|
Family history assessment to detect increased risk for colorectal cancer: conceptual considerations and a preliminary economic analysis |
Ramsey S D, Burke W, Pinsky L, Clarke L, Newcomb P, Khoury M 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 study considered family history assessment at age 40 years to detect increased risk of colorectal cancer, followed by screening colonoscopy at age 40 for those with a suggestive (positive) history. A positive history was defined as those with .one or more first-degree relatives who were diagnosed with colorectal cancer before age 60 years or with two first-degree relatives diagnosed at any age/. The comparator technology was standard population screening at 50 years of age.
Economic study type Cost-effectiveness analysis.
Study population The study population comprised a hypothetical cohort of .people at ,moderate- risk of colorectal cancer/ due to family history.
Setting The setting was the community. The economic study was carried out in the USA.
Dates to which data relate The effectiveness data were taken from studies published between 1979 and 2003. The cost data related to 2001 and 2004. The price year was unclear.
Source of effectiveness data The effectiveness data were derived from a review and synthesis of previous studies, supplemented by modelling assumptions.
Modelling The authors created a decision analytic model to ascertain the clinical and economic impacts to the technology of interest. The model took the proportion of persons undergoing family history assessment into account, as well as whether the results were positive or negative. Positive persons received a colonoscopy at age 40 and negative people received a colonoscopy at age 50. Colonoscopies were then repeated every 10 years. Possible outcomes were colorectal cancer or no colorectal cancer, followed by living or dying.
Outcomes assessed in the review The authors did not carry out a systematic review of the literature. They assessed:
the proportion of persons with a positive family history;
the lead time for polyps;
the proportion of polyps that do not progress;
the risk of perforation with screening colonoscopy;
the relative risk of developing colorectal cancer in a person with family history, aged 40 - 50;
the probability of developing colorectal cancer at age 40 - 50; and
the life expectancy benefit for avoidance of colorectal cancer when aged 40 ; 50.
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 Eight primary studies were included in the review.
Methods of combining primary studies Although some data from the primary studies were combined, the method used was not reported.
Investigation of differences between primary studies Results of the review The proportion of persons with a positive family history was 9% (Range: 8 to 15).
The lead time for polyps was 6.4 years (Range: 4 to 9).
The proportion of polyps that do not progress was 0.95 (Range: 0.8 to 0.99).
The risk of perforation with screening colonoscopy was 1.96 per 1,000 (Range: 0.5 to 2.5).
The relative risk of developing colorectal cancer in a person with family history, aged 40 ; 50, was 2.25 (Range: 1.8 - 2.7).
The probability of developing colorectal cancer at age 40 - 50 was 0.185% (Range: 0.148 to 0.222).
The life expectancy benefit for avoidance of colorectal cancer when aged 40 - 50 was 10.35 years (Range: 5 to 12).
Methods used to derive estimates of effectiveness The decision model was supported by a number of modelling assumptions.
Estimates of effectiveness and key assumptions The model initially assumed that disease progression follows the .polyp age shift/ hypothesis.
There is general consensus that most colorectal cancers develop from adenomatous polyps.
The age shift hypothesis suggests that persons with a family history develop more polyps at an earlier age.
The authors also tested an assumed theory that persons with a family history develop polyps at the same rate as those without but have a greater risk of the polyps transforming into invasive cancer.
The proportion of persons assessed for family history at age 40 was assumed to be 50%.
The proportion of positives receiving colonoscopy (adherence) was assumed to be 75%.
The efficacy of colonoscopy for removing precancerous polyps was assumed to be 100%.
Measure of benefits used in the economic analysis The summary measure of health benefit was the life-years gained (LYG). The LYG were estimated from the review of the literature. The benefits were discounted at a rate of 3%.
Direct costs The costing was carried out from the perspective of the health insurer. As such, it focused on the direct costs associated with assessing family history, screening colonoscopy, treating perforations, polypectomy and pathologic evaluation, and the lifetime costs of treatment for colorectal cancer. The unit costs were based on actual data from published sources and were reported separately from the quantities. The quantities were determined via the decision model. The costs were discounted at a rate of 3%, which is appropriate for the long time horizon of the model. A price year was not reported.
Statistical analysis of costs The authors used a Monte Carlo simulation to propagate uncertainty through their decision model. Probability distributions were assigned to variables and the authors ran the model until the outcomes converged. This analysis was carried out for the 15 variables found to be most influential in the sensitivity analyses.
Indirect Costs The indirect costs were not relevant to the perspective adopted.
Sensitivity analysis One-way sensitivity analyses were carried out on all variables to assess the robustness of the results to variability in the data. The analysis focused on the importance of the discount rate and time horizon for the model. Ranges for the analysis were based on 95% confidence intervals, where available, and expert opinion.
Estimated benefits used in the economic analysis Assuming 22 million people entered the model, screening and polyp removal was reported to add 29,331 years of life (undiscounted) to this group.
Cost results In the study cohort, the total cost of family history screening (including colonoscopy, polypectomy and perforations) would be $890 million per year.
The average cost per precancerous polyp detected was $314,500 (undiscounted).
Synthesis of costs and benefits After discounting, this gives an incremental cost-effectiveness of family history assessment of $58,228 per LYG. The authors also reported a mean incremental cost-effectiveness ratio from a probabilistic sensitivity analysis of $94,428 per LYG. This is dramatically different, and it was unclear whether this was discounted or undiscounted.
The three most influential variables from the sensitivity analysis were life expectancy benefit associated with colorectal cancer avoidance, the cost of family history screening and the cost of colonoscopy.
Authors' conclusions .Family history assessment for colorectal cancer susceptibility is a costly but potentially beneficial intervention./
CRD COMMENTARY - Selection of comparators The authors compared family history assessment at age 40 to no family history assessment. The authors reported that these two alternatives cover the uncertainty in current guidelines over when to start screening. They also said that the alternatives reflected practice in the USA but there were other options available.
Validity of estimate of measure of effectiveness The authors did not carry out a systematic review of the literature. Their review informed a decision model, thus they appear to have selected references that provided data relevant to their model. The authors referenced the sources of their information, but did not discuss them in the text, and reported the inputs into their model thoroughly. Thorough reporting improves the readers' ability to transfer the results to their own setting. Some estimates from primary sources were combined. Further information on how the data were combined and an explanation for possible differences between sources would have been beneficial.
Validity of estimate of measure of benefit The estimation of benefits, LYG, was modelled in a decision analytic model and informed by data from published studies. This is an appropriate way to estimate benefit over the longer time horizon used. It also provides comparability with a wide range of health-related technologies. Future work should try to estimate the impact on quality of life, as the colonoscopy itself may have a shorter run detrimental effect but detection and early treatment may have significant longer run positive impacts.
Validity of estimate of costs Overall, the authors made a detailed report of their cost analysis that was appropriate to the health insurer perspective adopted. The source of the estimates was clear and base-case and sensitivity analysis data were well explained. However, the authors did not report a price year. This limits comparisons with the results of other studies as it would be unclear whether data from the same or different years were being compared.
Other issues The authors did not draw on other literature to make comparisons with similar studies and report whether their findings were consistent with others. However, they had reported a lack of evidence in this area, and it is possible that there was no appropriate existing literature with which to draw comparisons. The issue of generalisability to other settings was discussed, with the authors proposing alternative technologies that might be more suitable in other settings depending on the constraints faced elsewhere.
Several limitations to the analysis were presented. These focused on modelling assumptions, such as uncertainty in the number of persons who would visit a primary care physician at age 40, and whether information about individual risk would influence adherence to screening. The authors provided a range of methods to assess uncertainty in the analysis and, for instance, presented a cost-effectiveness acceptability curve. This analysis in particular might have been discussed in more detail to give the reader a better understanding of the implications. The authors could have provided a better explanation for the difference between the incremental cost-effectiveness ratios reported.
Implications of the study The authors claimed that family history assessment has .feasibility for implementation in clinical practice/. They highlighted the need for further research that better defines the benefits of colonoscopy in 40-year-olds, and clearer delineation of polyp behaviour.
Source of funding Supported by the National Cancer Institute and the National Human Genome Research Institute.
Bibliographic details Ramsey S D, Burke W, Pinsky L, Clarke L, Newcomb P, Khoury M J. Family history assessment to detect increased risk for colorectal cancer: conceptual considerations and a preliminary economic analysis. Cancer Epidemiology, Biomarkers and Prevention 2005; 14(11): 2494-2500 Other publications of related interest Sonnenberg A, Delco F, Inadomi JM. Cost-effectiveness of colonoscopy in screening for colorectal cancer. Ann Intern Med 2000;133:573-84.
Johns LE, Houlston RS. A systematic review and meta-analysis of familial colorectal cancer risk. Am J Gastroenteral 2001;96:2992-3003.
Indexing Status Subject indexing assigned by NLM MeSH Adult; Age Factors; Age of Onset; Colorectal Neoplasms /diagnosis /economics /epidemiology /genetics; Cost-Benefit Analysis; Decision Making; Female; Genetic Predisposition to Disease; Health Policy; Humans; Male; Mass Screening /economics; Middle Aged; Pedigree; Prevalence; Research Support, N.I.H., Extramural; Risk Factors AccessionNumber 22006000018 Date bibliographic record published 30/06/2006 Date abstract record published 30/06/2006 |
|
|
|