|
Preventive therapy for tuberculosis in HIV-infected persons: analysis of policy options based on tuberculin status and CD4+ cell count |
Sawert H, Girardi E, Antonucci G, Raviglione M C, Viale P, Ippolito G |
|
|
Record Status This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn. Health technology Preventive therapy for tuberculosis in HIV-infected persons. The following five intervention strategies were considered: Isoniazid preventive therapy (IPT) for tuberculin-positive cohort members only; IPT for tuberculin-positive cohort members and anergic cohort members with CD4+ cell counts lower than 0.20X10(9)/L; IPT for tuberculin-positive cohort members and anergic cohort members with CD4+ cell counts lower than 0.35X10(9)/L; IPT for tuberculin-positive cohort members and all anergic cohort members; and IPT for all HIV-positive patients regardless of purified protein derivative (PPD) status (response to skin test) and CD4+ cell count.
Study population A hypothetical cohort of 100,000 HIV infected adults was analysed over 10 years. The cohort was first stratified by CD4+ cell count and then on the basis of tuberculin status; which could be purified protein derivative (PPD) positive (tuberculin positive), PPD negative (tuberculin negative or nonanergic, i.e. reactive on skin tests to at least 1 recall antigen other than PPD), or anergic.
Setting The setting was community and hospital. The economic study was conducted in Italy.
Dates to which data relate Effectiveness data were derived from studies published between 1982 and 1997. Resource use and cost data were derived from studies published between 1991 and 1994. The price year was 1997.
Source of effectiveness data Effectiveness data were derived from a review of the literature.
Modelling A decision analytic model was used to model the costs and effects of IPT, as well as an IPT induced side effect, hepatitis, on each of the nine groups in the cohort. The output of the decision tree was a new distribution of tuberculosis infected and tuberculosis-uninfected cohort members in the three groups with different CD4+ cell counts, and the number of deaths from IPT induced hepatitis. Projections of further epidemiological developments for each group in the cohort were simulated using the Markov chain model.
Outcomes assessed in the review The outcomes assessed in the review were the distribution of HIV-infected patients by CD4+ cell counts, their distribution by CD4+ cell counts and by delayed-type hypersensitivity status (response to skin tests), the prevalence of tuberculosis infection by delayed-type hypersensitivity status, the compliance with IPT, the efficacy of IPT, the percentage of IPT-induced hepatitis, the efficacy of IPT after hepatitis, the percentage of deaths from IPT-induced hepatitis, the annual transition probabilities by CD4+ cell count (to active tuberculosis, to death from active tuberculosis, to lower CD4+ cell counts, to death from other causes), the percentage detection of active tuberculosis, the risk of new tuberculosis infection per year, the risk of tuberculosis after new infection per 100 person-years, new tuberculosis infections generated by treated or untreated persons, the prevalence of HIV among patients who acquire new tuberculosis infection outside the cohort, and the percentage of multi-resistant tuberculosis as a percentage of the overall tuberculosis.
Study designs and other criteria for inclusion in the review The main source of effectiveness data was a published prospective cohort study performed in Italy. Other effectiveness data were derived from the literature. The authors did not discuss study designs or inclusion and exclusion criteria.
Sources searched to identify primary studies Criteria used to ensure the validity of primary studies Methods used to judge relevance and validity, and for extracting data The judgement criteria applied to assess the validity of the primary studies were not specified.
Number of primary studies included The review used at least 12 primary studies.
Methods of combining primary studies Summary statistics from individual studies were used.
Investigation of differences between primary studies Results of the review The distribution of HIV-infected patients by CD4+ cell counts was 38% in the >0.35x10(9)/L group, 23.6% in the 0.35-0.20x10(9)/L group and 38.4% in the <0.20x10(9)/L group.
The distribution by delayed-type hypersensitivity status in the >0.35x10(9)/L group, was 10.7% tuberculin positive, 51.1% tuberculin negative, nonanergic, and 38.2% anergic. In the 0.35-0.20X10(9)/L group, the distribution was 8.8% tuberculin positive, 36.1% tuberculin negative, nonanergic, and 55.1% anergic. In the <0.20x10(9)/L group, the distribution was 3% tuberculin positive, 9.3% tuberculin negative, nonanergic, and 87.7% anergic.
The prevalence of tuberculosis infection was 0.9-1% in tuberculin positive patients, 0-0.05% in tuberculin negative patients and 0-20% in anergic patients.
The compliance with IPT was 75%, the efficacy of IPT was between 85 and 95%, the percentage of IPT-induced hepatitis was between 0.3 and 6.4%, the efficacy of IPT after hepatitis was 25%, the percentage of deaths from IPT-induced hepatitis was between 0.06 and 8.7%.
In the >0.35x10(9)/L group, the annual transition probabilities were 0.02 (range: 0.005 - 0.08) to active tuberculosis, 0.25 (range: 0.04 - 0.87) to death from active tuberculosis, 0.22 (range: 0.20 - 0.25) to the 0.35-0.20x10(9)/L CD4+ cell count group, 0.06 (range: 0.05 - 0.08) to the <0.20x10(9)/L CD4+ cell count group and 0.02 (range: 0.01 - 0.03) to death from other causes.
In the 0.35-0.20x10(9)/L group, the annual transition probabilities were 0.08 (range: 0.003 - 0.20) to active tuberculosis, 0.25 (range: 0.09 - 0.59) to death from active tuberculosis, 0.28 (range: 0.24 - 0.32) to the <0.20x10(9)/L CD4+ cell count group and 0.04 (range: 0.02 - 0.05) to death from other causes.
In the <0.20x10(9)/L group, the annual transition probabilities were 0.12 (range: 0.05 - 0.30) to active tuberculosis, 0.36 (range: 0.23 - 0.53) to death from active tuberculosis and 0.29 (range: 0.26 - 0.32) to death from other causes.
The percentage detection of active tuberculosis was between 80 and 100%, the risk of new tuberculosis infection was between 1.5 to 2.5 per year, the risk of tuberculosis after new infection per 100 person-years was between 4.7 and 13.9, the number of new tuberculosis infections generated by treated or untreated persons was between 2 and 10, the prevalence of HIV among patients who acquired new tuberculosis infection outside the cohort was between 0.2 and 20, the percentage of multi-resistant tuberculosis as a percentage of the overall tuberculosis was between 3 and 5%.
Measure of benefits used in the economic analysis The measures of benefit used in the economic analysis were Life Expectancy Gains and Quality-Adjusted-Life-Year (QALYs). Quality adjustment factors for the various levels of immunosuppression were taken from a published study. The method of health state valuation of the source was not discussed. The value of the health state utilities was not reported.
Direct costs Costs were discounted at 3%. Quantities and costs were not reported separately. The direct costs measured included general screening for HIV-infected individuals (tuberculin skin test and skin test with other antigens), specific screening for all patients eligible for IPT (chest radiography, serum alanine aminotransferase (ALT) and serum aspartate aminotransferase (AST), and serum bilirubin), preventive therapy for tuberculosis (one clinic visit, isoniazid, Pyrodoxine hydrochloride, ALT and AST X7), treatment of active tuberculosis (hospital room for 30 days, average diagnostic procedures and drugs), treatment of adverse effects (three clinical visits and 3 ALT and AST measurements, hospital room cost for 7 days and 3 ALT and AST measurements). The higher cost of treating multi-drug resistant (MDR) tuberculosis was also accounted for. The quantity/cost boundary adopted was that of the hospital. Quantities were estimated using published literature or by making assumptions. Costs for most screening and treatment procedures were estimated on the basis of a published analysis recently performed in Italy. For procedures for which an empirical cost analysis was unavailable, charges incurred within the Italian National Health System were applied. The price date was 1997.
Statistical analysis of costs No statistical analysis was performed on the costs, however the costs were reported with confidence intervals resulting from the probabilistic sensitivity analysis.
Indirect Costs Indirect costs were not included.
Currency Results were reported in US dollars ($). The original data were costed in Italian lira (L) that were converted into US Dollars at the rate of $1 = L1,700.
Sensitivity analysis A probabilistic sensitivity analysis was performed. Each variable derived from the empirical observation in the Italian prospective study was defined as a triangular probability distribution and uniform distributions were used for the remaining variables. Each analysis was based on 500 model simulations during which all variable distributions were sampled independently and simultaneously using the Latin Hypercube technique. A sensitivity analysis was performed to determine incremental gains for all possible combinations of compliance rates from 5% to 100% and levels of tuberculosis infection from 0% to 50%.
Estimated benefits used in the economic analysis Values reported were medians. No preventive treatment to the cohort resulted in a total of 313,475 QALYs. The incremental gain from providing IPT to tuberculin-positive patients compared to no IPT was 1,153 QALYs, (range: 1,026 - 1,245). The incremental gain from extending IPT to tuberculin-positive cohort members and anergic cohort members with CD4+ cell counts lower than 0.20x10(9)/L compared to the last strategy was 1,057 QALYs (range: 874 - 1,163). The incremental gain from extending IPT to tuberculin-positive cohort members and anergic cohort members with CD4+ cell counts lower than 0.35x10(9)/L compared to the last strategy was 528 QALYs (range: 386 - 593). The incremental gain from extending IPT to tuberculin-positive cohort members and all anergic cohort members compared to the last strategy was 460 QALYs (range: 245 - 539), and the incremental gain from extending IPT for all HIV-positive patients regardless of PPD status and CD4+ cell count compared to the last strategy was 192 QALYs (range: -166 - 305) QALYs. These estimated benefits were based on the assumption of 75% compliance with preventive therapy. Benefits were discounted at 3% and estimated for over 10 years. Side effects of the treatment (risk of hepatitis) were included in the analysis.
Cost results Values reported were medians (ranges), which is why some of the point estimates were outside the reported range. No preventive treatment to the cohort resulted in total costs of 61.9 million dollars. The incremental cost of providing IPT to tuberculin-positive patients compared to no IPT was -7.7 million dollars (range: -6.5 - 9.0). The incremental cost from extending IPT to tuberculin-positive cohort members and anergic cohort members with CD4+ cell counts lower than 0.20x10(9)/L compared to the last strategy was -0.8 million dollars (range: -2.3 - 0.5). The incremental cost from extending IPT to tuberculin-positive cohort members and anergic cohort members with CD4+ cell counts lower than 0.35x10(9)/L compared to the last strategy was -2.2 million dollars (range: -1.3 - 3.3). The incremental cost from extending IPT to tuberculin-positive cohort members and all anergic cohort members compared to the last strategy was -1.7 million dollars (range: -0.7 - 2.7) and the incremental cost from extending IPT for all HIV-positive patients regardless of PPD status and CD4+ cell count compared to the last strategy was 2.7 million dollars (range: 2.3 - 3.1). These estimated costs were based on the assumption of 75% compliance with preventive therapy. Costs were discounted at 3%. The duration of the intervention was 10 years. Costs of adverse effects (hepatitis and the more expensive to treat multi-drug resistant tuberculosis) were included in the analysis. Knock on costs of new cases of tuberculosis and HIV-infection were also accounted for.
Synthesis of costs and benefits Costs and benefits were combined into cost in US dollars per QALY gained. Costs and benefits were discounted at 3% and the price year was 1997. The incremental cost-effectiveness ratio between the intervention that extended IPT to tuberculin-positive cohort members and anergic cohort members with CD4+ cell counts lower than 0.20x10(9)/L compared to providing IPT to tuberculin-positive patients only was $578 per QALY gained. The incremental cost-effectiveness ratio between the intervention providing IPT to all HIV-positive patients compared to the strategy that extended IPT to tuberculin-positive cohort members and all anergic cohort members was $14,405 (range: $8,504 - $2,633,750) per QALY gained. Other incremental cost-effectiveness ratios were not defined, because there were gains in effectiveness and decreases in costs. Sensitivity analysis showed that, for a policy of providing IPT to tuberculin positive individuals only, a reduction in medical costs could be expected for compliance levels as low as 10%. For the policy of extending IPT to anergic patients, both compliance levels and prevalence of tuberculosis were varied and the results showed that with a tuberculosis infection level of 10%, additional costs would be incurred (from $6,035 to $9,000 per QALY) if compliance were higher than 70%. For infection levels lower than 10%, cost-effectiveness ratios were higher than $10,000 per QALY even with compliance rates at 100%. For a compliance rate of 75%, a reduction in medical costs could be expected only if the tuberculosis infection rate was higher than 17.5%.
Authors' conclusions IPT administered to tuberculin-positive HIV-infected patients increases life-expectancies and reduces medical costs. Its extension to anergic patients may be justifiable on economic grounds in populations with high prevalence of tuberculosis infection.
CRD COMMENTARY - Selection of comparators The choice of the comparators was justified by the need to analyse the cost-effectiveness of IPT on different risk groups. Care must be taken in interpreting the results, since these represent incremental costs and gains of successively wider applications of IPT (which are dependent on the relative distribution and relative numbers of patients in each risk-group) and, with the exception of the first comparison, are not comparisons between IPT and no preventive treatment.
Validity of estimate of measure of benefit The study contained a very thorough model of the intervention on different risk-groups, including side effects and knock on effects, as well as an investigation of the effects of different compliance levels and prevalence of tuberculosis levels through sensitivity analysis. However, the authors did not state that a systematic review of the literature was undertaken and no information was provided on the criteria for including the estimates from these studies in the model. Furthermore, no information was provided on the design of the Italian prospective study that provided the main estimates of effectiveness. The estimation of QALYs was justifiably modelled from the effectiveness results (gains in life expectancy), however the utilities used in the study were not provided in the article, which could be problematic in terms of generalising the study to other settings or for making comparisons with other studies.
Validity of estimate of costs Although the authors conducted the study from the point of view of the health service, it was not clear whether personnel costs had also been included in the costs of screening and treatment. The costs of side-effects and knock on effects were appropriately included in the model. However, it was not possible on the basis of the article to determine exactly which screening and treatment costs had been derived from the empirical study references and which costs had been derived from charges incurred within the Italian National Health System. This is problematic, since the categories of cost included may be different in both sources and it is difficult to assess whether this may impact on the authors' conclusions. Costs were correctly reported separately from quantities. The estimation of resource use seems to have been based on assumptions made by the authors rather than on empirical evidence from a study. However, as the resource use is explicitly defined in the article in terms of number and type of screening tests and treatments, comparisons with other settings are possible. Sensitivity analyses were not conducted on resource use and costs that could have facilitated comparisons. The authors performed appropriate currency conversions and applied discounting correctly. The rate of 3%, applied for both costs and benefits, however, might differ from the recommended discounting rates in other countries.
Other issues Appropriate comparisons were made with findings from other studies for the effectiveness results. Cost-effectiveness studies do not seem to have been previously conducted on this intervention. On the issue of generalisability, it was pointed out that comparisons with low income countries and with settings with different distributions of levels of immunosuppression would be inappropriate. The conclusions of the authors reflected the scope of the analysis in that the results concerned the potential application of IPT to HIV-positive patients. However, the authors did point out that the model variables were derived from a study carried out before the advent of combination antiretroviral therapy and protease inhibitors which may prolong the life expectancy of HIV-infected persons and affect the incidence of opportunistic infections.
Implications of the study Compared with costs of treating patients with active tuberculosis in a high-income country, IPT may result in savings for the health care provider when targeted on tuberculin positive individuals. The results of the study support the case for not providing routine IPT to HIV-infected patients on the basis of anergy skin test, but IPT provision could be justified in HIV-infected populations at high-risk of tuberculosis infection. More information of the feasibility and operational problems of IPT programmes in different settings, in particular those related to patient compliance, is needed.
Source of funding Supported by grants 8203.07 and 9401.09 from the Ministero della Santa-Progetto AIDS, Rome, Italy.
Bibliographic details Sawert H, Girardi E, Antonucci G, Raviglione M C, Viale P, Ippolito G. Preventive therapy for tuberculosis in HIV-infected persons: analysis of policy options based on tuberculin status and CD4+ cell count. Archives of Internal Medicine 1998; 158: 2112-2121 Other publications of related interest Antonucci G, Girardi E, Raviglione M C, Ippolito G. Risk factors for tuberculosis in HIV-infected persons: a prospective cohort study. JAMA 1995;274:143-148.
Holtgrave D R, Pinkerton S D. Updates of cost of illness and quality of life estimates for use in economic evaluations of HIV prevention programs. Journal of Acquired Immune Deficiency Syndromes and Human Retrovirology 1997;15:54-62.
Indexing Status Subject indexing assigned by NLM MeSH AIDS-Related Opportunistic Infections /economics /microbiology /prevention & Antitubercular Agents /economics /therapeutic use; Cost-Benefit Analysis; Humans; Isoniazid /economics /therapeutic use; Italy; Life Expectancy; Patient Compliance; Prospective Studies; Quality-Adjusted Life Years; Survival Analysis; Tuberculosis /economics /etiology /prevention & control; control AccessionNumber 21998008256 Date bibliographic record published 31/03/2001 Date abstract record published 31/03/2001 |
|
|
|