The potential demand for an AIDS vaccine in Thailand

Health Policy 57 (2001) 111– 139 www.elsevier.com/locate/healthpol

The potential demand for an AIDS vaccine in Thailand Viroj Tangcharoensathien a,*, Wiput Phoolcharoen a , Siriwan Pitayarangsarit a, Sukhontha Kongsin b, Vijj Kasemsup a, Sripen Tantivess a, Chutima Suraratdecha c a Senior Research Scholar Program in Health Economics and Financing, Health Systems Research Institute, 5 th floor Mental Health Dept Bldg., Nonthaburi 11000, Thailand b Faculty of Public Health, Mahidol Uni6ersity, Bangkok 10400, Thailand c School of Economics, Sukhothai Thammathirat Open Uni6ersity, Nonthaburi 11120, Thailand

Received 28 September 2000; accepted 5 January 2001

Abstract The recent ongoing phase III clinical trial of a preventive vaccine in Thailand has prompted studies on potential demand for the vaccine among public, employers and households. This study aims to demonstrate the impact of HIV/AIDS, estimate the AIDS vaccine budget required and design the vaccination strategies for different population groups. The analysis is based on available secondary data and several assumptions on levels of secondary infections among various risk groups. Among 15 groups, we identified eight groups as potential vaccinees: Direct CSW, IDU in treatment, IDU out of treatment, male STD, transport workers, CSW indirect, conscripts and prisoners. The vaccine budget, excluding other operating expenditure, was estimated based on a single dose regimen ranging from 100 Baht (US$3) to 1000 Baht (US$29) per dose. A total of US$1.8– 17.7 million is required for non-infected catch-up population and US$0.2– 1.9 million for the maintenance population in the subsequent year. We foresee a relative inefficient and inequitable consumption of AIDS vaccine, which requires proper policy analysis and government interventions. Before vaccine adoption, strong preventive measures must be in place. AIDS vaccine could play an additional, not a substituting, role. A thorough understanding, a wide consultation with stakeholders and public debates are crucial steps for sound policy formulation. © 2001 Elsevier Science Ireland Ltd. All rights reserved.

* Corresponding author: Tel.: + 66-2-951128693; fax: +66-2-9511295. E-mail address: [email protected] (V. Tangcharoensathien). 0168-8510/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S0168-8510(01)00119-1

112

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

Keywords: AIDS vaccine; Cost and benefit; Potential demand

1. Introduction The AIDS epidemic in Thailand had tremendous negative impacts on the households, demographic structure as well as economic and health sectors. The government, non-governmental sector and general public had responded to the epidemic through major public health preventive programmes, notably the use of condoms, fostering education aiming at safer sex and behavioural changes, blood screening and recently the use of antiretroviral drugs (ARV) for maternal transmission prevention. The initial AIDS control approach using ‘‘medical model’’ has gradually transformed to a comprehensive ‘‘civil empowerment’’. The government also invested in the treatments of opportunistic infections (OI) among AIDS patients including other community palliative care. The recent ongoing phase III clinical trials of a preventive vaccine in the USA among gay men and in Thailand among intravenous drug users (IDU) will produce a result by Fall 2002. Economic perspectives on AIDS vaccine, therefore, should be investigated thoroughly for vaccine policy adoption, for example: financial requirement, who will pay for the vaccine — public and private; which groups should be the priority and the most cost-effective for vaccination; at what level the public and private household are willing to pay for the vaccine. Most importantly, before the adoption of AIDS vaccine, thorough studies of cost and outcome comparing vaccine with existing preventive strategies are needed for policy adoption of vaccine. Although Thailand1 has maintained her efforts on primary prevention with a reputable record, the Thai Government may need to see how a new AIDS vaccine could play an additional role. Thailand has a higher ability to pay compared to other developing countries. The current economic crisis, however, placed a hard burden on both public and private finance. Therefore, all aspects must be considered seriously before an adoption of an AIDS vaccine. Since Thailand is currently the only developing country with an ongoing Phase III AIDS vaccine efficacy trial, the study on the potential market of AIDS vaccine and what specific group is the most-cost effective candidate for vaccinations is a crucial input for policy analyses. Consequently, the specific objectives of this study are: 1. To demonstrate the costs and impact of HIV/AIDS in terms of disease and economic burden to individuals and society. This is a proxy for the benefits of current prevention measures, i.e. information, education and communication on behaviour changes, condoms, vertical transmission, blood screening, and of a new alternative AIDS vaccine. 1 A middle-income developing country with a 1999 GNP per capita of US$2,045 down from 2740 in 1997 [1].

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

113

2. To estimate the AIDS vaccine budget required and the vaccination strategies for different population groups who potentially might be vaccinated, as well as prioritize based on their relative risk behavior (a proxy for the potential benefits of vaccination) and ease of access (a proxy for the cost of vaccination). This demonstrates the potential role of vaccination strategies. The remainder of the paper has four sections. First, an impact of AIDS on health and development is reviewed and investigated. Then the vaccination strategies are explored. The key factors relating the ability and willingness to pay for an AIDS vaccine are outlined in Section 4. Finally, conclusions are drawn, and policy implications are provided. 2. Health and development impact of AIDS AIDS is not only a health concern, but also a major development problem. It has had a serious impact on life expectancy achievement in the past two decades in developing countries especially several sub-Saharan African countries. This section contributes to the assessment of health and development impacts of HIV/AIDS in Thailand whereby the morbidity and mortality affected the prime age group especially men. The devastating effects of AIDS on health, mortality, life expectancy, economic and development are, therefore, demonstrated to convince policy makers to further commit efforts and resources on primary prevention including the potential role of AIDS vaccine.

2.1. Impacts on health, mortality and life expectancy at birth Table 1 provides HIV infection rates by risk groups from the latest 16th round of HIV sentinel in June 1998 [2]. Intravenous drug users (IDU) had the highest infection rate of 47.46%, followed by brothel-based commercial sex workers (direct CSW) at 21.05%. Fishermen and male STD had an alarming rate of 10.61 and 8.5%, respectively. The infection rate among non-brothel-based sex workers (indirect CSW) was one third of that of direct CSW. Pregnant women and blood donors were proxies of general population and had infection rates of 1.53 and 0.39%, respectively. There was a notable regional difference in infection rates among direct CSW, indirect CSW and antenatal care (ANC). The Northern region was severely affected. The cumulative number of HIV infections in Thailand was estimated at more than one million by 2000 [3]. Trend analysis of HIV infection rate among selected population from the first to 16th round of sero-surveillance found a steady increasing trend from 1989 to 1995, and then, a promising reverse trend was observed among four groups, namely direct and indirect CSW, ANC and blood donors. The achievement of infection rate among IDU, 30% in December 1994 showed an upward trend towards 47.5% in June 1998. The infection rate among male STD was more or less stabilized for some years (Figs. 1– 3). The campaigns on condom use among high-risk groups and on safer sex practices are important comprehensive control measures explaining a decreasing trend.

114

Region

Direct CSW

Indirect CSW

IDU

Male STD

Donor

ANC

Fishermen

Male CSW

Thailand North Central Northeast South

21.05 28.68 21.05 15.79 17.45

6.67 6.86 7.56 6.74 6.43

47.46 30 46.67 20.24 50

8.5 10.87 8.33 5.72 11.67

0.39 0.37 0.42 0.25 0.43

1.53 2.03 2.06 1.24 1.62

10.61

15.2 11.84 15.20

10.61

21.39

a

Source: The 16th round sero-surveillance, June 1998. Division of Epidemiology. MOPH.

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

Table 1 HIV prevalence median percent, by risk groups and regions, Thailanda

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

115

Fig. 1. HIV infection rate among direct, indirect CSW and male STD, June 1989 to 1998.

During the period of 1984 to August 1999, there were 120000 reported AIDS cases (including those who died) [4]. AIDS morbidity and mortality affect men and women at their prime age 25– 29, which accounts for 28.5% of total cumulative AIDS cases from 1984 to August 1999. The most economic productive prime age group of 20– 39 years consists of 78% of total AIDS cases (Fig. 4). One small peak was observed among babies under 5 years old; this group accounts for 10% of total female AIDS and 3% of total male AIDS cases. Not only affected prime age groups, AIDS was predominant among general workers (43%) and agriculturist (21%) that are among the poorer bands of the society. This could have grave economic implications among these poorer households. Among risk groups of AIDS, male heterosexuals constituted 63% of the total, followed by 17% female heterosexuals, whereby IDU and vertical transmission equally shared 5% of total cases (Fig. 5). The World Bank’s estimate of the impact of AIDS on life expectancy at birth in six countries in 1997 (Fig. 6) showed that life expectancy at birth in Zimbabwe and

Fig. 2. HIV infection rate among IDU, June 1989 to 1998.

116

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

Fig. 3. HIV infection rate among ANC and donated blood, June 1989 to 1998.

Cote d’Ivoire was 22 and 11 years shorter. The impact on Brazil, Haiti and Guyana was 5 years shorter. If there were no AIDS epidemic, Thai people would have gained two more years. Since it took two decades (1976–1996) for Thailand to achieve some 12 more years for both sexes (from 58 and 63.8 for male and female in 1976 to 69.9 and 74.9, respectively, in 1996), then a 2 year loss is significant [5]. Deaths due to AIDS were grossly under-reported: only deaths in hospitals whereby causes were verified by professionals and reported to Epidemiology Division. The cause of deaths at home was verified by a lay-person to the Central Registry. Using a demographic projection on AIDS death by Wongboonsin et al. [3] and Im-em [6], we calculated a magnitude of 11.9– 13.6 times under-reporting of AIDS deaths. A verification of death certificate in Nan, one of the northern provinces hit hard by AIDS, was carried out by a comprehensive verbal autopsy for all deaths during July 1997 to June 1998 in every district [7,8]. There were 10.8 times under-reporting by death certificates. If we adjusted AIDS mortality by a factor of 10 to the number of reported deaths to Epidemiology Division, alarmingly, AIDS became the first leading cause of death in Thailand since 1995,

Fig. 4. Percentage AIDS cases by sex and age group, 1984 – 1999 (Aug).

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

117

Fig. 5. AIDS case distribution by risk factors, 1984 – Aug 1999.

surpassing the big three killers, namely heart diseases, accidents and injuries, and cancers (Fig. 7). Unfortunately, macro-level information on AIDS morbidity and estimation in terms of DALY was not available.

2.2. De6elopment impact According to Wongboonsin et al. [3], 550000 cumulative AIDS deaths were estimated by 2000, of which 78% of total deaths were in the prime age group of 20– 39. This 429000 prime adult premature death toll not only posed a tremendous economic loss to the country, but also caused a great amount of direct and indirect suffering and both tangible and intangible impacts to family members and dependants.

Fig. 6. Impact of AIDS on life expectancy. Source: World Bank, 1997

118

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

Fig. 7. Leading causes of death.

2.2.1. Access to AIDS care Care2 for AIDS, for fear of unlimited high cost and long-term financial burden, was excluded from the benefit package in most insurance schemes, notably social health [9] and Social Welfare for the poor and voluntary Health Card Scheme. Access to AIDS care by those who are and are not covered by an insurance scheme was theoretically subsidized by the MOPH earmarked budget for HIV/AIDS, but in reality, the budget was very limited compared to potential demand. There is no official policy statement precluding Civil Servant Medical Benefit Scheme (CSMBS) beneficiaries from access to ARV. CSMBS beneficiaries, therefore, enjoyed access to ARV [10]. This posed an equity problem on equal access to care by other groups. CDC Department resources are limited. The budget was earmarked for 2500 vertical transmission prevention in fiscal year 1998; this could only accommodate 14% of potential demand estimated at 18000 infections in pregnancies [11]. The Thai Red Cross Society donation campaign could hardly meet the potential demand especially after the economic crisis in mid-1997. The AIDS budget allocation to hospital is inadequate compared to demand for care. For example, Phayao provincial hospital received an allocation of 3.2 million Baht for OI drugs in 1997, but the total consumption of opportunistic infection (OI) drugs was 6.9 million Baht. The deficit was absorbed through hospital non-budgetary revenue and other budget lines, especially free care for the low-income scheme. The cost of OI treatment (excluding the use of ARV) for AIDS was estimated at US$800 – 1500 (average US$1150) per person per annum. There was a potential need of 920 million Baht (calculated at 40 Baht per US$1) for OI treatment in 1998, but only 166 million Baht budget were available. Resources could meet 18% of the potential demand [11]. As a result of resource constraints, especially during the period of economic downturn, several responses by hospitals especially in the northern province were 2 Interpreted as the use of all ARV. The treatment of opportunistic infections was included in the package.

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

119

adopted: capping of OI drug expenditures, supportive and palliative replacing definitive treatment among selected cases. District hospitals referred patients to upper levels of care. The provincial AIDS team developed guidelines on case selection for OI treatments. Hospital counsellors provided death counselling — to be well prepared for death among terminal cases and advocate alternative medicine, e.g. herbals and meditation [11]. In summary, it is very frustrating among health personnel confronted with ethical dilemmas and resource constraints for AIDS treatment. This evidence strongly supports further efforts for primary prevention strategies including the future potential role of vaccines.

2.2.2. Direct and indirect costs per AIDS case Vongsaroj [12] estimated an overall cost of 128 million Baht for admission and 920 million Baht for outpatients for a whole province of Phayao — a very high HIV prevalence province; this does not match with the 23 million Baht AIDS budget. Income foregone from premature death was estimated at 573 million Baht at 1998 prices. The overall economic cost was 1621 million Baht (US$45 million at a 36 Baht exchange rate); this accounts for 9.02% of gross provincial product. This loss excludes intangible costs and other consequences on the demographic structure, household composition and allocation of labours, caregivers, orphans, stigmatization, grief and sufferings. In a Chiangrai district hospital, AIDS death contributed to 34–60% of the total number of hospital deaths in 1995–1998. It took 4–7% of total inpatients but 10– 14% of total bed days due to longer stays. Increasingly, AIDS placed severe constraints on the limited resources, from 20% in 1995 to 40% in 1997 and 27% of total expenditure in 1998 [13], mainly on inpatient care. This micro-level analysis provided insights on how AIDS consumed vast amounts of scarce health resources. 2.2.3. Public and pri6ate spending on HIV/AIDS care The government invested in HIV/AIDS prevention and control, during the first 6 initial years, with a massive increase [14]. A real term reduction was observed even pre-economic crisis in 1993–1994 and especially after 1996–1997. It must be noted that in 1999, two other loans (Social Investment Programme, 49 million Baht and Miyazawa loan 31 million Baht) of 80 million Baht retarded the reduction rate from 32 to 1%. The loan under Social Investment Programme was earmarked for community empowerment on AIDS prevention, and Miyazawa was for psychosocial services and curative care for AIDS patients. A 20% real-term reduction on MOPH AIDS budget between 1997 and 1998 was observed. The use of ARV and treatments for OI consuming 55% of total were protected (fewer cuts than average). ARV for vertical transmission prevention was cut by 78% but possibly replaced by public donations. More than half of the MOPH AIDS budget was for curative (ARV and OI), and the other 45% was for prevention (vertical transmission, breast-milk replacement, blood screening, universal precaution, NGO subsidy, laboratory test and condom distributions). The analysis indicates that non-MOPH budgets were spent on programme activities

120

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

related to prevention, community empowerment, IE&C and other campaign. Then, more than 70% of total national AIDS budget was aimed for prevention and disease control. It is difficult to estimate expenditure per capita since the magnitude of cross-subsidy using public hospital non-budgetary source of finance was not available. Information on private spending by AIDS patients was not available. Several estimations were made. The estimate by Khongsin [15] was based on the choice of care sought between public (hospital for inpatient and outpatients) and private (hospital, clinics, home care, pharmacies, traditional healers and community care), and unit costs at various levels of care. The total 1315 million Baht was estimated at 1997 prices in Chiangrai province, of which 43% was shouldered by private households. Vongsaroj [11] estimated private spending for AIDS inpatient care at 65% of the total expenditure in Phayao.

2.3. Summary remark We have demonstrated that AIDS had a serious negative impact on health and life expectancy. After adjustment, AIDS mortality could surpass the previously big three killers and rank the highest since 1995. Premature deaths among prime-age males have posed massive economic losses as well as demand for care and placed a strain on the limited health-care resources. Intangible costs were uncountable, especially for the poorest rural households experiencing AIDS deaths, and also placed an additional burden on the limited health-care resources. Evidence supports the calls for policy makers to further commit to bring down the AIDS burden through cost-effective, affordable and sustainable preventive measures.

3. Vaccination strategies for HIV/AIDS preventive vaccine Among several key questions is: which is the most cost-effective population group to be vaccinated, given limited resources on vaccine and its delivery as well as the impact of vaccination on overall epidemic? Policy-makers should note that neither the efficacy nor the price of vaccine is known at the moment, but we can discuss vaccination strategy based on a hypothetical vaccine cost, relative costs of delivery and relative benefits in terms of preventing primary and secondary infections among different groups of population. When the cost of delivery is weighed against the potential benefit, the most cost-effective target for vaccination is that with the highest benefit and lowest cost. Vaccine is a new alternative among current preventive measures such as the use of condoms and changes in sexual behaviour. The National AIDS Committee’s concern is the false sense of security among vaccinees, low frequency of condom use, and unsafe sex practices. This phenomenon is counter-productive to epidemic control. Furthermore, since we cannot anticipate that either all high-risk will lie up for a vaccine or a vaccine will be fully effective, we then put two scenarios in our analysis, vaccine strategy whereby condom use could and could not be sustained.

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

121

The main objective of this section is to answer two questions: the cost of vaccination (vaccine and delivery cost) and the potential benefit of vaccination strategy (not vaccine efficacy and effectiveness per se). Whereas it is relatively expensive to reach the informal sex workers, child prostitutes and IDU who are outside treatments as they are unlawful and at risk of prosecutions, it is relatively cheaper to access vocational schoolers and conscripts. The cost of delivery for childhood vaccines (lower cost of separate DTP and HB) is significant, 38% of total cost. When the higher cost of combined DTP+ HB vaccine was administered, the cost of vaccine delivery decreased to 25% [16]. The role of delivery cost was reduced when the vaccine cost increased, but it still played a role. In our conceptual thinking, the cost of vaccine delivery is weighed against potential benefit in terms of total preventable cases. Given that the vaccine efficacy is acceptable by the Thai Food and Drug Authority, the vaccination outcome depends on the potential number of total (primary and secondary) infections averted, the higher infection risk — the more benefits gained from vaccination, and, vice versa, the higher condom use rate and sustainable sexual behavior changes, the lower the potential benefits gained from the AIDS vaccine.

3.1. Conceptual framework A four-quadrant figure was constructed as a result of relative low and high cost of vaccination and potential benefits (Fig. 8). Sub-population groups categorized in Quadrant II (low cost, high benefits) are the most cost-effective for vaccine strategies, given limited resources. In contrast, Quadrant IV — high cost but low benefits — is not worth the investment. Quadrants I (low cost, low benefits) and III (high cost, high benefits) are subsequent priorities for policy consideration when a vaccination of target population in Quadrant II has been achieved. The reference line on cost of vaccine delivery can be moved to the right if the economic situation

Fig. 8. Conceptual framework.

122

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

is favourable and vice versa for more limited resources. The reference line on the potential benefit could be shifted upward if a stringent policy is adopted to gain the maximum benefit on epidemic control or shifted downward if resources are more favourable. 3.2. Ranking of risk groups In Table 2, specific population groups were identified on its catch-up and maintenance size quoted by various published sources, authors’ estimations and the latest 1998 HIV prevalence. The prevalence of a similar group was used when specific data were not available. Note that while we are awaiting results from vaccine trial, our analysis does not take into account a vaccine efficacy. The level of effectiveness will only affect the size of the total infections of the target population proportionally. As a result, the number of lifetime secondary infections, i.e. the number of infections averted, was estimated based on the level of sexual contacts and other risk activities (sharing needles) among different groups. The condom use rate was retrieved from various available secondary data, and when data were not available, some assumptions were made (see Tables 3 and 4). The total number of preventable cases is the sum of secondary (Column F) and primary preventable cases (Column D). The number of secondary infections prevented (column F) is the multiplication of prevalence per capita (Column D) and number of lifetime secondary infections (column E). The scenario in Column F does not take into account the use of condoms, thus yielding significantly high benefits. However, it is possible that the first generation of an AIDS vaccine will be only partially effective; the prevailing risk reduction strategies such as condom use must still play an important role to avoid the abandonment of risk-reduction attempts. When the condom use rate is taken into account, the number of lifetime secondary infections in Column H is lower than that in Column E. The total number of infections prevented when condom use rate was sustained at current level is calculated in Column I. Detailed values of basic parameters on secondary infections and condom use rate from various sources and assumptions made by researchers are listed in Table 3. Regarding the costs, our estimation excludes costs of vaccine delivery that could be associated with the routine EPI infrastructure. Furthermore, vaccine budgets will increase proportionally when subsequent booster doses are required. Since it is unknown how many boosters are required, only one dose was estimated. Table 4 provides a relative cost of vaccine delivery ranging from the lowest (0) to highest (1). Those who are relatively easy to access, e.g. conscripts in military barracks and prisoners, are less costly than MSM and IDU out of treatment. Fig. 9 plots column F (Table 2) on the Y axis [in a log scale] against column J on the X axis [in normal scale] for a no-condom-use scenario. Fig. 10 depicts a scattergram with the Y-axis for Column I, and the X-axis for Column J was constructed when the current condom use rate was taken into account. In both figures, two reference lines were drawn, 0.5 person prevented on the Y axis, and 0.5 relative cost of vaccine delivery on the X axis. This provides four quadrants with similar results in the two figures.

Table 2 Size, prevalence, number of life time secondary case prevented and relative cost of vaccination among 15 population groups

2.

Catch-up sizea

Maintenance size per annumb

Prevalence June98 (per capita)

Life time secondary infections without condom

A Sex workers directc

B 57 600

C 1440 [assume 2.5% of catch-up]

D 0.2105

E F = (E*D)+D Regular 92.0306 customer 80%*1.7/night* 20d*12m*4yr/ 12= 108.8 Casual customer 20%*1.7/night *20d*12m *4yr =326.4 Husband =1 Total 436.2 Regular 22.1444 customer 25%*0.6 /night*10d*12m* 6yr/18 repeated= 6 ’Casual customers: 75% *0.6*10d*12 m.*6yr =324 Husband =1 Total 331

Sex workers indirectd

76 800

2304 [assume 3% of catch-up]

0.0667

Total infection Condom use prevented rate without condom

G With customer 92%

Life time secondary infections when condom was used

Total infection Relative prevented with vaccination condom cost (0 to 1)

H (109+326) *0.08 +(1*.885) =total 35.7

I =(H*D)+D 7.7256

J 0.45

(324+6)*.08 +(1*.885) =27.3

1.8866

0.7

With husband 11.5%

with customer 92%

With husband 11.5%

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

1.

Group

123

124

Table 2 (Continued) Catch-up sizea

Maintenance size per annumb

3.

Male STD patients

0.305 per 1000 pop. = 18 300

11 895 [assume 0.0850 65% of catch-up]

4.

Transport workers

507 000e

2535 [assume 0.5% of catch-up]

Prevalence June98 (per capita)

0.0850

Life time secondary infections without condom Similar to transport workers Total 54 To CSW: 1/m*12m.*4yr =48

Total infection prevented without condom

Condom use rate

Life time secondary infections when condom was used

Total infection Relative prevented with vaccination condom cost (0 to 1)

4.6750

Similar to conscript

27.7

2.4359

0.4

4.6750

With CSW 52.4%

(48*0.476) +(5*0.785) +(1*.885) =total 27.7

2.4359

0.65

(48*0.476) +(5*0.785) +(1*.885) =total 27.7

0.6305

0.1

2.8476

0.2

To other female 5 To spouse 1

5.

Conscripts

50 000 peryr. *2 years, 100 000

50 000 [new annual recruits]

0.0220

Total 54 To CSW: 1/m*12 m.*4yr = 48

With other female 21.5% With spouse 11.5% 1.2100

To other female 5 To spouse 1

6.

IDU in treatment

38,860f

972 [2.5% of catch-up]

0.4746

Total 54 50% share 2.8476 needle among in a group of 10= 5 Total 5

With CSW 52.4%

With other female 21.5% With spouse 11.5% Not applicable 5

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

Group

Table 2 (Continued) Catch-up sizea

Maintenance size per annumb

Prevalence June98 (per capita)

Life time secondary infections without condom

7.

MSM

3750 [0.5% of catch-up]

0.0850

Total 6

8.

IDU out of treatment Civil servants

2.5% of male pop. 750 000 38 860g

3886 [10% of catch-up] 5600 [0.5% growth rate set by government]

0.4746 0.0153

9.

1 120 000h

Total infection prevented without condom

Condom use rate

Life time secondary infections when condom was used

Total infection prevented with condom

Relative vaccination cost (0 to 1)

0.5950

57.1%

2.6

0.3038

0.75

Total 5

2.8476

Na

5

2.8476

0.85

Set norm as general population: To CSW 1, Other 1, Spouse 1 Total 3

0.0612

With CSW 89%

(0.11*1) +(.216*1) +(.885*1) =1.2

0.0338

0.2

Same as civil servants

0.0800

1.2

0.0442

0.15

With other 78.4% With spouse 11.5% Same as Civil servant

10.

Policemen +military personnel

230 000 +670 000 =900 000i

1150+3350 = 4500 [0.5% annual growth rate]

0.0200

11.

Prisoners

130 997i

0.1700

Total 3 Total 2

0.5100

0%

2

0.5100

0.1

12.

University students

1 033 325j

6550 [assume 5% of catch-up] 258 000 [one fourth of catch-up]

0.0010

To CSW 1

0.0030

With CSW 89%

0.11+0.216 =0.33

0.0013

0.2

Total 2

With other 78.4%

125

Other 1

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

Group

126

Table 2 (Continued) Catch-up sizea

Maintenance size per annumb

Prevalence June98 (per capita)

Life time secondary infections without condom

Pregnant women

1 000 000

850 000 [actual figure]

0.0153

To spouse 1

To 0.3 child, through vertical transmission Total 1.3 Spouse 1

14.

Health workers

180 000i

900 [0.5% growth rate]

0.0153

15.

High-school Students

926 288k

308 763 [one-third of catch-up]

0.0010

Total 1 To CSW 0.025 To other female.047

Total infection prevented without condom

Condom use rate

Life time secondary infections when condom was used

Total infection prevented with condom

Relative vaccination cost (0 to 1)

0.0352

6.29%

1.3*.937 = 1.2

0.0339

0.35

0.0306

With spouse 95%

1*0.05=0.05

0.0161

0.1

0.0011

With CSW 44%

(.56*.025) +(.75*.047) =0.05

0.0010

0.2

With female 25%

Total 0.072 a

The number of the whole group for mob-up vaccination. New entry into the group in the subsequent year. c Direct CSW refers to a brothel base according to the Thai definition. The official figure for the 1998 commercial sex premise census was 64 000 CSWs throughout the country. Source: The 1998 commercial sex premise census, CDC, Thailand. Assume 60% are direct CSW, then 38 400 blow up of under-reporting by a factor of 1.5 times. d Indirect CSW includes non-brothel based (e.g. massage parlour, cafe´ , cocktail lounge, telephone arrangement and other entertainment premises). We estimated 40% of 64 000 CSW = 25 600. We blow up by a factor of 3 to include those informal arrangements that are not enumerated by the census. e Source: Department of Land Transport, 1997 number of transport vehicles. f Source: Department of Medical Service database on addiction under treatment, Ministry of Public Health, 1999. g Assume half of total IDU are not under treatments. h Source: Civil Service Commission Office, 1999. i Source: Department of Corrections, MOI, 1997. j Source: Ministry of University Affairs, 1999. k Source: Ministry of Education, 1999. b

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

13.

Group

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

127

Table 3 Value of basic parameters on secondary infections and condom-use rate, various sources and assumptions Group Sex workers direct

Sex workers indirect

Male STD patients Transport workers Conscripts

IDU in treatment MSM IDU out of treatment Civil servants Policemen+military personnel Prisoners University students Pregnant women Health workers High-school students

a

Basic parameters and assumptions for calculation of life time secondary infection Chance to provide service to regular customer 80 and 20% to casual customers Same regular customer had 12 visits during the 4 years Number of customers per night 1.7a Working day per year = 20/month*12 months = 240 days Working year per life time assumed at 4 years Assume every CSW gets at least one husband Condom use with customer 92% with husband 11.5%a Chance to provide service to regular customer 25 and 75% to casual customers Same regular customer had 18 visits during the six years Number of customers per night 0.6a Working day per year = 10/month*12 months = 120 days Working year per life time assumed at 6 years Assume every CSW gets at least one husband Condom use with customer 92%, with husband 11.5%a No secondary data; we use information on sexual activities and condom use rate among conscripts as a reference No secondary data; we use information on sexual activities and condom-use rate among conscripts as a reference Chance to visit CSW, once a month for 4 years, though conscript serves 2 years, but we add another 2 years after conscription for disease spreading Assume spreading infection to 5 other females and to one spouse Condom use rate with CSW 52.4%, with casual female partner 21.5% and with spouse 11.5%b Assume 50% in a group of a clan of 10 IDU sharing syringe and or needles Assume a total of six secondary infections per lifetime Condom use rate 57.1%c Assume 50% in a group of a clan of 10 IDU sharing syringe and or needles Assume the following: infect to CSW 1, to other partner 1, to spouse 1 Condom-use rate among CSW 89%, among other female 78.4%, and with spouse 11.5%c Assume similar to civil servants Condom use rate among CSW 89%, among other female 78.4%, and with spouse 11.5%c Assume two secondary cases spread through MSM. Assume condom use rate 0% Assume the following: infect to CSW 1, to other partner 1 Condom use rate among CSW 89%, among other female 78.4%, and still have no spousec Assume spreading infection to husband 1, 0.3 vertical transmission Condom use rate with spouse 6.29%b Assume transmission to spouse 1 Assume condom use rate with spouse 95% Assume transmission to CSW 0.025 and other female 0.047 Condom use rate among CSW 44%, among other female 25%, and still have no spouseb

Source: Epidemiology division, 1999. Source: Sex behavior survey, 1998. c Source: Chamratrithirong et al. The success of the 100% condom promotion programme in Thailand: Evaluation of the 100% condom promotion programme and the validation of the decline in trends for selected STDs. Mahidol University, Institute for Population and Social Research, Thailand, 1999. b

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

128

In Fig. 9, there five groups fall into Quadrant II, namely, direct CSW, male STD, IDU in treatment, conscript and prisoners, respectively. Four groups were in Quadrant III namely, indirect CSW, transport workers, IDU out of treatment and MSM. The rest (six groups) was in Quadrant I. No group in this study falls into Quadrant IV. If resource is abundant, policy-makers could shift the X reference line to the right to filter more risk groups in Quadrant III to Quadrant II, e.g. transport workers and indirect CSW. In Fig. 10, if the condom-use rate were sustained at the 1998 level when vaccine was adopted, we would find that the benefits in terms of the number of primary and secondary cases prevented would be much lower than when condoms were totally dis-regarded. However, the scattergram does not change a decision on the priority of risk groups to be vaccinated, except MSM that falls into Quadrant IV. Note that the number of secondary cases prevented played a significant role in determining which groups are categorized into which Quadrant. When resources are limited, the number of secondary preventions rather than the personal protection should be included in the key decision criteria. Thus, future fine-tuned values of these parameters help to identify target groups for vaccination.

3.3. Le6els of spending for risk groups and hypothetical 6accine prices The vaccine strategy will have little impact on AIDS epidemic if it is taken up by low-risk individuals who have an ability to pay but not those with a high risk and are unable to pay. This scenario is likely to appear if the vaccine is available through the market mechanism. By simultaneously identifying key risk groups for subsidized immunization programme, the government, especially in developing countries, can avoid such harm. Table 4 Relative costs of vaccination, sorted by relative ease of accessa Group

Relative vaccination cost (0–1)

Conscripts Prisoners Health workers Policemen+military personnel IDU in treatment Civil servants University students High-school students Pregnant women Male STD patients Sex workers direct Transport workers Sex workers indirect MSM IDU out of treatment

0.1 0.1 0.1 0.15 0.2 0.2 0.2 0.2 0.35 0.4 0.45 0.65 0.7 0.75 0.85

a

Source: Authors’ estimation.

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

129

Fig. 9. Condom is dis-regarded when vaccine is available.

If the government decides to give a higher priority to epidemic control, i.e. outcome oriented on the ground that the cost of vaccination is affordable at a marginal level as it may attach with the existing extensive public health infrastructure, there will be an additional potential for efficiency improvement of delivery system. From Fig. 9, when 0.5 case prevented is the cut-off point, eight groups were filtered as vaccination candidates regardless of the cost of vaccine delivery (Direct CSW, IDU in and out of treatment, male STD, transport workers, indirect CSW, conscripts and prisoners). Table 5 shows the budget requirement for vaccine costs among 15 groups with varying levels of vaccine coverage if the vaccine cost ranges from 100 Baht (US$2.6) to 1000 Baht (US$26.3) per dose. Note that the infected portion was deducted from the catch-up population, as this vaccine will not be given to infected persons. If only eight high-risk groups that are above 0.5 averted cases were vaccinated, the vaccine cost alone required a total budget of 67–672 million Baht for one dose of AIDS vaccine or US$1.8– 17.7 million (at 38 Baht per US$). For the maintenance population in the subsequent year, the total budget requirement for vaccine will be 7.2– 72 million Baht or US$0.2 to 1.9 million. The likely scenario for AIDS vaccine is a mix financing strategy whereby public funds are used to support the high-risk low-ability-to-pay groups and private funds for the rest while maintaining the current low-cost personal protection efforts.

130

Table 5 Budget requirement, million Baht and US$, varying coverage and vaccine cost per dosea Six groups of vaccine candidates

9. 10. 11. 12. 13. 14. 15.

1. 2. 3. 4. 5. 6. 7. 8.

Sex workers direct IDU in treatment IDU out of treatment Male STD patients Transport workers Sex workers indirect Conscripts Prisoners Subtotal 8 groups MSM Police/ military Pregnant women Civil servants Health workers University students High school Students Total Baht Total $US Sex workers direct IDU in treatment IDU out of treatment Male STD patients Transport workers Sex workers indirect Conscripts Prisoners Subtotal 8 groups

0.8 0.8 0.5 0.7 0.75 0.6 1 1

Non-infected catch-up*

45 475 20 417 20 417 16 745 463 905 71 677 97 800 108 728

0.5 1 0.8 1 1 1 1

686 250 882 000 984 700 1 102 864 177 246 1 032 292 925 362

0.8 0.8 0.5 0.7 0.75 0.6 1 1

Maintenance 1440 971.5 3886 11 895 2535 2304 50 000 6 550

Number vaccinated

Hypothetical vaccine cost per dose 1000

500

100

36 380 16 334 10 209 11 721 347 929 43 006 97 800 108 728 672 107 343 125 882 000 787 760 1 102 864 177 246 1 032 292 925 362 5 922 755

36.4 16.3 10.2 11.7 347.9 43.0 97.8 108.7 672 343.1 882.0 787.8 1102.9 177.2 1032.3 925.4 5922.8 160.6

18.2 8.2 5.1 5.9 174.0 21.5 48.9 54.4 336.2 171.6 441.0 393.9 551.4 88.6 516.1 462.7 2961.4 80.3

3.6 1.6 1.0 1.2 34.8 4.3 9.8 10.9 67.2 34.3 88.2 78.8 110.3 17.7 103.2 92.5 592.3 16.1

1152 777 1943 8327 1901 1382 50 000 6550 72 032

1.2 0.8 1.9 8.3 1.9 1.4 50.0 6.6 72.1

0.6 0.4 1.0 4.2 1.0 0.7 25.0 3.3 36.2

0.1 0.1 0.2 0.8 0.2 0.1 5.0 0.7 7 .2

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

1. 2. 3. 4. 5. 6. 7. 8.

Vaccine coverage

Six groups of vaccine candidates

Vaccine coverage

Non-infected catch-up*

Number vaccinated

Hypothetical vaccine cost per dose 1000

9. 10. 11. 12. 13. 14. 15.

a

MSM Police/ military Pregnant women Civil servants Health workers University students High school Students Total Baht Total US dollar

0.5 1 0.8 1 1 1 1

3750 4500 850 000 5600 900 258 000 308 762.7

1875 4500 680 000 5600 900 258 000 308 763 1 331 670 US$

1.9 4.5 680.0 5.6 0.9 258.0 308.8 1 331.7 35.0

Source: Authors’ calculation. Note: non-infected catch-up population is adjusted by (1−prevalence rate)*catch-up population.

500 0.9 2.3 340.0 2.8 0.5 129.0 154.4 665.8 17.5

100 0.2 0.5 68.0 0.6 0.1 25.8 30.9 133.2 3.5

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

Table 5 (Continued)

131

132

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

4. Ability and willingness to pay for an AIDS vaccine Issues on the source of vaccine financing and willingness to pay by the public sector, e.g. MOPH, Bureau of the Budget, the National Economic and Social Development Board and Ministry of Finance as well as the private sector, e.g. employers, firms and households are the founding stone for policy development in AIDS vaccination. This section contributes to the discussion and exploratory work using Hepatitis B Vaccine (HBV) as a proxy to understand public and private perception on vaccine. Issues on whether the vaccine was seen as public or a private goods are discussed.

4.1. Key factors affecting public and pri6ate demand Financing of vaccines in Thailand, especially those in the national immunization programme3 was almost solely footed by the public sector — the MOPH. Table 6 draws public and private roles in financing and provision of vaccine and applies to the country situation. The household played an insignificant role (particularly on adult vaccine and payment for vaccine delivery costs at public and private outlets). Donors and others such as UNICEF contributed only to the National Immuniza-

Fig. 10. Condom is sustained when vaccine is available. 3

Childhood vaccine, e.g. BCG, DTP, OPV, Hepatitis B, MMR, Japanese Encephalitis vaccine.

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

133

Table 6 Vaccine financing and service provision Provision of services

Public financing

Private financing

Public sector

Universal Vaccine Coverage policy: BCG, DTP, OPV, HB, MMR, JE Vaccine budget and handling 724 million Baht in 1999FY, excluding all other non-vaccine recurrent cost incurred by health centre, district and provincial hospitals, as EPI was integrated into an extensive health delivery network, not a vertical programme

Expenses for adult vaccine, e.g. HB, rabies Accessibility cost and some other antipyretics for vaccinee

Private sector

Estimate cost of access to national EPI services 60 million Bahta Vaccine provided to private clinics and Adult vaccine, e.g. chicken pox, hospitals to help provide coverage in influenza urban setting Estimate 7% of 724 million Bahtb in Accessibility cost and others 1999FY Unknown portion but estimated at negligible fraction

a Estimate at 10 Baht transportation cost per visit and 10 Baht for drugs, e.g. Paracetamol*3 visit/yr*1 million vaccinees. b Estimated at 7% of total vaccine services provided by the private sector.

tion Day for Polio eradication. The MOPH plays a pivotal role in financing and provision of vaccine under the National Immunization Programme. Moreover, the 1999 thirty-cluster vaccine coverage survey for the majority of vaccine services in 12 provinces throughout the country showed that 93% of total clients’ services were provided by the public sector [17]. The Thai household hardly spent anything on vaccines compared to other health expenditures. Socio-economic surveys conducted by the National Statistical Office demonstrated that a majority of household health expenditure was spent at private clinics and hospitals mainly on curative service [18]. Since 1990, the proportion of household spending at private institutions was higher than that of the public, reflecting more spending on curative care. The consumption expenditure pattern, however, changed after the recent economic crisis in 1997. Who will finance an AIDS vaccine depends very much on the society’s perception of the AIDS vaccine. If a provision of any vaccine was perceived by general public as the Government’s Constitutional responsibility according to Article 52 of 2541BE Constitution4, it is very likely that any pressure would force the govern4 For the prevention and elimination of serious communicable diseases, the government must provide to its citizens free of charge and timely, according to related laws rules and regulations.

134

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

ment to adopt an AIDS vaccine when it becomes available in the world market. This notion and the historical evolution of vaccination in Thailand shape our analysis toward public demand instead of a private demand for an AIDS vaccine [19]. We, however, could not explore the magnitude of household ability and willingness to pay for vaccine as data on price of vaccine in the market (non-EPI) and private household spending on other adult vaccines are very limited. There is a wide range of interpretation of AIDS care as to whether it is purely a public or private good, or something between. For example, a treatment for OI was provided free or exempted in public outlets (unless the patients can pay); in reality, free ARV prescription is hardly covered by public funds and seems to be a private responsibility. A STD treatment was lay-perceived as a personal responsibility, and patients are willing to receive treatment especially in private outlets since external benefit of treatment was not well understood by the public. In contrast, technocrats advocate, on an external benefit ground, that treatment of STD is the public responsibility, albeit subsidized prices were charged in public settings. Consequently, several parameters that come into the decision-making process for the government adoption of a new AIDS vaccine include: “ Vaccine efficacy and safety. “ Cost of vaccine and its operations. “ Cost and effectiveness comparison of AIDS vaccine and current alternatives (counseling, IE&C, community empowerment, condoms, vertical transmission, blood screening). “ The country economy and disease burdens. “ Social pressure and political dimension of decision process. “ Government and societal opinion on whether an AIDS vaccine is a public or private concern. Historically, the procedure of adoption of new vaccines into the national EPI started firstly with a proven high safety and efficacy (immunogenicity) and disease burdens followed by a small field trial on cost and sero-conversion measurements. The cost of vaccine and its operation could be one of the most crucial determinants for adoption. For example, the MOPH adopted HB vaccine into EPI at no more than the affordable level of US$2 per dose [19]. The decision on HIV antigen, in addition to HIV antibody blood donor screening to detect window period blood in 1992 for all blood units throughout the country, was mainly due to social pressure and political decision. An adoption of National Immunization Day for polio eradication was mainly pushed by international organizations such as WHO and UNICEF as well as an achievement of polio-free Americas. In the meantime between vaccine availability in the market and the government decision, there is definitely a demand by private individuals, which is likely determined by the ability to pay and risk of HIV infections. We believe that an initial private demand will permeate for some time, followed by public demand and adoption for some specific risk groups. During the initial phase, access to vaccine is likely to be rationed by an ability to pay. We foresee a relative inefficient consumption whereby the lower risk group is vaccinated, and there is a potential

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

135

market failure due to agency relationship and imperfect information. Evidence showed that the high-risk group confined in the poorer section of the society will be inadequately covered if vaccine costs are high. These inefficiency and inequity problems require proper policy analysis and government interventions. In summary, based on the pre-requisite of vaccine efficacy and safety, key factors affecting public and private demand are shown in Table 7. Private demand is determined by costs of vaccine, household income, personal risk, cost of alternatives such as the use of condoms and personal perception and experience on AIDS. Public demand is similarly determined by costs of vaccine and its operations, cost effectiveness in comparison with existing alternatives, the country economy, social pressure and political decisions. These are circumscribed by the interpretation of vaccine as public or private goods.

4.2. Lessons from other 6accines As described above, public financing is the major source of vaccine financing. Though evidence is inadequate, in-depth interviews with national EPI managers revealed that private expenditure played a minimum role, for example, on HB and JE vaccines in northern endemic provinces. There was a comprehensive study on cost and effectiveness comparing separate doses of DTP and HB and a combined dose of DTP-HB. The cost per fully immunized child was 264 and 380 Baht for separate and combined regimens, respectively, at 1995 constant prices. Vaccine costs had 61 and 75% share of the two regimens; vaccine operation and capital depreciation were 38 and 25%, respectively [16]. As vaccine costs take the lion’s share of the total costs, purchasing costs and wastage reduction had the highest potential for programme efficiency. However, a combined regimen provided a higher outcome in terms of Hepatitis B sero-conversion, 94.08% compared to 88.77% for separate regimens. The cost-effectiveness values were 297.4 and 403.9 Baht for separate and combined regimens, respectively [16]. Table 7 Factor affecting public and private demand for AIDS vaccine Public

Private

Cost of vaccine and its delivery Cost and effectiveness comparison of AIDS vaccine and current alternatives Country economy Social pressure and political dimension of the decision Government and societal opinion on whether AIDS vaccine is a public or private concern

Cost of vaccine Household income Personal risk of infections Cost of alternatives such as condoms Perception or personal experience on morbidity and mortality with AIDS

136

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

4.3. E6idence of willingness to pay for HIV/AIDS pre6ention and treatment Evidence on public and private spending level on HIV prevention and AIDS treatment could be used as a proxy for willingness to pay for aversion of HIV/AIDS. Spending on condoms reflects a willingness to pay for primary HIV prevention. Condom purchase by the CDC Department showed a slightly increasing trend before the crisis in 1997 and major reduction of 58% in nominal terms in 1996 – 1997. Monopsonistic purchaser, CDC Department, got a very competitive price of average 1.28 Baht per piece during 1994–1998 [20]. Data are inadequate on the proportion of condom distribution in private market going for HIV prevention versus family planning purposes. During 1996– 1998, public condom consumption purchased by the CDC department was 134.5 million Baht in nominal term. Household consumption was 35 times higher than CDC distribution channels, 4694 million Baht in nominal terms. We estimated that 20 million male adults aged 15–60 years old would use 57 condoms per year or 1.1 per week; condom consumption was 78 Baht per adult per year or 1.5 Baht per week. We were confident when our estimate was cross-validated by the wholesale price per piece we calculated. Whatever the purpose of using a condom is, it could serve as a prevention measure against HIV infections. Condom expenditure, therefore, is a proxy for the willingness to pay for HIV prevention. Although the budget for free condom distribution by CDC was severely cut in 1998, the condom use rate showed a sustainable increasing trend, from 93.8 and 90.9% among brothel-based and non-brothel-based commercial sex workers in 1993 to 97.5 and 94.1%, respectively in 1998 [2]. This is the proxy evidence on the willingness to pay for AIDS prevention among customers of CSW. Regarding the public and private resources contributed to the prevention of vertical transmission, it was evident that public sources dominated private, NGO and donations (95 versus 5%). There was a significant increase in public resources, 23 and 165% during 1997– 1998 and 1998– 1999, respectively [21]. This reflects the government’s perception on the importance of this investment. There are also resources from the MOPH Department of Health for breast-milk substitution for newborns of HIV infected mothers. Data from the MOPH-FDA during 1994–1998 showed that the total national expenditure on opportunistic-infection (OI) drugs dominated total spending, i.e. 82– 71% in 1994– 1996 whereby anti-retroviral (ARV) drugs spending caught up to 48% of the total in 1998. OI and ARV increased steadily from 537 in 1994 to 1546 million Baht in 1998. A sharp increase during the economic boom (1994– 1997) was followed by a 13% drop in 1998 [22]. In summary, there is a significant amount of public concern over the AIDS problem and government investment on primary prevention and care. The public adoption of AIDS vaccine is determined by vaccine efficacy, cost and economic environment, risk and potential benefit comparing existing strategies. However, social pressure and political dimensions may supersede these factors. Imperfect information and market failure may lead to an inefficient private consumption and

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

137

may have little epidemic impact, which requires a proper policy analysis and a participatory process among major stakeholders.

5. Conclusion We demonstrated that AIDS has had devastating long-term effects on health and development, severely affecting the poorer sections of society and placing a harsh burden on households. The Thai AIDS prevention and control programme has achieved the goal of halting infection to some extent among some high-risk groups especially direct and indirect CSW, but not for IDU. The infection index among the general population using ANC as a proxy showed a recent downward trend. This paper aims to answer the question: if an AIDS preventive vaccination strategy is adopted in the future, which groups are the most cost-effective for vaccine targets? We use two parameters for assessment, relative cost of vaccine delivery and number of total infections averted as an outcome of vaccine strategy. The number of secondary-infection averted cases is the multiplication of the number of lifetime spreading and prevalence per capita. For the scattergram of cost and outcome, we employed a cut-off point of 0.5 total case prevented and 0.5 score of relative cost of vaccine delivery to identify which groups should be targeted. Based on the potential benefit regardless of the cost of vaccine delivery, eight groups were identified as potential targets (Direct CSW, IDU in and out treatment, male STD, transport workers, indirect CSW, conscripts and prisoners). As the number of secondary infection prevention strongly determined the outcome of our analysis on identification of priority target; future fine-tune of this parameter provides a more solid foundation for policy formulation in this regard. Other risk groups such as fishermen and taxi drivers should be identified in the portfolio of potential target for vaccination. We also need to understand better the actual costs of delivery via non-traditional channels and the incentives for people to misrepresent themselves when the vaccine is made available only for the high-risk groups. The budget requirement for vaccine costs excluding delivery ranged from US$1.8 to 17.7 million (1999 prices). This was estimated, given vaccine costs of 100–1000 Baht per dose for a single dose regimen for a total 672106 vaccinees in the catch-up population excluding those who are infected (as vaccine is not given to infected persons) and various vaccine coverage rates. The budget for maintenance groups with a similar coverage required US$0.2–1.9 million. There is a dynamics and interaction between public and private demand for vaccine and the private response to not only a pricing policy and vaccine effectiveness, but also incentives provided by public targeting the high risk and maximum potential benefit on epidemic controls. The number of doses purchased privately is small and the number of secondary infections prevented per vaccinee is lower. Also, the total household willingness to pay for vaccine may exceed that for the public sector if vaccine costs are affordable. If vaccine costs are high, the coverage tends to be low. Then, immunization among relatively lower-risk individuals will not

138

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

contribute to the overall impact of epidemic. This posed an inefficient consumption by lower-risk individuals and an inequable problem when high-risk but low-income groups (e.g. direct, indirect CSW, prisoners and conscripts) who yield higher benefit will inadequately consume vaccine privately. This requires prudent public interventions through financing and/or subsidies to the risk group with a higher externality and providing public information on the vaccine costs and efficacy to those who may consume privately. Acknowledgements We are grateful for data and information provided by Drs Kamnuan Ungchusak and Chaiyos Kunanusont, MOPH Epidemiology Division, FDA Medical Device Division, Department of Health and Communicable Disease Control. This work was financially supported by the World Bank AIDS Vaccine Task Force. We highly appreciate support by Martha Ainsworth, comments and recommendations made by members of a round table discussion in Bangkok, David Bichai, Alan Hinman and Jane Rowley. We acknowledge the full support by Health Systems Research Institute and Thailand Research Fund to Senior Research Scholar Program. References [1] National Economic and Social Development Board. The Economic Developments in 1999 and Outlook for 2000. Bangkok: Office of the Prime Minister, 1999. [2] Epidemiology Division. The Sixteenth Round of HIV Sero-surveillance. Bangkok: Ministry of Public Health, 1999. [3] Wongboonsin K, Kiranandana S, Surasiengsunk S, van Griensven G, Gamett G, Anderson R. Demographic Impact of the HIV/AIDS Epidemics in Thailand: Mathematical and Statistical Projections. Bangkok: Chulalongkorn University Institute of Population Studies, 1997. [4] Epidemiology Division. The Report on AIDS Cases and Symptomatic HIV. Bangkok: Ministry of Public Health, 1999. [5] National Statistical Office. The 1995 – 1996 Survey of Population Changes. Bangkok: Office of the Prime Minister, 1997. [6] Im-em V. Mortality Trends and Levels to Verify the AIDS Epidemic in Thailand: Analysis from Death Registration Statistics, 1984 – 97. Bangkok: Mahidol University Institute for Population and Social Research, 1999. [7] Nan Provincial Health Office. Studies on Cause of Death in Nan Province, July 1997 to June 1998. Mimeograph in Thai, 1999. [8] Choprapawan C. Personal communications on verbal autopsy in six provinces (Nan, Ranong, Nakornsrithammarat, Khonkaen, Nakornsawan and Bangkhen district of Bangkok), Research paper, 1999. [9] Social Security Office. Description of Benefit Package for Social Security Workers. Bangkok: Ministry of Labour and Social Welfare, 1996. [10] Tangcharoensathien V, Lertiendumrong J, Saranasathaporn S. Research report on the research for the reform of Civil Servant Medical Benefit Scheme. Bangkok: Health Systems Research Institute. Mimeograph in Thai, 1999. [11] Pothisiri P, Tangcharoensathien V, Lertiendumrong J, et al. Funding priorities for the HIV/AIDS crisis in Thailand. Paper presented at the ‘‘Funding and Policy’’ session at the World AIDS Conference in Geneva, 1998:1998.

V. Tangcharoensathien et al. / Health Policy 57 (2001) 111–139

139

[12] Vongsaroj R. The analysis of economic loss of HIV/AIDS in Phayao. Bangkok: Chulalongkorn University Graduate School, Master Thesis in Economics, 1998. [13] Maejan Hospital. Report on Hospital Expenditure and AIDS Caseload. Chiangrai: Chiangrai Provincial Health Office, 1999. [14] National Economic and Social Development Board. The National HIV/AIDS Programme Budget, Several Years. Bangkok: Office of the Prime Minister, 1999. [15] Khongsin S. Formative Research Report on Household Impact of AIDS in Thailand, Phaoyao Province. Bangkok: Mahidol University Faculty of Public Health, 1999. [16] Tharmapornphilas P, Thamapalo S, Chunsuthiwat S, et al. Cost Effectiveness of Combined and Separate DTP and HB Immunization in Thailand. Bangkok: Ministry of Public Health Epidemiology Division, 1999. [17] Department of Communicable Diseases Control. Budget on Expanded Programme on Immunization, Several Years. Bangkok: Ministry of Public Health, 1999. [18] National Statistical Office. The Report on Household Socio-Economics Surveys, Several Years. Bangkok: Office of the Prime Minister, 2000. [19] Chunhasuthiwat S. Personal Communication on Historical Evolution of Vaccine Adoption into National EPI Programmes. Bangkok: Ministry of Public Health CDC Department, 1999. [20] Department of Communicable Diseases Control. Budget on Condom Purchasing, Several Years. Bangkok: Ministry of Public Health, 1999. [21] Department of Health. Budget on HIV/AIDS Vertical Transmission Prevention, Several Years. Bangkok: Ministry of Public Health, 1999. [22] Food and Drug Administration. Figures on Opportunistic Infection Drugs and Anti-retroviral Drugs Import and Local Production by Government Pharmaceutical Organization. Bangkok: Ministry of Public Health, 1999.

.