臺灣即食生菜沙拉中金黃色葡萄球菌中毒之微生物風險評估

金黃色葡萄球菌 (Staphylococcus aureus) 因具有產生腸毒素的能力且在惡劣環境下仍可以存活，對消費者健康存在潛在的危害風險，是臺灣地區近年來即食生菜沙拉中高風險病原菌。預測微生物學 (predictive microbiology) 是定量微生物風險評估 (quantitative microbial risk assessment, QMRA) 之暴露評估步驟不可或缺的工具，其以數學模式描述整個食物供應鏈過程中微生物消長的狀態。基於考量消費者的飲食習慣，本研究以4、18、25、30和40°C等溫度環境建立即食生菜沙拉中金黃色葡萄球菌之預測生長模式，並使用 RMSE、Af 和Bf 等統計數值作為驗證指標，比較實驗菌數值與預測數值之間的差異；同時以金黃色葡萄球菌最適生長與產毒溫度35°C進行外部驗證，判斷模式的可靠性。此外結合冷鍊物流時間溫度資料、國人膳食數據、微生物檢驗報告以及金黃色葡萄球菌食品中毒閾值105 MPN/g，利用蒙地卡羅模擬法 (Monte-Carlo simulation) 一萬次拉丁方抽樣，模擬即食生菜沙拉在整個供應鏈過程及不同儲存環境下，金黃色葡萄球菌菌數的變化，推估各階段菌量分布和消費者食用後中毒風險，以了解消費者購買臺灣優良農產品標章 (CAS) 驗證和未經驗證 (如早餐店、夜市攤販等) 之即食生菜沙拉產品受金黃色葡萄球菌影響的中毒風險，期望透過風險評估結果，找出食品供應鏈過程之重要管制點 (critical control point, CCP)，提供衛生主管機關和食品業者進行風險管理之科學證據，並助於喜好即食生菜沙拉消費者的風險溝通，確保整個食物供應鏈的食品安全。結果顯示，消費者每年攝入未經驗證的即食生菜沙拉之金黃色葡萄球菌中毒風險是食用 CAS 驗證產品的82倍 (第95百分位數)；影響此風險最重要的因子為消費者飲食習慣 (包括儲存時間和溫度)，未經驗證產品之初始汙染水平 (包括汙染菌量和汙染率)，以及零售店架上陳列之溫度。因此本研究建議政府持續推動 CAS 優良驗證制度，協助即食產業提升；鼓勵業者比照優良驗證系統，加強低溫物流管理；並透過風險溝通強化消費者的風險感知能力，鼓勵民眾購買經政府驗證的 CAS 產品，且於購買後盡快食用完畢，以減少於室溫下放置時間，保障自我飲食的安全。

關鍵字

金黃色葡萄球菌；預測微生物學；定量微生物風險評估；即食生菜沙拉；食品安全

並列摘要

Staphylococcus aureus, which is able to grow and produce toxin in a wide variety of foods, is the second common etiologic agent of foodborne illness in Taiwan. Recently, it was reported to pose the highest risk with the consumption of ready-to-eat (RTE) lettuce salads, especially foods incompliant with the Certified Agricultural Standards (CAS). Predictive modeling, a promising field in food microbiology, can be used to describe the behavior of microorganisms at different temperature conditions. It is an indispensable tool for exposure assessment and used to find critical control point (CCP) in the process by estimating changes in microbial numbers to ensure food safety from farm to human beings. Therefore, the objective of this study was to develop mathematical models for predicting the growth of S. aureus in raw lettuce under different storage temperatures and to perform quantitative microbial risk assessment of S. aureus in consumers in Taiwan. S. aureus isolated from RTE lettuce salads sold in local markets was inoculated onto pre-cut lettuce to achieve an initial inoculation level of 103-104 MPN/g. Based on Taiwanese dietary habit, the inoculated raw lettuce was stored at 4, 18, 25, 30, and 40°C to observe the growth of S. aureus. The populations of S. aureus during storage were determined, and the growth curves were analyzed using the USDA-Integrated Pathogen Modeling Program 2013 (IPMP 2013) to develop predictive models for quantitative microbial risk assessment (QMRA). The models were validated with statistic index and additional samples stored at 35°C. The results showed that No lag phase model, Huang rate model, and Polynomial model were more suitable for describing the growth of S. aureus than other models in IPMP 2013. Additionally, the QMRA with several scenarios estimated that the probability of S.aureus-related foodborne illness for annual intake of CAS certification RTE lettuce salad is 7.05×10-7 per person (95th percentile); however, that of raw lettuce without CAS certification is much higher than former by 82 times with a value of 5.78×10-5 (95th percentile). The regression coefficient values of sensitivity analysis showed that the major factors influencing S. aureus risk of RTE lettuce salads in Taiwan were consumer dietary habit, including home storage time and temperature, initial contamination level of S. aureus, and retail storage temperature. In conclusion, the risk of S. aureus-related foodborne illness through RTE lettuce salads consumption is low in Taiwan, and the developed predictive models in this study could be useful for quantitative microbial risk assessment.