社區發作肺炎病人之抗生素治療分析

A retrospective study of antibiotics treatment for patients with community-onset pneumonia

陳翰霖(Han-Lin Chen)

指導教授 ： 張上淳

國立臺灣大學/醫學院/臨床藥學研究所/碩士(2014年)

https://doi.org/10.6342/NTU.2014.00227

摘要

研究背景：
肺炎是最常見的下呼道感染之一，傳統上可分類為社區型肺炎和院內型肺炎。然而由於醫療照護系統的進步，院外發生的肺炎病人可能因密切接觸或曝露於醫療照護環境中而有多重抗藥性致病菌風險；美國胸腔學會及美國感染症學會於2005年發表的治療指引將這類病人歸類為「醫療照護相關型肺炎（healthcare-associated pneumonia）」，並建議使用2到3種廣效抗生素，而與臺灣的指引有所出入。但近來研究認為醫療照護相關型肺炎病人的異質性相當高且不同國家、地區致病菌種差異相當大，因此有必要了解當地社區型肺炎與醫療照護相關型肺炎基本特性的差異、致病菌種，期待能找出適當的經驗性抗生素，尤其是在人滿為患的醫院急診部。
研究目的：
瞭解臺大醫院急診部之社區發作肺炎之特性、抗生素處方型態、致病菌種分布、治療結果，並探討治療失敗之相關因子，期待能找出臺大醫院急診社區發作肺炎病人適當的初始經驗性抗生素。
研究方法：
本研究為病歷回溯性研究，病人群為2013年1月1日至2013年12月31日至臺大醫院急診部診斷碼為肺炎之隨機抽樣所得之病人。記錄並分析此病人群中的社區型肺炎及醫療照護相關型肺炎的基本特性、醫師處方之經驗性抗生素型態、菌種分布及抗藥性情形等。以治療結束時、第30天及病人出院時之死亡率來分析治療結果。最後用羅吉斯迴歸分析找出第30天死亡之風險因子。
研究結果：
在本研究時段之一年間，臺大醫院急診部共計有2963名符合診斷碼之肺炎病人，隨機抽樣之500名病人中，171名符合納入條件並有完整之資料，其中150名為住院病人。在150名住院病人中，97名屬社區型肺炎，53名屬醫療照護相關型肺炎。醫療照護相關型肺炎病人以曾住院過佔最多數為66.0%，其次為曾於醫療院所接受過靜脈輸注之抗生素佔20.8%。醫療照護相關型肺炎病人比起社區型肺炎病人，來診時意識狀態較為不清、有較高的肺炎嚴重度、共病嚴重度及APACHE Ⅱ score。此外，醫療照護相關型肺炎病人較常有無法行走、管餵等問題、較常使用H2 blockers、PPI制酸劑；過去三個月較常使用過抗生素。但其他方面，無論在年齡、症狀及徵象、影像學上是否涉及肺之雙側等皆無統計學上顯著差異。
社區發作肺炎常以單一初始經驗性抗生素治療，約佔7成。社區型肺炎病人較常單用non-antipseudomonal β-lactams/β-lactamase inhibitors或ceftriaxone；醫療照護相關型肺炎病人則偏好用β-lactams/β-lactamase inhibitors（antipseudomonal β-lactams/β-lactamase inhibitors與non-antipseudomonal β-lactams/β-lactamase inhibitors）。而兩組皆沒有單用macrolides。至於併用抗生素方面，大多是併用macrolides，其中社區型肺炎病人最常與non-antipseudomonal β-lactams/β-lactamase inhibitors併用；醫療照護相關型肺炎病人最常與antipseudomonal β-lactams/β-lactamase inhibitors或non-antipseudomonal β-lactams/β-lactamase inhibitors併用；皆沒有併用aminoglycosides或vancomycin；但在7天內會有1.3%的病人會使用vancomycin。
以入院後一天內之檢體培養結果及血清學檢驗、尿液抗原檢驗結果分析，只有18.8%的致病菌診斷率。K. pneumoniae與P. aeruginosa為主要致病菌；而由於人數較少，不同致病菌種在社區型肺炎組或醫療照護相關型肺炎組之間並沒有統計學上顯著差異，需待未來更大型研究闡明。而抗藥性方面，菌種對於amoxicillin/clavulanate、ampicillin/sulbactam、piperacillin/tazobactam、cefotaxime、levofloxacin及ciprofloxacin的抗藥性比例介於0%至22%之間。多重抗藥性菌種佔致病菌檢驗陽性的社區發作肺炎病人13.3%。以醫療照護相關型肺炎預測多重抗藥性致病菌，其sensitivity為75.0%、specificity 73.0%；positive predictive value 30.0%、negative predictive value 95.0%。
對於社區型肺炎或醫療照護相關型肺炎的治療結果，在排除掉其他感染可能的影響後，住院天數、是否住進加護病房、加護病房天數及併發症方面皆沒有統計學上顯著差異；但在死亡率方面，無論病人出院時之死亡率或入院30天內死亡率皆以醫療照護相關型肺炎組較高。
30天內死亡之風險因子，在排除其他感染下，包含高Charlson comorbidity index以及有呼吸衰竭併發症。
結論：
醫療照護相關型肺炎之病人有較高的肺炎嚴重度、共病嚴重度、較常使用過抗生素。社區發作肺炎常以單一初始經驗性抗生素治療，但由於本研究致病菌診斷率低（18.8%）而無法清楚證實此種處方之合理性。雖然本研究為回溯性研究，K. pneumoniae、P. aeruginosa是最常見菌種且抗藥性並不高。30天內死亡之風險因子，包括高Charlson comorbidity index以及有呼吸衰竭併發症。
關鍵字：
肺炎、急診、社區發作、社區型、醫療照護相關型、經驗性抗生素

關鍵字

肺炎 ； 急診 ； 社區發作 ； 社區型 ； 醫療照護相關型 ； 經驗性抗生素

並列摘要

Background:
Pneumonia is one of the most common lower respiratory tract infections and traditionally can be categorized into community-acquired pneumonia (CAP) and nosocomial pneumonia. However, because healthcare systems have progressed a lot, patients with pneumonia developing outside the hospital were likely to have multidrug-resistant pathogens due to closely contacting or exposuring to healthcare systems and had been recategorized as “healthcare-associated pneumonia (HCAP)” by the American Thoracic Society (ATS) and the Infectious Diseases Society of America (IDSA) since 2005. A recommendation on 2 to 3 broad-spectrum antibiotics combinations for treatment of HCAP patients was made by the ATS/IDSA guidelines. Nevertheless, a recent review regarded HCAP patients as a heterogeneous group and pathogens pattern varied from region/country to region/country. Therefore, to know the characteristics of local patients of CAP and HCAP and pathogen patterns better and to treat CAP and HCAP properly in Taiwan are necessary, especially in the crowded emergency department of a hospital.
Objectives:
The aims of this study were to describe characteristics of patients with community-onset pneumonia, who visited the emergency department (ED) of the National Taiwan University Hospital (NTUH), the antibiotic prescription patterns, the distribution of pathogens, treatment outcomes, and risk factors of treatment failure and to find the appropriate initial empirical antibiotics for the community-onset pneumonia.
Methods:
This is a retrospective and chart-review study, which included randomly sampled patients who visited the ED of the NTUH with pneumonia-related diagnosis from January 1st, 2013 to December 31st, 2013. Information of the patients was collected and analyzed, including characteristics of the community-onset pneumonia, prescription patterns of empirical treatment, the distribution of pathogens and resistance patterns, definite antibiotics and treatment outcomes. D30 mortality was used as the primary endpoint of treatment outcomes. The logistic regression method was used to find the prognostic factors.
Results:
There were 2963 pneumonia patients diagnosed at the ED of NTUH during the one year study period. 500 of them were randomly selected and 171 of them were included in the analysis after excluding those with hospital-acquired pneumonia, those with incomplete data, and those who met other exclusion criteria; 150 needed to be hospitalized. Among 150 patients who needed to be hospitalized, 97 patients were categorized as CAP and 53 patients were HCAP. Most HCAP patients were recent hospitalized, accounted for 66.0%; patients recently receiving IV antibiotics were accounted for 20.8%. HCAP patients were more likely to be admitted with an altered mental status and had higher pneumonia severity index score, comorbidity severity score and APACHE Ⅱ score compared to CAP patients. Also, HCAP patients were more likely to be unable to walk, to have tube feeding, to take H2 blockers or PPI antacids, and to have taken antibiotics within previous 3 months. However, in other aspects, there were no statistically significant differences in ages, signs and symptoms, and bilateral lung involvement by imaging techniques between the two groups of patients.
Monotherapy was preferred to be used as empirical regimen for treatment of community-onset pneumonia, accounted for 70% of the patients. CAP patients were more likely to be treated with non-antipseudomonal β-lactams/β-lactamase inhibitors (39.2%) or ceftriaxone (9.3%); HCAP patients were more likely to be prescribed with β-lactams/β-lactamase inhibitors, both antipseudomonal (41.5%) and non-antipseudomonal (32.1%). No macrolide monotherapy was prescribed for community-onset pneumonia patients who need to be hospitalized. As to combination therapy, non-antipseudomonal β-lactams/β-lactamase inhibitors plus azithromycin was more commonly prescribed for CAP patients (19.6%), and antipseudomonal β-lactams/β-lactamase inhibitors plus azithromycin or non-antipseudomonal β-lactams/β-lactamase inhibitors plus azithromycin was more commonly prescribed for HCAP patients (both 5.7%). No aminoglycosides or vancomycin was used initially. But within 7 days, 1.3% of community-onset pneumonia patients would receive vancomycin treatment.
Specimens for culture sampled within 1 day, serologic tests, and urine or sputum antigen tests were included for pathogens detection for the patients, and the diagnostic rate was only 18.8%. Klebsiella pneumoniae and Pseudomonas aeruginosa were the most common pathogens; no significant difference was found between CAP and HCAP patients, but further studies with larger sample size are required due to small sample size of this study. As to the susceptibility results, the rates of resistance to amoxicillin/clavulanate, ampicillin/sulbactam, piperacillin/tazobactam, cefotaxime, levofloxacin, and ciprofloxacin were around 0% to 22%. Multidrug-resistant pathogens accounted for 13.3% of community-onset pneumonia patients with positive pathogen diagnostic tests. Prediction of multidrug-resistant pathogens by any factor of HCAP had a sensitivity of 75.0%, a specificity of 73.0%; positive predictive value of 30.0% and negative predictive value of 95.0%.
There were no statistically significant differences between CAP and HCAP patients in length of hospitalization, ICU admission, length of ICU hospitalization, and frequency of complications. The mortality rate of the HCAP group was higher than that of the CAP group, for both in-hospital mortality rate and 30-day mortality rate.
By multivariate logistic regression analysis, the risk factors for 30-day mortality included patients who had complication of respiratory failure and high Charlson comorbidity index score when patients with other infections were excluded in the analysis.
Conclusions:
HCAP patients were more likely to have higher pneumonia severity index score and comorbidity severity score and to have taken antibiotics recently. Monotherapy was preferred to be used to treat community-onset pneumonia. The pathogens were detected in very low rate because aggressive diagnostic tests were used in low percentage of patients. K. pneumoniae and P. aeruginosa were the most common pathogens, and the resistance rates to various antibiotics were not high. The mortality rate was higher in the HCAP group. The risk factors for 30-day mortality included patients who had the complication of respiratory failure and high Charlson comorbidity index.
Keywords:
Pneumonia, emergency department, community-onset, community-acquired, healthcare-associated, empirical antibiotic

並列關鍵字

pneumonia ； emergency department ； community-onset ； community-acquired ； healthcare-associated ； empirical antibiotic

參考文獻

3. Mandell LA, Wunderink RG, Anzueto A, et al. Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis 2007;44 Suppl 2:S27-72.

4. Ewig S, Welte T, Chastre J, Torres A. Rethinking the concepts of community-acquired and health-care-associated pneumonia. Lancet Infect Dis 2010;10:279-87.

5. Brito V, Niederman MS. Healthcare-associated pneumonia is a heterogeneous disease, and all patients do not need the same broad-spectrum antibiotic therapy as complex nosocomial pneumonia. Curr Opin Infect Dis 2009;22:316-25.

6. Wunderink RG, Waterer GW. Clinical practice. Community-acquired pneumonia. N Engl J Med 2014;370:543-51.

14. Herrera-Lara S, Fernandez-Fabrellas E, Cervera-Juan A, Blanquer-Olivas R. Do seasonal changes and climate influence the etiology of community acquired pneumonia? Arch Bronconeumol 2013;49:140-5.