|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 25, No 6 (February 20), 2007: pp. 617-624 © 2007 American Society of Clinical Oncology. DOI: 10.1200/JCO.2006.06.0210 What Is the Most Cost-Effective Population-Based Cancer Screening Program for Chinese Women?
From the Department of Community Medicine and School of Public Health, University of Hong Kong, Pokfulam, Hong Kong, China; and the Department of Health Policy and Management, Harvard School of Public Health, Boston, MA Address reprint requests to Gabriel M. Leung, MD, Department of Community Medicine, 21 Sassoon Rd, Faculty of Medicine Bldg, University of Hong Kong, Pokfulam, Hong Kong, China; e-mail: gmleung{at}hku.hk
PURPOSE: To develop a policy-relevant generalized cost-effectiveness (CE) model of population-based cancer screening for Chinese women. METHODS: Disability-adjusted life-years (DALYs) averted and associated screening and treatment costs under population-based screening using cervical cytology (cervical cancer), mammography (breast cancer), and fecal occult blood testing (FOBT), sigmoidoscopy, FOBT plus sigmoidoscopy, or colonoscopy (colorectal cancer) were estimated, from which average and incremental CE ratios were generated. Probabilistic sensitivity analysis was undertaken to assess stochasticity, parameter uncertainty, and model assumptions. RESULTS: Cervical, breast, and colorectal cancers were together responsible for 13,556 DALYs (in a 1:4:3 ratio, respectively) in Hong Kong's 3.4 million female population annually. All status quo strategies were dominated, thus confirming the suboptimal efficiency of opportunistic screening. Current patterns of screening averted 471 DALYs every year, which could potentially be more than doubled to 1,161 DALYs under the same screening and treatment budgetary threshold of US $50 million with 100% Pap coverage every 4 years and 30% coverage of colonoscopy every 10 years. With higher budgetary caps, biennial mammographic screening starting at age 50 years can be introduced. CONCLUSION: Our findings have informed how best to achieve allocative efficiency in deploying scarce cancer care dollars but must be coupled with better integrated care planning, improved intersectoral coordination, increased resources, and stronger political will to realize the potential health and economic gains as demonstrated.
Most previous studies have focused on the cost effectiveness (CE) of screening for specific cancers independently without due regard for the overall policy coherence of recommendations between different cancers or the total budgetary constraints faced by policymakers. This disconnect between scientific research and the day to day imperatives of policymaking is at least partly responsible for the low uptake of CE results into policy formulation.1 Internationally, among others,2-4 the WHO has begun to address this gap by developing new methods for generalized CE analysis (CEA).5-13 Generalized CEA combines a competing choice analysis (where a number of available, mutually exclusive screening programs for each cancer are compared and the decision between different strategies depends on whether the extra benefit is worth the extra cost evaluated on the margin or incrementally) with a model with a limited budget cap to a range of possible independent health programs, and the objective is to maximize the total net effectiveness or health benefit of the interventions selected, where a set of related interventions is evaluated with respect to the counterfactual of the null set of the related interventions.14 This provides a complete set of information for evaluating both independent and mutually exclusive options to identify the health-maximizing combination of interventions for any given budget.5 Often, cancer screening comes under a separate budget line item in ministries of health, and policymakers look for evidence-based guidance on how best to allocate this money to minimize the population cancer burden overall. Thus, in the Westernized Chinese city of Hong Kong, we developed an integrated, multiple cancer screening generalized CEA model for the three leading causes of female cancer mortality for which some form of effective screening is available.15-17
Specification of Screening Strategies We specified screening strategies to be examined in terms of testing frequency, target age group, and screening modality according to current, best evidence–based recommendations.15-18 From the societal perspective, we considered the following as independent options: annual or biennial mammography for women aged 40 to 74 years or 50 to 74 years; cervical cytology every 3, 4, or 5 years among those aged 25 to 64 years; and annual fecal occult blood testing (FOBT) or sigmoidoscopy every 3 or 5 years or annual FOBT plus sigmoidoscopy every 5 years or colonoscopy every 5, 7, or 10 years in those aged 50 to 74 years. We compared 100% with status quo coverage as mutually exclusive choices within each cancer screening strategy. The comparator was set as the null scenario of no screening.5 All interventions and combinations were assessed assuming they were implemented for 30 years. Costs, therefore, were only incurred over 30 years, but all benefits accruing because of actions taken during this period were included even if they would have been derived beyond the time horizon evaluated. Figure 1 summarizes the decision model structure.
Effectiveness We quantified the potentially avertable cancer burden for different cancers by age-weighted, disability-adjusted life-years (DALYs), which is a composite statistic of health losses through mortality in terms of years of life lost as a result of premature death and morbidity in terms of years lived with disability.19,20 Years of life lost were estimated using the Hong Kong standard life-table21; years lived with disability were calculated based on the incidence perspective, in which the incidence of the disease, average duration of disability, and associated (cancer stage–specific) disability weights were multiplied. Both the duration and disability weights were adapted from the WHO Global Burden of Disease studies for an established market economy region.22 Detailed methods for estimating the reduction in DALYs for each cancer can be found in the Appendix (online only). Base case estimates of clinical variables and associated ranges adopted in the probabilistic sensitivity analysis are listed in Table 1.
Costs We included the following costs: screening, associated travel (round trip by public transportation), and total time (0.5 hour for FOBT, 3 hours for Pap or mammogram, 4 hours for sigmoidoscopy, and 5 hours for colonoscopy) costs; direct medical (including the cost of confirmatory tests and for treating preinvasive lesions), travel, and time costs of managing both true- and false-positive screens (see Appendix for detailed assumptions); and cancer treatment costs derived from the Hospital Authority's Patient-Related Group (PRG) costing models.25 Stage-specific PRG cost categorizations for the three cancers are listed in Table A1 of the Appendix (online only). In calculating time costs, we assumed a 48.4% female labor force participation rate, median monthly personal income of Hong Kong $8,900 for local women,23 and a 44-hour work week. Costs for the various screening tests and procedures were derived from government gazetted fees and charges of local private hospitals (Table 1).24 All costs were converted from Hong Kong dollars (US $1 = Hong Kong $7.8) and expressed in 2001 US dollars. Further details on the specification, derivation, and calculation of costs can be found in the Appendix as well as in Appendix Table A3 (online only).
CE Finally, we carried out a probabilistic sensitivity analysis using second-order Monte Carlo simulation to assess stochasticity and uncertainty in the DALY, cost, and CE ratio estimation. Each input parameter was assigned a probability distribution (Appendix Tables A2 and A3, online only), from which a random value was sampled to generate an outcome value. We used the percentile method to estimate the uncertainty intervals in which the fifth and 95th percentile values were taken from 1,000 Monte Carlo simulations. All analyses were performed using R version 2.1.1 and Microsoft Excel (Microsoft, Redmond, WA).
Table 2 lists the estimated cancer burden under different screening scenarios for Hong Kong Chinese women. As of 2001, cervical, breast, and colorectal cancers together were responsible for about 13,000 DALYs (in a 1:4:3 ratio, respectively) annually in Hong Kong, which has a total female population of approximately 3.4 million. Assuming 100% coverage, triennial Pap screening could potentially save 594 DALYs (range, 412 to 635 DALYs), or 433 more DALYs (range, 300 to 457 DALYs) compared with the status quo, every year. Opportunistic mammographic screening averted 86 DALYs (range, 83 to 92 DALYs) but could potentially reduce a further 389 DALYs (range, 387 to 514 DALYs) if a policy of biennial screening for all women aged 50 to 74 years was implemented. Colonoscopy screening every 7 years was the most effective strategy in terms of DALYs averted, although it, like all other colorectal screening strategies, carried an expected number of iatrogenic deaths associated with endoscopy-related perforations. Of note, colonoscopy screening every 5 years would be expected to generate the largest number of iatrogenic deaths.
Table 3 lists total annualized costs, health effects (in terms of DALYs averted), and average CE ratios for each of the 18 interventions compared with the null scenario of no screening, assuming steady-state screening. The incremental CE ratios and associated 5% and 95% intervals (generated by the probabilistic sensitivity analysis) for those options that were not dominated are also tabulated. All status quo cancer screening strategies were dominated, confirming the suboptimal efficiency of opportunistic screening. Colonoscopy screening every 5, 7, or 10 years dominated all other colorectal cancer screening strategies. Cervical cancer screening at any of the three screening intervals would be more cost effective than mammographic or colorectal screening. This is illustrated by the position of the three Pap strategies in the expansion path diagram (Fig 2). Each cancer-specific expansion path joins the interventions that would be selected for increasing availability of resources. The slopes at different segments in the expansion paths represent the incremental CE ratios as given in Table 3, where the steeper the slope is, the more expensive the intervention per DALY averted.
Finally, we summarized our results in a generalized CE league table (Table 4) where the incremental costs, effects, and CE ratios of the nondominated strategies on the expansion paths for the three cancers are compared. In order of purchase preference, the first strategy entering the table is Pap every 5 years, followed by an incremental increase from Pap screening every 5 years to every 4 years. The third intervention to buy would be colonoscopy every 10 years, followed by mammography for women 50 to 74 years old every 2 years, and so on. The health-maximizing combination for any budget level can be determined by assuming that all screening programs are divisible with proportional costs and effects and any combination of programs is feasible.
Current total (ie, both public and private) annual screening and treatment expenditure in Hong Kong totals about US $50 million, according to the Hospital Authority PRG cost schedule and our own microcosting estimates, including travel and time costs to be consistent with our analytic framework. Therefore, at the current budgetary threshold, Pap every 5 years, Pap every 4 years to Pap every 5 years, and 0.30 unit of colonoscopy every 10 years can be funded, with a maximum effect of 1,161 DALYs averted. If this current budget is doubled, the health-maximizing combination will be Pap every 5 years, Pap every 4 years to Pap every 5 years, and 0.73 unit of colonoscopy every 10 years, with 2,061 DALYs saved. Of note, mammography would not be a CE option unless the budgetary threshold is at least tripled or quadrupled to US $150 to 200 million.
Our results show that the status quo of opportunistic screening for cervical, breast, and colorectal cancer is inefficient and that organized, population-based screening programs for all women in Hong Kong could substantially increase benefits and reduce costs while maintaining the same level of expenditure. Specifically, the current pattern of investment averts 471 DALYs every year, which could potentially be more than doubled to 1,161 DALYs with the same budgetary constraint of US $50 million according to our generalized CEA. For Hong Kong Chinese women, who are perhaps the most Westernized women in all of East Asia, our findings suggest that full liquid Pap screening coverage at every 4 years (notwithstanding the availability and uptake of an effective human papillomavirus vaccine), followed by colonoscopy screening every 10 years and then biennial mammography screening starting at age 50 years, with escalating resource requirements, could achieve optimal allocative efficiency in the deployment of the cancer care budget. By extension, given the geo-ethnic, epidemiologic, and socioeconomic similarities between Hong Kong and the other East Asian countries including Taiwan, Singapore, Korea, and Japan, these conclusions are likely applicable to a much wider regional population. Moreover, as mainland China rapidly transits through socioeconomic development (starting from the eastern seaboard to the inner western provinces) and thus experience the attendant epidemiologic changes over the next 20 years, Hong Kong can be a reliable sentinel harbinger population and provide forward intelligence about how best to prevent cancer through screening as described here. Of the three cancers evaluated in this report, current government screening guidelines18 only list cervical cytology as a routine population-based strategy, whereas the recommendations are equivocal for colorectal screening (citing insufficient evidence), and mammography is contraindicated for the general population. At the present expenditure level, these recommendations are mostly consistent with our generalized CEA, which suggests 100% Pap coverage every 4 years plus 0.3 of a 10-yearly colonoscopy program (or equivalently, 30% coverage of the general population assuming program divisibility, constancy of effect for fractional implementation, and nonbiased coverage by risk category). Status quo colonoscopy coverage as of 2003 to 2004 was approximately 3.3% ever screened, whereas the prevalence was 4.7% for FOBT, sigmoidoscopy, or colonoscopy.27 Mammographic screening was the least CE strategy examined, which is in contrast to recent findings from models based on US women that claimed that approximately half of the mortality reduction observed for breast cancer was a direct result of screening.28 This is unsurprising given the prior probability of disease compared with white women is only approximately half in Hong Kong Chinese women, who already present with one of the highest breast cancer incidence rates in all of East Asia.29,30 This immediately translates into a very low positive predictive value (between 0.018 and 0.134), resulting in a large number of false positives that require further invasive and expensive confirmatory procedures.31 The cost of managing false-positive mammograms amounted to 19% of the estimated total costs of a biennial mammographic screening strategy starting at age 50 years in our microcosting exercise. Given the current lack of consensus about the preferred tool for colorectal cancer screening, our findings give efficiency-based guidance on a population screening strategy. Our model, parameterized using trial efficacy data including both sexes, indicates a preference for colonoscopy over flexible sigmoidoscopy alone or in combination with FOBT. This is consistent with a recent report by Schoenfeld et al32 showing that the majority of cases of advanced neoplasia in women would be missed, disproportionately more than in men, if they underwent flexible sigmoidoscopy alone. Specifically, only 35.2% of women with advanced neoplasia would have had their lesions identified if they had undergone flexible sigmoidoscopy alone compared with 66.3% of matched men from the Veterans Affairs Cooperative Study 380 (P < .001).32 This latest result from the United States lends further support to our CE recommendations for population cancer screening in women. Several caveats should be mentioned. First, we did not incorporate new technologies such as human papilloma virus DNA testing for cervical cancer screening, new human papilloma virus vaccines for cervical cancer prevention, or virtual colonoscopy for colorectal cancer screening in our evaluation exercise. Instead, we focused on quantifying the additional benefits and cost savings that could occur with modest changes to an organized screening program assuming no new technology diffusion. The relative performance of these emerging screening options would be best tested first in randomized trials. Second, we assumed perfect compliance and follow-up for all women, which would be highly unlikely in practice. There could be instances where this may alter the CE ordering within or between cancer-specific screening interventions. For instance in the United States, follow-up after FOBT is substantially less than it is after Pap smear or mammography. Because the cost of FOBT screening strategies is predominantly in the follow-up, an empirical CE calculation for FOBT may well show it to be substantially less effective, but also substantially less expensive, than the idealized case modeled here, thus affecting the internal preference ordering within colorectal screening strategies, which may or may not influence the between-cancer comparisons (where colonoscopy has been modeled to be the most CE). On a related issue, our analysis was intended to broadly inform population-based screening policy and did not comprehensively capture the heterogeneous behavior of clinicians and women. There is evidence suggesting that women at higher risk for cancers may be less likely to get screened, whereas women at lower risk are more likely to be screened.27,33,34 If this were true, we may have underestimated the gains in health effects and economic cost savings if Hong Kong were to shift from the inefficient status quo to organized screening. We did not include administrative costs associated with a population-based screening program or intangible costs arising from clients' psychological burden when dealing with positive screens and so on, and therefore, total costs would have been underestimated. Furthermore, we costed the analysis in 2001 dollars to be consistent with the latest cancer incidence and mortality data available at the time of the study. Given that there was no major change to the relative cost structure or epidemiology in the three cancers examined, the rank order of the buying options would, therefore, not have differed significantly, and our conclusions would hold. Finally, we give a generic note on the potential limitations of this type of generalized CEA. CEA is an inexact science, particularly when data are limited, notwithstanding the probabilistic sensitivity testing of plausible ranges of parameter values that we carried out.7 Randomized controlled trials, supplemented by postimplementation prospective evaluations, must remain the gold standard to definitively address questions of efficacy and effectiveness. Next, the issue of divisibility of interventions remains intensely debated and is as yet unresolved among experts in the field.5 A further challenge of the approach is to sort out technical inefficiencies in the production of a given intervention with questions of allocative efficiency that are the focus of concern.5 Although the interventions covered here will go some way in informing how best to achieve allocative efficiency in deploying scarce cancer care dollars, they alone are insufficient to realize the potential health and economic gains. A coordinated response between public health agencies (which issue clinical practice guidelines), health service providers in the private and public sectors and at all levels of care, and nongovernmental organizations (which are influential advocates to improve knowledge translation and social marketing) is needed. Rather than exclusively concentrating on individual cancers, all of these organizations must set aside vested interests, broaden traditionally narrow service foci, and take a gestalt approach to planning and promoting cancer screening for women. If we were to increase the number of life-years saved through screening, not insubstantial additional resources are required. However, the interventions reflected by the expansion paths are reasonably cost effective, which highlights the importance of overcoming other barriers to averting the most DALYs at the lowest cost overall not only through rethinking our approach to cancer screening but also by strengthening the moral and political will to do so. Finally, broadening the policy perspective, decision makers should consider a very wide basket of health services that crosses not just specific diseases but even disease categories and apply the approach used in this article to rationally allocate the resources available within their populations.
The authors indicated no potential conflicts of data.
Conception and design: Jane J. Kim, Gabriel M. Leung Financial support: Gabriel M. Leung Collection and assembly of data: Pauline P.S. Woo Data analysis and interpretation: Pauline P.S. Woo, Jane J. Kim, Gabriel M. Leung Manuscript writing: Pauline P.S. Woo, Gabriel M. Leung Final approval of manuscript: Pauline P.S. Woo, Jane J. Kim, Gabriel M. Leung
Methodology. Formulas for disability-adjusted life-year (DALY) estimation (Murray CJ, Lopez AD. Bull World Health Organ 72:481-494, 1994; Murray CJ, Lopez AD, Jamison DT. Bull World Health Organ 72:495-509, 1994; Mathers CD, Bernard C, Iburg K, et al. The Global Burden of Disease in 2002: Data Sources, Methods and Results. Geneva, Switzerland, World Health Organization, 2003) are as follows:
Applying age-weighting and discounting,
Estimating DALYs associated with cancer screening.
Similarly for colorectal cancer, we applied the following incidence and mortality reduction rates to compute avertable DALYs as a result of different screening strategies: 20% and 33% from annual fecal occult blood testing (FOBT) screening; 40% and 44% from sigmoidoscopy screening every 5 years; 60% and 70% from annual FOBT plus sigmoidoscopy screening every 5 years; and 75% and 80% from colonoscopy screening every 10 years, respectively (Pignone M, Rich M, Teutsch SM, et al. Ann Intern Med 137:132-141, 2002; Winawer SJ, Zauber AG, Ho MN, et al. N Engl J Med 329:1977-1981, 1993; Winawer SJ, Fletcher RH, Miller L, et al. Gastroenterology 112:594-642, 1997; Frazier AL, Colditz GA, Fuchs CS, et al. JAMA 284:1954-1961, 2000; Anderson WF, Guyton KZ, Hiatt RA, et al. J Natl Cancer Inst 94:1126-1133, 2002; Pignone M, Saha S, Hoerger T, et al. Ann Intern Med 137:96-104, 2002; Walsh JME, Terdiman JP. JAMA 289:1288-1296, 2003; Winawer SJ, Fletcher R, Rex D, et al. Gastroenterology 124:544-560, 2003; Huang CS, Lal SK, Farraye FA. Cancer Causes Control 16:171-188, 2005). These rates are referenced from the screening guidelines recommended by the American Cancer Society for those average-risk adults starting at age 50 years (http://www.cancer.org/docroot/PED/content/PED__X_ACS_Cancer_Detection_Guidelines_36.asp?sitearea=PED). Other intervals of screening were also tested in the model, and their effectiveness is described in Table A4. For breast cancer, we adopted a mortality reduction of 20% in the 50 years or older age group given biennial mammography screening (Leung GM, Lam TH, Thach TQ, et al. Am J Public Health 92:1841-1846, 2002). Annual compared with biennial mammography was associated with a 1.2% absolute increase in survival (Wai ES, D'yachkova Y, Olivotto IA, et al. Br J Cancer 92:961-966, 2005), whereas mortality reduction in the 40 to 49 years age group came out to be smaller than in the greater than 50 years age group (Table A4) (Smith RA, Duffy SW, Gabe R, et al. Radiol Clin North Am 42:793-806, 2004). We assumed that there would be no change in incidence from mammography screening. In the model, screening reduces morbidity by shifting cancer stage at diagnosis earlier. We accounted for this expected change in years lived with disability as a result of screening by adjusting the disability weights based on historical prescreening versus postscreening stage distributions in the United States, by referencing Surveillance, Epidemiology, and End Results data from 1975 to 1979 (prescreening) and 2002 (postscreening) (Surveillance, Epidemiology, and End Results Program public use CD-ROM [1973-2002]. Bethesda, MD, National Cancer Institute, 2005). The current local stage distribution for 2001 to 2004 was adopted in estimating the status quo scenario (Clinical Management System and the Clinical Data Framework. Breast Cancer Stage Distribution 2001-2004. Hong Kong, People's Republic of China, Hospital Authority, Government of the Hong Kong Special Administrative Region, 2004). We also accounted for the expected adverse iatrogenic health disbenefits (specifically deaths), averaged over the 50- to 74-year-old screened age group, associated with colorectal cancer screening. The sigmoidoscopy complication rate was assumed to be three per 10,000 tests over a 5-year period, 2.8% of which would result in death (Winawer SJ, Fletcher RH, Miller L, et al. Gastroenterology 112:594-642, 1997). Colonoscopy was assumed to produce 8.5 complications per 1,000 tests, including 1.3 perforations, 0.2 deaths, and 3.4 and 3.6 major and minor bleeding episodes, respectively, over a 10-year period (Winawer SJ, Fletcher RH, Miller L, et al. Gastroenterology 112:594-642, 1997).
Cost of managing false-positive screens and related complications.
We thank Sue Goldie for her invaluable help in the planning stages of the study, John Howes for facilitating access to Hospital Authority cost data, and Hextan Ngan, Annie Cheung, and Kar-Fai Tam for helpful advice regarding cervical cancer care.
Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Neuman PJ, Rosen AB, Weinstein MC: Medicare and cost-effectiveness analysis. N Engl J Med 353:1516-1522, 2005 2. Oregon Health Services Commission: Oregon Medicaid priority setting project. Portland, OR, Oregon State Government, 1991 3. Jamison DT, Mosley WH, Measham AR, et al: Disease Control Priorities in Developing Countries. New York, NY, Oxford University Press, 1993 4. Tengs TO, Adams ME, Pliskin JS, et al: Five-hundred life-saving interventions and their cost-effectiveness. Risk Anal 15:369-390, 1995[CrossRef][Medline] 5. Murray CJL, Evans DB, Acharya A, et al: Development of WHO guidelines on generalized cost-effectiveness analysis. Health Econ 9:235-251, 2000[CrossRef][Medline] 6. Evans DB, Adam T, Tan-Torres Edejer T, et al: Time to reassess strategies for improving health in developing countries. BMJ 331:1133-1136, 2005 7. Evans DB, Tan-Torres Edejer T, Adam T, et al: Methods to assess the costs and health effects of interventions for improving health in developing countries. BMJ 331:1137-1140, 2005 8. Evans DB, Lim SS, T Adam T, et al: Evaluation of current strategies and future priorities for improving health in developing countries. BMJ 331:1457-1461, 2005 9. Adam T, Lim SS, Mehta S, et al: Achieving the millennium development goals for health: Cost effectiveness analysis of strategies for maternal and neonatal health in developing countries. BMJ 331:1107, 2005 10. Tan-Torres Edejer T, Aikins M, Black R, et al: Achieving the millennium development goals for health: Cost effectiveness analysis of strategies for child health in developing countries. BMJ 331:1177, 2005 11. Morel CM, Lauer JA, Evans DB, et al: Achieving the millennium development goals for health: Cost effectiveness analysis of strategies to combat malaria in developing countries. BMJ 331:1299, 2005 12. Baltussen R, Floyd K, Dye C: Achieving the millennium development goals for health: Cost effectiveness analysis of strategies for tuberculosis control in developing countries. BMJ 331:1364, 2005 13. Hogan DR, Baltussen R, Hayashi C, et al: Achieving the millennium development goals for health: Cost effectiveness analysis of strategies to combat HIV/AIDS in developing countries. BMJ 331:1431-1437, 2005 14. Hunink M, Glasziou P, Siegel J, et al: Decision Making in Health and Medicine: Integrating Evidence and Values. Cambridge, United Kingdom, Cambridge University Press, 2001 15. International Agency for Research on Cancer Working Group on Evaluation of Cervical Cancer Screening Programs: Screening for squamous cervical cancer: Duration of low risk after negative results of cervical cytology and its implication for screening policies. BMJ 293:659-664, 1986[Medline] 16. Humphrey LL, Helfand M, Chan BKS: Breast cancer screening: A summary of the evidence for the US Preventive Services Task Force. Ann Intern Med 137:347-360, 2002 17. Pignone M, Rich M, Teutsch SM, et al: Screening for colorectal cancer in adults at average risk: A summary of the evidence for the US Preventive Services Task Force. Ann Intern Med 137:132-141, 2002 18. Cancer Expert Working Group on Cancer Prevention and Screening.Report of Cancer Expert Working Group on Cancer Prevention and Screening. Hong Kong, People's Republic of China, Department of Health, Government of the Hong Kong Special Administrative Region, 2004 19. Murray CJ, Lopez AD: Quantifying disability: Data, methods and results. Bull World Health Organ 72:481-494, 1994[Medline] 20. Murray CJ, Lopez AD, Jamison DT: The global burden of disease in 1990: Summary results, sensitivity analysis and future directions. Bull World Health Organ 72:495-509, 1994[Medline] 21. Hong Kong Census and Statistics Department: Hong Kong life tables 1996-2031. Hong Kong, People's Republic of China, Government of the Hong Kong Special Administrative Region, Government Printers, 2004 22. Mathers CD, Bernard C, Iburg K, et al: The Global Burden of Disease in 2002: Data Sources, Methods and Results. Geneva, Switzerland, WHO, 2003 23. Hong Kong Census and Statistics Department: 2001 population census: Main report. Hong Kong, People's Republic of China, Government of the Hong Kong Special Administrative Region, Government Printers, 2001 24. Hospital Authority: List of charges: S.S. No. 4 to Gazette No. 44/1996. Hong Kong, People's Republic of China, Government Printers, 1996, 2003 25. Hospital Authority: Patient-related group (PRG) costing for breast and rectum cancers (1998/99), cervical cancer (1999/2000). Hong Kong, People's Republic of China, Government of the Hong Kong Special Administrative Region, 2001 26. Tan Torres-Edejer T, Baltussen R, Adam T, et al: Making Choices in Health: WHO Guide to Cost-Effectiveness Analysis. Geneva, Switzerland, WHO, 2003 27. University of Hong Kong: Population health survey report. commissioned by Department of Health. Hong Kong, People's Republic of China, Government of the Hong Kong Special Administrative Region, 2005 28. Berry DA, Cronin KA, Plevritis SK, et al: Effect of screening and adjuvant therapy on mortality from breast cancer. N Engl J Med 353:1784-1792, 2005 29. Leung GM, Thach TQ, Lam TH, et al: Trends in breast cancer incidence in Hong Kong between 1973 and 1999: An age-period-cohort analysis. Br J Cancer 87:982-988, 2002[CrossRef][Medline] 30. Leung GM, Thach TQ, Chan E, et al: Short-term, medium-term, long-term and lifetime risks of developing and dying of breast cancer in a Westernised Chinese population: Evidence from Hong Kong between 1976 and 2000. Cancer 103:501-508, 2005[CrossRef][Medline] 31. Leung GM, Lam TH, Thach TQ, et al: Will screening mammography in the East do more harm than good? Am J Public Health 92:1841-1846, 2002 32. Schoenfeld P, Cash B, Flood A, et al: Colonoscopic screening of average-risk women for colorectal neoplasia. N Engl J Med 352:2061-2068, 2005 33. Adab P, McGhee SM, Yanova J, et al: The pattern of cervical cancer screening in Hong Kong. Hong Kong, People's Republic of China, Health Services Research Committee/Health Care and Promotion Fund Report, 2001 34. Wong BC, Chan AO, Wong KW, et al: A pilot study of participation in faecal occult blood testing and screening colonoscopy after health education in Hong Kong. Eur J Cancer Prev 14:181-184, 2005[CrossRef][Medline] Submitted February 3, 2006; accepted August 15, 2006. Related Editorial
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2007 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
10 mm, villous adenoma, adenoma with moderate or severe dysplasia, or invasive cancer) was specified according to a study among local Chinese asymptomatic adults older than 50 years (Sung JJ, Chan FKL, Leung WK, et al. Gastroenterology 124:608-614, 2003). These values were used as the upper bound because the study was conducted among adults who had never been screened and, thus, the prevalence of polyps would have been higher than in a steady-state screening setting. This assumption was tested in the probabilistic sensitivity analysis subsequently. We estimated the expected number of diagnostic colonoscopies and related complications generated by false-positive FOBT and FOBT plus sigmoidoscopy screens. The number of perforations as a result of sigmoidoscopy and colonoscopy were estimated by using rates of 0.3 and 1.3 per 1,000 tests, respectively (Winawer SJ, Fletcher RH, Miller L, et al. Gastroenterology 112:594-642, 1997). We also accounted for the costs of colonoscopic surveillance for individuals after removal of high-risk adenomatous polyps by assuming that those with advanced lesions would be followed up with routine colonoscopy every 5 years (Atkin WS, Saunders BP. Gut 51:v6-v9, 2002 [suppl V]). 
