泛抗藥性不動桿菌菌血症病人預後之研究：著重於藥物治療分析

Abstract

目的 探究泛抗藥性不動桿菌菌血症(pandrug-resistant Acinetobacter baumannii bacteremia , PDRAB bacteremia)在不同治療下之流行病學及治療結果，以期找尋最合適之治療策略及預後因子。 設計 從西元二００一年一月至二００四年九月的回溯性研究。 地點 國立台灣大學附設醫院——位於台灣北部的一家教學醫院 對象 在研究期間內發生泛抗藥性不動桿菌菌血症之成人病患 方法 從病歷取得病人的臨床資料，如住院基本資料，發生泛抗藥性不動桿菌菌血症前的潛在疾病、臨床表徵及移生或感染之記錄，發生菌血症當時之細菌學資料及臨床症狀，發生菌血症後之併發症、相關因子及處置，初始治療及確定治療之各個藥物及組合，和治療後第二天、第七天及第三十天之治療反應及結果。 所有病人描述性資料的比較由卡方檢定、費氏精確檢定和T檢定。治療評估及第三十天死亡率、第七天死亡率和第七天反應率之預後分析則由單變相和多變相邏輯式迴歸分析。存活曲線則由Kaplan-Meier method繪製，以Log-rank test比較。 結果 在研究期間內有八十九位病人產生泛抗藥性不動桿菌的血液培養，其中七十七位成人患者納入研究分析。其中，五十六位（佔72.7%）為泛抗藥性不動桿菌單一菌種的菌血症感染而被納入後續的治療與預後分析。 所有研究的菌血症患者都為院內感染，平均Pitt bacteremia score為4.3 ± 2.5，Acute Physiology Score and Chronic Health EvaluationⅡ (APACHE Ⅱ) score為 20.8 ± 9.3，同時有許多合併症。第二天、第七天及第三十天及出院的死亡率分別為24.7%、32.5%、50.7%及62.3%。在單一泛抗藥性不動桿菌菌血症的病人中，二十七位（48.2%）在三十天內死亡。Pitt bacteremia score四分以上(Odds ratio (OR) = 18.53, 95% Confidence Interval (CI) = 3.10-110.72)及免疫抑制狀態(OR = 8.06, 95%CI = 1.87-34.84)為預測三十天死亡率的重要因子。而APACHE Ⅱ score十七分以上(OR = 13.95, 95%CI = 2.45-79.39)及免疫抑制狀態(OR = 11.81, 95%CI = 2.30-60.78)則是預測第七天反應結果的預測因子。而只有免疫抑制狀態(OR = 25.00, 95%CI = 4.79-130.52)為第七天死亡率的預測因子。並沒有任何初始或確定治療在分析中顯著有效。 結論： 病患在發生泛抗藥性不動桿菌菌血症當時的嚴重程度較大者以及處於免疫抑制狀態者（尤其是持續免疫低下者）之第三十天死亡率及第七天反應結果較不樂觀。並且屬於免疫低下者在第七天死亡率也較高。顯示宿主的免疫力在發生此種泛抗藥性伺機性病原體菌血症時扮演很重要的角色；不論使用何種藥物可能都無法影響泛抗藥性不動桿菌菌血症之治癒率。 關鍵詞： 泛抗藥性不動桿菌，菌血症，治療，預後

Objectives This study aims at investigating the epidemiology and treatment outcomes of pandrug-resistant Acinetobacter baumannii (PDRAB) bacteremia with different treatment regimens in order to know the appropriate treatment strategies and prognostic factors for PDRAB bacteremia. Design A retrospective analysis from January 2001 to September 2004 Setting National Taiwan University Hospital (NTUH)--a university teaching hospital in northern Taiwan Population Adult patients with PDRAB bacteremia Methods Clinical data were obtained from medical records, including patients’ profiles, underlying diseases and/or predisposing factors, colonization/infection of bacteria and interventions before the onset of PDRAB bacteremia; microbiological data and clinical manifestations at the onset of PDRAB bacteremia; clinical parameters, colonization/infection of bacteria after the onset of PDRAB bacteremia, initial and definite antibiotic treatment regimens, managements, and treatment response and clinical outcome. The differences were compared by chi-square test, Fisher’s exact test, and t test separately. Treatment evaluation and analyses of prognosis of Day 30 mortality, Day 7 response, and Day 7 mortality were conducted only in monomicrobial PDRAB bacteremia by univariate and multivariate logistic regression analysis. Survival curves shown by Kaplan-Meier method were compared with log-rank test. Results There were eighty-seven patients with PDRAB bacteremia at NTUH during the study period. Seventy-seven of them meet the inclusion criteria were included in this study. Of these 77 patients, fifty-six patients (72.7%) were monomicrobial PDRAB bacteremia and they were analyzed further for treatment response and prognosis. All 77 patients with PDRAB bacteremia were hospital-acquired. They were generally critical (Pitt bacteremia score = 4.3 ± 2.6) and severe (Acute Physiology Score and Chronic Health Evaluation Ⅱ (APACHE Ⅱ) score = 20.8 ± 9.3), and had many comorbid diseases. The crude mortality rate on Day 2, Day 7, Day 30, and at discharge were 24.7%, 32.5%, 50.7%, and 62.3%, respectively. For those 56 patients with monomicrobial PDRAB bacteremia, 27 (48.2%) died within 30 days. Critical patients (Pitt bacteremia score ≧4, odds ratio (OR) = 18.53, 95% confidence interval (CI) = 3.10-110.72), and immune suppression status (OR = 8.06, 95%CI = 1.87-34.84) were the most important predictors for Day 30 mortality. APACHE Ⅱ Score ≧ 17 (OR = 13.95, 95%CI = 2.45-79.39) and immune suppression status (OR = 11.81, 95%CI = 2.30-60.78) were significant factors for prediction of poor response on Day 7. As for Day 7 mortality, immune suppression status (OR = 25.00, 95%CI = 4.79-130.52) was the only significant predictive factor. None of the initial or definite therapy regimens significantly influenced the response and outcome. Conclusion The disease severity of patients at the onset of PDRAB bacteremia and immune suppression status were two significant factors for prediction of Day 30 mortality and Day 7 response. Patients with immune suppression status also might have poor outcome of Day 7 mortality. The immune status of host was the most important factor influencing the treatment response and outcome in patients PDRAB bacteremia. Antibiotic regimens did not influence the outcome. Keywords PDRAB, Bacteremia, Prognosis, Treatment