70例真菌血症的临床特点分析

The clinical characteristics of 70 cases of fungemia

目的评价北京协和医院真菌血流感染患者病原学及临床特点。方法回顾性分析北京协和医院2008至2010年真菌血流感染患者的微生物学和临床资料。统计分析采用SPSSl7．0软件。结果本组70例均存在基础疾病和潜在危险因素的患者确诊为真菌血流感染，40例（57．1％）为单数菌感染，30例（42．9％）住院期间合并细菌性血流感染。从70例患者的血培养中共分离122株病原体，72株真菌，50株细菌；念珠菌61株（84．7％），占绝对优势，白念珠菌31株（50．8％），非白念珠菌30株（49．2％）。分析真菌血症的原发感染来源，35例（50．0％）原发真菌血症，18例（25．7％）同时存在下呼吸道定植，10例（14．3％）引起中心静脉导管相关真菌血症，3例（4．3％）继发于腹腔真菌感染，另有4例（5．7％）存在多个部位相同真菌定植。住院期间死亡37例，粗病死率为52．9％，真菌血症所致的直接病死率为32．9％。单因素分析显示，ICU治疗（χ2=15．136，P〈0．001）、30d内手术史（χ2=3．540，P=0．060）、有创机械通气（χ2=4．450，P=0．035）与住院期间死亡相关；住院期间合并细菌性血流感染（χ2=5．657，P=0．017）、循环系统疾病（χ2=3．399，P=0．065）、ICU治疗（X2=4．955，P=0．026）增加归因病死率。多因素分析中，ICU治疗增加真菌血流感染患者住院期间死亡风险，30d内手术史与住院期间死亡负相关；ICU治疗和住院期间合并细菌性血流感染增加真菌血症患者的归因病死率。结论真菌血症多发生于有一定基础疾病的患者，常伴随细菌血流感染。出现真菌血症的患者预后差，半数以上病人死亡，ICU治疗同时增加真菌血流感染的住院期间死亡和归因死亡风险，住院期间合并细菌性血流感染增加归因病死率，30d内手术史与住院期间死亡负相关。

Objective To evaluate the etiological and clinical characteristics of fungemia in Peking Union Medical College Hospital. Methods Microbial and clinical information of patients with fungemia consulted in Peking Union Medical College Hospital during 2008 to 2010 were retrospectively analyzed. Results A total of 70 patients were diagnosed with fungemia, and 100% of them had underlying diseases or potential risk factors. Of them, 40 （ 57.1% ） patients were monomicrobial fungemia, and the other 30 （42. 9% ） patients with positive blood cultures were caused by at least two different microbes during hospitalization. Among 122 strains of microbes isolated from blood cultures, 72 were fungi and 50 were bacteria. Among the isolated fungi, 61 （84. 7% ） were Candida species, 31 （50. 8% ） were Candida albicans and 30 （49.2%） were non-albicans. According to the colonization or infected sites other than blood of the isolated fungi, 35 cases （50. 0% ） were primary fungemia; 18 （25.7%） were colonizing at lower respiratory tract simultaneously; 10 （ 14. 3% ） caused central-line related fungemia; 3 （4. 3 % ） were secondary to intra- abdominal fungal infection; and another 4 （5.7%） isolates had multiple colonization sites. During hospitalization, 37 cases died with a crude mortality rate of 52.9%, and 22 （32. 9% ） died of fungemia itself. In single factor analysis, ICU hospltalizahon（X2= 15. 136, P 〈 0. 001 ）, operation history within 30 days （X2 =3. 540, P =0. 060） and invasive mechanical ventilation（X2 =4. 450, P =0. 035） were related to crude mortality. Bacteremia during hospitalization （X2 = 5. 657, P = 0. 017）, circulatory underlying diseases（X2 = 3. 399, P = 0. 065） and ICU treatment （X2 = 4. 955, P = 0. 026） increased attributable mortality. In the multivariate analysis, ICU history increased mortality during hospitalization, however, the operation history within 30 days was independently irrelevant to crude mortality during hospitalization. ICU history and bacteremia during hospitalization were independently correlated to attribution mortality of the patients with fungemia. Conclusions Fungemia, usually accompanied with bacteremia, occurs often in the patients with underlying diseases. Patients with fungemia have poor prognosis and more than 50% patients die. ICU history increases the risk both to crude and attributable mortality. The patients with fungemia who had polymicrobial bloodstream infection have a higher attribution mortality. Operation history within 30 days is independently negatively correlated to attributable mortality.