對於懷疑感染的血液透析患者，比較qSOFA以及SIRS scores 對其預測菌血症的能力

研究目的：血液透析患者免疫功能相較於一般成人不佳，因此一旦感染之後易造成嚴重的併發症甚至死亡，其中菌血症是常見嚴重併發症且造成死亡的原因之一。過往曾有諸多研究，致力於找尋能良好預測菌血症的工具，其中廣為大家所接受的包括了SIRS criteria. 然而對於感染的洗腎患者，卻有研究顯示SIRS criteria未必能在這類族群患者身上呈現相同的預測能力。在2016年Sepsis-3國際共識大會Seymour等學者提出了新的臨床工具qSOFA score，對於非加護病房照護下的感染患者，qSOFA score可有效評估其不良預後。在此我們試圖對於感染的洗腎患者，比較qSOFA score與SIRS criteria對其菌血症的預測能力。 研究方法及資料：本研究為一回朔性研究，收集了中山醫學大學附設醫院自2017年1月1日至2017年12月31日，前來急診就醫的洗腎患者( ICD-10-CM code N18.6 ); 收案條件為在急診檢傷時體溫大於攝氏38度，排除條件則有1. 到院前死亡2. 換腎手術後的患者。分別統計其qSOFA score and SIRS criteria並利用MedCalc 軟體來比較是否qSOFA score 比 SIRS criteria 對於感染的洗腎患者，有更佳預測菌血症的能力。研究結果：我們共收錄了783人次的急診洗腎病患，符合檢傷體溫大於攝氏38度共107人。扣除了1人為到院前死亡以及6人為換腎患者，最後總計100人，其中血液細菌培養呈現陽性結果的共有21人次( 21%)。對於菌血症的預測能力qSOFA (AUC, 0.53; 95% CI, 0.44 to 0.64)的表現略勝過SIRS (AUROC, 0.52; 95% CI, 0.418 to 0.621; P = 0.78)，但統計學上並無差異。 結論與建議：根據我們的研究可以得知雖然qSOFA score(比上SIRS criteria)對於感染洗腎病患擁有較佳的預測菌血症能力，但兩者在統計學上並無差異，且AUROC of qSOFA只有0.54 (95% CI, 0.43 to 0.63)。因此對於感染洗腎患者，我們不建議將qSOFA score拿來當作預測其菌血症的工具。

Objective： Infection is a critical issue for patients with hemodialysis. Bacteremia is also a common and serious complication. The outcome is usually poor and bacteremia accounts for 75% of infection-associated deaths in hemodialysis patients. In Japan and United State, infection is the 2nd cause of death in hemodialysis patients. In Taiwan, infection is the leading cause of death in hemodialysis patients and it accounts for 49.6% of deaths. SIRS criteria was proposed to have a good ability to predict bacteremia. However, the performance is not as good as in hemodialysis patients. In 2016, according to the 3rd International Consensus Definition for Sepsis and Septic Shock Task Force, qSOFA score was introduced by Seymour et al. and it has significant predictive validity for non-ICU encounters mortality. Thus, we try to demonstrate that if qSOFA score is a good predictor for bacteremia in hemodialysis patient with infection compared with SIRS criteria.Methods and Materials：We reviewed and analyzed the electronic medical records of patients with hemodialysis presenting to the emergency department of Chung Shan Medical University Hospital, a tertiary medical center in central Taiwan, between 2017/01/01 and 2017/12/31. Inclusion criteria: HD patients with fever ( >38°C on ED triage). Exclusion criteria: 1. OHCA and 2. Status post renal transplantation surgery. Results：We enrolled 107 HD patients with fever on ED triage. 6 of them had received renal transplantation surgery and 1 of them was OHCA. Total 100 HD patients with suspected infection was analyzed and 18 of them was diagnosed with bacteremia. The AUC of qSOFA score (0.53, 95% CI 0.43-0.63) was slightly greater than the AUC of SIRS criteria (0.51, 95% CI 0.42-0.62). However, the p value is 0.78, insignificantly. Conclusion and Suggestion：qSOFA score is not a good predictor to detect bacteremia in patients with hemodialysis. However, more efforts need to be done in the future because of the poor outcome and high mortality rate.