2008年的三聚氰胺事件，為一全球性食品安全風險的重大案例，並引發各界對於其他潛在食品安全風險的廣泛關注。雖然有食品安全標準和相關法規的訂定，食品安全風險管理常見的問題是欠缺作為食品安全風險評估和優先控制的預警指標。
本研究一共提出三個演算法：
一、配適分佈與模擬演算法，估計在對數常態分佈下三聚氰胺左刪失數據濃度的參數。
二、風險指標演算法，在未考量檢驗不確定度下篩選含較高三聚氰胺濃度之食品類別。
三、誤判機率演算法，計算考量檢驗不確定度下食品類別含三聚氰胺但被評為合格的機率。
本研究從香港食物安全中心所公佈三聚氰胺檢測結果，進行了實證分析。並針對多個食品類別所估計之風險指標（RI）和消費者風險指標（CR）進行了比較與討論。同時根據RI和CR指標，提出了危機評估程序，以幫助決策者確定三聚氰胺問題上的危機評估與管理的優先次序。本研究所建立之風險矩陣與危機評估程序可應用於其他化學污染物問題，如塑化劑（鄰苯二甲酸酯鹽類），萊克多巴胺(培林)等。

The melamine incident 2008 was a global food crisis and drew attentions to other potential food safety risks. Although there are regulations and standards for food safety, but one common problem in food risk management is that it lacks on hazard indicators - indicators in ranking of food risk and control.
The three algorithms developed in this article were:
1. A distribution fitting algorithm proposed to estimate population parameters for left-censored melamine data under log-normal assumption.
2. A risk index algorithm proposed to screen out food product categories with higher concentration without considering measurement uncertainty.
3. A misjudgment probability algorithm proposed to calculate the probability that food categories containing melamine more than legal limit but classified satisfactory under consideration of measurement uncertainty.
The test results on melamine collected from the website of the Centre for Food Safety of Hong Kong are empirically analyzed by the proposed algorithms. The risk index (RI) and the consumer’s risk (CR) of multiple food categories are discussed and compared in details. Based on risk index (RI) and consumer’s risk (CR), we build a risk assessment process to help assess melamine risk and make sample strategy in surveillance programme. The proposed risk assessment process can be applied to other chemical contaminant problems such as plasticizer (phthalate esters) and ractopamine (paylean), etc.

誌 謝 i
摘 要 ii
Abstract iii
Contents iv
List of Figures vi
List of Tables viii
Abbreviations used in the dissertation ix
CHAPTER 1 INTRODUCTION 1
1.1 Overview of the melamine incidents 1
1.2 Research motive 5
1.3 Research purpose 6
1.4 Research limit 7
CHAPTER 2 LITERATURE REVIEW 8
2.1 Risk management system 8
2.2 Risk assessment 9
2.3 Risk matrix 9
2.4 New Zealand’s food safety risk management framework 10
2.5 Melamine 10
2.6 Methods for the analysis of melamine 11
CHAPTER 3 RESEARCH DESIGN 14
3.1 Methodology 14
3.2 Research procedure 16
CHAPTER 4 DISTRIBUTION FITTING ALGORITHM 19
4.1 Distribution fitting algorithm 20
4.2 Empirical analysis 24
4.4 Parameter estimated using substitution method 29
4.5 Verification with different log transformation 31
4.6 Findings 32
CHAPTER 5 RISK INDEX ALGORITHM 33
5.1 Methodology 33
5.2 Empirical analysis 35
5.3 Findings 38
CHAPTER 6 MISJUDGEMENT PROBABILITY ALGORITHM 40
6.1 Methodology 43
6.2 Empirical analysis 47
6.3 Findings 52
CHAPTER 7 RISK ASSESSMENT PROCEDURE 54
7.1 Risk assessment procedure 54
7.2 Risk management for food surveillance procedure 59
CHAPTER 8 CONCLUSION 61
8.1 Review of research findings 61
8.2 Future works 63
REFFERENCES 64
APPENDIX A 68
APPENDIX B 72
APPENDIX C 76
SAS programs for analyses of melamine test data 76
SAS lifereg syntax with dataset and upper endpoint taken log 78
SAS lifereg syntax with dataset and upper endpoint taken natural log 79

Bell, S. (1999). A beginner’s guide to uncertainty of measurement, Measurement good practice guide No.11 (Issue 2).Teddington, UK: National Physical Laboratory.
Centre for Food Safety (CFS). (2008a). Melamine tainted milk and milk products. Retrieved November 8, 2010, from
Centre for Food Safety (CFS). (2008b). Notes of the thirteenth meeting of the trade consultation forum. Retrieved November 8, 2010, from http://www.cfs.gov.hk/english/committee/files/Notes%20of%2013th%20trade%20consultation%20forum.pdf
Centre for Food Safety (CFS). (2009). Test results on melamine. Retrieved November 8, 2010, from http://www.cfs.gov.hk/english/whatsnew/whatsnew_fstr/melamine_result.html
Chan M., Lo C.K., Cheng L.S., Cheung T.C. & Wong Y.C. (2009). Evaluation of testing capabilities for the determination of melamine in milk through an inter laboratory proficiency test programme during. Food Additives and Contaminants, 26 (11), 1450-1458.
Chen, J. (2011, Augst). The challenges of effective surveillance in the event of a food incident in China, Regional symposium: From food incidents to crisis management. Presented at APEC Food Safety Cooperation Forum, Hong Kong.
Cook, R. (2008). Simplifying the creation and use of the risk matrix. London: Springer.
Crawford C. (2003). Qualitative risk assessment of the effects of shellfish farming on the environment in Tasmania, Australia. Ocean & Coastal Management, 46, 47-58.
Croghan, C. & Egeghy, P.P. (2003, September). Methods of dealing with values below the limit of detection using SAS. Presented at Southern SAS User Group, St. Petersburg, FL.
Crossley, S. (2010).Benefits and Risks of the Use of Chlorine-containing Disinfectants. Presented at IAFP’s 6th European Symposium on food safety, Dublin, Ireland.
Embarek, P.K.B. (2009, July). Melamine contamination of milk: WHO perspective on food safety issues. Presentation at the meeting of 4th Dubai International Food Safety Conference, Dubai, United Arab Emirates.
Faculty of Sexual and Reproductive Healthcare (FSRH). (2010). Service Standards for Risk Management. Retrieved June 8, 2012, from http://www.fsrh.org/pdfs/SSRiskManagement.pdf
FDA. (2007). Melamine pet food recall of 2007. Retrieved November 8, 2010, from http://www.fda.gov/animalveterinary/safetyhealth/recallswithdrawals/ucm129575.htm
FDA. (2008). Interim method for determination of melamine and cyanuric acid residues in foods using LC-MS/MS: Version 1.0. Retrieved November 8, 2010, from http://www.fda.gov/Food/ScienceResearch/LaboratoryMethods/DrugChemicalResiduesMethodology/ucm071673.htm
Fisher, R. (1984). Information Systems Security. New Jersey, USA: Prentice-Hall.
Food & Health Bureau. (2009). Third Report of the Expert Group on Melamine Incident. Retrieved November 8, 2010, from http://www.fhb.gov.hk/download/press_and_publications/otherinfo/081020_melamine/e_report20090415.pdf
Gillespie, B. W. (2011). Data Analysis Methods when Some Data are Below a Limit of Detection, Center for Statistical Genetics, University of Michigan, Ann Arbor, MI, USA. www.misug.org/uploads/8/1/9/1/8191072/bgillespie_left_censoring.pdf
Goel, S. & Chen, V. (2005). Information security risk analysis - A matrix-based approach. Presented at Information Resource Management Association (IRMA) 16th International Conference, San Diego, CA.
Gossner, C.M., Schlundt, J., Embarek, P.B., Hird, S., Wong, D.L.F., Beltran, J.J.O., et al. (2009). The melamine incident: implications for international food and feed safety [Electronic version]. Environmental Health Perspectives, 117, 1803-1808.
Government Laboratory. (2008). Test methods. Retrieved November 8, 2010, from http://www.govtlab.gov.hk/text/english/development.htm
Hayes K.R. (1997). A review of ecological risk assessment methodologies. CRIMP Technical Report Number 13, CSIRO Division of Marine Research, Hobart, Australia.
Hewett, P. & Ganser, G.H. (2006). A Comparison of Several Methods for Analyzing Censored Data. The Annals of Occupational Hygiene, 51(7), 611-632.
Ho, H.C. and Yang, P. (2007). A note on the sharp ratio for a class generalized stochastic volatility process [Electronic version]. Journal of the Chinese Statistical Association, 45, 340-354.
Ibáñez, M., Sancho, J. V., and Hernández, F. (2009). Determination of melamine in milk-based products and other food and beverage products by ion-pair liquid chromatography-tandem mass spectrometry [Electronic version]. Analytica Chimica Acta, 649, 91–97.
Jin, Y., Hein, M.J., Deddens, J.A. and Hines, C.J. (2010). Analysis of lognormally distributed exposure data with repeated measures and values below the limit of detection using SAS [Electronic version]. Annals of Occupational Hygiene, 55(1), 97-112.
Johnson, N.L. and Kotz, S. (1970). Distributions in Statistics: Continuous Univariate Distributions vol.1. New York: Houghton Mifflin.
Keith R. H. (2001). Biosecurity and the role of risk-assessment, Technical Report No. 17, CSIRO Division of Marine Research, Hobart, Australia. Retrieved November 8, 2010, from ftp://www.cmar.csiro.au/pub/hayes/Ballast%20water%20references/Bio-security%20and%20the%20role%20of%20risk%20assessment.pdf
Lam, W. (2008). Milk scandal sours China's 'soft power', Asia Times. Retrieved November 8, 2010, from http://www.atimes.com/atimes/China/JJ10Ad02.html
Lau, H.Y., Wong, C.S., Ma, J.K.F., Kan, E. and Siu, K.L. (2009). US findings of melamine-related renal disorders in Hong Kong children. Pediatric Radiology, 39(11), 1188-1193.
Lui, S. F. (2008). Melamine tainted milk products-situation in Hong Kong [Electronic version]. Hong Kong Medical Diary, 13(11), 31-32.
Mandel J. (1984). The Statistical Analysis of Experimental Data, NY, USA: Dover Publishers.
Mauer, L.J., Chernyshova, A., Hiatt, A., Deering, A. & Davis, R. (2009). Melamine detection in infant formula powder using near- and mid-infrared spectroscopy[Electronic version]. Journal of Agricultural and Food Chemistry, 57(10), 3974-3980.
Moskowitz, H., Plante, R. and Tsai, H. T. (1991). Single-sided economic screening models incorporating individual unit misclassification error and risk preference. European Journal of Operational Research, 53(2), 228−243.
Moskowitz, H., Plante, R. and Tsai, H. T. (1993). A multistage screening model for evaluation and control of misclassification error in the detection of hypertension. Management Science, 39(3), 307−321.
National Air and Radiation Environmental Laboratory (NAREL). (2009). MARLAP Measurement Uncertainty. Retrieved November 8, 2010, from http://www.epa.gov/radiation/marlap/index.html
New Zealand Food Safety Authority (NZFSA). (2010). New Zealand's food safety risk management framework. Wellington, NZ: Author.
Oak Ridge National Laboratory (ORNL). (2005). Statistical methods and software for the analysis of occupational exposure data with non-detectable values. Retrieved November 8, 2010, from http://www.hss.doe.gov/healthsafety/WSHP/be/docs/statmethods.pdf
Owen, D. B., Mcintire, D. and seymour, E. (1975).Table using one or two screening variables to increase acceptable product under one-sided specification. Journal of Quality Technology, 7(3), 127-138.
Pajek, M., Kubala-Kukus A., Banas, D., Braziewicz, J. and Majewska, U. (2004). Random left-censoring: a statistical approach accounting for detection limits in x-ray fluorescence analysis. X-RAY SPECTROMETRY, 33(4), 306-311.
Pickering, A. and Cowley, S. 2010. Risk Matrices: implied accuracy and false assumptions. Journal Health & Safety Research & Practice, 2(1), 9-16.
Product Quality Research Institute (PQRI). (2011). HAZOP training guide, Product Quality Research Institute. Retrieved November 8, 2010, from http://www.pqri.org/pdfs/mtc/hazop_training_guide.pdf
SAS Institute. (2009). Base SAS® 9.2 Procedures Guide: Statistical Procedures (3rd Ed.). NC, USA: Author.
Sharpe, W.F. (1966). Mutual fund performance. The Journal of Business, 39(1), Part 2: Supplement on Security Prices, 119-138.
Sharpe, W.F. (1994). The Sharpe ratio. Journal of Portfolio Management, 21(1), 49-58.
Shia, J., Mallet, C., Young, M., Li, J., Meng, Y. & Qi, C. (2008). Rapid specific analysis of melamine contamination in infant formula and liquid milk by UPLC/MS/MS, Food Safety Magazine. Retrieved November 8, 2010, from http://www.foodsafetymagazine.com/article.asp?id=2774&sub=sub2#Shia
Standards Association of Australia (SAA). (1999). Risk Management AS/NZS 4360:1999, Strathfield, NSW, Australia: Author.
Tang, K. (1987). Economic design of a one-sided screening procedure using a correlated variable. Technometrics, 29(4), 477−485.
Taylor, H.C., and Russell. J.T. (1939). The relationship of validity coefficients to the practical effectiveness of tests in selection. Journal of Applied Psychology, 23, 565-578.
Tsai H.T. & Moskowitz. H. (1986). A single screening procedure using individual misclassificalion error under one sided specification. Institute Paper No. 896, Krannert Graduate School of Management, Purdue University.
Tsai, H.T. and Yeh, H.C. (1999). A two-stage screening procedure for mailing credit assessment. IMA Journal of Mathematics Applied in Business and Industry, 10(4), 317-329.
Turnbull, B.W. (1976). The empirical distribution function with arbitrarily grouped, censored and truncated data. Journal of the Royal Statistical Society, Series B (Methodological), 38(3), 290-295.
WHO/HSE-FOS. (2008). Melamine: Analytical Methods. Retrieved November 8, 2010, from http://www.who.int/entity/foodsafety/fs_management/Melamine_methods.pdf
World Health Organization. (2008a). Melamine contamination event, China, Retrieved November 8, 2010, from http://www.who.int/foodsafety/fs_management/infosan_events/en/index.html
World Health Organization. (2008b). Expert meeting to review toxicological aspects of melamine and cyanuric acid. Retrieved November 8, 2010, from http://www.who.int/foodsafety/fs_management/Exec_Summary_melamine.pdf
World Health Organization. (2008c). Experts set tolerable level for melamine intake. Retrieved November 8, 2010, from http://www.who.int/mediacentre/news/releases/2008/pr48/en/
World Health Organization. (2008d). Melamine and cyanuric acid: toxicity, preliminary risk assessment and guidance on levels in food. Retrieved November 8, 2010, from http://www.who.int/entity/foodsafety/fs_management/Melamine.pdf