|
Colorectal cancer screening in Australia: an economic evaluation of a potential biennial screening program using faecal occult blood tests |
Stone C A, Carter R C, Vos T, St John J |
|
|
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 cost-effectiveness of a biennial screening programme for colorectal cancer (CRC) using faecal occult blood tests (FOBTs) was examined. All positive FOBTs would be followed by colonoscopy.
Economic study type Cost-effectiveness analysis and cost-utility analysis.
Study population The study comprised the Australian population of 1996 with an average risk of CRC. "Average risk" refers to asymptomatic individuals above the age of 50 years, with no family history of CRC, and no other special factors. It also includes, at a slightly higher risk of up to double the average, individuals with one first-degree relative above the age of 55 years diagnosed with CRC.
Setting The setting was primary care. The economic study was carried out in Australia.
Dates to which data relate The effectiveness evidence was derived from studies published between 1994 and 1996. Epidemiological data used in the study were taken from studies published in 1999. The price year was 1996. The date to which the resource use referred was not reported.
Source of effectiveness data The effectiveness data were derived from a review of published randomised controlled trials (RCTs). National data were used for cancer incidence, mortality, age of onset and mean survival time for each age group. The Cancer Strategies Group (CSG) working party and DJB StJ provided the "expert opinion" when there was no published evidence available and/or when the overseas evidence needed modification for the local context. Data from the South Australian Cancer Registry provided the stage distribution and 5-year survival rates.
Modelling A model was used to determine the years of life lost (YLLs), the disability-adjusted life-years (DALYs) averted and the health service costs associated with the intervention. Separate models were used for each of the Duke's Stages (A to D) of CRC.
Outcomes assessed in the review The outcomes assessed in the review were:
the participation rate,
the FOBT positivity rate/colonoscopy rate,
the number of deaths from colonoscopies,
the proportion of de facto screening colonoscopies, and
the reduction in mortality with the biennial screening programme.
In addition, the distributions of patients in Stages A to D at the time of diagnosis were reported, considering the status quo versus the biennial screening programme.
Study designs and other criteria for inclusion in the review RCTs and epidemiological studies reporting Australian data were included in the review. However, no criteria were reported to have been used to include studies 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 No information was reported on the methods used to judge the relevance and validity of the data extracted. The authors excluded one RCT because a high positivity rate of their FOBT resulted in 28% of the patients having one or more colonoscopies.
Number of primary studies included At least 6 studies were included in the review, of which three were RCTs.
Methods of combining primary studies The authors used a meta-analysis to combine the data from the RCTs.
Investigation of differences between primary studies Results of the review The participation rate with the proposed biennial screening programme was 66% biannual, or 33% per annum (range: 40 - 82).
The FOBT positivity rate/colonoscopy rate was 2% (range: 0.015 - 0.07).
The number of deaths from colonoscopies was estimated at 2/10,000 (range: 1/10,000 - 3/10,000)
The proportion of de facto screening colonoscopies was 15% (range: 0 - 15).
The reduction in mortality with the biennial screening programme, compared with the status quo, was 14% (mean 13, standard deviation 3.6).
The distributions of patients in Stages A to D at the time of diagnosis under the status quo were 0.14, 0.33, 0.31 and 0.21, respectively.
The distributions of patients in Stages A to D at the time of diagnosis under the biennial screening programme were 0.26, 0.28, 0.32 and 0.15, respectively.
Methods used to derive estimates of effectiveness Expert opinion, combined with some data used to populate the model, was used to derive some of the estimates of effectiveness.
Estimates of effectiveness and key assumptions On advice from the CSG Working party, the 5-year survival rates from each stage were increased by 5% to reflect current improved survival rates.
The current rate of opportunistic FOBT screening was assumed to be 2% (range: 0.015 - 0.07).
The proportion of patients requiring follow-up surveillance of polyps was 25% (range: 7 - 30).
Measure of benefits used in the economic analysis The measures of benefits used were the YLL and the DALYs. The DALYs combined YLLs and years lived with disability (YLDs). The YLDs were calculated as the product of the incidence and average duration of the health state, weighted for severity. The number of YLLs was modelled from predicted mortality based on the 5-year survival rate. For each death averted, the YLLs recovered were estimated from the difference between the discounted life expectancy for that age group and the discounted mean survival time for CRC patients in each age group. The health benefits were discounted at a rate of 3%, according to the rate used in the burden of disease studies.
Direct costs All costs were presented for the 1996 reference year, with future costs discounted to present value at a rate of 3% per annum. The costs were estimated according to a life-time horizon, which was longer than 2 years for some of the patients. Therefore, discounting was appropriate. The costs and the quantities were reported separately. Gross costs included infrastructure, FOBT screens, diagnostic work-up, and the cost of complications arising from the diagnostic procedures resulting from the screening programme. Net costs included projected treatment savings, savings from reduced de facto screening by colonoscopy, and the additional expense anticipated from increased follow-up activity. The cost estimates came from a variety of sources and appear to have been based on actual data. The infrastructure costs came from a personal communication with the Commonwealth screening programme. The remaining costs came from various publications. It was unclear whether the quantities came from the same source as the costs or from an alternative source. The costs reported were the average costs.
Statistical analysis of costs Point estimates were used for the costs. No statistical analyses of the costs were reported.
Indirect Costs No indirect costs were included in the economic analysis.
Currency Australian dollars (Aus$).
Sensitivity analysis A multiway probabilistic sensitivity analysis, which used Monte Carlo simulation, was performed to assess the robustness of the base-case results.
The input uncertainty distributions were based on a combination of the reported confidence intervals for the variables, the range of reported variables in the literature, and expert opinion on the range of likely values under Australian conditions. The output distributions reported were the 2.5 and 97.5 percentiles of the 2,000 Monte Carlo simulations, because of their skewed distribution. The software used identified major influential factors, as well as the input uncertainty distributions that had the greatest impact on the results, by regression and correlation of inputs and outputs for each iteration of the simulation.
Estimated benefits used in the economic analysis The base programme (screening 55- to 69-year-olds) would have prevented 250 deaths and avoided 3,200 YLL and 3,200 DALYs due to CRC in 1996. Extension of the screening programme to older age groups (70 to 74 years, or 70 years and over) would have avoided significantly more YLL and DALYs (i.e. 1,100 DALYs and 1,300 DALYs, respectively) than extension to younger age groups (50 to 54 years, or 45 to 54 years; i.e. 730 DALYs and 490 DALYs, respectively).
Cost results The gross annual cost of the base programme in 1996 would have been Aus$55 million, including infrastructure costs of Aus$7.9 million.
The estimated net annual costs associated with the biennial screening programme were Aus$39 million. This included Aus$7.6 million of anticipated treatment savings from less advanced disease, plus Aus$13 million savings from a reduction in de facto screening colonoscopies, offset by Aus$4.3 million of additional costs of follow-up surveillance.
The addition of older age groups would cost significantly less than the addition of younger age groups. More specifically, Aus$13 million of incremental gross costs (Aus$5.7 million of incremental net costs) for extending the screening to those aged 70 to 74 versus Aus$22 million of incremental gross costs (Aus$18 million of incremental net costs) for extending the screening to those aged 50 to 54.
Synthesis of costs and benefits The costs and benefits were combined by estimating the cost per DALY saved. The costs per DALY saved were estimated incremental to the status quo of minimal opportunistic screening.
The baseline estimated health benefits and costs for the status quo were not reported, although they could be derived from the information reported in the paper.
From the perspective of the Australian Government as the third-party founder, the incremental cost-effectiveness ratio (ICER) would be $17,000/DALY gross ($12,000 net). Extending the programme to 70- to 74-year-olds reduced the ICER to $12,000/DALY gross ($5,300 net).
Extending the programme to include 50- to 54-year-olds would increase benefits, but at a substantial cost of $29,000 DALYs gross ($24,000 net).
The results of the Monte Carlo simulation showed that the point estimate for the number of deaths avoided was 250 for the base programme, with lower and upper limits of 99 and 400. The gross costs may be as low as $46 million or as high as $96 million. The corresponding ICER ranged from a low of $13,000/DALY to a high of $52,000/DALY.
The size of the mortality reduction and the screening participation rate were the parameters that introduced most uncertainty into the estimated health benefits. FOBT positivity and the participation rates had the greatest impact on the cost estimates. The major influences on the uncertainty of the ICER were the size of the mortality reduction and the FOBT positivity rate. Other influential parameters were the participation rate and the expected reduction in the number of de facto screening colonoscopies due to the presence of a screening programme.
Authors' conclusions The findings of the study supported the case for a national programme directed at the 55- to 69-year-old age group, with extension to 70- to 74-year-olds if there were sufficient resources.
CRD COMMENTARY - Selection of comparators The reason for the choice of the comparator was clear. The comparator was chosen because it represented the status quo in terms of screening for CRC in Australia. You should decide if the comparator represents current practice in your own setting.
Validity of estimate of measure of effectiveness The authors did not state that a systematic review of the literature had been undertaken. They identified four RCTs concerned with FOBT screening, but only included three of these in the estimate of effectiveness. It was not stated how the studies were identified and selected for inclusion in the review. Therefore, some relevant studies might not have been included, consequently biasing the estimation of the effectiveness parameters. One study was excluded because of a high positivity rate for FOBT. It was unclear why this study found such a high rate (real or artificial), thus it is not possible to objectively asses the impact of its exclusion on the estimate of effectiveness. The three RCTs were combined using a meta-analysis, but there were no details of whether or how the other studies were combined. The CSG working party provided advice on increasing survival rates to reflect improvements in recent years. Estimates of survival were investigated in a sensitivity analysis, using ranges that appear to have been appropriate.
Validity of estimate of measure of benefit The estimation of benefits was modelled. The summary measures of health benefit used in the economic analysis were appropriate, as they would allow the study findings to be compared with those for other interventions.
Validity of estimate of costs All the categories of cost relevant to the perspective adopted appear to have been included in the analysis. The costs and the quantities were reported separately, which would facilitate reflation exercises to other settings. A sensitivity analysis of the quantities was not conducted. However, a probabilistic sensitivity analysis was performed to assess the uncertainty surrounding the study results. Appropriate discounting was performed. Most of the sources used for the cost estimation and the price year were identified.
Other issues The authors presented an extensive description of prior economic evaluations on CRC screening using FOBT. Appropriate comparisons of their findings with those from other studies were made. The authors stated that their study differed from other studies in several respects, the main difference being the age group at which to start screening. In additionally, the study results were compared with those from other studies evaluating different primary prevention programmes. The issue of generalisability to other settings was not discussed. Nevertheless, the cost and efficacy data appear to have been based on Australian evidence, or were amended to reflect the Australian context. This may limit the generalisability of the study results beyond Australia. However, the extensive sensitivity analysis undertaken improves the generalisability of the findings. The authors do not appear to have presented their results selectively. The study involved patients of different age categories and this was reflected in the authors' results. The authors reported an additional limitation of their study. They excluded the costs of future medical care unrelated to CRC, which will occur because of the decrease in premature mortality.
Implications of the study The authors reported that the findings of their study carry a number of policy implications for countries recommending or considering implementing screening programmes for CRC. If less than $50,000/DALY is used as the yardstick for acceptable cost-effectiveness, then a biennial population-screening programme using FOBT provides value for money, particularly if screening starts at the age of 55 rather than the most commonly advocated age of 50.
Source of funding Jointly funded by the Commonwealth Department of Health and Aged Care and the Victorian Department of Human Services, on behalf of the Cancer Strategies Group of the National Health Priorities Committee.
Bibliographic details Stone C A, Carter R C, Vos T, St John J. Colorectal cancer screening in Australia: an economic evaluation of a potential biennial screening program using faecal occult blood tests. Australian and New Zealand Journal of Public Health 2004; 28(3): 273-282 Other publications of related interest Frazier A, Colditz G, Fuchs C, et al. Cost effectiveness of screening for colorectal cancer in the general population. JAMA 2000;284:1954-61.
Sonnenberg A, Delco F, Inadomi JM. Cost-effectiveness of colonoscopy in screening for colorectal cancer. Annals of Internal Medicine 2000;133:573-84.
Vijan S, Hwang EW, Hofer TP, et al. Which colon cancer screening test? A comparison of costs, effectiveness and compliance. American Journal of Medicine 2001;111:593-601.
Indexing Status Subject indexing assigned by NLM MeSH Australia /epidemiology; Colorectal Neoplasms /diagnosis /epidemiology; Humans; Mass Screening /economics /standards /utilization; Occult Blood; Sensitivity and Specificity AccessionNumber 22004008994 Date bibliographic record published 31/01/2006 Date abstract record published 31/01/2006 |
|
|
|