# Further details of the methods used

Technical Appendix (as supplied by author):

Modelling Costs and Effects of TB Control

Introduction

This technical appendix explains the estimation of costs and health effects of TB control strategies in 14 epidemiological sub-regions. It starts with a detailed description of the model, and then discusses the calibration of the model, the projections, and the estimation of costs and health effects at the population level.

Model

The study employs the simplified TB-HIV model as developed by John Stover of Futures Group International (Stover et al. 1998), and as applied by Dye et al. (1998). The model structure is a state-transfer compartmental model, includes age-structures and allows for exogenous reinfection (Fig. 1). The present analysis differs from previous applications in two respects. First, its focus is on the evaluation of both costs and effects of interventions, whereas previous applications assessed the impact of interventions in effects only. Second, it employs different epidemiological sub-regions (i.e. 14 epidemiological regions of the world, developed for the assessment of a wide range of health interventions by the WHO-CHOICE project) and a different, more recent, base-year (year 2000). Definitions of variables and transition parameters are in Table 1.

Figure 1. Flow diagram of the age-structured compartmental model for tuberculosis

The flow chart in Fig. 1 represents the following set of difference equations. Variables and parameters are defined in Table 1. For brevity, we write S(t,a) as S, and S(t+1, a+1) – S(t,a) as S¢ , with similar mappings leading to L and L’, Ti and Ti,, and so forth:

(1)

(2)

(3)

(4)

i (5)

(6)

(7)

(8)

(9)

Equation (9) represents immunity in the simplest way. The justification is that we are not primarily concerned with the effect of acquired immunity on TB incidence, but rather with the way in which immunity might influence the impact of DOTS.

The birth of susceptibles requires a boundary condition, S(t,0) = 1, and the maximum lifespan is taken to be 80 years, after which everyone dies. Equations (1)-(9) exclude deaths from causes other than TB because these have no influence on incidence and prevalence rates by age. However, population age structure is an important determinant of transmission, and is therefore included in the force of infection:

(10)

where .

The case detection ‘rate’ (d) is the ratio of the treated cases/yr to incident cases/yr. From equations (1)-(9), for example, the incidence rate of infectious cases at age a is:

(11)

Table 1. Definitions of variables and parameters in the age-structured TB model.

 Symbol Interpretation S(t,a) Never before infected, susceptible to infection L(t,a) Latently infected, or cured of TB under good chemotherapy Ti(t,a) Infectious (smear positive) TB; primary, endogenous, exogenous or relapse Tn(t,a) Non-infectious (smear negative) pulmonary and extra-pulmonary TB Ni(t,a) Self-cured, from infectious TB; non-infectious Nn(t,a) Self-cured, from non-infectious TB; non-infectious Fi(t,a) Proportion of Ti whichis not cured under treatment (classed as having ‘failed’, ‘defaulted’ or ‘transferred out’ in cohort analysis) Fn(t,a) As Fi, but from Tn M(t,a) Immune to infection, naturally (MOTT) or following vaccination I(t,a) Incidence rate of infectious (sub I) or non-infectious (sub n) TB λ (t) Incidence rate (all rates per capita) or force of infection, or annual risk of infection (ARI) β (t) Per capita contact rate between Tiand other individuals θ Exponential rate of decline in β , reflecting ‘socio-economic improvement’ π (t,a) Proportion of population in age class a at time t
 m+(a) Rate at which immunity is acquired by S as a result of non-specific natural infection (age-independent) or vaccination (age-independent, or children < 1 yr) m- Rate at which protective immunity is lost μ Death rates; subscripts i, n, HIV and TB/HIV refer to different rates for Ti , Tn, THIV and TTB/HIV f(a) Proportion of progressive primary cases which becomes infectious φ Proportion of Fiwhich is infectious n(a) Rate of natural cure for Ti and Tn p(a) Proportion of infected S which develop progressive primary TB (within 1 yr), infectious or non-infectious r Rate of relapse from F to T rn Rate of relapse after self-cure, from N to T v(a) Rate at which L progress to TB by endogenous reactivation w Rate of smear conversion, from non-infectious to (Tn ) to infectious TB (Ti) x(a) Proportion of (exogenously) re-infected L which is susceptible to developing TB within 1yr d Rate at which TB cases are found and treated k Proportion of treated cases given curative chemotherapy ε Relative case detection rate of non-infectious cases δ Proportion of failed treatment cases that are multi-drug resistant l Rate of cure for multi-drug resistant cases y Rate at which multi-drug resistant cases are found and treated

Table 2. Estimates for transition parameters in Table 1.

 Parameters Value (range) Sources μI 0.3 (0.2-0.4) Rutledge & Crouch 1919, Berg 1939, Drolet 1938, Thompson 1943, Tatersall 1947, Lowe 1954, Springett 1971, NTI 1974, Grzybowski & Enarson 1978 μn 0.21 (0.18-0.25) Lindhart 1939, Murray et al 1993 μHIV 0.25 (0.1-0.33) Nunn et al 1992, Nunn & Felten 1994, Edlin et al 1992, Allen et al 1992, Mulder et al 1994, Perriens et al 1995, Whalen et al 1995 μTB/HIV 1.0 (0.75-1.0) f(£ 15) 0.08 (0.012-0.1) Styblo 1977, Murray et al 1993, Barnett & Styblo 1991 f(>15) 0.65 (0.5-0.65) f(HIV) 0.3 (0.19-0.4) Colebunders et al 1989, Meeran 1989, DeCock et al 1991, Githui et al 1992, Elliot et al 1993, Sassan Morokro et al 1994, Nunn et al 1994, Cauthen et al 1996, Espinal et al 1996 n 0.2 (0.15-0.25) Springett 1971, Olakowski 1973, NTI 1974, Enarson & Rouillon 1994, Grzybowski & Enarson 1978 p(£ 15) 0.04 (0.015-0.14) Sutherland 1968, 1976 Ferebee 1970, Comstock 1982, Sutherland et al 1982, Styblo 1986, Krishnamurthy et al 1976, Krishnamurthy & Chaudhuri 1990, Vynnycky 1996, Vynnycky & Fine 1997, Dye et al 1998 p(>15) 0.14 (0.08-0.25) p(HIV) 0.67 (0.36-0.8) DiPerri et al 1989, Daley et al 1992, Edlin et al 1992, Coronado et al 1993 r 0.3 (0-0.5) Grzybowski et al 1965, Horwitz 1969, Ferebee 1970, Chan-Yeung et al 1971 rn 0.03 (0.02-0.04) Springett 1961, Grzybowski et al 1965, Ferebee 1970, Chan-Yeung et al 1971, Campbell 1974, Nakielna et al 1975, Styblo 1986 v(£ 15) 5’ 10-5 (0-10-4) Horwitz et al 1969, Barnett et al 1971, Sutherland et al 1982, Styblo 1991, Vynnycky 1996, Vynnycky & Fine 1997, Dye et al. 1998 v(>15) 1.13’ 10-4 (10-4-3’ 10-4) v(HIV) 0.17 (0.04-0.2) Schulzer et al 1992 x(£ 15) 1.0 (0.5-1.0) Sutherland 1968, Sutherland et al 1982, Vynnycky 1996, Vynnycky & Fine 1997, Dye et al 1998 x(>15) 0.35 (0.1–0.6) x(HIV) 0.75 (0.5-1.0) Assumed (no data) w 0.015 (0.007-0.02) Ferebee 1970, HKCS 1974 m+(a) 0.2 (0-0.5) for infant vaccination; natural immunity excluded from calculations in this paper Colditz et al 1994, Fine 1994, 1995 m- 0.15 (0.06-0.2) MRC 1972, Hart & Sutherland 1977, BTTA 1975 δ 0.58 (0..29-0.8) Suarez et al 2002 l 0.48 (0.3-0.7) Suarez et al 2002, Leimane et al. 2005 y 1 Assumed (no data)

Note: All rates are per capita per year; ranges have been subjected to (one way) sensitivity analysis (see Main text)

Calibration

We calibrated the model to produce TB incidence, prevalence and mortality rates for each region for the period 1950-2000 that matched those actually observed in the same period, based on the best and most recent available evidence (WHO 2004). This was done by establishing equilibrium rates by age (0-80 years) in the starting year of analysis by dropping the time dependence in equations (1)-(9), and then solving numerically. The equilibrium value of β was calculated from the relation between force of infection and the prevalence of smear positives (equation 10). We chose 1950 of starting year of analysis for all regions (approximately when drugs became widely available). We applied similar transition parameters and regional case detection and cure rates for the period 1950-2000 to those specified in Dye et al. (1998), with 1995 values to be similar to year 2000 values. Regional population estimates, including background mortality rates, were based on WHO estimates (WHO 2000). Regional HIV/AIDS estimates of incidence, prevalence and case-fatality for the period were based on UNAIDS internal projections, and were applied to each of the 9 classes of individuals in Figure 1. We then have a parallel 9-class TB/HIV model, with the different transfer rates indicated in Table 2. A number of indicators can be used to describe the model, and are presented in Table 3.

Table 3. Model indicators, year 2000.

 Epidemiological sub-regions AFRO E SEARO D Indicators Annual risk of infection (ARI %) 2,5 2,6 Incidence rates (new cases per 105 per year) All forms 369 155 Infectious cases 180 107 Ratio incidence rate infectious cases/ARI 73 42 Prevalence rate (infectious cases per 105 per year) 236 161 Death rate (all forms per 105 per year) 361 96 Change in annual risk of infection (% per year) 7,5 -1,7 Change in incidence rate (% per year) 8,5 -1,1 Change in contact rate (% per year) -0,3 -0,6

Projection

We used the calibrated model to make a number of projections. To allow the effectiveness of current practices to be evaluated, we first estimated what would happen to transmission, morbidity and deaths if all current interventions ceased. By then introducing all possible interventions against this background of no interventions being implemented, we estimated the population health effects (and costs) of those interventions.

Each projection simulated the change from equilibrium (steady state with respect to time), accounting for (a) the initial incidence rate in each country, (b) the duration and background rate of decline in TB (based on ARI data and modelled by reducing β ) (Cauthen et al 1988, Murray et al 1993), (c) the recent history of, and prospects for improving, case finding and cure rates, (d) demography. We assume that DOTS programmes for treatment of smear-postive and smear-negative cases provides cure rates of 85% for everyone treated (the same for HIV-positive and negative; Grosset 1992, Harries 1997), and that case detection increases from year 2000 levels to 70% (except for Western European regions that have higher rates) over a period of ten years (Evans et al. 2005). We assumed that 100% of treatment failures with MDR are identified and that the cure rate is 48% over the period (Suarez et al. 2002).

Numerical simulations of equations (1)-(9) were carried out for a population of constant, arbitrary size, in each of the 14 epidemiological sub-regions. Population age-structures for the 14 epidemiological sub-regions were then used to convert rates (incidence, prevalence etc) to numbers; these different age-structures reflect births and deaths (all causes, including TB).

Estimating population health and costs

On the basis of the projected incidence, prevalence and mortality data, we estimated the population impact of the different scenarios in terms of healthy years lived (HYL). To do so, we used a population model PopMod (Lauer et al. 2003). Health state valuations were taken from the Burden of Disease study (Murray et al. 1996). The model was run for 100 years, i.e. the length of time necessary for all people affected by the interventions over the ten years of the analysis to have died. The difference between the HYL in each intervention scenario and the no-intervention scenario is the health gain of the intervention. This can also be interpreted as the number of disability adjusted life-years (DALYs) averted.

To estimate total patient costs, the project numbers of patients for each intervention were multiplied by the cost per patient. Details on the calculations of patient costs are provided in Tables 4 and 5. Programme costs were estimated independent of the model, and its details are reported elsewhere (WHO-CHOICE website).

Table 6 shows the results for patient and programme costs.

Table 4. Resource use patterns for treatment of smear-positive, smear-negative, and MDR cases

 Intervention component Resource use Volume/costs* DIAGNOSTICS † ‡ Treatment of smear-positive cases # smears to detect one smear+ 30 # X-rays per smear+ to detect 1 smear- 9 DRUGS§ Treatment of smear-positive cases Intensive phase regimen 2/HRZE(3) # weeks 8 # days per week 3 cost per regimen \$4.39 Continuation phase regimen 4/HR(3) # weeks 16 # days per week 3 cost per regimen \$3.45 total doses 72 Treatment of smear-negative cases Intensive phase regimen 2/HRZ(3) # weeks 8 # days per week 3 cost per regimen \$2.33 Continuation phase regimen 4/HR(3) # weeks 16 # days per week 3 cost per regimen \$3.45 Treatment of MDR cases regimen # weeks 78 # days per week 7 cost per regimen \$264.49 HEALTH CENTRE VISITS Treatment of smear-positive cases Intensive phase # weeks 8 days per week 3 # visits 24 Continuation phase # weeks 16 days per week 1 # vists 16 Intensive + Continuation phase 40 Monitoring visits # visits 3 Total (supervision + monitoring) 43 monitoring diagnostics smears 6 Treatment of smear-negative cases Intensive phase # weeks 8 days per week 3 # visits 24 Continuation phase # weeks 16 days per week 0.25 # visits 4 Intensive + Continuation phase 28 Treatment of MDR cases Intensive phase # weeks (3 months) 12.5 # days per week 6 # visits 75 Continuation phase # weeks (15 months) 62 # days per week 6 # visits 375 intensive + Continuation phase 450 Monitoring visits # visits 18 Total (supervision + monitoring) 468 monitoring diagnostics # smears 18 # culture 9 # X-ray 3

* Costs are presented in I\$. Because drugs are traded goods, these costs are equivalent to US\$. Details of this approach are reported elsewhere.12

Based on the assumption that one per 10 TB suspects presenting to health centres is tested smear-positive. Since each suspect is tested with three smear tests, this means that 30 smears are used to detect one smear-positive case. The remaining nine suspects remain suspect cases and are tested with an X-ray. This means that for every case that is tested smear-positive, 9 X-rays are performed which will detect x number of smear-negative cases. The number of X-rays used for case-detection of x smear-negative cases is then equal to 9*smear-positive cases detected (and is then independent of x).

‡ #=number

§ H=Isoniazid R= Rifampicin Z=Pyrazinamide E=Ethambutol K=Kanamycin Cx=Ciprofloxacin Et=Ethionamide

Table 5. Unit costs (I\$)*

 Region Coverage Health facility† Laboratory tests Outpatient visit Inpatient day Smear test X-ray Culture test Afr-E ‡ 50% \$3.98 \$15.94 \$1.99 \$34.55 \$24.68 80% \$5.84 \$15.94 \$2.00 \$34.67 \$24.77 95% \$7.18 \$15.94 \$2.01 \$34.80 \$24.86 Sear-D § 50% \$3.85 \$14.95 \$1.14 \$23.91 \$17.08 80% \$3.85 \$14.95 \$1.14 \$23.91 \$17.08 95% \$3.85 \$14.95 \$1.14 \$23.91 \$17.08 * Costs are expressed in I\$ and can be converted in US\$ for a reference country in a region. For example, cost estimates in Afr-E in I\$ should be divided by a factor 4.5 to obtain US\$ cost estimates for Kenya. In Sear-D, this factor is 5.2 to obtain US\$ cost estimates for India. Details of this approach are discussed elsewhere .12 †Estimates may seem low because of the assumption of technical efficiency.26 ‡ Costs of smear tests and X-rays based on 1998 Malawi data (converted to year 2000 prices using GDP deflators), costs of culture test based on 2000 Peru data. All cost estimates are converted to country-specific estimates using purchasing power parity exchange rates. Differences in costs vary by coverage level because of local distribution costs. § Costs of smear tests based on 2000 China data, costs of X-rays based on 1998 Malawi data (converted to year 2000 prices using GDP deflators), costs of culture test based on 2000 Peru data. All cost estimates are converted to country-specific estimates using purchasing power parity exchange rates.

Table 6 Cost (in international dollars (\$Int)) and DALYs averted per patient for different strategies for tuberculosis control in the two regions Afr-E and Sear-D

 Region and coverage level Patient costs for treatment (\$Int)* Programme costs for treatment (\$Int)† DALYs averted by intervention strategy‡ Smear-positive cases Smear-negative and extra-pulmonary cases After treatment failure Smear-positive cases Smear-negative and extra-pulmonary cases After treatment failure Minimal DOTS Full DOTS Minimal DOTS plus resistant cases Full combination Afr-E: 50% 209 327 2284 266 32 711 72 42 72 42 80% 280 374 3068 257 45 934 72 42 72 42 95% 332 408 3633 300 86 1703 72 42 72 42 Sear-D: 50% 186 321 2170 236 75 214 55 45 51 43 80% 186 321 2171 476 70 1427 55 45 51 43 95% 186 321 2171 493 77 1551 55 45 51 43

Costs are given in international dollars (a hypothetical unit of currency that has the same purchasing power that the US\$ has in the United States at a given point in time). Details of this approach are discussed elsewhere.18

*Patient cost may differ by coverage level because of differences in costs of outpatient visits and local distribution costs of goods.

†Programme costs may differ by coverage level because of differences in cost functions and number of patients treated.

‡See methods section for details of interventions.

References

Allen S, Batungwanayo J, Kerlikowske K, et al. Two-year incidence of tuberculosis in cohorts of HIV infected and uninfected urban Rwandan women. American Review of Respiratory Diseases 1992; 146:1439-44.

Barnett G, Grzybowski S, Styblo K. Present risk of developing active tuberculosis in Saskatchewen according to previous tuberculin and X-ray status. Bulletin of the International Union Against Tuberculosis 1971; 45: 51-74.

Barnett, G. and K. Styblo. Bacteriological and X-ray status of tuberculosis following primary infection acquired during adolescence or older. Bulletin of the International Union Against Tuberculosis 1977; 52: 5-16.

Berg, G. The prognosis of open pulmonary tuberculosis. A clinical statistical analysis. Acta Tuberculosea Scandinavica 1939; Suppl IV: 1-207.

British Thoracic and Tuberculosis Association. Present effectiveness of BCG vaccination in England and Wales. Tubercle 1975; 56: 129-137.

Campbell AH. Relapse in patients with tuberculosis. Bulletin of the International Union of Tuberculosis 1974; 1: 219-22.

Cauthen G, Pio A, ten Dam H. Annual risk of tuberculosis infection. Geneva, WHO 1988; WHO/TB/88.154.

Cauthen GM, Dooley SW, Onorato IM et al. Transmission of Mycobacterium Tuberculosis from Tuberculosis Patients with HIV infection of AIDS. American Journal of Epidemiology 1996; 44: 69-77.

Chan-Yeung M, Galbraith JD, Schulson N et al.. Reactivation of inactive tuberculosis in Northern Canada. American Review of Respiratory Diseases 1971; 104: 861-65.

Colditz GA, Brewer TF, Berkey CS et al. Efficacy of BCG vaccine in the prevention of tuberuclosis. Journal of the American Medical Association 1994; 271: 698-702.

Colebunders R, Ryder R, Nzilambi N et al. HIV infection in patients with tuberculosis in Kinshasha, Zaire. American Review of Respiratory Disease 1989;139:1082-85.

Comstock G. Epidemiology of tuberculosis. American Review of Respiratory Disease 1982; 125 :8-16.

Coronado V, Beck-Sague C, Hutton M et al. Transmission of multidrug-resistant Mycobacterium tuberculosis persons with human immunodeficiency virus infection in an urban hospital: epidemiological and restriction fragment length polymorphism analysis. The Journal of Infectious Diseases 1993; 168: 1052-55.

Daley C, Small P, Schecter G et al. An outbreak of tuberculosis with accelerated progression among persons infected with the human immunodeficiency virus. New England Journal of Medicine 1992; 326: 231-35.

De Cock KM, Gnaore E, Adjorlolo G, et al. Risk of tuberculosis in patients with HIV-1 and HIV-II infections in Abidjan, Ivory Coast. British Medical Journal 1991; 301: 496-99.

Di Perri G, Gruciani M, Danzi M et al. Nosocomial epidemic of active tuberculosis among HIV - infected patients. Lancet 1989; 2: 1502-04.

Drolet G. Present trend of case fatality rates in tuberculosis. American Review of Tuberculosis 1938; 37:125-51.

Dye C, Garnett GP, Sleeman K, Williams BG. Prospects for worldwide tuberculosis control under the WHO DOTS strategy. Lancet 1998; 352:1886-91.

Edlin B, Tokars J, Grieco M et al. An outbreak of multidrug-resistant tuberculosis among hospitalised patients with the acquired immunodeficiency syndrome The New England Journal of Medicine 1992; 326: 1514-21.

Elliot A, Hayes, R, Halwiindi B et al. The impact of HIV infectiousness of pulmonary tuberculosis: a community study in Zambia AIDS 1993; 7: 981-87.

Enarson D, Rouillon A. The epidemiological basis of tuberculosis control. In: Davies, PDO (ed) Clinical Tuberculosis. Chapman and Hall, London1994.

Espinal M, Reingold A, Perez G et al. Human immunodeficiency virus infection in children with tuberculosis in Santo Domingo, Dominican Republic: Prevalence, clinical findings, and response to antituberculosis treatment. Journal of Acquired Immunodeficiency Syndromes and Human Retroviriology 1996; 13: 155-59.

Evans D, Tan-Torres Edejer T, Adam T, Lim SS and the WHO-CHOICE MDG Team. Achieving the Millennium Development Goals for health: methodology for cost-efefctiveness analysis. Methods for assessing the costs and health impact of interventions. BMJ 2005

Ferebee S. Controlled chemoprophylaxis trials in tuberculosis a general review. Advances in Tuberculosis Research 1970; 17: 28-106.

Fine PEM, Sterne JAC, Ponninghaus JM, Rees, RJW. Delayed-type hypersensitivity, mycobacterial vaccines and protective immunity. Lancet 1994; 344: 1245-1249.

Fine PE. Variation in protection by BCG: implications of and for heterologous immunity. Lancet 1995; 346: 1339-45.

Githui W, Nunn P, Juma E et al. Cohort study of HIV-positive and HIV-nagtive tuberculosis, Nairobi, Kenya: comparison of bacteriological results. Tubercle and Lung Disease 1992; 73: 203.

Grosset JH. Treatment of tuberculosis in HIV infection. Tubercule and Lung Disease 1992; 73: 378-83.

Grzybowski S, McKinnon N, Tuters L et al. Reactivations in inactive pulmonary tuberculosis. American Review of Respiratory Diseases 1965; 352: 361-209.

Grzybowski S, Enarson D. The fate of cases of pulmonary tuberculosis under various treatment programmes. Bulletin of the International Union Against Tuberculosis 1978; 53: 70-75.

Harries AD. Tuberculosis in Africa: clinical presentation and management. Pharamacology and Therapeutics 1997; 73: 1-50.

Hart PD’A, Sutherland I. BCG and vole bacillus vaccines in the prevention of tuberculosis in adolescence and early adult life; final report to the Medical Research Council. British Medical Journal 1977; 56: 293-295.

Hong Kong Chest Service/Tuberculosis Research Centre. A controlled trial of 2-month, 3-month, and 12-month regimens of chemotherapy for sputum smear negative pulmonary tuberculosis. American Review of Respiratory Diseases 1984; 130: 23-28.

Horwitz O. Public health aspects of relapsing tuberculosis. American Review of Respiratory Diseases 1969; 99: 183-93.

Horwitz O, Wilbek E, Erickson PA. Epidemiological basis of tuberculosis eradication. 10. Longitudinal studies on the risk of tuberculosis in the general population of a low-prevalence area. Bulletin of the World Health Organization 1969; 41: 95-113.

Krishnamurthy V, Chaudhuri K. Risk of pulmonary tuberculosis associated with exogenous reinfection and endogenous reactivation in a south Indian rural population - A mathematical estimate. Indian Journal of Tuberculosis 1990; 37: 63-67.

Lauer JA, Murray CJL, Roehrich K, Wirth H. PopMod: A longitudinal four-state population model with two disease states and comorbidity, Cost Eff Resour Alloc 2003; 1: 6.

Leimane V, Riekstina V, Holtz TH, et al. Clinical outcome of individualised treatment of multidrug- resistant tuberculosis in Latvia: a retrospective cohort study. Lancet 2005; 365: 318-326.

Lindhart M. The statistics of pulmonary tuberculosis in Denmark, 1925-1934. A statistical investigation of the occurrence of pulmonary tuberculosis in the period 1925-34, worked out on the basis of the Danish National Health Service File of notified cases and deaths. Ejnar Munksgaard, Copenhagen, 1939.

Lowe CR. Recent trends in survival of patients with respiratory tuberculosis. British Journal of Preventive and Social Medicine 1954; 8: 91-98.

Meeran K. Prevalence of HIV infection among patients with leprosy and tuberculosis in rural Zambia. British Medical Journal 1989; 298: 364-65.

MRC Tuberculosis Vaccines Clinical Trials Committee. BCG and vole bacillus vaccines in the prevention of tuberculosis in adolescence and early adule life. Bulletin of the World Health Organization 1972; 46: 371-385.

Mulder D, Nunn A, Kamali A et al. Two-year HIV-1 associated mortality in a Ugandan rural population cohort. Lancet 1994; 343: 1021-23.

Murray C, Styblo K, Rouillon A. Tuberculosis. In: Jamison DT et al. Disease Control Priorities in Developing Countries. Oxford University Press, 1993.

Nakielna E, Cragg R, Grzybowski S. Lifelong follow-up of inactive tuberculosis: its value and limitations. American Review of Respiratory Disease 1975; 112: 765-772.

National Tuberculosis Institute, Bangalore. Tuberculosis in a rural population in South India: a five year study. Bulletin of the World Health Organization 1974; 51: 473.

Nunn P, Felten M. Surveillance of resistance to antituberculosis drugs in developing countries. Tubercle and Lung Disease 1994; 75:163-67.

Nunn P, Brindle R, Carpenter L. Cohort study of human immunodeficiency virus infection in patients with tuberculosis in Nairobi, Kenya: analysis of early (6 month) mortality. American Review of Respiratory Diseases 1992; 146: 849-54.

Nunn P, Mungai M, Nyamwaya J et al. The effect of human immunodeficiency virus type-1 on the infectiousness of tuberculosis. Tubercle and Lung Disease 1994; 75: 25-32.

Olakowski T. Assignment report on a tuberculosis longitudinal survey, National Tuberculosis Institute, Bangalore. Geneva, WHO 1973.

Periens JH, St Louis ME, Mukadi YB, et al. Pulmonary tuberculosis in HIV-infected patients in Zaire: a controlled trial of treatment for either 6 or 12 months. New England Journal of Medicine, 1995; 332: 779-84.

Rutledge C, Crouch J. The ultimate results in 1 694 cases of tuberculosis treated at the Modern Woodmen Sanitorium of America. American Review of Tuberculosis

1919; 2: 755-6.

Sassan Morokro M, De Cock KM, Ackah A et al. Tuberculosis and HIV infection in children in Abidjan, Cote d'Ivoire. Transactions of the Royal Society of Tropical Medicine and Hygiene 1994; 88: 178-81.

Springett V. Ten-year results during the introduction of chemotherapy for tuberculosis Tubercle 1971; 52: 73-87.

Stover J. TB-HIV spreadsheet model. A model for illustrating the effects of the HIV epidemic on tuberculosis. The Futures Group International, The POLICY project 1998.

Styblo K. Tuberculosis control and surveillance. In: Flenley D, Retly J (eds). Edinburgh, Churchill Livingstone. Advances in Respiratory Medicine 1986: 77-108.

Sutherland I. The ten-year incidence of clinical tuberculosis following ‘conversion’ in 2,550 individuals aged 14 to 19 years. Tuberculosis Surveillance and Research Unit Progress Report 1968. The Hague, Royal Netherlands Tuberculosis Association (KNCV).

Sutherland I. Recent studies in the epidemiology of tuberculosis, based on the risk of being infected with tubercle bacilli. Advances in Tuberculosis Research 1976; 19: 1-63.

Sutherland I, Svandova E, Radhakrishna S. The development of clinical tuberculosis following infection with tubercle bacilli. Tubercle 1982; 63: 255-68.

Tattersall W. The survival of sputum positive consumptives. A study of 1192 cases in a county borough between 1914 and 1940. Tubercle 1947; 28: 85.

Thompson B. Survival rates in pulmonary tuberculosis. British Medical Journal 1943; 2: 721.

Vynnycky E. An Investigation of the Transmission Dynamics of M. tuberculosis.

PhD thesis, University of London, 1996.

Vynnycky E, Fine PEM. The natural history of tuberculosis: the implications of age-dependent risks of disease and the role of reinfection. Epidemiology and Infection 1997; 119: 183-201.

Whalen C, Horsburgh C, Hom D et al. Accelerated course of human immunodeficiency virus infection after tuberculosis. American Journal of Respiratory and Critical Care Medicine 1995; 151: 129-35.

World Health Organization. Burden of Disease 2000. Available from: www.who.int/evidence.

World Health Organisation. Global tuberculosis control - surveillance, planning, financing WHO Report 2004, WHO/HTM/TB/2004.331

WHO-CHOICE website at www.who.int/choice