Usefulness of Gram staining of tracheal aspirates in initial therapy for ventilator-associated pneumonia in extremely preterm neonates

Transcription

1 (2010) 30, r 2010 Nature Publishing Group All rights reserved /10 $32 ORIGINAL ARTICLE Usefulness of Gram staining of tracheal aspirates in initial therapy for ventilator-associated pneumonia in extremely preterm neonates Y Katayama, H Minami, M Enomoto, T Takano, S Hayashi and YK Lee Division of Neonatology, Department of Pediatrics, Takatsuki General Hospital, Takatsuki City, Osaka, Japan Objective: To determine the effectiveness of Gram staining of tracheal aspirates for predicting causative microorganisms and guiding appropriate initial antibiotic therapy to treat ventilator-associated pneumonia (VAP) in extremely preterm neonates. Study Design: A prospective study was conducted on all episodes of presumed VAP in neonates with a gestational age below 28 weeks at birth. Result: Gram-positive cocci and Gram-negative bacilli were the causative microorganisms in 11 and 38 confirmed episodes of VAP, respectively. The sensitivity and specificity of Gram staining were 82 and 100% for Grampositive VAP and 100 and 82% for Gram-negative VAP, respectively. The initial antibiotic therapies based on Gram-stain findings were appropriate for 96% (47/49) of patients with VAP. Conclusion: Gram staining of tracheal aspirates is useful in predicting classes of causative microorganisms and for guiding appropriate initial antibiotic therapy for VAP in extremely preterm neonates. (2010) 30, ; doi: /jp ; published online 1 October 2009 Keywords: preterm neonate; bacterial pathogens; nosocomial infection Introduction Ventilator-associated pneumonia (VAP) is a common hospitalacquired infection among neonatal intensive care unit (NICU) patients. 1,2 The rates of VAP are high, especially in extremely preterm neonates (<28 weeks). 3 However, no gold standard exists for defining or diagnosing VAP in these infants 4 and consensus guidelines for empirical antibiotic therapy to treat presumptive VAP have not been published. Quantitative bacteriological cultures, microscopic identification of polymorphonuclear cells containing intracellular bacteria, and Gram staining of bronchoalveolar lavage or protected specimen brushes have relatively high Correspondence: Dr Y Katayama, Division of Neonatology, Department of Pediatrics, Takatsuki General Hospital, Kosobe-cho, Takatsuki City, Osaka , Japan. yoshinoli@nifty.com Received 5 July 2009; revised 20 August 2009; accepted 23 August 2009; published online 1 October 2009 sensitivity and specificity for a diagnosis of VAP 5,6 in older children. However, these procedures might be too invasive for extremely preterm neonates. Tracheal aspirate cultures that can be collected less invasively than bronchoscopy or non-bronchoscopic deep pulmonary aspiration closely correlate with protected specimen brushes cultures from adult patients with clinical pneumonia who are on long-term mechanical ventilation. 7,8 Thus, we postulated that Gram-stained pulmonary specimens obtained from tracheal aspiration might represent a reliable method of arriving at an etiological diagnosis of VAP and contribute to the appropriate selection of initial therapeutic strategies for neonates. To our knowledge, the value of Gram staining of tracheal aspirates to guide etiological diagnoses and initial antibiotic therapy for VAP in extremely preterm neonates has not been reported. The anatomy and pathophysiology of premature neonates differ from those of adults and older children, and thus diagnosis and treatment for VAP should be specifically studied in such infants. We conducted a prospective study to assess the usefulness of Gram staining of tracheal aspirates for the rapid estimation of bacterial pathogens and to select the appropriate initial antibiotic therapy for VAP in extremely preterm neonates. Methods Setting This study was conducted at Takatsuki General Hospital (Osaka, Japan), which has a level 3, 51-bed NICU that annually admits 500 patients. We prospectively studied all extremely preterm neonates with clinically suspected VAP that were admitted to the NICU for >48 h between April 2007 and December The study was approved by the Ethics Committee of the hospital. Informed consent was obtained from parents at the time when an attending physician indicated the occurrence of VAP. Sputum studies All neonatal physicians at our NICU were trained in Gram staining by medical technologists at our hospital before starting this study.

2 271 Nurses directly collected sputum specimens by endotracheal aspiration into sterile specimen containers (Nippon Sherwood Medical Industries Ltd., Tokyo, Japan) using a closed aspiration system (Trach Care; Ballard Medical Products, Draper, UT, USA). Purulent portions of sputum specimens were immediately Gram stained and examined by a neonatal physician. The presence of few squamous epithelial cells (<10 in a 100 microscopic field) was confirmed to avoid contamination with upper respiratory or oropharyngeal secretions. Smears were screened for predominant bacterial morphological types in an oil immersion field ( 1000 microscopic field). Pathogenic bacteria were estimated based on the results of the Gram stain (Gram positive or Gram negative) and on morphology. Sputum cultures were also immediately processed on blood agar, chocolate agar, Columbia agar, and bromothymol blue lactose agar media. Definitions VAP is defined as nosocomial pneumonia in patients that develops >48 h after starting mechanical ventilation. Our patients were diagnosed with VAP when ventilator demand increased with increasing amounts of endotracheal aspirates and when microorganisms and polymorphonuclear leukocytes were identified in Gram-stained smears of aspirates. Increased serum C-reactive protein (CRP), intracellular bacteria on Gram-stained smears, or both confirmed a diagnosis of VAP. The etiological diagnosis of VAP was considered as definitive when any microorganism was isolated from tracheal aspirates at concentrations of X10 3 cfu ml 1. 9 Antibiotic therapy Antibiotic therapy was administered after assessing the Gram-stain results. Monotherapy with piperacillin (PIPC) or a combination of PIPC and amikacin comprised the initial antibiotic therapy for VAP with Gram-negative bacilli (GNB), and vancomycin was selected for VAP with Gram-positive cocci (GPC). When staining revealed both Gram-positive and Gram-negative microorganisms, antibiotic therapy was targeted to the more predominant microorganism (vancomycin for GPC>GNP, PIPC, or PIPC and amikacin for GNP >GPC). Statistical methods Correlations between the results of the Gram-stained tracheal sputum smears and cultures were evaluated by analyzing the sensitivity, specificity, positive, and negative predictive values of the findings considering the causative culture results as reference standards. Results Over the study period, 50 episodes of presumed VAP occurred in 36 patients. One, two, and seven patients had and four, three, and two episodes of VAP, respectively, which fulfilled the two major criteria described. Serum CRP was increased in 35 (74%) of 47 episodes of VAP. CRP was not tested in three episodes of VAP. Intracellular bacteria on Gram-stained smears were detected in 30 (60%) of 50 episodes of VAP. One episode of presumed VAP, without an accompanying increase in serum CRP and no intracellular bacteria on Gram stain, could not be confirmed, as no microorganisms were isolated from endotracheal aspirates and antibiotic therapy did not affect the respiratory condition. Finally, 49 episodes in 35 patients were confirmed as VAP. The mean (s.d.) gestational age and birth weight of the patients were 25 (1.6) weeks and 656 (155) g, respectively. The mean (s.d.) duration from starting mechanical ventilation to the onset of VAP was 32 (19) days. Microbiology of VAP Table 1 summarizes the microbiology of VAP and indicates that Pseudomonas aeruginosa was the most frequently isolated (18 of 49 episodes of VAP; 37%). Klebsiella pneumoniae and Serratia marcescens were both isolated from 7 (14%) of 49 VAP episodes, respectively, and Gram-positive bacteria accounted for 10 (22%) of 49. Correlation analysis Table 2 shows correlations between Gram-stain results and final causative organism findings. The causative microorganism of all eight episodes with GPC only in Gram-stained sputum was GPC. Of the 37 episodes with GNB only in Gram-stained sputum, GNB was the causative-cultured organism in 36 (97%). Gram staining revealed both Gram-positive and Gram-negative microorganisms in four episodes of VAP (GPC>GNB, 1; GNB>GPC, 3). The one episode with GPC>GNB on Gram staining had Gram-positive VAP. Table 1 Microbiology of isolated organisms Organisms No. (%) of cases Gram-positive cocci Methicillin-resistant Staphylococcus aureus 5 (10) Methicillin-susceptible Staphylococcus aureus 5 (10) Coagulase-negative Staphylococcus aureus 1 (2) Gram-negative bacilli Pseudomonas aeruginosa a 18 (37) Klebsiella pneumoniae 7 (14) Serratia marcescens 7 (14) Escherichia coli 2 (4) Enterobacter cloacae 2 (4) Enterobacter aerogenes 1 (2) Acinetobacter baumannii 1 (2) a Includes four patients with mixed infections comprising Pseudomonas aeruginosa (Klebsiella pneumoniae, n ¼ 3; Serratia marcescens, n ¼ 1).

3 272 Table 2 Correlation between Gram stain findings and final causative organism cultures Gram stain results Causative culture Of the three episodes with GNB>GPC on Gram staining, the causative-cultured organisms were GNB in two (67%). To evaluate the clinical usefulness of Gram-stained tracheal aspirates for determining the causative microorganisms of VAP, sensitivity and specificity, as well as positive and negative predictive values, were analyzed (Table 3). Gram staining indicated that GPC was highly (100%) specific for Gram-positive VAP and the sensitivity was 82%. For Gram-negative VAP, sputum Gram stain showed high sensitivity (100%) and specificity (82%) to GNB. Antibiotic therapy All nine patients with sputum predominately showing GPC (GPC only, eight and GPC>GNB, one) were initially treated with vancomycin. Eight patients recovered from VAP without further modifications. In one patient, PIPC was added 3 days after starting vancomycin therapy, because GNB appeared on sputum Gram stains and the serum CRP level had not decreased (1.53 and 1.56 mg per 100 ml) by that time. Thirty-two of 40 episodes of VAP with sputum predominately showing GNB (GNB only, 37 and GNB>GPC, 3) were initially treated with PIPC. This strategy was ineffective in 2 of the 32 patients, because GPC was finally detected as the causative organism in cultures. The respiratory condition of these two patients did not improve during PIPC administration. They recovered from VAP after an antibiotic change from PIPC to GPC GNB GPC only (n ¼ 8) 8 (100) 0 (0) GNB only (n ¼ 37) 1 (3) 36 (97) GPC>GNB (n ¼ 1) 1 (100) 0 (0) GNB>GPC (n ¼ 3) 1 (33) 2 (67) GNB, Gram-negative bacilli; GPC, Gram-positive cocci; GPC>GNB, Gram-positive cocci predominant compared with Gram-negative bacilli; GNB>GPC, Gram-negative bacilli predominant compared with Gram-positive cocci. Values within parentheses under Causative culture are %. Table 3 Usefulness of Gram staining of tracheal aspirates for identifying final findings of cultured causative organisms Causative culture results GPC (%) GNB (%) Sensitivity Specificity Positive predictive value Negative predictive value Abbreviations: GNB, Gram-negative bacilli; GPC, Gram-positive cocci. VCM 2 days after initial treatment based on the second finding of sputum Gram stain or culture results. The remaining 8 of the 40 episodes were treated with a combination of PIPC and amikacin. No further modifications were applied to this group of patients. Overall, initial antibiotic therapy according to findings in Gram-stained tracheal aspirates was effective in 96% (47/49) of VAP episodes. Discussion The present findings showed that Gram staining of tracheal aspirates was useful for predicting the causative microorganisms of VAP in extremely preterm neonates. Gram staining of tracheal aspirates was highly sensitive (82 and 100%) and specific (100 and 82%) for Gram-positive and Gram-negative VAP, respectively. Thus, Gram staining of tracheal aspirates was also useful for guiding initial antibiotic therapy, which was appropriate for 96% of VAP episodes. Earlier studies have reported conflicting findings from Gram-stain analysis of pulmonary specimens obtained by various means for guiding initial antibiotic therapy for VAP in adult patients. Kopelman 10 found that the accuracy of Gram-stained bronchoalveolar lavage in diagnosing VAP for any organism was limited (58%) and concluded that altering antibiotic therapy based on Gram staining could be inappropriate. Albert et al. 11 found poor agreement between tracheal aspirates or bronchoalveolar lavage Gram-stain class and results of pathogenic cultures. However, Veinstain et al. 12 reported that Gram staining of protected telescope catheter specimens or endotracheal aspirates contributes to a high rate (86%) of initial appropriate management of confirmed VAP. Namias et al. 13 found that Gram staining of tracheal aspirates showing GNB predicts GNB organisms in 89% of surgical ICU cultures and 91% of trauma ICU cultures. Rea-Neto et al. 14 reviewed studies published between January 1966 and June 2007 that compared cytological examinations with quantitative cultures. Their summary of the findings indicated that the rapid availability of cytological data, including inflammatory cells and Gram stains, might support initial therapeutic decisions for patients with suspected VAP. To our knowledge, the usefulness of Gram-stain analysis of pulmonary specimens for guiding initial antibiotic therapy of VAP in neonates has not been documented. The present findings might represent the first information about the usefulness of Gram staining of tracheal aspirates for guiding initial antibiotic therapy of VAP in extremely preterm neonates. The rate of appropriate initial antibiotic therapy in this study (96%) was higher than that reported for adult patients. We usually Gram stain specimens immediately after collection, whereas specimens described earlier were stained at microbiology laboratories. The decreased interval between sample collection and Gram staining might have contributed to our higher rate of success with antibiotic therapies. In addition, the difference in anatomy and physiology

4 273 between adults and preterm neonates might also have been a factor in this discrepancy. Understanding the microbiology of VAP is critical for choosing empirical antibiotic therapy. The most commonly isolated causative organisms in this study were P. aeruginosa (n ¼ 18; 37%), K. pneumoniae (n ¼ 7; 14%), and S. marcescens (n ¼ 7; 14%). Methicillin-resistant and methicillin-susceptible Staphylococcus aureus were each isolated from 5 (10%) patients. These data are similar to an earlier report of VAP in 19 extremely preterm neonates, 3 which reported a principal distribution of causative microorganisms, including P. aeruginosa (n ¼ 8; 42%), Enterobacter spp. (n ¼ 8; 42%), K. pneumoniae (n ¼ 5; 26%), and S. aureus (n ¼ 5; 26%). Consensus guidelines addressing empirical coverage of suspected VAP in NICU patients are not available. 15 Empirical therapy for VAP in premature neonates should be based on flora that is indigenous to the particular NICU. The Gram-stain results of GNB at our NICU led to the selection of PIPC, which is very active against P. aeruginosa and effective against the other GNB that caused VAP at our NICU, namely, K. pneumoniae, S. marcescens and Escherichia coli. Amikacin was combined with PIPC to treat severe VAP. When the Gram-stain results indicated GPC, vancomycin was started because of the high incidence of methicillin-resistant S. aureus as a causative organism of GPC-related VAP at our NICU. No current gold standard exists for diagnosing VAP in neonates. The Centers for Disease Control and Prevention published recommendations in 2008 that included special criteria for health care-associated pneumonia among infants aged p1 year as part of the National Healthcare Safety Network surveillance. 16 However, this criterion does not refer to VAP. Few studies of VAP in neonates have been performed because it is difficult to diagnose. This study did not evaluate the accuracy of our diagnostic criteria for VAP, but rather examined the importance of Gram staining of tracheal aspirates to the initial choice of treatment strategy for confirmed VAP in extremely preterm neonates. New infiltrates in chest radiographs of an intubated premature neonate might represent the development of bronchopulmonary dysplasia, atelectasis, or pulmonary edema. 4 Cordero et al. 17 compared VAP diagnosed by general hospital radiologists, a panel of experts, and a pediatric radiologist from another hospital and found inter-reader discrepancy. We did not use radiographic findings as a diagnostic criterion for VAP, because their evaluation is too subjective for use as a decision-making tool. However, the concordance of the final microbial results, the effectiveness of the selected antibiotics, and subsequent improvements in clinical symptoms revealed the accuracy of our diagnostic criteria for VAP in extremely preterm neonates. Although the present results indicated that Gram staining of tracheal aspirates is highly reliable for rapidly estimating bacterial pathogens associated with VAP in extremely preterm neonates, the concordance rates between Gram-stain results and final causative organism findings were not 100%. When Gram staining of tracheal aspirates mis-identifies the causative microorganisms, antibiotics should be tailored based on the final culture findings of aspirate samples. Careful evaluation of clinical responses to initial treatment is important for successful VAP management in premature neonates. Conclusion We showed that rapid Gram staining of tracheal aspirates by trained neonatal physicians is useful for predicting the class of final causative microorganisms of VAP and for guiding initial antibiotic therapy in extremely preterm neonates. We also identified the largest distribution of causative microorganisms of VAP in such infants to date. A prospective, randomized study of the outcome of administering initial antibiotics based on the results of Gram-stained tracheal aspirates compared with broad-spectrum antibiotics is required to establish consensus guidelines regarding antibiotic therapy for VAP in premature neonates. Conflict of interest The authors declare no conflict of interest. Acknowledgments We thank the nursing and medical staff of the NICU of Takatsuki General Hospital and the laboratory staff for excellent technical assistance. References 1 Gaynes RP, Edwards JR, Jarvis WR, Culver DH, Tolson JS, Martone WJ. Nosocomial infections among neonates in high-risk nurseries in the United States. National Nosocomial Infections Surveillance System. Pediatrics 1996; 98: National Nosocomial Infections Surveillance (NNIS). System report, data summary from January 1992 through June 2004, issued October Am J Infect Control 2004; 32: Apisarnthanarak A, Holzmann-Pazgal G, Hamvas A, Olsen MA, Fraser VJ. Ventilatorassociated pneumonia in extremely preterm neonates in a neonatal intensive care unit: characteristics, risk factors, and outcomes. Pediatrics 2003; 112: Baltimore RS. The difficulty of diagnosing ventilator-associated pneumonia. Pediatrics 2003; 112: Labenne M, Poyart C, Rambaud C, Goldfarb B, Pron B, Jouvet P et al. Blind protected specimen brush and bronchoalveolar lavage in ventilated children. Crit Care Med 1999; 27: Gauvin F, Dassa C, Chaibou M, Proulx F, Farrell CA, Lacroix J. Ventilator-associated pneumonia in intubated children: comparison of different diagnostic methods. Pediatr Crit Care Med 2003; 4: Rumbak MJ, Bass RL. Tracheal aspirate correlates with protected specimen brush in long-term ventilated patients who have clinical pneumonia. Chest 1994; 106: Kirtland SH, Corley DE, Winterbauer RH, Springmeyer SC, Casey KR, Hampson NB et al. The diagnosis of ventilator-associated pneumonia: a comparison of histologic, microbiologic, and clinical criteria. Chest 1997; 112: Carvalho EM, Massarollo PC, Levin AS, Isern MR, Pereira WL, Abdala E et al. Comparative study of etiological diagnosis of nosocomial pneumonia. Braz J Infect Dis 2008; 12:

5 274 Gram stain in ventilator-associated pneumonia 10 Kopelman TR. Can empiric broad-spectrum antibiotics for ventilator-associated pneumonia be narrowed based on Gram s stain results of bronchoalveolar lavage fluid. Am J Surg 2006; 192: Albert M, Friedrich JO, Adhikari NK, Day AG, Verdant C, Heyland DK. Utility of Gram stain in the clinical management of suspected ventilator-associated pneumonia. Secondary analysis of a multicenter randomized trial. J Crit Care 2008; 23: Veinstein A, Brun-Buisson C, Derrode N, Alvarez A, Pinsard M, Robert R et al. Validation of an algorithm based on direct examination of specimens in suspected ventilator-associated pneumonia. Intensive Care Med 2006; 32: Namias N, Harvill S, Ball S, McKenney MG, Sleeman D, Ladha A et al. A reappraisal of the role of Gram s stains of tracheal aspirates in guiding antibiotic selection in the surgical intensive care unit. J Trauma 1998; 44: Rea-Neto A, Youssef NC, Tuche F, Brunkhorst F, Ranieri VM, Reinhart K et al. Diagnosis of ventilator-associated pneumonia: a systematic review of the literature. Crit Care 2008; 12: R Foglia E, Meier MD, Elward A. Ventilator-associated pneumonia in neonatal and pediatric intensive care unit patients. Clin Microbiol Rev 2007; 20: , table of contents. 16 Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance definition of health careassociated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control 2008; 36: Cordero L, Ayers LW, Miller RR, Seguin JH, Coley BD. Surveillance of ventilatorassociated pneumonia in very-low-birth-weight infants. Am J Infect Control 2002; 30: