prevalence geographic areas, aimed at the highest risk partnerships or private health service delivery have groups and "bridge" populations (who have sexual targeted low-income and socially vulnerable populations partners in high risk groups and the general population). are generally marred by two main constraints: limited scale After consulting with the Government, Avahan was (they are usually done as pilots and in a limited geographic launched in 2003. area), or where programs have grown in scale, they have taken a long time to do so. And many of these programs to Agreeing on this clear program focus was the first and serve the poor focus on treatment rather than prevention. most critical step towards success. If Avahan had trod the same path as many others--trying to spread services One program that has applied private sector management across all of India, the results would likely be very different. practices to scale up a public health intervention that targets marginalised groups stands out. The Avahan Targeting program is an HIV initiative funded by the Bill & Melinda Avahan targets female sex workers, high-risk men who Gates Foundation. Its focus is on preventing HIV have sex with men, the transgender (hijra) community, transmission by targeting high risk groups -- female sex injecting drug users, and clients of sex workers. These workers and their clients, men who have sex with men, groups and their sexual partners likely account for over and injecting drug users -- in the highest prevalence six 85% of HIV transmission in India (Arora et al 2004). states in India (Avahan, 2008). This is not a new program Avahan's geographic focus is in six states (combined concept for India or for other countries with concentrated population of 300 million): Andhra Pradesh, Karnataka, HIV epidemics. What is new is the way this program has Tamil Nadu, Maharashtra, Manipur and Nagaland, which used business practices typically found in high performing accounted for 83% of infections in 2002, and now companies to implement its HIV prevention program. contribute 65% of all new infections (Gates Foundation 2007, NACO 2008). Within the six states, HIV is Before exploring the business practices Avahan concentrated in a small set of high prevalence districts in "borrowed" that might have made the difference, we look three key geographic clusters: at the nature of the HIV/AIDS epidemic in India and how this shaped the Avahan strategy. The Northern Karnataka - Southern Maharashtra corridor India's HIV epidemic ­ and Avahan's response The Coastal Andhra Pradesh delta North-Eastern India Avahan was conceived at a time of debate about HIV transmission dynamics in India and the character of India's Figure 1: HIV prevalence by district HIV epidemic. One school argued that India's HIV epidemic was becoming (or was already) generalized. (A generalised epidemic is sustained and driven by sexual behavior in the general population, independently of transmission from high-risk groups.) Others argued that HIV in India was still clearly concentrated among high risk groups of sex workers, high risk men-who-have-sex-with- men, transgenders or hijras, clients of sex workers, and injecting drug users in the north-east (Dandona etal 2006). This debate had significant implications for how to best program funds for HIV prevention. A concentrated epidemic calls for a rigorous and intense focus on interventions among high risk groups at the centre of transmission; a generalized epidemic requires major investments in prevention programs for the general population as well. Faced with this controversy, many programs took the conservative decision to invest broadly Source: NACO, India, 2005 in interventions for the general population, as well as running some programs targeting high-risk groups. Setting up ­ clear aims, substantial funding The Gates Foundation convened an expert group to look The Gates Foundation committed $100 million to Avahan carefully at the emerging data on the epidemic. They were in November 2002, and $100 million more in April 2003 as 3 the Avahan implementation strategy unfolded. The stated risk MSM and transgender communities; 2) variable quality aim and scope of the project were to create a fully and access to clinics treating sexually transmitted integrated prevention program to scale up coverage infections; and 3) absence of systematic structural among populations at greatest risk in India's highest interventions to address issues like violence facing high prevalence states, complementing government and other risk populations. donor work. The latter part of this is critical -- there were already some sound programs targeting high risk groups, Figure 2 shows Avahan and other program coverage at but few had been scaled up. Overall coverage of high risk the end of 2007 after 48 months of implementation; careful groups was variable and low. And there were other critical joint planning ensured complementarity and greatly gaps in the HIV prevention landscape that Avahan sought expanded access to services among the targeted to fill, namely: 1) very little coverage or programs for high populations. Figure 2: Complementary HIV Programming Manipur 62% 26% 12% IDU 35,000 est. Nagaland 53% 26% 22% 28,000 est. Karnataka 28% 72% 68,000 est. Andhra Pradesh 29% 61% 11% FSW 115,000 est. Maharashtra 72,000 est. 26% 74% Tamil Nadu * 38% 36% 26% 84,000 est. Karnataka 23% 77% 23,000 est. Andhra Pradesh 19% 76% 5% HR- 46,000 est. MSM Maharashtra 64% 36% 27,000 est. Tamil Nadu * 24% 49% 26% 21,000 est. Gov ernment of India & Others Av ahan Uncov ered Percentages indicate intended coverage though establishment Percent of Mapped Urban of services in specific geographic area. * Includes districts with no intended coverage. ** Mapping and size estimation quality varies by state. Key Population covered** Does not include rural areas. Source: Avahan and State AIDS Control Society program data Building the team Choosing a person with these qualifications to head up a public health program was unusual. When setting up the Avahan program, the Gates Foundation believed that selection of its director would be First on his agenda was to build a team with strong one of the most important decisions. Many articles in the management experience and most importantly, problem Harvard Business Review point to leadership as the most solving skills. The team he chose to roll-out the world's important factor in driving performance of organizations. largest HIV prevention program was unconventional to say The Foundation decided that the chief operating officer of the least. One of the three senior managers had been the Avahan would be selected for his/her leadership, business CEO of a successful software technology company, one management skills, and demonstrated skill in manuevering had co-founded a financial services investment company in the political landscape of India. Ashok Alexander was and the third had a global management consulting known throughout India as an innovative private sector background. Avahan also recognized the need for leader with experience and know-how in finding solutions scientific leadership to provide overall epidemiological and to difficult problems. A former Director at McKinsey and programmatic experience and credentials, and soon Company, Mr. Alexander headed its India office and had brought in a senior HIV leader. Three additional people tackled scores of difficult problems facing the Indian were recruited with private sector backgrounds in Government, including restructuring steel companies. 4 marketing and sales, consumer products, and operations. organization, making design and implementation They all brought critical ingredients: a consuming passion modifications as necessary. and sense of mission, rigorous focus on tracking outcomes and meeting targets, and an intrinsic comfort with solving Box 2. Watches to condoms? problems related to scaling up. Given their newness to the field they did not assume certain things often taken for After eleven years working in sales for one of India's largest watch manufacturers, Jonty joined Avahan, granted in the public health sector -- like prolonged periods India's largest HIV prevention program. Several skills of time to set up infrastructure and find partners and from her watch selling days proved valuable in HIV generally slow implementation. Targets and indicators prevention: focused on outcomes, not inputs. For example, when reviewing progress with NGOs they did not look at how Community-centric -- She had designed sales strategies many people were trained, but rather focused on how with one target in mind, the watch customer, who was much traffic STI clinics were seeing--the outcome they reference point for all decisions and policies. Now her needed. Helping change purchasing and other behavior main consumers are sex workers, MSMs and IDUs, amongst rural and urban populations, locating and whose needs, habits and predispositions drive the design deploying outlets, driving towards 100% market share and and roll-out of her strategy. meeting goals quarter after quarter without fail, were well- Scale -- The success of her former company's business honed skills they brought to Avahan. And they began to was dependent on its `footprint' or scale of operations employ these skills, and marketing and business strategies and ability to target the right areas. Similarly, Avahan's (see Box 2) to generate demand for HIV prevention success depends on rolling out a standard set of services services among their `customers' who happen to be female and infrastructure at scale across the high prevalence and male sex workers, transgenders, other men who have districts with market saturation as the goal. sex with men, and injecting drug users. Integration -- The "four Ps" of marketing (price, product, place and promotion) that are critical elements for a new From their collective experience in business they watch rollout, also guided the roll-out of HIV commodities understood that new programs have a high degree of early and services. failure. Successful ventures share similar characteristics, including developing a strong design that can be tested to Flexibility -- Staying ahead of the game requires ensure its appropriateness to solving the problem, flexible constant review and adjustment of strategies. While execution (implementation) with a "learning as we go" Avahan's basic program structure and goals stayed model, and high use of data to feed learning back into the constant, there was a similar emphasis on monitoring and adjusting the program to meet changing needs. Recruiting Partners Figure 3: The Avahan Organisation A key aspect of Avahan's strategy THE AVAHAN ORGANIZATION is to form partnerships with organizations that are already successfully implementing Foundation essential activities such as State-level staff in 5 9 Voluntary Counseling and Testing Strategy Lead locations Partners (VCT) and providing antiretroviral treatment and other drugs. 24 grantees, District-level 134 Grassroots 31 grants Planning Indian NGOS In the first six months of 6 states, operations, Avahan made major 83 districts implementation grants to nine Hotspot-level 7,458 Implementation Peer Educators and Outreach lead NGOs who would run the Workers programs for high risk groups and ~279,000 clients in the six target states. Individual-level Female Sex Workers High-Risk Men who have sex with men These large NGOs sub-granted to Tracking Injecting drug users ~5,000,000 Men at Risk about 134 grass-roots NGOs, Receiving Services which then worked with over 7,000 peer educators and outreach workers who became Cross Cutting Support: the face of the program. Figure 3 Capacity building, Advocacy, Monitoring and Evaluation, Knowledge Building presents the Avahan December 2007 data organizational structure. 5 The roll-out In order to keep close tabs on the rollout, Avahan staff Avahan's focused prevention program offers a package of spent most of the first year in field visits, and conducting services consisting of three main elements that had proven mapping and size estimations with partners to gather effective in reducing HIV transmission in high risk groups: information on its market, and establish a denominator ­ Peer-led outreach and behavior change i.e. total population needing coverage ­ against which it communication: Members of the high risk could plan and measure service uptake. It also worked communities are the ones reaching out to other high with partner NGOs to make sure they were on target sub- risk individuals to discuss strategies for negotiating granting to smaller NGOs, hiring staff and outreach condom use, information about STI treatment and a workers, and getting their procurement systems up and host of other HIV-related information. running. Avahan did not pilot first and scale later, but took Clinical services: offered at convenient centers the approach that with the solutions largely known, in order where key populations can be diagnosed and treated to achieve impact quickly, it had to scale rapidly, and fine- for sexually transmitted infections (STIs), rather than tune as it went along. Twenty four months into implement- just being referred or advised to seek care. ation, 83 percent of targeted high risk population members had been contacted by a peer educator at least once. Commodity distribution: Distribution of condoms to sex workers, their clients, MSM and hijras as well as free needles to injecting drug users in the northeast. Cutting across these core services, the program also focused on community mobilization interventions to address structural and environmental impediments such as violence, migration, mobility and barriers to accessing entitlements. These elements were critical for community buy-in to the program and stronger uptake of services. With the agreed elements for HIV prevention in place, Avahan moved quickly to create its `footprint'--which meant setting up a vast infrastructure of drop-in centers, STI clinics (where location proved to be key), and most importantly, its `sales force' which eventually comprised over 7,000 peer educators and outreach workers who A community meeting in Andhra Pradesh with Avahan staff became the front line of service provision (figure 4). Figure 4: Roll-out of infrastructure and services Avahan Reached Its Performance Targets was rapid SERVICES Table 1 compares Avahan with other excellent HIV Dec 03 Year 1 Year 2 Year 3 Year 4 prevention programs that also have contributed to checking India's HIV epidemic. The data underscore 489 Avahan's scale and rapidity, even benchmarked against other leading programs. Precisely comparable data are not available from all 4 programs, so the authors identified 30 basic measures of program performance and asked each Towns covered 7,458 program for aggregate data for each indicator from the beginning of the program until present. These rough measures show how rapidly Avahan eclipsed the scale of 240 others. For example, it quickly established programs in 4 Peer educators to 12 times as many districts, mounted 10 to 15 times as 278,000 many interventions, reached 6 to 20 times as many sex workers and 20 to 80 times as many male clients, treated 53,000 4 to 40 times as many STI clients and distributed 10 to 70 times as many condoms. Avahan provides convincing Denominator covered (FSW, MSM, IDU) proof that rapid, dramatic scale-up of prevention effort is 14.7 mil feasible. This result is resoundingly important in Asia, where scale and coverage of known proven interventions remains the greatest challenge. Figures 5-8 show the 0.8 mil dramatic increases in key coverage and scale indicators Condoms distributed and sold per month (in millions) from the very beginning of the program's implementation. Source: Avahan's routine monitoring data 6 Table 1. Avahan compared with other programs1 Program Aspect APAC Avert Sonagachi Avahan Time frame 1995­2008 (12 yrs) 2001­2008 (6 yrs) 1992­2008 (16 yrs) 2003-2008 (5 yrs) No. of Districts 22 7 14 83 No. of Interventions 64 52 43 650 Current No. of listed 50,600 SW, etc. 13,600 SW, etc. 34,000 SW 287,000 SW/MSM/IDU HRG (est.) 10,000,000 Male Current No. of male 500,000 Male clients, 300,000 (est.) 100,000 male clients/ clients, truckers clients covered others babus Current No. of STI 233,000 20,000 (est.) 30,000 885,000 consult. /year Number of Condoms 17 million distributed 3 million 4 million 180 million distributed distributed annually 12,000 outlets 147,000 outlets Note: HRG: high risk groups, SW: Sex workers, MSM: men having sex with men, IDU: injecting drug users Source: Internal Routine Monitoring data provided by each program; websites www.apacsvhs.org, www.avertsociety.org, www.durbar.org Figure 5: Condoms Distributed Annually Figure 6: High Risk Group Individuals Reached Figure 7: Number of Bridge Population Figure 8: Number of STI Clients Serviced Reached ­ Truckers, Male clients (The dip in year 5 is because Avahan's male client STI programs ended in early 2008 when they were assumed within the Government (NACO) male client STI program.) 7 Figure 9: A comparison of the increases in prevention service coverage achieved by Avahan with the (limited) behavioural and biological data available shows strong increases in use of condoms and reductions in STI rates in the high risk groups targeted by Avahan (Figures 9 and 10). The results are not attributable only to Avahan interventions. In particular, Avahan covers only a small percentage of East Godavri, where the Andhra Pradesh government also supports interventions. The STI decline reflects the combined effect of Avahan and Government of India interventions. Figure 10: 8 Substantiating these encouraging intermediate trends, Avahan acknowledges that their initial cost structure, outcome data support the hypothesis that Avahan with its emphasis on urgency and the best personnel contributed to accelerated and deepened declines in HIV and partners, was higher than comparable Indian norms. transmission, shown below among pregnant women in They believed that an early investment in strong Dharmapuri district in Tamil Nadu (Figure 11). management was needed to establish a program that would focus on quality and give attention to hands-on Figure 11: implementation. These costs are being brought down as Avahan transitions the program to government ownership. Four Elements of Success It doesn't matter how sound the policy or how well intentioned the program is. If an organization doesn't pay close attention to management details ­ and how it will work on the ground--the result will be program failure. Successful programs need a little luck, but they need a lot of attention to management detail Professor Billy Hamilton, LBJ School of Government, The University of Texas Organizations that meet their goals and out-perform competitors typically have a number of characteristics in common. They have a clear purpose, flexible but consistent execution, use data for decision making, and are willing to admit mistakes and correct them (Rist 2005). Avahan's scale and speed distinguish it from Evidence from Karnataka state is even more impressive ­ other programs, and a closer examination shows how it shows no clear decline in HIV prevalence among young these four elements of success were embraced by pregnant women in non-Avahan districts, contrasted with a Avahan and were integral to its implementation. steep decline in Avahan districts (Figure 12). (i) Clarity on what success looks like Figure 12: Decline in HIV in young pregnant High-performing organizations know what success looks women attending ANC clinics in Karnataka ­ like and can define it simply and clearly. Avahan comparing Avahan and other districts believed that HIV in India could be reduced substantially by rapid, high coverage, high quality implementation of targeted interventions for sex workers, men-having-sex-with-men and drug users in India's highest prevalence states and districts. Furthermore, effective organizations are able to communicate their purpose and objectives so that all members of the organization understand their role in contributing to the common purpose (often called mission). For Avahan, this extended beyond the members of the Avahan team. Unlike a traditional donor-grantee structure where the donor and grantee maintain distinct structures and goals, Avahan sought to bring all members of the implementation pyramid (the 9 lead grantees, 6 capacity building grantees and 134 grassroots NGOs) under one umbrella, creating a unified sense of mission for the entire endeavour. Figure 13 (reading from 9 the bottom up to the top) represents Avahan's clear Box 3. Avahan's Strategy goals, strategy for attaining them, and enabling and The Avahan project design involved a four-pronged sustaining factors. Because Avahan opted not for a strategy: general population HIV program but for strategic strong focus in geography and target populations, it was able to 1. Simultaneous setup across regions/grantees of 3 articulate its priorities clearly, and translate these into phases of service delivery ­ first establishing the uniform objectives for the entire implementation pyramid. infrastructure, then increasing the scope and quality of coverage, and finally including additional services Figure 13: (e.g. community mobilization) into the core services; 2. Customizing services to client populations ­ adapting the standard package for local conditions to maximize uptake and coverage; 3. Addressing barriers to service uptake and generating demand ­ which often involved going against conventional wisdom and resulted in community mobilization becoming a key strategy; 4. Execution focus at every level using a structured management approach that combined design, data and ground reality: Programming standards which set guidelines, minimum quality and expected milestones through the Common Minimum Program (see While the goal was clear, the geographical scale of what Box 4) which could be adapted to local context. was required to rapidly reduce HIV transmission in India Field engagement. Regional deployment of was no small task. Avahan clearly articulated a definition people and partners enabled frequent field trips of success that required simultaneous and rapid scale up to see grassroots NGO efforts as well as (not piloting), and a massive footprint that covered the government and local influencers. These were entire geographic priority area of India's highest working visits to contextualize the monitoring prevalence states and districts (Box 3). data, understand practical implementation issues and anticipate future ones, and validate claimed Given the priority placed on scaling up, Avahan correctly successes. The Common Minimum Program identified management rather than technical expertise as even includes field engagement rules for all the most important skill set for the first phase. Recruitment levels of program participants. focused on finding a competent manager and supervisors, Analysis and use of monitoring data. Routine each of whom was given a defined geographic area in monthly reports are scrutinized for adherence to which to work and focus on standardizing services the planned service rollout and utilization levels. provided by the grassroots NGOs who were the Further analysis of anomalies often uncovered implementing partners. The early priority given to mapping program deficiencies and led to improvement. and enumerating the target populations ­ sex workers, Over time, this has decentralized to peers and clients, high risk MSM, hijras and IDUs ­ enabled accurate NGO staff to improve data accuracy and targets to be set against which to measure progress. Only stimulate detailed local problem solving. with an accurate denominator could all members of the Program reviews. With each partner, Avahan has project ­ from Avahan managers to outreach workers ­ formal 6-monthly program reviews to assess know their targets, monitor their progress and adjust progress, address issues and set future course tactics to achieve rapid service uptake and high coverage corrections and milestones. Each implementation or "market share". partner holds similar reviews with their sub- contracted NGOs. At an overall program level, With a clearly defined standard operating procedure (the "all partner meets" are also held to review overall "common minimum program" explained in Box 4) and a and regional performance (a strong disincentive to lag behind), share best practices which can management information system (MIS) that was simple then evolve into common standards, and refine and easy to use, a unified sense of purpose was ingrained and agree program direction. The core Avahan in the program--everyone, from the director down to each team interacts more intensively through weekly outreach worker has a clear understanding of what they review calls, multiple team meetings, and are there to do. constant sharing of issues and solutions. 10 Box 4: The Common Minimum Program The second characteristic of success is staying flexible The program struggled for the first year to get the during implementation while applying clearly defined multiple, diverse and spread-out entities and people in policies and practices in a consistent manner across the the virtual Avahan organization to work as one, until it program. This enables decision makers to use what they developed and adopted its common vision and operating learn about what works and what does not work during standards -- the "Common Minimum Program". Practical implementation, to make constant program-wide learning from the ground combined with high level improvements. The first success element stresses the priorities created a guide which encompassed: importance of a clear and common purpose; this element reinforces the importance of maintaining a flexible results- 1. Standards for Programmatic and Technical focused strategy for achieving it. approaches in interventions ­ including STI Clinical Operating Guidelines and Standards (COGS) and The two main approaches Avahan took to ensure guidelines and tools for all services, outreach, Behaviour consistency with flexibility were (1) designing systems to Change Communication, community participation, etc. guide implementation, and (2) creating incentives to 2. Key Project Milestones to capture pace of reward innovation. Once implementation began, Avahan infrastructure and service roll-out with desired sequence developed a set of guidelines called the Common and quantitative progress or service uptake targets. Minimum Program (CMP) which were meant to drive These evolved, and at each stage set direction and priorities and share learning across all implementing clarified priorities across the program to drive in a partners and grassroots NGOs. The CMP is an operating common direction. manual that combines technical standards, templates/tools for implementation and key milestones for program 3. A common Program Management Framework to success (see Box 4). The standards are improved and drive execution. This defined relationships across the adapted to local conditions where appropriate ­ relying on extended organization, specified ownership of areas feedback from peers' micro-planning to determine the across the partners/ NGOs/ people, expectations of optimal combination and intensity of services. support and field engagement intensity, review processes, etc. Examples of how the CMP enhanced Avahan's program 4. Data collection for decision-making with clearly performance abound. The following examples illustrate defined tools, processes and specific metrics for data how customized delivery approaches and interventions collection and analysis at all levels. These include enable Avahan to maximize its reach with available grassroots (peer-level) to program-wide routine resources: monitoring metrics and indicators, qualitative and Most NGOs follow the uniform STI management quantitative assessments/ surveys, repeat mapping and protocols in the STI Clinic Operating Guidelines and denominator size estimation, condom (or needle/syringe) Standards (COGS). But the location of service points requirements, etc. is influenced by peer input, which helps decide whether to establish fixed clinics close to hotspots, or mobile clinics in vans or tents. To cater to the (ii) Consistent but Flexible Execution constraints of sex-workers in brothels, satellite clinics Peter Drucker, the father of modern management science, are held there on pre-scheduled days. In areas with once said "There is nothing so useless as doing efficiently low numbers of high-risk group members, private that which should not be done at all." The design of many, providers are trained and contracted to provide clinical if not most, development programs is based on a theory of services. change that is agreed upon before the program starts. Centres to serve marginalized sex workers and MSM Unfortunately, programs are rarely evaluated during their are nondescript (to avoid drawing adverse attention), execution to determine if the logic of the original design but where it made sense, a distinct and visible brand remains valid and is really working. Sometimes this type of was promoted. The 736 KEY clinics in the PSI evaluation is conducted after the program has ended, but program targeting male clients and 17 KHUSHI clinics is seldom done during its execution (World Bank, IEG in the TCIF program targeting truckers were uniformly 2008). Thus, programs are often implemented using branded and conveniently located for easy untested designs, with little incentive to restructure the identification and access. The clinics that targeted program or change its original design despite lessons that male clients use a franchise model and enrolled may be learned along the way. The incentives are usually physicians who were already seeing a high number of to meet time-bound targets in implementing program male STI patients. activities, without really knowing whether the activities help to achieve key performance targets. Avahan leadership Avahan partners and their staff also demonstrated believed in testing their design against expected goals, their commitment to flexible implementation during and did so. regular program reviews. Regular informal and periodic formal reviews were held off and on-site to 11 discuss implementation progress and to arrive at customer paradigm one step further ­ and made course corrections (Avahan 2008) and necessary community members, with their inside knowledge and shifts in implementation focus. The reviews provided access, an integral part of program design and execution. lessons to be shared across the different sites, and feedback and support to parts of the program that Avahan's Central Management Information System might be experiencing implementation challenges. (CMIS) plays a key role in tracking the program at all levels - from individual customers, to NGOs, up to the In the early 1980's Tom Peters made management history foundation management team. The CMIS data allow real with his book "In Search of Excellence". Focused on time decision making. Performance is evaluated across all improving the quality of organizations, the book levels and quick corrective action is taken when underscores the need to continuously improve how an opportunities are noted. The partners have devised organization delivers its goods and services because one innovative mechanisms for ensuring quality, using these cannot assume that when reforms are introduced they will data. For example, the Foundation that runs programs for "stick" or that they will remain adequate in the days ahead. truckers evaluates and rates its grantee NGOs in detail Avahan's commitment to flexible management was based twice a year on seven parameters (Staff retention, on a similar philosophy. By allowing their competent and Interpersonal communication, etc.) (Figure 14). committed team the freedom to test the design and Performance below 60% of target is coded red for innovate based on learning, and by providing strong immediate correction. Items for which performance is 60- continual management of the Avahan program with few 80% are coded amber or needing improvement; things formal chains of approval, reaching targets quickly became above 80% make the cut and are coded green. NGOs possible (see figure). rated green on all seven aspects get a prestigious "GOLD Standard" certification for the year. This encourages holistic quality, and also builds motivation and self-esteem Unique while fostering excellence through friendly competition. People Formula At the next level, Avahan's MIS collects indicators on Free- service provision, uptake and community activities across dom to 134 grassroots NGOs. Aggregate data reveal the target innovate communities' interactions with peer educators and Active Management utilization of the STI clinics. Data generated by the client of & across community also feeds into the MIS, not just to report Programs upwards, but to inform and prioritize outreach by peers. Avahan's information system is distinguished from those of Achieve Results many public health projects in two critical ways: first, the system is designed to provide real-time information, to (iii) Use of Data for Decision Making enable constant performance management and tactical Understanding the market it serves is fundamental to an changes as required. Second, it is designed primarily for organization's ability to meet performance goals and hands-on data-driven project management rather than targets. Examples of market information questions are: for just reporting upward and outward. 1. Who are our customers? 2. Where are they located and how can we reach (iv) Willing to Admit Mistakes and Correct Them them? In order to provide the right incentives for innovation, the 3. What quantity and quality of services do they Avahan leadership kept its eye on the impact it hoped to want? achieve, and maintained a "build the ship as we sail" motto 4. When is the best time to provide them with the to encourage managers to look closely at findings and not services they need and want? be afraid to try new things. This extended to relationships with grantees also. Going back to the drawing board with a Customer paradigm: Avahan program managers' grantee was not viewed as a failure, nor were they deliberate strategy is to treat its target populations of sex penalized for not getting the design `right' at the outset. workers as customers to be sold a specific product: safe Rather it was viewed as requisite midcourse correction to sex/safe behaviour. In implementing this strategy, achieve maximum impact (Box 5). In day to day consumer needs and viewpoints are central. The team operations, program managers were given incentives to worked to adapt product marketing, build brand loyalty and remain flexible and spend a lot of time in the field looking develop specific techniques for increasing uptake. Some for structural barriers impeding program uptake, closely time into the program, they observed that community-led monitoring monthly data from across the program and programs like Sonagachi had similar power and making adjustments where needed. sustainability to direct-marketing. So Avahan took the 12 Figure 14: The Management Information System flags sub-par performance for immediate correction, and creates pride in strong performance Box 5: Refocusing the Truckers Program ­ the 80-20 rule in action The truckers program is run by the foundation arm of India's largest trucking company ­ Transport Corporation of India Foundation (TCIF). TCIF initially set up intervention sites at 36 locations along major national highways. Within two years, however, program data indicated that despite a national presence, critical program gaps remained. A behavior tracking survey in mid-2005 revealed that program awareness among the target population was only 12 percent and service uptake was only 7 percent. Other data revealed that approximately 40-50 percent of services were inadvertently directed at individuals other than the highest risk long-distance truckers (such as short distance truckers and other people working at the transshipment locations). The grantee partner and Avahan managers were faced with a choice: should they end the investment in this poorly performing program, or go back to the drawing board? Realizing that there was much to gain from a successful truckers program, they stepped back, and redesigned based upon the 80-20 rule in business ­ 80% of all sales or profits typically emanate from about 20% of outlets. The result was a halving of intervention sites - from 34 to 17 hubs. In addition, the team focused on improving accessibility and visibility of services at each location through intelligent placement of services. Other change created a trucker-friendly standardized interface staffed with trucker peers in all locations to increase brand recall and credibility, and the (unchanged) budget was used to intensify services in the smaller number of locations. This soon paid off: outreach/clinic services uptake doubled, condom sales increased 50%, and over 85-90% services reached long-distance truckers. Table 2: Approximate comparative costs per person reached Costs (estimated, nominal) APAC Avert DMSC Avahan Total Intervention spending (direct implementation costs (Sonagachi) and associated costs for advocacy, capacity building etc) Cumulative Program life-time cost per registered HRG ~$320 ~$450 ~$300 $418 member Cumulative Program life-time cost per Bridge Group ~$30 ~$60 ~$20 $7 member reached Source: Public data (www.usaid.gov/in, www.apacvhs.org, www.avertsociety.org, www.durbar.org) and internal documents (routine monitoring data, budgets) provided by the respective organizations Note: The programs' duration, scope and approaches vary. Limited data are available (especially for other entities' cost-breakdowns). Given this paper's primary focus on large, fast impact, per-capita cost over the program lifetime (rather than annual cost) was deemed most relevant. These were calculated separately for High Risk Groups (HRG ­ sex-workers, transgenders, MSMs & IDUs) and bridge groups (male clients like truckers, migrants, slum/street-dwellers, youth, industrial workers and Babus) because of vastly differing intervention packages and costs (much higher cost package of STI treatment, individual tracking, advocacy, capacity building, community mobilization, etc. for HRG). Costs shown are total program prevention-oriented direct intervention costs (total grant amount less estimated indirect costs ­ Knowledge Building, M&E and Govt. Support ­ and deducting Care & Support costs). Indirect costs were split between HRG and Bridge Population interventions. Per-capita costs were estimated for the latest number of people in HRG or bridge populations contacted or registered with each program. 13 Can sound business practices help public health Table 2 shows that Avahan's cumulative program costs programs achieve desired change faster, and at per sex worker are at the high end of comparable comparable cost? programs in India. As noted, Avahan invested heavily in creating a well-managed, high quality program. Also, there Dr. Robert Childs of the Wharton School of Business was early agreement that the core Avahan team would believes that in order for an organization or program to have a hands-on approach to overseeing implementation achieve its goals it must pay attention to 17 elements of with many visits to the field. This type of implementation is strategic execution. Beyond strong teams, knowledge and more costly due to the intensity of contact and learning feedback, this business model has shown that comprehensiveness of Avahan's sex worker program, with good implementation determines success or failure. its strong focus on rights, empowerment and community Avahan is closely aligned to the Wharton model through ownership as critical intervention elements. However, strong design focus, commitment to engaging and Avahan's bridge group client costs are the lowest, motivating its team, and most certainly in picking the right reflecting the ability to reach large numbers of male clients people for the right jobs. Avahan turned traditional of sex workers efficiently (and offering indirect support for management of public health programs on its head by the proposition that higher sex worker costs are a function treating management skills as the scarce resource most of intervention intensity and comprehensiveness). needed for success, and technical skills as supportive skills that could be acquired more easily. Throughout its Having established scale-up and impact, Avahan now execution, Avahan has challenged conventional wisdom faces the challenge of aligning its cost structure with on how to manage public health programs by bringing in national norms. This is important because one of Avahan's business practices such as market research and analysis, early goals that has not been achieved yet is to turn the along with traditional public health technical skills. Avahan project over to it natural owners, the Government of India also used strong social marketing skills, leveraging a and the target communities. This likely would require the "sales force" of the affected communities to "sell" safer costs to be absorbed and sustained by the national sexual practices such as regular use of condoms to groups government. Avahan has shown that cost alignment is most vulnerable to contracting HIV. feasible, as illustrated in Figure 15. Phased reductions in the intensity and scope of intervention components could Questions asked by many who have studied Avahan is bring Avahan's intervention cost structure quite close to whether or not it is replicable in other locations, how government costs, based on AP's cost structure. Avahan Avahan's costs compare with similar programs, and can achieve both rapid, early, high impact scale-up and a whether the program's method for achieving change and transition to sustainable national cost standards. strong focus on implementation could be adapted to addressing other public health challenges. Figure 15: Potential savings identified to align Avahan costs to national norms Cost comparisons Avahan's initial funding of $200 million seems large at first glance; for example, most World Bank health projects are significantly smaller than this. However, the amount is not overly large for a health program in India. In 2007 the World Bank committed a similar amount through a loan agreement with the Government of India for a third HIV/AIDS program. HIV/AIDS grants to India from the Global Fund to Fight AIDS, Tuberculosis and Malaria between 2003 and 2007 total $351 million (four grants range from $30 to $141 million). 14 Beyond HIV/AIDS in India The central questions we posed were (1) whether the There are numerous persistent public health challenges in business model that underpinned Avahan was the reason India. Faster progress might be made by selecting other behind the rapid results observed, and (2) whether it would significant public health problems for which effective be possible to adopt and adapt these practices to address solutions are known, and identifying the roadblocks. other public health problems. Avahan's achievements are Applying the four elements of Avahan's success described impressive, and comparison with other programs suggests above, using business management practices, might help that the program has been able to achieve significantly improve other health outcomes in India. greater expansion in access, behaviour change, and impact than similar programs over a similar time, and with fairly comparable costs. Members of MSM HIV prevention NGO References, further information Arora, P, A Cyriac and P Jha, India's HIV-1 epidemic. Dandona, L, V Lakshmi, G Anil Kumar and R Canadian Medical Association Journal, Nov. 2004, 171 Dandona, Is the HIV burden in India being (11); 1337-8 overestimated? BioMedCentral Public Health 2006, 6:308 doi:10.1186/1471-2458-6-308 Avahan: India AIDS Initiative, page on Bill and Melinda gates Foundation website NACO India, Note on HIV Sentinel Surveillance and http://www.gatesfoundation.org/avahan/Pages/overvie HIV Estimation 01Feb08. On line at: w.aspx http://www.nacoonline.org/Quick_Links/Publication/ME _and_Research_Surveillance/Reports_and_Surveys/N Avahan--The India AIDS Initiative: The business of ote_on_HIV_Sentinel_Surveillance_and_HIV_Estimati HIV prevention at scale. Bill & Melinda Gates on_01_Feb_08/ Foundation. New Delhi, India. 2008. http://www.gatesfoundation.org/avahan/Documents/Av Steen R. et al., "Pursuing Scale and Quality in STI ahan_HIVPrevention.pdf Interventions with Sex Workers: Initial Results from Avahan India AIDS Initiative," Sexually Transmitted Claeson, M. and A. Alexander, Tackling HIV In India: Infections 82, no. 5 (2006): 381­385. Evidence-Based Priority Setting And Programming, Health Affairs 27 (4) 1091-1102. On line at: http://siteresources.worldbank.org/EXTSAREGTOPHI VAIDS/Resources/TacklingHIVIndia.pdf 15 Acknowledgments: About the authors: The authors would like to acknowledge all the people who Jody Zall Kusek is the Lead Monitoring and Evaluation supported the research of this paper, met with us and Specialist in the World Bank Global HIV/AIDS Program. generously provided information, including Hari Menon, jkusek@worldbank.org Alkesh Wadhwani, Aparajita Ramakrishnan, Jayanti David Wilson is Lead Specialist on HIV prevention in Rajagopalan, Aparajita Bhalla, Stephen Moses, Vandana the World Bank Global HIV/AIDS Program. Gurnani, Sushena Reza-Paul, Sanjeev Gaikwad, Abhishek Jain, Amit Srivastav, Tarun Vij, Bimal Charles and J. Austin Thomas worked on this evaluation as a consultant Pratheeba of APAC, Anna Joy and Ernest Noronha of to the World Bank Global HIV/AIDS Program Avert and Samarjit Jana and Shyamal Ghosh of DMSC. We would like to thank the Government of India National AIDS Control Organisation, in particular K. Sujatha Rao, and G. Ashok Kumar. We wish to acknowledge the time and effort taken by Padma Chandrasekaran and Gina Dallabetta in the early review of this paper and for their helpful comments. We would personally like to thank and acknowledge Negar Akhavi for her new found friendship Photographs: and continued counsel in the development of this work and Page 2: © World Bank, photographer John Isaac. other matters on AIDS in India. Finally we wish to say All others: ©copyright Bill & Melinda Gates Foundation thank you to Ashok Alexander for showing us that the solutions to intractable public sector problems can come from those with open minds about finding new solutions and who never say "it won't work"...... A member of the transgender community For more topics in the "HIV//AIDS - Getting Results" series, please go to www.worldbank.org > Getting Results March 2009 16