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.

The use of a single, low-dose computed tomography (CT) scan to screen high-risk individuals for lung cancer.

Screening.

Cost-effectiveness analysis.

The population comprised patients aged 60 years and over with at least 10-pack-years of smoking, no prior history of cancer, and who were suitable for thoracic surgery.

The setting was secondary care. The economic study was conducted in New York, USA.

The effectiveness data related to 1997, 1999 and 2000. The resource use data related to 1997 to 1999. The prices related to 1997 to 1999 and were adjusted to year 2000 US dollars using the medical consumer price index.

The effectiveness data were derived from a single study.

The resource use data were gathered prospectively and were derived from the same sample used in the effectiveness analysis. The data were taken from the computer records of the hospital at which the patients were treated.

The study sample comprised volunteers aged 60 years or older with a history of at least 10-pack-years of smoking, with no prior history of cancer (except non-melanoma skin cancer), and who were suitable for thoracic surgery. The study did not have a control group. See the 'Other Publications of Related Interest' section (Henschke et al.).

This was a non-comparative, prospective, cohort study that was conducted at a single centre.

The primary health outcomes used were:

the probability of a positive screen result,

the probability of diagnosing a non-small-cell lung cancer, and

the stage distribution of screen-detected lung cancer.

The prevalence of lung cancer in the study sample was 2.7%. The incidence of non-calcified nodules was 23%.

Eight-five per cent of screen-detected cancers were Stage I, 4% were Stage II, 7% were Stage IIIA, 4% were Stage IIIB, and 0% were Stage IV.

A higher percentage of cancers were detected at an earlier stage in the study sample in comparison with published data on cancers detected in usual care. The data also suggested that annual CT scan screening might translate into a reduction in lung cancer mortality. See the 'Other Publications of Related Interest' section (Henschke et al.).

A deterministic, decision tree model was constructed to calculate the costs and life-years associated with low-dose CT scan screening and usual care. Following a positive screen result, patients with no growth detected underwent a series of high-resolution CT scans (HRCT). Patients with signs of growth after the diagnostic or follow-up CT scans were investigated using fine-needle aspiration (FNA) biopsy. The life expectancy depended on the stage of cancer detected

The measure of benefits used in the economic analysis was the life-years saved (LYS). This was calculated from the National Center for Health Statistics for individuals without lung cancer (and for those cured of lung cancer) to be 16.3 years. The cure rate for Stage I and II cancers was calculated according to a published formula. Stage-specific estimates of lead-time were added to the life expectancy of unscreened individuals. The estimates were 1.5 years for Stage I cancer, 2.5 years for Stage II cancer, 3.5 years for Stage IIIA cancer, 4.0 years for Stage IIIB cancer and 4.5 years for Stage IV cancer. The stage-specific survival was estimated using data from the National Cancer Institute Surveillance and End Results Registry and a formula that fitted a declining exponential approximation to life expectancy. The estimates were 4.5 years for Stage I cancer, 3.5 years for Stage II cancer, 2.6 years for Stage IIIA cancer, 1.6 years for Stage IIIB cancer, and 1.0 years for Stage IV cancer. Further details of the formulae used were provided elsewhere (Beck et al., see Other Publications of Related Interest).

The study included the physician- and hospital-based costs incurred in the first year following the diagnosis of lung cancer, as recorded by the New York-Presbyterian Hospital's financial system cost database and the Faculty Practice Plan database. The study also included the cost of terminal care for patients with advanced stage cancer by assuming that it was equal to the costs of care for Stage IV cancer. The cost of the screening programme included low-dose CT scan, HRCT, limited CT and FNA. The resource quantities used were not reported separately from the unit costs. The model did not extrapolate the costs beyond one year, but the future costs were discounted at a rate of 3% per annum. The study reported the average costs as year 2000 US dollars. The costs were adjusted for inflation using the medical consumer price index.

The costs were treated deterministically. No tests to assess the difference in costs between the treatment groups were performed. This was appropriate given that the costs were measured in only one group using direct data.

The indirect costs were not included in the analysis, which was consistent with the perspective adopted.

US dollars ($).

Several one-way sensitivity analyses were performed. These assessed the effect of changes in the cost of low-dose CT scans, diagnostic CT scans, the treatment of Stage I cancer and terminal care. They also investigated changes in the probability of overdiagnosis, the prevalence of lung cancer, the stage-specific lead-time, and the cure rates for Stage II non-small-cell lung cancer. The authors stated that clinically plausible ranges were used for the analyses, but these were not reported. The sensitivity analysis addressed variability in the data.

The average life expectancy with screening was estimated to be 16.15 years per patient, compared with 16.05 years with no screening. Thus, screening was expected to increase survival by 0.1 years. The life-years were not discounted.

The average cost of low-dose CT scan screening was $1,174 per patient, compared with $942 for no screening. Hence, low-dose CT scan screening is expected to increase the costs by 230. The costs were discounted at a rate of 3% per annum. The excess costs associated with investigating overdiagnosed lung cancers were included in a sensitivity analysis.

The costs and benefits were synthesised to provide an estimate of the costs per LYS. The incremental cost per LYS saved with low-dose CT scan screening, compared with no screening, was $2,500 in year 2000 US dollars. A sensitivity analysis addressed the issue of overdiagnosis of lung cancer by CT scan screening. It found that the incremental cost-effectiveness ratio of screening versus no screening exceeded $50,000 per life-year when the probability of overdiagnosis reached 50%. As the analysis employed LYS, it did not account for the effects of overdiagnosis on quality of life.

The cost-effectiveness of low-dose computed tomography (CT) scan screening for lung cancer in high-risk individuals compared favourably with other screening programmes, and is likely to be a practicable and acceptable policy.

The study employed no screening as the comparator, as this was current practice in the authors' setting. You must decided whether the detection of lung cancers by symptoms or signs, or incidentally as a result of investigation for other reasons, represents current practice in your own setting.

The study design was a voluntary, prospective, non-comparative cohort study. The authors acknowledged that such a design is subject to volunteer bias, lead-time bias and the problem of the overdiagnosis of lung cancers that would otherwise have never been detected or would not have been fatal. These biases were addressed in the analysis, but mainly through the use of modelling assumptions. The study sample consisted of volunteers for screening. It is therefore unlikely to have been representative of all persons aged 60 and over with a 10-pack-year history of smoking, no prior history of cancer, and who were suitable for thoracic surgery.

The measure of benefit was the LYS. This was calculated by assuming that some Stage I and II cancers would be cured, so by detecting cancer at an earlier stage screening could improve mortality. Life expectancy was calculated according to published formulae. The authors stated that the primary study was small in size, and a more accurate estimate of benefit could be gained with a larger sample.

All the categories of costs relevant to the perspective adopted were included in the analysis. However, they were limited to the 1 year following the diagnosis of lung cancer. This may have biased the cost results toward screening as, although Stage I and II cancers are cheaper to treat over 1 year than stage III to IV cancers, the greater life expectancy suggests that they will require treatment for a greater number of years. The costs and the quantities were not reported separately. The costs were taken directly from hospital and family practice databases in the authors' settings. This may reduce the generalisability to other settings. The authors acknowledged that extending the screening programme to hospitals without the equipment or facilities to perform CT scans might entail extra costs that were not reflected in the current analysis.

The authors made appropriate comparisons of their findings with those from other cancer screening programmes. They stated that the results were not suitable for generalising to other age or risk groups, and acknowledged that the cost estimates were specific to the hospital used in the study. The authors do not appear to have presented their results selectively and their conclusions were within the scope of their analysis.

The authors recommended that further research be conducted to obtain more accurate estimates of the costs and effects of low-dose CT scan screening for lung cancer in high-risk individuals, but that such a programme appears to be cost-effective based on their analysis.

None stated.

Henschke CI, McCauley DI, Yankelevitz, et al. Early Lung Action Project: overall design and findings from baseline screening. Lancet 1999;354:99-105.

Berkson J, Gage RP. Survival curve for cancer patients following treatment. Journal American Statistical Association 1952;47:501-51.

Beck JR, Kassirer JP, Pauker SG. A convenient approximation of life expectancy (The "DEALE"): I. Validation of the method. Use in medical decision making. American Journal of Medicine 1982;73:883-8.

Beck JR, Pauker SG, Gottlieb JE, et al. A convenient approximation of life expectancy (The "DEALE"): II. Validation of the method. Use in medical decision making. American Journal of Medicine 1982;73:889-97.