本研究為一前瞻性調查研究，於台灣南部某醫學中心新生兒加護病房進行使用氧氣噴霧器 (nebulizer) 之保溫箱內音量及突發性噪音事件分佈之評估。以立意取樣方式，於研究期間選擇連續一週的時間，以電子噪音計持續監測保溫箱內之音量。並隨機自調查期間選取其中假日及非假日之白班、小夜及大夜各一班共48小時，在床邊觀察 ?d 65加權分貝（dBA）之突發性噪音事件。結果發現，保溫箱之音量分佈 ?T59 dBA者僅佔所有觀察時間之24.8%，介於60-64dBA者佔58.9%，65-69 dBA者佔10.7%，?d70 dBA者則佔5.6%。而在48小時的觀察中，?d65 dBA之突發噪音共947件，以人為因素佔最多 (79%)。其中開關或碰撞保溫箱佔最多，其次分別為工作人員談話聲音穿透保溫箱、在保溫箱內執行護理活動過程所產生之聲音、紙張或封袋撕開聲、工作人員開關垃圾桶及碰撞不?袗?工作車或器械所造成。而在21%之非人為因素中，以監視器警示聲及保溫箱馬達啟動聲佔最多。此研究結果反應出保溫箱內音量分佈遠超過美國小兒科醫學會（1974）所建議之加護環境噪音應低58 dBA之標準；然而大部分突發性高噪音事件的發生是可經由工作人員之行為修飾來改善，因此此噪音分佈調查結果及突發性噪音來源之確認，可做為改善使用氧氣噴霧器之保溫箱噪音的依據。

　　　　　The purpose of this study was to investigate the noise distribution andsources of peak noise inside an incubator with a nebulizer at a neonatal intensive careunit of a medical center in Southern Taiwan. Sound levels were monitoredcontinuously with an electronic sound-meter for 24 hours daily over a one-weekperiod. Three working hours (day, evening, and night hours) in the weekday andweekend (total 48 hours) were selected randomly from the one-week period of noisesurvey to observe peak noise at levels 65 dBA. Results revealed that 24.8% of thetotal monitoring period had sound levels at 59 dBA, 58.9% at 60-64 dBA, 10.7% at65-69 dBA, and 5.6% at 70 dBA. Furthermore, a total of 947 peak noises 65 dBAwere found within the 48 hours, of which 61.5% were in a range of 65-69 dBA, 24%of 70-74 dBA, 9.8% of 75-79 dBA, and 4.8% 80 dBA. Human-related sources,equaling 79%, were the dominant peak noises. These noises included opening andclosing doors, banging the incubator hood, conversation among staff, nursing activityinside the incubator, tearing and opening paper or bags, opening and closing trash canlids, and bumping metal carts or other apparatus. Nonhuman-related sources were21% including alarms of monitors and running of the incubator motor. Results of thisstudy showed that the noise distribution in the incubator with nebulizer was far abovea protective limitation of 58 dBA, suggested by the American Academy of Pediatricsin 1974. However, most peak noises could be reduced by modification of staffbehavior. Therefore, determinations of noise distribution and sources of peak noise inthis study are useful for further noise reduction programs.