|
Cost effectiveness analysis of isoniazid preventive therapy to the contacts of tuberculosis patients under Japanese settings |
Yoshiyama T |
|
|
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 The use of isoniazid as preventative therapy for those who have had contact with tuberculosis (TB) patients.
Economic study type Cost-effectiveness analysis.
Study population The study population comprised Japanese men and women, stratified according to age (20, 30, 40, 50, 60 and 70 years). Since the attack rates and prevalence values of TB tend to differ with age, this enabled the relative cost-effectiveness of therapy to be studied for each age group. The therapy was targeted at people with a tuberculin test result showing a red mark of at least 30 mm in diameter.
Setting The setting was the community. The economic study was conducted in Tokyo, Japan.
Dates to which data relate The effectiveness data were obtained from studies published between 1970 and 1999, both in and outside of Japan. The dates during which the resource use data were gathered were not stated. The price year was not reported.
Source of effectiveness data The effectiveness data were derived from a review and synthesis of completed studies.
Modelling Decision trees were used to compare the therapy group and the non-therapy group in terms of the costs, the number of resulting TB cases, the number of TB-related deaths, the number of deaths due to side-effects, the loss of disability-adjusted life-years (DALYs), and the loss of healthy life-years.
Outcomes assessed in the review The parameters in the model were derived using the outcomes from the literature, although not all the data from the literature were reported. The outcomes were:
the transition of TB infection rate in Japan,
the secondary infection rate,
the infection rate for the medical profession,
the death rate among TB patients across age groups,
the sensitivity and specificity of the tuberculin tests,
the risk of TB among the infected population,
the effectiveness of isoniazid,
the compliance rate of the patients in taking medication,
the rate of isoniazid-resistant TB,
the rate of side-effects due to isoniazid, and
the death rate due to isoniazid side-effects.
Study designs and other criteria for inclusion in the review 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 Number of primary studies included Twenty primary studies were included.
Methods of combining primary studies Investigation of differences between primary studies Results of the review The parameters for the decision tree were calculated from the outcomes reviewed.
The infection rates for patients in a non-medical profession were: 0.013 for the 20-year age group, 0.031 for the 30-year age group, 0.085 for the 40-year age group, 0.231 for the 50-year age group, 0.482 for the 60-year age group, and 0.767 for the 70-year age group.
The infection rates for those in the medical profession were: 0.013 for the 20-year age group, 0.065 for the 30-year age group, 0.149 for the 40-year age group, 0.350 for the 50-year age group, 0.0.640 for the 60-year age group, and 0.875 for the 70-year age group.
The death rate due to hepatitis, resulting from isoniazid therapy, was 0.00001 for those aged below 35 years and 0.00002 for those aged above 35 years.
The rate of hepatitis that needed treatment ranged from 0.003 to 0.023.
The rate of infection in patients who completed the isoniazid therapy was 0.8.
The effectiveness of isoniazid therapy among those who completed the therapy was 0.85. The effectiveness among those who do not complete the therapy was 0.15.
The rate of infection in patients who did not take the isoniazid therapy was 0.05.
The rate of TB cases among those who were newly infected and who did not take the isoniazid therapy was 0.07 (within two years of infection). The rate of TB cases among those who were already infected, but who did not take the isoniazid therapy, was 0.0008 (annual rate).
The sensitivity of the tuberculin tests ranged from 0.3 to 0.7.
The specificity of the tuberculin tests was 0.85.
The death rates due to TB for those who did not take the therapy were: 0.003 for the 20-year age group, 0.004 for the 30-year age group, 0.009 for the 40-year age group, 0.018 for the 50-year age group, 0.034 for the 60-year age group, and 0.074 for the 70-year age group.
The rate of patients who suffered from the after-effects of TB ranged from 0.01 to 0.03.
The rate of side-effects from the TB treatment ranged from 0.03 to 0.12.
The secondary infection rate was 1.2.
The weightings used for the calculation of the DALY were:
for the treatment of hepatitis resulting from the isoniazid therapy, 0.4 (3 months);
for death due to hepatitis, 1;
for TB cases, 0.2 (6 months);
for death due to TB, 1 ; and
for after-effects of TB, 0.4 (rest of life).
The loss of healthy life-years was calculated using the following values:
for isoniazid preventative therapy, 3.35 days;
for hepatitis, 3 months;
for TB, 2.4 months;
for the side-effects of TB treatment, 3 months; and
for the after-effects of TB, 0.4 multiplied by the life expectancy.
Measure of benefits used in the economic analysis The measures of benefit used were the number of TB cases prevented, the number of deaths prevented, the loss of DALYs prevented, and the loss of healthy life-years prevented. No summary benefit measure was derived and utilised in the economic analysis. The calculation of DALYs was based on published methodologies, which involved weighting initial outcome values (see the 'Effectiveness Results' section). The authors addressed the controversial issue of weighting healthy life-years to derive DALYs in their sensitivity analysis.
Direct costs The direct costs were for the tuberculin test, preventive therapy, hepatitis treatments, and TB treatments. The cost of preventative therapy included the cost of monthly examination, isoniazid, and liver function tests. The cost of hepatitis treatment included the cost of hospitalisation, intravenous drips, and lung function tests. The cost of TB treatment included the cost of hospitalisation, medication, and examinations.
The quantities and costs were not reported separately. The cost estimates were based on insurance points for each of the conditions included. A discount rate of 3% was used. The price year was not stated.
Indirect Costs The indirect costs were not included.
Sensitivity analysis A multi-way sensitivity analysis was conducted to examine the positive hit-rate of tuberculin tests, by varying the new TB infection rate for the six age groups. Sensitivity analyses were also reported to have been carried out to investigate the effect of healthy life-years and weighting values on the calculation of DALYs.
Estimated benefits used in the economic analysis Full details of the health benefits for each age group were provided in the paper. A summary of the principal findings is presented here.
Across the non-clinical population, the rates of TB cases and TB-related death were smaller among the therapy group than among the non-therapy group, even when the probability of new infection was zero. However, the loss in DALYs and healthy life-years became smaller among the therapy group than the non-therapy group, for those aged younger than 70 years, when the probability of new infection was increased to a value between 0.1 and 0.21.
For the 70-year-old population, the therapy only seemed to become favourable in terms of DALYs when the probability of new infection reached 0.23. For the therapy to become favourable in terms of healthy life-years, the probability needed to reach 0.37.
Similar results favouring the therapy group were also reported for the clinical population. with or without baseline tuberculin testing.
Further improvements in the effectiveness of therapy, in terms of loss in DALYs and healthy life-years, were also found when considering the probability of a second infection
Cost results The cost of the tuberculin test was 3,806 Y per person, while the cost of preventive therapy was 17,198 Y per person.
The cost of treatment was 1,100,000 Y per person for hepatitis, and 1,600,000 Y per person for TB.
For the non-clinical population, the total direct costs were smaller among the therapy group than the non-therapy group when the probability of new infection was equal to or above 0.16 for patients aged 20, 0.19 for patients aged 30, 0.29 for patients aged 40, and 0.56 for patients aged 50.
For those patients aged above 60 years, the costs became higher for the therapy group regardless of the probability of new infection.
Similar differences were observed between the therapy and non-therapy groups among the clinical population. However, having the baseline tuberculin tests was found to lead to higher costs for the therapy group, compared with the non-therapy group, regardless of the new infection rates.
The probability of second infection was considered in the calculation of the costs. For the population aged less than 50 years, the costs became lower for the therapy group if the probability of new infection was 18%. However, for the older population, the costs become higher for the therapy group unless the probability of new infection was either 40% (60-year age group) or 83% (70-year age group).
Synthesis of costs and benefits The costs and benefits were not combined. The sensitivity analyses showed that isoniazid preventative therapy became an effective strategy for TB when the probability of new infection reached about 0.2 for people aged younger than 70 years. The costs were found to be highly affected by the probability of new infection. When the probability was below 0.2, the total costs became lower for the therapy group, compared with the non-therapy group, for people aged less than 30 years old. When the probability was about 0.3, the cost of therapy became favourable for patients aged less than 40 years. When considering the probability of second infection, the cost-effectiveness of the preventative therapy was improved for the population aged less than 50 years.
Authors' conclusions The effectiveness of isoniazid preventative therapy was highly dependent on the probability of new infection. The effectiveness was measured in terms of the number of cases of tuberculosis (TB) prevented, the number of TB-related deaths prevented, the loss of disability-adjusted life-years (DALYs) prevented, and the loss of healthy life-years prevented. The use of preventive therapy following tuberculin testing was found to be effective for people aged younger than 70 years, when the probability of new infection was higher than 0.2.
The total medical costs were lower for those undergoing the preventative therapy, when the probability of new infection was greater than 0.29 for people aged less than 40 years. When the probability of second infection was taken into consideration, the cost-effectiveness became more favourable for the younger generations. This was not the case for the older generations due to the high expected costs, unless the probability of new infection became very high. The precision in determining the probability of new infections has yet to be clarified. There is a need for further studies to look into this aspect, in relation to the tuberculin testing criteria.
CRD COMMENTARY - Selection of comparators The rationale for the selection of comparator was clear. The cost-effectiveness of the preventative therapy, compared with 'no therapy', was evaluated across different age populations in order to examine the current criteria for the use of isoniazid preventative therapy in Japan. In addition, the possibility of expanding the criteria was considered.
Validity of estimate of measure of effectiveness The effectiveness measures were derived from a review of the literature, but the approach adopted meant that it was difficult to comment on the validity of the estimates. In particular, little information was provided on how these studies were identified, or the criteria on which they were selected and assessed. In addition, the methods used to derive the estimates were not provided, and other relevant studies may have been omitted. To mitigate these potential shortcomings, the authors conducted a good range of sensitivity analyses. They also provided a comprehensive, and clearly presented, set of results.
Validity of estimate of measure of benefit The measures of benefit used were appropriate. These were derived directly from the effectiveness results using a modelling process.
Validity of estimate of costs The cost data appeared to include all the cost categories relevant to the perspective adopted, i.e. the Japanese health care system. Appropriate discounting was performed. However, the mechanism used to convert from insurance points to costs (which may have been charges) was not fully transparent from the paper. Indirect costs (productivity losses) may also have been relevant for the groups studied, and for a societal perspective these would need to be included. The price year was not reported, which made reflation exercises problematic. It was also difficult to generalise the results to other settings.
Other issues The authors referred to several other studies but did not compare their overall results with similar research. The issue of generalisability was not specifically addressed, although it was clearly considered; data from the United States was utilised in the decision tree, suggesting that this was also relevant for the Japanese population. The principal concern of this paper was that the methods used to find and select the primary studies were unclear. There was, therefore, a possibility of selection bias. The sensitivity analyses that were conducted tended to limit this disadvantage.
Implications of the study The authors suggested that the current criteria for the use of preventative therapy for TB are cost-effective for older members of the population, who are not included under the current policy (patients aged less than 30 years). Consideration should, therefore, be given to expanding this programme.
Bibliographic details Yoshiyama T. Cost effectiveness analysis of isoniazid preventive therapy to the contacts of tuberculosis patients under Japanese settings. Kekkaku. Tuberculosis 2000; 75(11): 629-641 Indexing Status Subject indexing assigned by NLM MeSH Adult; Aged; Antitubercular Agents /economics /therapeutic use; Chemoprevention /economics; Cost-Benefit Analysis; Disease Transmission, Infectious /prevention & Humans; Isoniazid /economics /therapeutic use; Japan /epidemiology; Middle Aged; Models, Theoretical; Tuberculosis /epidemiology /prevention & control; control /transmission AccessionNumber 22001006095 Date bibliographic record published 30/04/2002 Date abstract record published 30/04/2002 |
|
|
|