Methods: Financing DOTS expansion
Background and objectives
This series of annual reports on global TB control included financial analyses for the first time in 2002.17 In the 2002 report, we presented annual financial requirements and funding gaps in the 22 HBCs for 2002 and for the period 2001-5, based on 5-year plans and costing studies. A full analysis of financial needs and gaps was published as a scientific paper.18 In last year's report,5 we continued to focus on the 22 HBCs. We analysed the funding requirements, funding sources and funding gaps for the calendar year 2003, based on data collected from a standardized form that was sent to all HBCs as part of a new WHO financial monitoring system. We also included an assessment of new funding made available between March 2002 and December 2002, and revised estimates of funding gaps for the planning period 2001--2005, based on data obtained via the new monitoring system and from donor agencies, and on a review of GFATM proposals and World Bank project documents.
This year's report has 4 objectives:
- For HBCs in fiscal year 2003, to quantify total and per patient NTP budgets and TB control costs (i.e. including costs not reflected in NTP budget data), and the funding sources and gaps related to these budgets and costs;
- For HBCs in fiscal year 2002, to quantify total and per patient NTP expenditures and TB control costs, and the funding for these expenditures and costs;
- For HBCs in fiscal years 2004 and 2005, to estimate the total resources required to meet global targets for case detection and cure;
- For other countries in fiscal year 2003, to quantify NTP budgets and funding gaps.
Data collection
We collected data from 4 main sources: NTPs, GFATM proposals, the WHO CHOICE website,19 and costing guidelines developed for the Disease Control Priorities in Developing Countries project (DCPP).20 Data provided directly by countries were collected by means of a 1-page questionnaire included in the annual WHO data collection form (http://www.who.int/gtb/publications/globrep). NTP managers were asked to complete 2 tables. The first table concerned the NTP budget for fiscal year 2003 in US$, and the funding and funding gaps related to this budget. The second concerned NTP expenditures in US$, and the source of funds for these expenditures, for fiscal year 2002. The form also requested information about dedicated TB control infrastructure and the way in which general health infrastructure is used for TB control -- for example, the number of dedicated TB beds that exist, the number of visits that patients need to make to a general health facility during treatment, and the average number of days for which patients are hospitalized. We also asked for an estimate of the number of patients that would be treated in 2003. Compared to data collection in 2002, we requested data on fiscal rather than calendar years because it is the fiscal year for which budget and expenditure data are routinely recorded in countries' financial systems. We searched the GFATM website for each HBC and downloaded any proposals that were identified for TB control. We used the WHO CHOICE website to identify the average costs, in international dollars (I$), of a hospital bed day and an outpatient visit to a clinic in each HBC. The costing guidelines for the DCPP were used to identify the purchasing power parity exchange rates required for conversion of I$ costs to costs in US$ (for consistency with budget and expenditure data reported on the surveillance form).
Data analysis: high-burden countries
For each country, we used the data provided on the WHO data collection form to complete the budget and expenditure sections of two sets of standardized tables. One set covered NTP expenditures, costs not covered in NTP expenditure data, and total TB control costs in 2002; the other covering NTP budgets, costs not covered in NTP budget data, and total TB control costs in 2003 (see Budgets and expenditures in country profiles, Annex 1). For NTP budgets and expenditures, these tables were designed to show totals, and to give a breakdown by line item, as well as to show funding sources and funding gaps. Both sub-sections of these tables replicated the format in which data were requested on the surveillance form. When data were incomplete or included some apparent inconsistencies (e.g. as compared with data that were included in GFATM proposals), we followed up with WHO and NTP staff in the relevant countries and regions, and made the necessary revisions. We did not adjust data reported for 2002 to 2003 prices because it was not clear what exchange rates had been used for conversion of local costs to US$, and any adjustment would make only a small difference to the values reported.
Costs not reflected in NTP budget and expenditure data were defined as: (a) days spent in hospital during treatment, and (b) outpatient visits to health facilities for DOT and monitoring. These costs were estimated in four steps. First, we converted the international $ prices of bed days and clinic visits reported on the WHO CHOICE website into US$ prices using the exchange rates provided in the DCPP costing guidelines. Second, we multiplied the average number of hospital days and visits required per patient (estimated on the WHO surveillance form) by the average cost in US$ for a bed-day and a clinic visit, to give the total cost per patient treated. Third, we multiplied the cost per patient treated by the number of patients notified in 2002 to estimate total costs in 2002. Fourth, we estimated total costs for 2003 as the number of patients that NTPs expected to treat in 2003 multiplied by the cost per patient treated.
We used these data to complete the "costs not reflected in NTP budgets" and "costs not reflected in NTP expenditures" sub-sections of the 2 sets of standardized tables described above. Finally, we summed all costs to calculate the total estimated costs of TB control in 2002 and 2003. The total cost per patient was calculated as the total cost divided by the total number of notifications (for 2002) or the total number of patients expected to be treated (for 2003). We then compared the total government contribution to TB control costs with total government health spending to estimate the percentage of total government health expenditures used for TB control. Total government spending on health was estimated by multiplying the year 2000 government health spending per capita in US$ as estimated in the World Health Report 200221 by population size. We also explored the association between GNI per capita and (a) government contributions to total NTP budgets and TB control costs, and (b) the cost per patient treated. Data on GNI were taken directly from World Development Indicators 2003.22
To estimate funds needed in 2004 and 2005, we updated the analysis of resource requirements previously published for the period 2001--5.18 We used the same methods described in this paper and related supplementary material, but revised the analysis to include new information wherever this was available. The main methodological points are:
1. The number of cases to be treated in 2004 and 2005 was estimated by assuming that the global targets for case detection and cure will be reached in 2005, and that there is constant progress towards these targets from 2002 (the most recent year for which notification are available; in the original analysis, the number of cases to be treated was projected from 1999 notification data).
2. Three categories of resource requirements were considered: those required by NTPs, those required within the general health services to support treatment of patients (e.g. the staff and infrastructure needed for inpatient care and outpatient visits for DOT and monitoring), and those required to operate dedicated TB hospitals (important only in the Russian Federation). The resources required by NTPs were generally estimated by multiplying the number of patients to be treated by either the NTP expenditures per patient in 2002 or the NTP budget per patient for 2003, whichever was higher. This implicitly assumes that the cost per patient treated remains constant as the number of patients treated expands in 2004 and 2005. Based on 2002 and 2003 data, this appears to be a realistic assumption for India and the Philippines. However, since comparison of 2002 and 2003 data suggest an increase in the cost per patient treated between 2002 and 2003 for most other countries, this assumption may lead to under-estimates of resource requirements in other countries. The one country for which this method was not applied was the Russian Federation. Here, the budget requirements included in a recently developed 5-year plan were used. Resource requirements for general health services were estimated by multiplying the cost per patient treated (estimated as explained above for costs beyond NTP expenditures/budgets in 2002 and 2003) by the number of patients to be treated. Requirements for dedicated TB hospitals were estimated by multiplying the cost per bed-day by the number of beds.
3. Sources of funding were defined as governments' regular budgets, loans, grants from the GFATM, and grants from other donors. When up-to-date information on projected funding from governments, loans and donors other than the GFATM, and remaining funding gaps, was not available for 2004 and 2005 (this applied to most countries since the WHO data collection form requested data for 2002 and 2003 only), we assumed that the 2003 level of funding would be in place in 2004 and 2005. For the resources required within the general health services, government regular budgets were assumed to be the only source of funding (as these resources are primarily staff, buildings, and the non-personnel inputs associated with operating facilities e.g. electricity and water, which are typically not funded by other sources). For GFATM funding, we used proposals, available in the public domain, to estimate the funds that would be available in both 2004 and 2005. For example, for a country that had a proposal approved in late 2003, we assumed that the funds requested for year 1 of the proposal would be available in 2004. Crucially, this assumes that funds will become available relatively quickly after proposals have been approved. If this does not happen, then the funds projected to be available from the GFATM will become a funding gap. We then defined the difference between total resource requirements and total projected funding as a "possible gap".
Data analysis: other countries
The data provided by countries other than the HBCs were less complete, and our analyses to date are more superficial. We used the data provided on the WHO surveillance form to calculate the total NTP budget and funding gap for each country submitting data, and summed the totals for each WHO region. We also assessed GFATM funding for both HBCs and other countries, as announced after the first 3 rounds of funding.
17 WHO. Global Tuberculosis Control: Surveillance, Planning, Financing. WHO Report 2002. WHO, 2002. WHO/CDS/TB/2002.295.
18 Floyd K, Blanc L, Raviglione M, Lee JW. Resources required for global tuberculosis control. Science 2002; 295: 2040 2041.
19 www3.who.int/whosis/cea/prices/unit
20 Disease Control Priorities Project. Guidelines for authors (unpublished), pp 71–74. Washington DC, World Bank/NIH, 2003.
21 WHO. The World Health Report 2002: Reducing Risks, Promoting Healthy Life. Geneva, WHO 2002.
22 World Bank. World Development Indicators. Washington DC, World Bank, 2003.