Pattern of sputum bacteriology in acute exacerbations of chronic obstructive pulmonary disease

Background: Chronic obstructive pulmonary disease (COPD) is a major cause of chronic morbidity and mortality worldwide. Acute exacerbation of COPD is redefined as a sustained worsening of a patient′s condition from a stable state (beyond normal day-to-day variations) that is acute in onset and that may warrant additional treatment in a patient with underlying COPD. Aim: This study aimed at searching for a pattern of sputum bacteriology and antibiotic sensitivity for acute exacerbation of COPD in patients admitted to Abbassia Chest Diseases Hospital. Patients and methods: This study included 110 patients who presented with acute exacerbation of COPD. The patients were classified into several groups according to different variables, such as severity, respiratory acidosis, and smoking habits. Bacteriological investigations were performed for all patients including Gram stain examination together with culture and sensitivity testing after proper processing of sputum or endotracheal samples. Results and conclusion: Klebsiella pneumoniae and Acinetobacter spp. were the most common isolates in patients with mild to moderate COPD admitted to the respiratory ICU and to the ward. Each had an incidence of five (15.15%) isolates in the ICU, whereas in the ward there were 13 (14.9%) isolates of Klebsiella spp. and seven (8.04%) isolates of Acinetobacter spp. Acinetobacter spp., however, was the most common isolate in patients with severe to very severe COPD, with an incidence of five (17.9%) isolates. Imipenem was the most sensitive antibiotic in all patient groups in the ICU and ward.


Introduction
Chronic obstructive pulmonary disease (COPD) is a major cause of morbidity and mortality worldwide [1]. COPD is characterized by progressive airfl ow limitation caused by chronic infl ammation of the airways and lung parenchyma [2].
The definition of acute exacerbation of COPD is a major point of criticism in many of the studies dealing with that issue. Recently, acute exacerbation of COPD was redefined as a sustained worsening of a patient's condition from a stable state (beyond normal day-to-day variations) that is acute in onset and that may warrant additional treatment in a patient with underlying COPD. Exacerbations are also associated with considerable physiologic deterioration and increased airway inflammatory changes that are caused by various factors such as viruses, bacteria, and possibly common pollutants [3].
Many studies have been conducted on the role of bacterial infection in COPD and have isolated bacteria in signifi cant numbers from patients with clinically stable COPD, indicating the presence of lower airway bacterial colonization. Th e presence of bacteria in the lower airway can result in a range of important eff ects on the lungs, including activation of host defenses with release of infl ammatory cytokines and subsequent neutrophil recruitment, mucus hypersecretion, impaired mucociliary clearance, and respiratory epithelial cell damage [4].

Aim
Th e study aimed at searching for a pattern of sputum bacteriology and antibiotic sensitivity for acute exacerbation of COPD in patients admitted to Abbassia Chest Diseases Hospital as a representation of the Egyptian population. A correlation was also determined between sputum bacteriology with severity, respiratory acidosis, and smoking pattern.

Patients and methods
Th is study included 110 patients admitted to Abbassia Chest Diseases Hospital who presented with acute exacerbation of COPD between September 2010 and July 2013. Th e patients were classifi ed into several groups according to diff erent variables.
All patients were subjected to the following: (1) Th orough history taking.

Sputum
Th e specimen for culture was collected before antibiotic therapy was initiated. Th e patient was instructed to rinse his or her mouth with water to decrease mouth bacteria and dilute saliva. Patients were instructed to take a deep breath, hold it momentarily, and then cough vigorously into a cup. Specimens were transported to the laboratory within minutes of collection. Sputum was collected in sterile sputum cups. If coughing up sputum was diffi cult, the patient was instructed to breath in a sterile hypertonic saline produced by a nebulizer.

Endotracheal suctioning
Endotracheal aspirates were performed using a sterile catheter. Th e suction tube was blindly introduced through intubation. Th e patient received hyperoxygenation by delivery of 100% oxygen for more than 30 s before the suctioning event. Th e procedure was performed by placement of a suction catheter through the artifi cial airway into the trachea and the application of negative pressure as the catheter was being withdrawn. Th e duration of each suctioning event was ∼10-15 s and the suction pressure was set as low as possible.

Gram stain
A Gram stain of the sputum was examined for polymorphonuclear leukocytes and epithelial cells. Leukocytes and squamous epithelial cells were counted.
Only sputa showing fewer than 10 squamous epithelial cells and more than 25 leukocytes per low-power fi eld (×100) were accepted for culture examination.

Sputum culture
Sputa were cultured on blood agar, MacConkey's medium, and chocolate agar. On the second day fi lms stained by Gram's stain were made from diff erent types of colonies. On the third day, sensitivity was evaluated from the suspected pathological colonies [5]. All these steps were performed inside a biological safety cabinet. Identifi cation of isolated bacteria was carried out through: (1) Microscopic examination.

Statistical analysis
(1) All data were collected, summarized, presented, and analyzed by using an appropriate statistical package for the social sciences program (SPSS, version 10; SPSS Inc., Chicago, Illinois, USA). (2) Quantitative data were summarized as mean and SD. (3) Qualitative data were summarized as number and percentage. (4) Th e test of signifi cance used for qualitative data was t he χ 2 -test.
Th e test of signifi cance used for quantitative data for two groups was the T-test and that for more than two groups was the F-test, whereas the post-hoc test (least signifi cant diff erence) was used for within-group comparisons.
Level of signifi cance P value more than 0.05 was considered nonsignifi cant (NS); P value less than 0.05 was considered signifi cant (S); and P value less than 0.01 was considered highly signifi cant (HS) [6]. Table 1 shows that the patients with AE-COPD included 110 patients: 100 (90.8%) were male and 10 (9.2%) were female. Table 2 shows that the age of these patients ranged from 40 to 78 years, with a mean of 54.88 ± 8.82 years. Table 3 shows that the most prevalent organisms in both the ICU and the ward were Klebsiella pneumoniae and Acinetobacter spp. [fi ve (15.15%) isolates each in the ICU], whereas their incidence in the ward was 13 (14.9%) isolates of Klebsiella spp. and seven (8.04%) isolates of Acinetobacter spp. Although there was no statistically signifi cant diff erence between the incidence of Acinetobacter spp. in the ward and that in the ICU, it was higher in the ICU than in the ward. Th ere was a statistically signifi cant diff erence between the incidence of Enterobacter spp. and Proteus spp. in the ICU and their incidence in the ward, with higher incidence of both in the ward. Table 4 shows that there was a statistically signifi cant diff erence in the sensitivity rates of imipenem, meropenem, tetracycline, vancomycin, kanamycin, cefadroxil, and ciprofl oxacin between the ICU and the ward, with higher sensitivity rates of imipenem, meropenem, tetracycline, and vancomycin in the ICU and higher sensitivity rates of kanamycin, cefadroxil, and ciprofl oxacin in the ward. Th e most sensitive antibiotics in the ICU were imipenem (14 cases, 60.9%), followed by meropenem (10 cases, 43.5%), levofl oxacin (nine cases, 39.1%), doxycycline, and amikacin (eight cases each, 34.8%), and cefotaxime (seven cases, 30.4%).

Results
Th e most sensitive antibiotic in the ward was imipenem (29 cases, 33.3%), followed by levofl oxacin (28 cases, 32.2%) and doxycycline and amikacin (27 cases each, 31%). Table 5 shows that there was a statistically signifi cant diff erence in the incidence of Acinetobacter spp., Pseudomonas spp., and Enterobacter spp. between mild to moderate COPD and severe to very severe COPD, with higher incidence in severe to very severe COPD. Klebsiella spp. is common in both groups [14 (15.22%) isolates in mild to moderate COPD vs. four (14.3%) isolates in severe to very severe COPD]. Pseudomonas aerugenosa 3 (9.1) 6 (6.9) 9 (7.5) 0.582 (NS) The number of isolates in ICU patients is 33, whereas the number of isolates in the ward is 87; S, signifi cant.    Table 6 shows that there was a statistically signifi cant diff erence in the sensitivity rates of imipenem and meropenem among severity groups, with higher sensitivity rates of both antibiotics in severe to very severe COPD than in mild to moderate COPD. Th e most sensitive antibiotic in severe to very severe COPD was imipenem (14 cases, 60.9%), followed by meropenem (10 cases, 43.5%) and levofl oxacin (nine cases, 39.1%). Th e most sensitive antibiotic in mild to moderate COPD was imipenem (29 cases, 33.3%), followed by levofl oxacin (28 cases, 32.2%) and amikacin and doxycycline (27 cases, 31%). Table 7 shows that there was a statistically signifi cant diff erence in the incidence of Acinetobacter infection among pH groups, with higher incidence in acidotic patients than in those without acidosis [six (16.22%) isolates vs. six (7.23%) isolates, respectively]. Klebsiella spp. is common in both groups [fi ve (13.51%) isolates vs. 13 (15.66%) isolates]. Table 8 shows that there was a statistically signifi cant diff erence in the sensitivity rates of imipenem, levofl oxacin, and meropenem among pH groups, with higher sensitivity rates of these antibiotics in acidotic patients than in patients without acidosis. Th e most sensitive antibiotics in patients without acidosis were amikacin and doxycycline (27 cases, 31%), followed by levofl oxacin (26 cases, 29.9%) and imipenem (25 cases, 28.7%). Th e most sensitive antibiotic in patients with acidosis was imipenem (18 cases, 78.3%), followed by levofl oxacin (11 cases, 47.8%) and meropenem (10 cases, 43.5%). Table 9 shows that there was a statistically signifi cant diff erence in the incidences of Klebsiella spp., Acinetobacter spp., Pseudomonas spp., Enterobacter spp., Proteus spp., and Streptococci spp. between ex-smokers and smokers, with a higher incidence of Klebsiella spp., Acinetobacter spp., Pseudomonas spp., and Enterobacter spp. in ex-smokers and a higher incidence of Proteus spp. and Streptococci spp. in smokers. Th e most prevalent organism in ex-smokers was K. pneumoniae (18 isolates, 17.6%), followed by Acinetobacter spp. (12 isolates, 11.8%). Th e most prevalent organism in smokers was Streptococcus pneumoniae (three isolates, 16.6%), followed by Proteus spp. (two isolates, 11.11%). Table 10 shows that there was a statistically signifi cant diff erence in the sensitivity rates of imipenem, levofl oxacin, amikacin, doxycycline, meropenem, and cefotaxime, with higher sensitivity rates of these antibiotics in ex-smokers than in smokers. Th e most sensitive antibiotic in ex-smokers was imipenem (41 cases, 44.1%), followed by levofl oxacin (35 cases, 37.6%) and amikacin and doxycycline (33 cases, 35.5%). Th e most sensitive antibiotics in smokers were imipenem, levofl oxacin, amikacin, doxycycline, meropenem, and cefotaxime (each two cases, 11.8%). Table 11 shows that the most sensitive antibiotic for Pseudomonas spp. was levofl oxacin (nine isolates, 100%); the most sensitive antibiotics for Klebsiella spp. were imipenem and meropenem (each 15 isolates, 83.33%); those for Enterobacter spp. were amikacin and doxycycline (fi ve isolates, 100%); the most sensitive antibiotic for Acinetobacter spp. was doxycycline (eight  isolates, 66.7%); and that for Proteus spp. was imipenem (four isolates, 100%).

Discussion
COPD is a major cause of chronic morbidity and mortality throughout the world. Many people suff er from this disease for years and die prematurely from the disease or from its complications [1].
Th is study was conducted to search for the pattern of sputum bacteriology and antibiotic sensitivity for acute exacerbation of COPD in patients admitted to Abbassia Chest Diseases Hospital as a representation of the Egyptian population between September 2010 and July 2013 in order to correlate sputum bacteriology with severity, respiratory acidosis, and smoking pattern.
In our study, it was found that the most prevalent organisms in both the ICU and the ward were K. pneumoniae and Acinetobacter spp. [fi ve (15.15%) isolates each in the ICU, and 13 (14.9%) isolates of Klebsiella spp. and seven (8.04%) isolates of Acinetobacter spp. in the ward]. Although there was no statistically signifi cant diff erence between the incidence of Acinetobacter spp. in the ward and that in the ICU, it was higher in the ICU than in the ward. Th e most prevalent organism in the whole study was K. pneumoniae (18 isolates, 15%), followed by Acinetobacter spp. (12 isolates, 10%), Pseudomonas aerugenosa (nine isolates, 7.5%), and Enterobacter spp. and Escherichia coli (fi ve isolates each, 4.2%) ( Table 3).
K. pneumoniae was also the predominant organism in a study performed by Cukic [7]. Th ey assessed 75 patients wit h AE-COPD who were treated in the ICU of the Clinic for Pulmonary Disease. In their study 44 (58.66%) patients had normal, nonpathogenic, usual bacterial fl ora isolated in sputum cultures and 31 (41.34%) had pathogenic bacteria in their sputum culture as follows: eight had K. pneumoniae, seven had S. pneumoniae, four had E. coli, and the others had other bacteria.
Th ese results also agree with those of Hui et al. [8], who found that Klebsiella spp., P. aeruginosa, and Acinetobacter spp. constitute a large proportion of pathogens identifi ed in patients wit h AECB. Th ese results also coincide with those of Lin et al. [9], who found that the most prevalent microorganism in the sputum culture of patients with acute exacerbation of COPD was K. pneumoniae (19.6%), followed by P. aeruginosa (16.8%), Haemophilus infl uenzae (7.5%), and Acinetobacter baumannii (6.9%), of Enterobacter spp. In accordance with these results, Li et al. [10] concluded that K. pneumoniae and P. aeruginosa are the most common sputum pathogens in hospitalized patients with AE-COPD.      However, these results disagree with those of Fagon et al. [11], who found that the most prevalent microorganism in COPD patients was H. infl uenzae (39%), followed by S. pneumoniae (16%) and Moraxella catarrhalis (7%). Th is disagreement may be due to the diff erence in environment, timing of the study, number of cases, and the method of sample collection, such as bronchoalveolar lavage and use of a protective brush.
Th ese results also disagree with those of Monsó et al. [12], who found that the most prevalent microorganism was H. infl uenzae (58%), followed by M. catarrhalis and S. pneumoniae (each 10%).
As regards severity in relation to organisms, it was found that there was a statistically signifi cant diff erence in the incidence of Acinetobacter spp., Pseudomonas spp., and Enterobacter spp. between mild to moderate COPD and severe to very severe COPD, with a higher incidence of these organisms in severe to very severe COPD compared with mild to moderate COPD. Klebsiella spp. is common in both groups [14 (15.22%) isolates in mild to moderate COPD vs. four (14.3%) isolates in severe to very severe COPD] ( Table 5).
Lin et al. [9] and Li et al. [10] observed that K. pneumoniae was more frequently isolated in stage I COPD than in stages II , III, and IV.
Our results agree with those of Li et al. [10], Miravittles et al. [13], and Brunton et al. [14], who concluded that P. aeruginosa was associated with poor clinical outcome. In addition, Noweta et al. [15] found that the most prevalent microorganism in acute exacerbation of severe and very severe COPD was A. baumannii (21%).
However, these results disagree with those of Lior et al. [16], who found in 468 patients with moderate COPD that the most prevalent microorganism was S. pneumoniae (34.8%), followed by M. catarrhalis (23.9%) and H. infl uenzae (12.6%). Th is disagreement may be due to the large diff erence in the number of cases.
Th e previous results disagree with those of Lode et al. [17], Rosell et al. [18], and Miravittles et al. [13], who found that the most prevalent microorganism in acute exacerbation of severe COPD was H. infl uenzae, followed by S. pneumoniae and P. aeruginosa. Th e disparity may be due to the diff erence in environment.
As regards the presence or absence of respiratory acidosis, it was found that there was statistically signifi cant diff erence in the incidence of Acinetobacter spp. among pH groups, with a higher incidence in acidotic pH than in patients without acidosis [six (16.22%) isolates vs. six (7.23%) isolates, respectively]. Klebsiella spp. is common in both groups [fi ve (13.51%) isolates vs. 13 (15.66%) isolates, respectively] and the most prevalent organism in patients without acidosis was Klebsiella spp. (13 isolates, 15.66%), followed by Acinetobacter spp. (six isolates, 7.23%) and P. aerugenosa (fi ve isolates, 6.02%). Th e most prevalent organism in patients with acidotic pH was Acinetobacter spp. (six isolates, 16.22%), followed by Klebsiella spp. (fi ve isolates, 13.51%) ( Table 7). Hypercapnia, an elevation of the level o f CO 2 in blood and tissues, is a marker of poor prognosis in COPD and other pulmonary disorders. Hypercapnia inhibits the expression of tumor necrosis factor an d interleukin 6 and phagocytosis in macrophages in vitro [19].
As regards smoking habits, it was found that there was statistically signifi cant diff erence in the incidence of Klebsiella spp., Acinetobacter spp., Pseudomonas spp., Enterobacter spp., Proteus spp., and Streptococci spp. between ex-smokers and smokers, with higher incidence of Klebsiella spp., Acinetobacter spp., Pseudomonas spp., Enterobacter spp. in ex-smokers and higher incidence of Proteus spp. and Streptococci spp. in smokers. Th e most prevalent organism in ex-smokers was K. pneumoniae (18 isolates, 17.6%). Th e most prevalent organism in smokers was Streptococci spp. (three isolates, 16.6%), followed by Proteus spp. (two isolates, 11.11%) ( Table 9).
Th ese results agree with those of Monsó et al. [20], who found that excessive smoking and duration of smoking are associated with progressive deterioration in lung function and associated with infection with P. aerugenosa and other Gram-negative virulent strains in patients with acute exacerbation of COPD.
Destache et al [21] found that the effi cacy of trimethoprim-sulfamethoxazole, tetracycline, and Erythromycin was 81%, whereas the effi cacy of azithromycin, ciprofl oxacin, and amoxicillin-clavulinic acid was 93%. Th ese fi ndings disagree with the results of this study, in which the sensitivity rate was 19.1% for trimethoprim-sulfamethoxazole, 5.5% for amoxicillinclavulinic acid, 3.6% for each of erythromycin and ciprofl oxacin, and 2.75% for each of azithromycin and tetracycline.
Wilson et al. [22] found that the rate of bacterial eradication after treatment with amoxicillin-clavulinic acid was 76.7%, that after treatment with levofl oxacin was 96.3%, and that after treatment with azithromycin was 87.4%. Th ese fi gures mismatch with the ours, in which the sensitivity rate was 33.6% for levofl oxacin, 5.5% for amoxicillin-clavulinic acid, and 2.75% for azithromycin.
Erkan et al. [23] noted the poor effi cacy of penicillin, ampicillin, amoxicillin-clavulinic acid, tetracycline, and Erythromycin against most prevalent respiratory pathogens in acute exacerbation of COPD. Th eir results agree with the low sensitivity rates of these antibiotics in this study (5.5% for amoxicillin-clavulinic acid, 3.6% for erythromycin, 2.75% for tetracycline, 1.8% for penicillin, and 0.9% for ampicillin).
As regards the sensitivity rates of antibiotics in relation to severity, it was found that there was statistically signifi cant diff erence in the sensitivity rates of imipenem and meropenem among severity groups, with higher sensitivity rates of both antibiotics in severe to very severe COPD than in mild to moderate COPD. Th e most sensitive antibiotic in severe to very severe COPD was imipenem (14 cases, 60.9%), followed by meropenem (10 cases, 43.5%), levofl oxacin (nine cases, 39.1%), amikacin (eight cases, 34.8%), and cefotaxime (seven cases, 30.4%). Th e most sensitive antibiotic in mild to moderate COPD was imipenem (29 cases, 33.3%), followed by levofl oxacin (28 cases, 32.2%) and amikacin and doxycycline (27 cases, 31%) ( Table 6).
Fein and Fein [24] recommended doxycycline, levofl oxacin, and other drugs as a treatment strategy for mild acute exacerbation of COPD and recommended cefotaxime, levofl oxacin, and other drugs for severe acute exacerbation of COPD. Th is agrees with the previously mentioned susceptibility rates in our study .
GOLD guidelines [1] recommended β-lactam and other drugs as a treatment strategy for mild and moderate acute exacerbation of COPD and recommended imipenem, meropenem, and high dose of levofl oxacin for severe acute exacerbation of COPD. Th is agrees with the previously mentioned susceptibility rates in our study.
As regards the sensitivity rates of antibiotics in relation to pH, it was found that there was statistically signifi cant diff erence in sensitivity rates of imipenem, Comparison of sensitivity rates of highly effective antibiotics in relation to most prevalent organisms . levofl oxacin, and meropenem among pH groups, with higher sensitivity of these antibiotics in patients with acidotic pH than in those without acidosis. Th e most sensitive antibiotics in patients without acidosis were amikacin and doxycycline (27 cases, 31%). Th e most sensitive antibiotic in patients with acidotic pH was imipenem (18 cases, 78.3%), followed by levofl oxacin (11 cases, 47.8%) ( Table 8).
As regards the sensitivity rates of antibiotics in relation to smoking, it was found that there was statistically signifi cant diff erence in sensitivity rates of imipenem, levofl oxacin, amikacin, doxycycline, meropenem, and cefotaxime, with higher sensitivity rates of these antibiotics in ex-smokers than in smokers. Th e most sensitive antibiotic in ex-smokers was imipenem (41 cases, 44.1%), followed by levofl oxacin (35 cases, 37.6%). Th e most sensitive antibiotics in smokers were imipenem, levofl oxacin, amikacin, doxycycline, meropenem, and cefotaxime (two cases each, 11.8%) ( Table 10). To our knowledge, there are no studies with results comparable to our results.