(introduction...)

Once the question to be answered is clear, the purpose of the costing and the scope of the work have been defined, and it is apparent what alternatives are being costed, the exercise can proceed to describe each alternative in detail. To do this it is useful to be clear about how costs will be classified. To undertake cost analysis of HIV/AIDS prevention, a primary classification by input and organisational level is recommended. It is also useful to make a secondary classification of the strategies by the activities that are taking place within them. This helps ensure that costs are not duplicated or omitted between inputs and activities. It also provides a framework by which a strategy can be costed by activity if this is deemed useful for decision making.

Inputs

The most common inputs to be found across all strategies are listed in Table 2 below and on the work sheets. When costing a particular project, it is necessary to specify in detail the inputs within each category and this can be done on work sheet forms 'C'.

Table 2. Classification of Costs by Main Input Categories

Capital Costs	Recurrent Costs
Buildings Equipment Vehicles Consultancies (non-recurrent)	Personnel Supplies Vehicle operating and maintenance Building operating and maintenance Consultancies (recurrent) Other (including media fees)

It is important to remember that all items which have a life longer than a year are treated as capital items. This means that capital items need to be annualised over the number of years that the items are expected to last (annualisation will be discussed in chapter 4). Thus all IEC material, training and consultancies whose impact is expected to last more than a year need to be treated as capital items. In order to gauge whether an item lasts more than one year, it is important to look at the frequency of the item in the project - e.g. how often is refresher training provided.

Tip-of-the-Trade 1: Activities as Inputs Another cost which may appear as an input category is start-up activities. Although this is really an activity, adding activities to the input list can be justified in certain cases. “If one activity is clearly separate from the others, both financially and administratively, it may be easier not to attempt to break it down into its component physical inputs, but merely to record the total cost. For example, start-up, training and social mobilisation activities are treated as categories of inputs and included along with personnel, vehicles and the like. When this is done, it is assumed that all the resources required for the activity (e.g., personnel and vehicles) are included in that category (e.g. training) and not under the separate categories of personnel, vehicles, and so forth. Thus, the full cost of all inputs used for training is estimated and used as the value for that category” (PHC: 9). Start-up, training and social mobilisation activities whose impact is expected to last more than one year are treated as capital items. In these guidelines, consultancies are also regarded as an input for reasons explained later. It is, therefore, important to ensure that none of the consultancy costs are double counted. For example, fees and allowances paid to individuals working on a consultancy input should not also be costed in the personnel inputs.

Activities

Resource inputs combine to accomplish activities. It is useful to present cost data by activity. Bringing together measures of cost and output by ‘activity’ rather than by the more aggregated ‘strategy’ can provide an extra tool in project evaluation. Each HIV/AIDS prevention strategy is likely to encompass a range of activities, some of which will be common across all strategies and some of which will be unique to particular strategies. Table 3 displays a categorisation by activity for each of our nine sample strategies.

Table 3. Classification of strategies by activity.

STRATEGY	ACTIVITIES
1. HIV Blood Screening	Donor recruitment - including everything related to motivating, educating, recruiting, selecting, screening, counselling and retaining blood donors. Blood collection - including all costs of collecting blood from accepted donors. HIV testing - including the first and any subsequent confirmatory tests. Blood processing - including all costs (except HIV testing) related to the processing of blood, including other testing, blood grouping, preparation of blood products, and all measures taken to ensure safety of blood prior to storage Blood storage/distribution - including all steps taken in storing and transporting blood through to its final transfusion in hospital. Training - training of all personnel, including laboratory, medical, managerial and support staff - training of all personnel, including laboratory, medical, managerial and support staff Management and administration - including planning and supervision of the strategy.
2. Mass media	Development and production of IEC materials - often includes focus group discussions and market research; testing, revision and re-testing of messages with sample audiences; and translation of the materials into a range of local languages. The costs of the production of the IEC materials once they are designed should also be included here. Transmission/distribution of IEC materials and messages - modes and costs of transmission will vary according to the type of media being used. Sometimes these activities will include pre-publicity by advertising firms, for example prior to launching television or radio dramas. Management and administration - including costs of the overheads of the campaign, planning costs and the cost of evaluating the strategy (including baseline surveys and post-campaign evaluations).
3. HIV school education	Development and production of curriculum and educational materials- includes preliminary work such as focus group discussions and testing, revision and re-testing of messages with sample audiences, as well as the production of the final materials. There may also be continuous production of materials for the project, for example when an AIDS magazine or newsletter is established. Training - A variety of school staff may be taken away from their routine activities and given training to implement the new curriculum. Ideally, training should address teaching methods as well as provide information on HIV/AIDS. If the new curriculum becomes an on-going activity, staff should be given refresher training as well as initial training Implementation of curriculum - comprises in-school teaching activities of the staff responsible for teaching the HIV/AIDS curriculum Management and administration - day-to-day management and administration of the strategy by the overseeing department, for example the Ministry of Education or Ministry of Health and also include the costs of planning and evaluating the strategy.
4. Condom social marketing	Development and production of IEC materials - may include market research, testing, re-testing and any necessary translation. Although promotional materials may be produced for mass media transmission, some may also be produced for more local distribution, particularly true in countries where explicit advertising of condoms is restricted. Local promotional materials may include such items as calendars, diaries, key rings, stickers and T-shirts. Technical assistance from specialist marketing consultants is commonly bought in from time to time during the project. Training - includes retailers receiving some guidance on the correct use of condoms so that they can relay this information, and possibly other HIV prevention information, to their customers. Necessary training may be undertaken by the social marketing organisation or contracted to the commercial manufacturer. Transmission of IEC materials/messages - includes all advertising and promotion activities. Activities may include sponsored conferences and professional seminars, award ceremonies for achievement, use of mass media, face-to-face distribution of promotional materials through health professionals, pharmacists, midwives, traditional birth attendants, traditional healers, shopkeepers and itinerant salesmen, and mass distribution in hotels, bars and companies. Condom distribution - includes costs of storing and packaging the condoms as well as distributing them to wholesalers or retailers. Distribution channels may be private or public sector or both. Packaging and distributing may be contracted out. Management and administration - includes day-to-day activities, planning and evaluation activities and technical support from management consultants.
5. STD treatment	Development/distribution of diagnostic and treatment protocols - this may include strengthening of existing laboratory facilities Training - including clinical and support staff Diagnosis - includes both laboratory and syndromic diagnosis Counselling - including drug and condom provision Partner notification - partner tracing, notification and counselling Management and administration - including planning, supervision and evaluation
6. Sex worker peer education	Development and production of IEC materials - for the target population Recruitment of peer educators - can be thought of as a social mobilisation component in the project. This activity will occur throughout the lifetime of the project, although the main recruitment will occur at the beginning. Both staff and other peer educators could do this. Training - of all staff and particularly of the CSW peer educators who will have responsibility for conducting the educational sessions includes refresher training Transmission of IEC materials/messages - to the target population, through individual and/or group sessions. Condom distribution/sales Management and administration - including planning, supervision and evaluation of the strategy.
7. Voluntary counselling and testing	Development and production of IEC materials - Clients may be shown a video, or handed educational materials at the pre-test counselling session; materials require development, testing, and production. Training - counsellors need to be trained in appropriate techniques, with fully trained counsellors available to conduct the training and supervise staff with lesser levels of training Counselling - pre- test - done with individuals, couples, or groups. Individual counselling may be followed or preceded by a group session, sometimes involving showing a video. Condoms may be distributed free of charge or sold. Blood may be drawn immediately for those who decide to proceed with the test, or at a separate visit. Testing - laboratory testing of blood may be done on site or at a central facility Counselling - post-test - all individuals may be given their results at a post-test counselling session, or only those found positive Management and administration - especially for the free-standing centres, there will be a number of management activities to be carried out.
8. Harm reduction	Development and production of IEC materials - often includes focus group discussions and market research; testing, revision and re-testing of messages with sample audiences; and translation of the materials into a range of local languages. The costs of the production of the IEC materials once they are designed should also be included here. Transmission/distribution of IEC materials and messages - modes and costs of transmission will vary according to the type of media being used. Recruitment of peer educators - can be thought of as a social mobilisation component in the project. This activity will occur throughout the lifetime of the project, although the main recruitment will occur at the beginning. Both staff and other peer educators could do this. Counselling - designed to decrease risky behaviour, such as sharing equipment Syringe/Needle Exchange - exchange of used needles/syringes for clean ones Disposal of contaminated needles - this could include provision of sharps containers, transport, biohazard disposal and possible incineration. Bleaching/Distribution of substances to clean equipment - distribution and advice of cleaning solutions Recruitment of peer educators/volunteers - can be thought of as a social mobilisation component in the project. This activity will occur throughout the lifetime of the project, although the main recruitment will occur at the beginning. Both staff and other peer educators could do this. Condom distribution - includes costs of storing and packaging the condoms as well as distributing them to wholesalers or retailers. HIV testing - including the first and any subsequent confirmatory tests and counselling pre and post-test. Referral to other services - includes referral to other medical services(e.g. STI treatment), social welfare and legal support Training - training of all personnel, including laboratory, medical, managerial and support staff Management and administration - including planning and supervision of the project
9. Mother-to- child transmission	Strengthening Antenatal facilities - including provision for HIV testing Training - of staff Purchasing and storage of ARV drugs and replacement feeding Counselling - before notification of results, during ARV regimen and for follow-up Management and administration - including planning and supervision of project and follow-up of mothers and infants

In practice, all cited activities may not be happening within the implementation of a particular strategy. It is, nonetheless, likely that implementation of each strategy entails some combination of these activities. In some cases a resource input, as specified in the primary classification, will be used solely for one activity and in other cases it will be shared between activities. For example in a government STD clinic, drug inputs may be entirely consumed by the 'STD treatment' activity, whilst a clinician's time may be split between ‘STD treatment’, ‘counselling’, ‘condom distribution’ and ‘management and administration’.

Some specific comments about activities in different HIV prevention strategies:

HIV Blood Screening - In a developed BTS the majority of these activities will have been taking place prior to implementation of HIV blood screening. It might therefore be tempting to assume that an HIV screening strategy can be costed by measuring only the incremental costs of adding the HIV blood testing. The situation is, however, more complicated than this. In areas where HIV prevalence is high, replacement of discarded blood can constitute the main cost of HIV safety ⁽²¹⁾. The cost of HIV blood screening can, therefore, be dependent more upon the costs incurred on pre-testing activities than on the costs of the HIV testing itself. For this reason it is not sufficient to measure only the incremental costs of the HIV blood testing activity. Rather, the costs of implementing an HIV blood screening strategy will be the costs of all resources used in undertaking the HIV testing plus the full replacement cost of the blood that tests HIV positive and is discarded. The unit costs of this blood will be made up of the costs of donor recruitment, selection and counselling, blood collection, tests conducted before the HIV test, and a proportionate share of overheads such as administration and staff training, supervision and management.

To obtain such unit costs, the costs of some activities will have to be obtained for the wider BTS and apportioned to the strategy activities in question. For example, management and administration costs are likely to be shared between pre-testing and post-testing activities and laboratory costs may be shared between HIV tests and other tests. Having embarked this far into a costing exercise it might be pragmatic to obtain costs for the BTS as a whole, rather than go half way and measure only those costs relevant to the HIV screening strategy. How complete a costing exercise to conduct will depend upon the objectives of that exercise and the audience interested in the results. The manager of a BTS may be interested not only in the cost of making blood HIV-safe but also in more comprehensive and more generally applicable cost data. By costing the whole BTS, total costs can be presented by activity and the proportion of costs attributable to the HIV blood screening strategy can also be disaggregated. If a full costing is deemed appropriate, detailed guidance can be sought from WHO's Costing of Blood Transfusion Services ⁽³⁷⁾.

CSW - There may be other complementary activities taking place such as initiatives to reach clients for educational sessions. Other activities for CSW projects may not be directly related to HIV prevention, e.g. cre and income support activities.

VCT - Some centres provide support to HIV positive groups, but this is not considered here as a core activity of a VCT service.

IDU - In practice, all cited activities may not be happening within the implementation of a particular strategy. It is, nonetheless, likely that implementation of each strategy entails some combination of these activities. In some cases a resource input, as specified in the primary classification, will be used solely for one activity and in other cases it will be shared between activities. For example a clinician’s time may be split between ‘STd treatment’, ‘counselling’, ‘condom distribution’ and ‘management and administration’.

Tip-of-the-Trade 2: What to include if calculating costs in a clinical trial Frequently, cost and cost-effectiveness analyses of HIV prevention strategies are done in the context of ongoing clinical trials. In order to consider the feasibility of delivering these services, it is very important to include all costs that were necessary to ensure the functioning of the project. For example, if the project had to provide additional basic infrastructure before looking at the efficacy of a particular intervention, the costs of this basic infrastructure should be included. Similarly, it is important to exclude associated with the research dimension of the intervention (e.g. some components of monitoring and evaluation).

Organisational Level

Some of the HIV/AIDS prevention strategies may have activities at a number of organisational levels. National programmes may operate from the field through districts and regions up to central level administration. As one moves up organisational levels in the hierarchy, away from the point of service delivery, it can become increasingly difficult to obtain data and to tease out relevant costs. Which costs at each organisational level are to be included in the costing study will depend upon the scope of responsibility of the decision maker who is expected to use the results and the way in which the results will be applied.

It is recommended that a full costing beis conducted at the main service delivery level. For higher organisational levels, it is recommended that personnel inputs are costed and any other additional or incremental costs incurred by adding the HIV/AIDS prevention strategy to existing work are included. Where HIV/AIDS prevention is integrated with other work, the incremental cost of its addition may be very small at regional and national levels and taking this approach means that the main inputs to be costed at the higher organisational levels are personnel, vehicle use and supplies. Overheads at those organisational levels, such as buildings, office equipment, utilities and general administration, can generally be excluded. Where, however, new HIV/AIDS prevention strategies require substantial new administrative, logistic or technical support from more central levels, these incremental costs will involve a wider range of inputs and be more substantial. These inputs should be costed and appropriately allocated to the HIV/AIDS strategy. For example, the cost of strengthening the National AIDS Control Programme to support a new preventive strategy should be costed in that strategy.

In order to determine the organisational level, it is important to consider the range of service providers available for a project or programme. Again these will differ by strategy:

HIV blood screening: The more localised the costing study, the more likely that it will be concerned with just one provider. In contrast, a national strategy may include a number of providers. For example, in Zaire the mining industry is responsible for a significant proportion of transfusions but transfusions are also available in the public and private sectors. Which providers to include should be clear from the objectives of the study and, in particular, from consideration of the decision-making audience.

Decisions about the organisational levels from which costs should be recorded are likely to be dictated by the objectives of each particular costing exercise. If the BTS to be costed is localised, for example HIV blood screening at one mission hospital, it may be necessary to cost only one service delivery level.

If the study is looking at national costs it will probably need to consider costs at all organisational levels from the centre to the periphery. Starting from the centre and moving down through the system, cost information for each organisational level should be sought from expenditure records. If such expenditure records are not available, costs may need to be built up from quantity and unit cost information collected during field visits. Fieldwork may need to be restricted to costing a representative sample of facilities at each of the organisational levels.

Mass media: At any one time there may be a number of HIV/AIDS media campaigns being conducted in a country by both government and non-government organisations. More so than in some of the other strategies, there can be an array of providers involved in the strategy, or even in one campaign. For example, a radio drama in Zambia ⁽⁴⁷⁾ entailed the involvement of the Health Education Unit of the Ministry of Health, the Government National AIDS Prevention and Control Programme, the National Radio Corporation, a number of colleges and NGOs, drama groups, and the USAID public health communication support programme, AIDSCOM. An array of implementers and organisations will need to be assessed for cost information. Some activities may be contracted out to commercial enterprises, for example IEC production to professional advertising agencies. This may simplify cost collection as the fees charged are likely to include the agency's own overheads and support costs for the work conducted. In this case the expenditure for the activity in question can be treated as an aggregate input rather than broken down into its constituent inputs.

By definition, a mass media campaign is often implemented at the national level and activities centralised. Costs may, therefore, need to be obtained from only one organisational level, although there may be a number of providers involved at that level. Costs collected will be those directly incurred by a particular campaign plus an appropriate share of the overhead costs of the organisation with primary responsibility for supervising, managing and administering the campaign or strategy. This might, for example, be a share of the running costs of the Health Education Unit of a Ministry of Health or its equivalent.

It is recommended that, apart from the lead agency in the campaign, efforts are not made to cost the overheads of other organisations with partial involvement in the strategy. A share of the personnel costs of those organisations should, however, be included, and allocated according to the share of staff time spent on the mass media campaign.

HIV school education: The main provider of the strategy is likely to be a government ministry, such as the Ministry of Education or Ministry of Health. Curriculum development and material production may be undertaken by them or contracted out to a specialised agency. Similarly, training of staff may be undertaken by the Ministry or by a contracted agency. The actual teaching of HIV/AIDS education is undertaken by staff in the schools included in the strategy. Donors may also provide inputs to the strategy, for example consultancies during the start-up phase. If the project is a more localised NGO initiative, the only providers involved may be the NGO and the schools in which they are implementing the strategy.

Most of the costing data should be available centrally from the organisation with overall responsibility for the strategy, for example, the relevant Ministry. Costs of the time that school staff spend on the strategy during training and teaching may need to be obtained from a sample of schools involved in the strategy. If any donors provide assistance to the project, for example start-up technical assistance, it may be necessary to approach them directly for details of the costs of their inputs. If the project is a more localised NGO initiative, cost data should be collected directly from the NGO and personnel and other costs from the schools in which the strategy is being implemented

Condom social marketing: There are usually a number of institutions involved in a social marketing project. A foreign donor may provide consultancies in the start-up phase and on-going financial assistance. Either the donor or a commercial manufacturer may be the source of condoms. In-country, most activities tend to be co-ordinated by a social marketing organisation. They may contract other organisations for marketing, promotion, and distribution. The Ministry of Health or Family Welfare may be involved in project policy, monitoring and distribution of the condoms through its own outlets. At the periphery, individual retailers will be purchasing and selling the condoms. In most instances the social marketing organisation will have central expenditure records that collate the financial costs of these decentralised activities. To obtain the economic costs of the strategy, it may be necessary to deal with some of the other agencies directly. For example, the social marketing organisation may not have information on the costs of donor inputs.

Due to the social marketing organisation’s role in centralising expenditure data from decentralised project activities, the majority of costing information should be available from that organisation’s office. Field visits may, however, be necessary to obtain extra data, for example, costs of any inputs that do not entail project expenditure. It may be necessary to approach the organisation’s headquarters overseas as well as foreign donors for details of the costs of external inputs, for example condoms and consultancies. Only direct contributions to the project should be included; overheads that are not specific to the country project can be excluded.

Tip-of-the-Trade 3: Gathering data from CSM organisations Key information to obtain from the social marketing organisation includes the total cost of the project (and an idea of the methods used to calculate this figure); the number of condoms sold; project expenditure on condoms; and revenue returning to the project from condom sales. Although social marketing organisations are usually prepared to share information, it may be in their own format which is not ideal for cost analysis. It may, therefore, be necessary to adapt such information as is available. For example, in a costing of a Futures Group project where condoms were part of a broader contraceptive social marketing project, it was found that costs were not tracked by contraceptive method (61). Costing, therefore, entailed taking a share of direct project expenditures allocated in proportion to Couple Years of Protection provided by condoms, plus the cost of the condoms themselves

STD treatment: Any primary level STD treatment service is likely to be provided by a single organisation, be that a government, private for profit or NGO provider. Occasionally, however, private or NGO facilities may receive government support for supervision, training or drug supplies. Some private and NGO facilities without their own laboratory or referral services may rely on government facilities for these.

· Government: If the study is looking at national costs it should consider costs at all levels from the centre to the periphery. Starting at the centre, cost information for levels below should be sought from expenditure records. If such expenditure records are not available, costs will need to be built up from information collected through field visits to the lower levels. A sample of representative primary level facilities should be included in the costing study. The costs of laboratory and referral services should be included in the costing study and these may be incurred at higher organisational levels.

Overhead and support costs for managing and administering the service may need to be collected at a number of levels. In vertical systems, relevant costs at the different levels are likely to be well defined and should be included in their entirety. There may, for example, be specific STD project offices at the national and provincial levels. Where STD services are integrated horizontally with other services, it will be harder to determine the costs that relate to support of the STD services. In this situation, only the incremental costs incurred in supporting the strategy should be included, together with the costs of the time that personnel spend working on the strategy.

· Private for profit and NGO: Some projects may consist of only one specialised STD clinic or one general clinic that provides STD treatment and the costs for these may be available from a visit to one organisational level only. Elsewhere, the primary STD service may be linked to a private/NGO referral hospital and two organisational levels may have to be visited to obtain the necessary costing data.

Sex worker peer education: This type of strategy is normally implemented by NGOs. A lead NGO may co-ordinate support from other NGOs and from national or international donors. Occasionally government institutions may provide some inputs to the project, such as condom supplies, building space or salary support. Donors may also provide condoms, as well as technical assistance. Condoms may also be sold by peer educators rather than just distributed for free.

It may be necessary to deal with only one or two organisational levels to cost this type of project. The main source of data will be the NGO’s management office. The second organisational level, for information that the NGO office is unable to provide, is the community level where peer educators conduct educational activities, although at this level, the organisation may be much more informal. Government contributions may need to be investigated separately, for example at the district or municipal level. It may also be necessary to approach donors for the costs of their inputs.

VCT: Where VCT is provided by a variety of agencies, the objectives of the study will determine whether a representative sample is required, or a study of the service thought to be most replicable, or that thought to be most effective.

Where VCT centres are provided routinely by governments, and if the study is concerned to look at costs nation-wide, costs will have to be considered at all levels from the centre to the periphery. Starting from the centre, cost information for levels below should be sought from expenditure records. If these are not available, costs will need to be built up through visits to lower levels. A sample of VCT facilities should be selected for detailed costing. The incremental costs of support given to the VCT clinics by higher levels should be included.

If a VCT service provided by an NGO is being costed, there may be only one level of interest, or at most the level of the service plus the incremental costs of support provided from a head office. Since VCT services may be funded from a number of different sources, including NGOs, government subsidies, and external donors, a variety of contacts may need to be made with funders at different levels to obtain comprehensive cost data.

IDU: An IDU project is typically implemented by NGOs. A lead NGO may co-ordinate support from other NGOs and from national or international donors. Occasionally government institutions may provide some inputs to the project, such as condom supplies, building space or salary support. Donors may also provide condoms, as well as technical assistance. Condoms may also be sold by peer educators rather than just distributed for free.

It may be necessary to deal with only one or two organisational levels to cost this type of project. The main source of data will be the NGO’s management office. The second organisational level, for information that the NGO office is unable to provide, is the community level where peer educators conduct educational activities, although at this level, the organisation may be much more informal. Government contributions may need to be investigated separately, for example at the district or municipal level. It may also be necessary to approach donors for the costs of their inputs.