| 摘要: | 隨著環境污染問題日益增加與民間之環保意識的抬頭，對於醫療體系所產出之廢棄物已成為敏感的問題，而感染性廢棄物之設計規劃則需較準確的產出量推估模式作為基礎。 本研究之目的，除探討相關因子對感染性廢棄物產出之影響外，並以統計方法分析各項因子與感染性廢棄物產出量之關連性，根據其關連性導引出感染性廢棄物產出量之預估模式，並建立桃園綜合醫院之感染性廢棄物產出量預估模式，供環保單位及國內醫療體系做為統計分析或參考之依據。 本研究所採用之統計分析方法為多元線性迴歸模式及主成份分析，考慮六項因子包括病床數、醫師人數、護理人數、門診人數、住診人數、急診人數等，分析各項自變數對於感染性廢棄物產出量之影響，以建立桃園綜合醫院之感染性廢棄物產出量預估模式。研究結果發現，病床數、醫師人數、護理人數、住診人數等四項自變數，相較於其他因子，其對於感染性廢棄物產出量之影響較為顯著，主成份分析其結果，病床數主要成份是3.001、醫師人數主要成份是1.654、護理人數主要成份是1.015，此主要成份集中在以上這三個因子中，且這三個主要成份累積比率為94%，已達到0.8以上的一個基準。該預測模型為感染性廢棄物的產生，得到： IWs=-70216.891+164.120 BN -1101.206 PN +483.213 NN +0.003 ON -1.470 IN +2.384 EN (1) 預測變數： 病床數(BN)、醫師人數(PN)、護理人數(NN) 門診人數(ON)、住診人數(IN)、急診人數(EN) 本研究開發了預測模型主成份分析和多變量分析。研究結果可能有助於設計和運行的處理設施和監管管理的感染性廢物的管理。 With the increasing environmental concerns on infectious wastes from medical systems, the economical collection, proper treatment, and efficient regulation of such infectious wastes have become an important issue. These tasks, however, all rely upon an accurate prediction for the infectious wastes (IWs) generation. This study tried to develop an predic-tion model for IWs generated from hospitals by applying principal com-ponent analysis and multivariate analysis. The former identified the im-portant factors that contribute to the generation of the IWs; and latter formulate a predict model from the related factors. Infectious wastes may be generated from medical research laboratory, general hospital, and small clinic. The prediction model may be various with the type of medical systems. Therefore, only the IWs generated from typical general hospitals was considered in this study. Principal compo-nent analysis and multivariate analysis were applied on the amounts of IWs generation from five regional general hospitals over a span of 5 years, regarding to six initially selected economical factors. The results indi-cated only four of them, including bed numbers(BN), physician num-bers(PN), nursing numbers(NN), and inpatient numbers(IN) were found significant contributing to the generation of IWs. The amount of informa-tion as shown by the principal component is 3.001, 1.654, and 1.015 for BN, PN, and NN, respectively; and the cumulative information amounts to 94%, a value more than a 0.8 benchmarks. The prediction model for IWs generation was obtained as follows: IWs=-70216.891+164.120 BN -1101.206 PN +483.213 NN +0.003 ON -1.470 IN +2.384 EN (1) Where, IWs=infectious generation (ton/yr) BN=bed number (bed/hospital) PN=physician number (cap/hospital) NN=nursing number (cap/hospital) ON=outpatient number (cap/hospital) IN=inpatient number (cap/hospital) EN=emergency number (cap/hospital) The model was further verified with generation data from a general hospital of of Taoyuan district and showed good prediction capacity. This study developed a prediction model by principal component analysis and multivariate analysis. The results may contribute to the de-sign and operation of the treatment facility and the regulatory manage-ment of the infectious wastes. |
