利用某醫院院內感染監測資料探討菌種感染率與死亡之時間趨勢相關分析

Time Trend in Mortality and Attack rate of Nosocomial Infection by specific pathogens :surveillence data between 1994-2003 in SKH Hospital ,Taiwan

沈淑惠(Shu-Hui Shen)

指導教授 ： 陳秀熙

國立臺灣大學/公共衛生學院/預防醫學研究所/碩士(2005年)

https://doi.org/10.6342/NTU.2005.10185

摘要

研究目的 本研究欲分析院內感染之侵襲率、致死率及相關死亡率之時間趨勢，並進一步分析此時間趨勢是否會隨多重感染發生型態、感染部門及感染菌種的不同而有差異。最後，我們希望評估不同感染菌種對住院日到感染發病日之時間間隔的影響。

材料與方法　本研究取材自某醫院自1994年到2003年十年間所有住院病人之入出院紀錄，及其是否院內感染之資料與院內感染者其相關的詳細資料暨死亡資料。排除19,271名因生產住院的個案，總計282,898名住院病人資料納入本研究進行分析。其中5330名病患(8885人次)曾發生院內感染。與死亡檔連結後找到1711名死亡個案(965名在醫院死亡，746名病危出院死亡)，其中1495 (87.4%)人是在感染之後三十天內死亡。進一步將死亡率分解為侵襲率的影響及致死率的影響並評估其時間趨勢，最後利用時間序列布瓦松迴歸模式探討上述三者之間的相關。並比較台灣及美國住院病患從住院日至感染日之間時程的比較。

結果　整體而言，院內感染侵襲率於十年內降低6%，由1994年的3.34%降至2003年的3.14%。然而，院內感染致死率卻增加了55%，由1994年的13.96%增加至2003年的21.6%。因院內感染死亡的死亡率時間趨勢則上升了46%，由1994年的0.46%上升至2003年的0.67%。

結論　利用某醫院十年內住院及院內感染資料，我們採用時間序列布瓦松迴歸分析將院內感染死亡率時間趨勢分解為侵襲率及致死率兩部份的影響，來自這兩部份影響的異質性除了可以推論在院內感染控制的測量上應用多重測量方法外，關於其成因則有待未來研究繼續深入探討。

關鍵字

院內感染 ； 侵襲率 ； 致死率 ； 死亡率 ； 布瓦松迴歸

並列摘要

Objective We aimed to (1) assess time trend in attack rate, case-fatality rate, and mortality rate; (2) assess whether time trend vary multiple episodes, infection site , and pathogen; and to evaluate the interval between admission and onset of infection by pathogen

Methods Data on surveillance of nosocomial infection in SKMH was available electronically. We collected information on discharged patients, episodes of nosocomial Infections, and deaths between 1994 and 2003. After excluding 19,271 deliveries in the contemporaneous period, a total of 282,898 discharged patients were included in the following analysis. Of these patients, we identified 5330 patients with 8885 episodes of NI during 10-year period. After linkage with Taiwan National Mortality Registry, 1,711 deaths were ascertained, including 965 in-hospital deaths and 746 critical against advice discharge (AAD) deaths. Among death cases, 1,495 deaths (87.4%) died within 30 days after onset of NIs. We used a decomposition method to divide mortality related to NI into attack rate and case-fatality rate and assess time trend for these three rates. Time-series Poisson regression model was used to assess the change of three rate after controlling for other significant factors. The comparison of interval between admission and onset of infection between the USA and Taiwan was also made.

Results The overall time trend in attack rate has declined by 6%, decreasing from 3.30% in 1994 to 3.14 in 2003. However, the overall case-fatality rate was elevated by 55%, increasing from 13.96% in 1994 to 21.6% in 2003. Both yielded 46% increase in mortality rate attributed to NI, from 0.46% in 1994 to 0.67% in 2003.pathogens has declined from 1996 but rebounded since 2000. Episodes without culture had a slight increase since 1996. For case-fatality, the majority of pathogens showed an increasing trend. Only coagulase-negative Staphylococcus s and K pneumoniae showed a decreasing trend.
The finding form attack rate together with case-fatality gives the results of mortality rate with an increasing trend for most pathogens. Only coagulase-negative Staphylococcus showed a decreasing trend. The mortality trend for P. aeruginosa, K.pneumoniae, and E. cloacae was fairly constant. For multiple pathogens, time trend in mortality trend was pursuant to that in attack rate with declining from the year of 1996 and resurged from the year of 2000.

By using time-series Poisson regression analysis, after controlling for gender, age, frequency of admission, length of stays, and infection site, an increase (per week) in attack rate was 0.04% for fungus, 0.001% for S.aureus, 0.1% for E.coli, 0.047% for Acinetobacter spp, 0.037% for Enterococcus, and 0.33% for Enterobacter spp. The remaining pathogens showed a decline trend including Pseudomonas aeruginosa , coagulase-negative Staphylococcus, and E.coloacae. Episodes with multiple pathogens has declined from 1996 but rebounded since 2000. The average interval between admission and onset of infection in our study was longer that that in the USA. This is particularly seen for S.aureus, Coagulase-nagative Staphylococcus, fungus, and E.coli.
Conclusions Nosocomial infection with time trend decomposed in to attack rate, case-fatality rate and mortality using data from a large medical center varied with age, pathogen and site of infection. Such heterogeneity of time trend in attack rate and case-fatality rate may imply different measures for control nosocomial infection and need to be clarified in future research.

並列關鍵字

nosocomial infection ； attack rate ； case-fatality rate ； mortality ； Poisson regression

參考文獻

Askarian M, Gooran NR. National nosocomial infection surveillance system-based study in Iran: additional hospital stay attributable to nosocomial infections. Am J Infect Control 2003;31:465-8.

Chen ML, Chen YC, Pan HJ et al. Secular trends in the etiology of nosocomial infection at a teaching hospital in Taiwan. Chinese Journal of Microbiology and Immunology 1995; 28:203-17.

Colpan A, Akinci E, Erbay A, Balaban N, Bodur H. Evaluation of risk factors for mortality in intensive care units: A prospective study from a referral hospital in Turkey. Am J Infect Control 2005;33:42-7.

Emory TG et al. National Nosocomial Infections Surveillance System. Description of surveillance methods. Am J Infect Control, 1991, 19:19–35.

Eggimann and Pittet ,National Nosocomial Infections Surveillance (NNIS) System Report, Data summary from January 1990–May 1999. Issued June 1999. Am J Infect Control, 1999, 27:520–532.