Table 2 Unit costs. For normal distributions, the first figure gives the mean and the second the variance. For uniform distributions, the two figures give the lower and upper limits.

TB diagnosis costs, existing level of case detection

Sputum smear positive case detected

101

Normal (101, 25)

Nganda et al [11]. For every sputum smear positive case detected, assume 10 TB suspects are seen. For each TB suspect, assume 3 sputum smears and 1 chest X-ray are done, in line with WHO guidelines.

TB diagnosis when case detection rate increased to 70%

Sputum smear positive case detected

303

Uniform (202, 404)

Recent financial data from Kenya's applications to the Global Fund to fight AIDS, TB and Malaria (Pennas T, written communication), data submitted to WHO by Kenya and other high burden countries [1], and financial analyses for the forthcoming second Global Plan to Stop TB (2006–2015) being prepared by the Stop TB Partnership. These indicate that the average cost per sm+ patient detected will increase 2–4 times when activities to improve case detection rates to 70% are implemented. Further details from authors upon request.

Treatment for sputum smear positive TB cases, existing cure rate

Patient treated

140

Normal (140, 49)

Nganda et al [11].

Treatment for sputum smear negative TB cases, existing cure rate

Person treated

130

Normal (130, 43)

Nganda et al [11].

Treatment for sputum smear positive TB cases if cure rates improved to 85%

Person treated

280

Uniform (210, 350)

Recent financial data submitted to WHO by Kenya and other high burden countries [1], and financial analyses for the forthcoming second Global Plan to Stop TB (2006–2015) being prepared by the Stop TB Partnership. These indicate that the average cost per patient treated will double when activities to improve cure rates are implemented. Further details from authors upon request.

Treatment for sputum smear negative TB cases if treatment completion rates improved to 85%

Person treated

260

Uniform (195, 325)

As above for treatment of sputum smear positive cases.

TLTI (6 months)

Person treated

32

Uniform (27, 37)

Bell et al [19], WHO estimates of population coverage of HIV/AIDS interventions [20]. Assume 13% adult population accesses VCT each year [20], and that 36% are HIV+, 100% are screened for TLTI, 43% start treatment of whom 38% complete treatment [21].

TLTI (lifetime)

Person year of treatment

64

Uniform (54, 74)

As above for TLTI for six months, plus assumption that treatment for one year is double the cost of treatment for six months.

Treatment for AIDS-related opportunistic infections and palliative care in absence of ART

Person year of treatment

211

Uniform (167, 323)

Unit costs used for Kenya and other low-income high HIV prevalence countries in Africa in recent estimates of the resources needed for a comprehensive response to HIV/AIDS, prepared by UNAIDS (Gutierrez JP, written communication). Kenya is in the middle of the range.

Cost of ART for a TB patient

Six person months of treatment when TB and ART overlap

495

Uniform (420, 544)

Unit costs used for Kenya and other low-income high HIV prevalence countries in Africa in recent estimates of the resources needed for a comprehensive response to HIV/AIDS, prepared by UNAIDS (Gutierrez JP, written communication). Kenya is in the middle of the range. Drop out rate on ART varies from 5% (optimistic scenario) to 20% [10][22].

ART, people without TB

Person year of treatment

640

Uniform (487, 743)

Sources and assumptions as stated above for ART for TB patients.