Cap --risk Factors And Treatment

Poster Discussion

Room C1e - 08:30-10:30 M ONDAY, S EPTEMBER 4 TH 2006

186. Community-acquired pneumonia: risk factors and treatment

P2082 Intrapleural heparin or heparin combined with human recombinant DNAase is not effective in the treatment of empyema in a rabbit model Oner Dikensoy 1,2 , Moon Jun Na 1 , Huai Liao 1 , Zhiwen Zhu 1 , Wonder Drake 3 , Edwin O. Donnelly 4 , Richard W. Light 1 . 1 Allergy and Pulmonary and Critical Care, Vanderbilt University, Nashville, TN, United States; 2 Pulmonary, Gaziantep University, Gaziantep, Turkey; 3 Infectious Diseases, Vanderbilt University, Nashville, TN, United States; 4 Radiodiagnostic, Vanderbilt University, Nashville, TN, United States Purpose: To investigate the effectiveness of intrapleural heparin or heparin combined with human recombinant DNAase (rhDNAase) in the treatment of empyema. Methods: Empyema was induced with the intrapleural injection of 109 Pasteurella multicoda organisms in infusion agar via a chest tube. There were three treatment groups each with six rabbits. Groups were randomly given 1000 IU heparin, or 1000 IU heparin plus 1mg rhDNAase, or saline via chest tube every 12 hours for a total of 6 treatments. The volume of each treatment was 3 ml. The animals were sacrificed at day 10 and the amount of empyema and pleural thickening was scored macroscopically on a scale of 0 to 6. All rabbits received 100,000 U procaine penicillin IM every 24 hours until sacrifice.

348s Abstract printing supported by Nonin Medical, Inc. Visit Nonin Medical, Inc. at stand C09

Poster Discussion

Room C1e - 08:30-10:30 M ONDAY, S EPTEMBER 4 TH 2006

Results: The total volume of pleural effusion aspirated was significantly higher in the heparin group (25.8± 10.7 mL) compared to either saline (8±8.9) or heparin plus human recombinant DNAase (6.8±6.1) groups (p=0.003). The mean empyema and pleural thickening scores did not differ significantly between the groups (p=0.8, p=0.5, respectively). A weak correlation was found between total volume of aspirated pleural fluid and pleural parameters of white blood cell counts and lactate dehydrogenase levels (r=0.546 and p=0.02, r=0.631 and p=0.02, respectively). Conclusion: The intrapleural administration of heparin alone or in combination with rhDNAase is no more effective than saline in the treatment of empyema in rabbits. Intrapleural heparin significantly increased the volume of pleural fluid drainage compared to combination and saline group.

P2083 Clinical cure rates and infection types in the MOxifloxacin Treatment IV (MOTIV) study in hospitalized patients with community-acquired pneumonia (CAP) Robert C. Read 1 , Antoni Torres 2 , Hartmut Lode 3 , Jean Carlet 4 , John H. Winter 5 , Javier Garau 6 , Thomas Welte 7 , Marie-Aude Le Berre 8 , Jane E. Ambler 9 , Shurjeel H. Choudhri 10 , Pierre Arvis 11 . 1 Academic Unit of Infection & Immunity, University of Sheffield Medical School, Sheffield, United Kingdom; 2 Servei de Pneumologia i Allèrgia Respiratòria, Hospital Clínic de Barcelona, Barcelona, Spain; 3 Department of Chest and Infectious Diseases, City Hospital Berlin-H-Heckeshorn, Berlin, Germany; 4 Intensive Care Unit and Infectious Diseases Department, Fondation-Hopital Saint-Joseph, Paris, France; 5 Department of Medicine, Ninewells Hospital and Medical School, Dundee, United Kingdom; 6 Department of Medicine, Hospital Mutua de Terrassa, Barcelona, Spain; 7 Pneumonology Department, Medizinische Hochschule Hannover, Hannover, Germany; 8 Biometry, Bayer HealthCare, Puteaux, France; 9 Global Clinical Development - Anti-Infectives, Bayer HealthCare Pharmaceuticals, West Haven, CT, United States; 10 Global Clinical Development - Anti-Infectives, Bayer HealthCare Pharmaceuticals, West Haven, CT, United States; 11 Medical Affairs, Bayer HealthCare, Puteaux, France Objective: To assess clinical cure at test of cure and bacterial eradication in MOTIV study subgroups. Hospitalized patients with CAP (PSI classes, III, IV or V) received 400 mg IV/PO moxifloxacin (MXF) q.d. or high dose ceftriaxone (CTX) 2 g q.d. and levofloxacin (LFX) 500 mg b.i.d. for 7–14 days. Patients and methods: 738 were randomized (MXF: 371, CTX/LFX: 367); perprotocol (PP) population: 569 (MXF: 291, CTX/LFX: 278) – 336 (59.1%) in PSI Classes IV–V. Organisms were identified by culture, urine antigen testing (S. pneumoniae, L. pneumophila) or serology (L. pneumophila, C. pneumoniae, M. pneumoniae). Results: 250 (43.9%) PP patients had baseline causative organisms (165 [49.1%] in PSI Classes IV–V); common CAP pathogens: 192 (33.7%), atypical pathogens: 86 (15.1%), mixed infections: 28 (4.9%). Clinical success rates in PP subgroups: microbiologically-documented infection – MXF: 114/127 (89.8%), CTX/LFX: 110/123 (89.4%); atypical organisms – MXF: 39/41 (95.1%), CTX/LFX: 41/45 (91.1%); S. pneumoniae – MXF: 69/77 (89.6%), CTX/LFX: 74/85 (87.1%); bacteremia – MXF: 15/20 (75.0%); CTX/LFX: 18/24 (75.0%). Bacteriological success rates (eradication + presumed eradication) in the MBV population were MXF: 45/54 (83.3%) and CTX/LFX: 46/54 (85.2%). Conclusion: In hospitalized patients with CAP, sequential monotherapy with IV/PO MXF 400 mg q.d. achieved clinical and bacteriological success rates similar to high-dose combination therapy with ceftriaxone plus levofloxacin.

P2084 In-vitro deposition study of a levofloxacin (LEV) solution into a novel human nasal cast model by the PARI VibrENT™ Uwe Schuschnig, Elisabeth Klopfer, Martin Luber, Titus Selzer, Manfred Keller. PARI GmbH, Aerosol Research Institute, Munich, Germany

Aerozolization of 3 ml LEV solution (15 mg) was conducted for 4 minutes into the left and right nostril, each. After the experiment the cast was dismantled and drug extracted with solvent from paranasal cavities including ostia, nasal cavity, from nebuliser and filter. LEV content of these solutions was assayed by HPLC. Results: About 3% of the LEV dose initially placed into the nebulizer was found in all sinus cavities. Deposition in the single sinuses ranged from 24 μg (frontal) up to 145 μg (maxillary). About 70% of the initial drug charge remained in the nebulizer while 6% were found in the nasal cavity and 22% on the exit filter. Conclusions: VibrENT delivers drug to the site of infection reducing potential systemic side effects.

P2085 Timing of antibiotic administration and outcomes for patients with community-acquired pneumonia (CAP). NAC-CV study Ada Luz Andreu Rodríguez, Rafael Blanquer Olivas, Estrella Fernández Fabrellas, Francisco Sanz, José Blanquer Olivas, Frederic Tatay. Pneumology, NAC-CV Group, Valencia, Spain Introduction: Guidelines have recommended antibiotic treatment within 8, 6, 4 even 2 hours of arrival at the hospital. Objective: To analyse timing of antibiotic administration (TAA) in patients of NAC-CV Study and its influence on mortality and length of stay. Method: NAC-CV is a prospective 12-moths multicenter study in 13 public hospitals, and 1314 patients were included.In this work, outpatients, patients with length of stay minor than 24 hours (10 patients, 3 of them died) and previously treated were excluded. A descriptive analysis of timing-adjusted mortality and length of stay was performed. Results. Seven hundred and forty five patients were included (32,6% women), with mean age 66,31(16,7), and classified by PSI (8,6%, 13,2%,22,4%,41,7% and 14,1% from I to V group), and 55 patients (7,4%) admitted in Intensive Care Units. Mean length of stay was 10,7 (8,5) days. Twenty six patients died (3,5%). Mean TAA was 304 (239,6) minutes and median 240 (10-1440), with P25 and P75 3 and 6 hours. Accumulated mortality was 3/310 (2.3%) if TAA≤2 h, 13/399 (3.3%) if ≤4 h, 20/578 (3.5%) if ≤ 6h and 25/660 (3.8%) if ≤8 h of arrival at the hospital, without difference for timing or age group. A significant difference was found for length of stay in eldest patients who were treated within 6 hours, particularly within 4 hours. Conclusions. 1. A half of our patients have been treated within 4 hours and 2/3 of them within 6 hours of arrival at the hospital. 2. Significant difference in timing-adjusted mortality was not found in any age group. 3. Length of stay decreased in eldest patients who were treated early.

P2086 Antibiotic prescribing in acute respiratory infections – are we following guidelines? B. Esdaile, R. Reddy, A. Sykes, B. Mann. Department of Respiratory Medicine, West Middlesex University Hospital, Middlesex, United Kingdom

Predict drug deposition in nasal cavities for topical sinusitis treatment with antibiotics. Nebulization efficiency was investigated using a cast model equipped with two cavities (sinuses) in frontal, maxillary and sphenoid position. Cavities as well as ducts (ostia) are exchangeable, allowing variation of the sinus volume and ostium diameter. Configuration in current in-vitro study can be seen from figure 2.

Introduction: Specific guidelines exist (BTS guidelines Thorax 2001; 56: (suppl IV) and Woodhead F et al., Eur Respir J 2005; 26:1138-1180) for the antibiotic management of Community Acquired Pneumonia (CAP), Lower Respiratory Tract Infections (LRTI) and acute exacerbations of Chronic Obstructive Pulmonary Disease (COPD). Inappropriate prescribing may contribute to the increasing incidence of Clostridium difficile associated disease (CDAD) and Methicillin Resistant Staphylococcus Aureus (MRSA). The aim of our study was to use these guidelines to assess whether intravenous antibiotics were being inappropriately prescribed. Design & Methods: A retrospective study of 50 consecutive admissions, coded as respiratory infections, during a six-week period from October 2004 and December 2004. Patient records were analysed and classified into three main categories: CAP, LRTI and COPD. Antibiotic administration was assessed for each patient and compared to the British and European Thoracic Society Guidelines. Results: 53% (8/15) of patients admitted with LRTI without chest radiograph changes received intravenous cephalosporins inappropriately. In those patients admitted with a CAP with a CURB-65 score of 0-1, 89% (8/9) were treated inappropriately with intravenous cephalosporins. 56% (5/9) of those admitted with infective exacerbations of COPD with 2 or more Anthonisen criteria without radiographic changes were treated with intravenous cephalosporins. Conclusions: Intravenous antibiotics (cephalosporins) are being inappropriately

349s Abstract printing supported by Nonin Medical, Inc. Visit Nonin Medical, Inc. at stand C09

Poster Discussion

Room C1e - 08:30-10:30 M ONDAY, S EPTEMBER 4 TH 2006

prescribed for the treatment of respiratory tract infections. It is important to further educate junior doctors as inappropriate prescribing has major infection control issues.

P2087 Antibacterial activity of telithromycin and comparators against pathogens isolated from patients with community-acquired pneumonia (CAP) in PROTEKT years 1–5 (1999–2004) David Felmingham 1 , Alkiviadis Vatopoulos 2 . 1 Surveillance Department, G.R. Micro Ltd, London, United Kingdom; 2 Department of Microbiology, National School of Public Health, Athens, Greece CAP has a worldwide annual incidence of 5–11 per 1000 adults and is a significant healthcare burden. Increasing rates of resistance among bacteria implicated in the pathogenesis of CAP may impose limitations on the treatment options for this respiratory tract infection. This analysis aimed to determine the susceptibility of Streptococcus pneumoniae (SP) and Haemophilus influenzae (HI) isolates collected from patients with CAP to a range of antibacterials, including telithromycin (TEL), over the first 5 years of the PROTEKT surveillance study. Minimum inhibitory concentrations and susceptibility rates were determined according to CLSI methodology and breakpoints. Antibacterial susceptibilities of these isolates for PROTEKT Years 1–5 combined were: S. pneumoniae (n=7722) Antibacterial Penicillin Ampicillin Amoxicillin–clavulanate Cefuroxime Erythromycin Clarithromycin Azithromycin Telithromycin

H. influenzae (n=4071)

S

MIC90

S

MIC90

(%) 63.6 – 93.4 71.4 65.6 65.7 65.6 99.8

(mg/L) 4 – 2 8 ≥128 ≥64 ≥128 0.12

(%) – 81.6 99.0 97.4 –a 84.9 99.6 99.6

(mg/L) – ≥32 1 2 8 16 2 2

MIC, minimum inhibitory concentration; S, susceptible. a No CLSI breakpoints

In total, 38.9% (3005/7722) of SP isolates exhibited multiple resistance phenotypes (MRP) and 15.4% (626/4071) of HI isolates were [Beta]-lactamase positive (BL+). Overall, 99.6% (2992/3005) of SP isolates exhibiting MRP and 99.4% (622/626) of BL+ HI isolates were fully susceptible to TEL. In vitro, TEL is active against common CAP pathogens including SP exhibiting MRP and BL+ HI.

P2088 Cost-efficacy analysis of cefixime switch therapy in the treatment of community acquired pneumonia Yubiao Guo, Canmao Xie, Mian Zeng, Jianqian Huang, Yifeng Luo. Department of Pulmonary & Critical Care Medicine, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou, Guangdong Province, China Objective: To compare and evaluate the efficacy, pharmaco-economics of ceftriaxone/cefixime switch therapy and ceftrioaxone alone in the treatment of community acquired pneumonia. Methods: 60 Chinese patients with community acquired pneumonia were randomize assigned into two treatment groups: one group (control group, n=30) received ceftriaxone 2g intravenous injection twice daily for 7days;The other group (switch group, n=30) received ceftriaxone 2g injection twice daily for 3 days and then switched to oral cefixime 200mg twice daily for 4 days. The efficacy and pharmacoeconomics of two groups were compared by using the minimal pharmacoeconomic cost-effect analysis. Results: The total effective rate of cefixime switch group and control group was 81.8% and 75% respectively, there was no significant difference between two groups (P>0.05).The drug-sensitivity rate of sputum-isolated bacteria and bacteria eradication rate were 81.5% and 74.1% in control group, while in switch group they were 78.2% and 70% respectively, there were no significant difference between two group (P>0.05). However, the mean total therapy cost and the antibiotic-related cost in switch group were less than those in control group (P<0.05). Conclusion: Cefixime switch therapy is an effective and economical approach for community acquired pneumonia.

P2089 Pneumonia as the reason for quitting smoking Malina Percinkovski, Mirjana Danilovic, Svetlana Krstic, Spasoje Popevic. Institute for Lung Diseases and Tuberculosis, Clinical Center of Serbia, Belgrade, Serbia & Montenegro The objective of the study was to establish the effect of the acute diseases of respiratory tract to quitting smoking. The out-patients - smokers treated for pneumonia, at the beginning and end of treatment as well as one year following the treatment were analyzed. Out of 84 patients, 69 (82%) were males and 15 (18%) were females. Bilateral pneumonia was manifested in 9 (10.7%) patients. Concurrent

manifestation of COPD was recorded in 54 (64%) patients, that is, severe condition in 28%, mid-severe in 49.3% and minor obstructive disorders were noted in 22.7% of cases. At the end of pneumonia treatment, 45 (53%) patients quitted smoking, i.e. 40 males and 5 females. Twenty-one smokers (25%) reduced smoking to 5-10 cigarettes per day. A year after the completed treatment, there were 78 smokers and 6 non-smokers among subjects, meaning that 7.5% of patients quitted smoking after the disease, while in spite of physician’s advice, 63% of them started smoking again and 21.5% never quitted smoking. Conclusion: Contracting the acute respiratory diseases (pneumonia) in smokers should guide us to more intensive and prolonged work on final quitting smoking.

P2090 Prescription of antibiotics for the treatment of community acquired pneumonia in a district general hospital in the UK David Owen, Tamara Shiner, Christopher Hilton, Ramachandran Sivakumar, Richard Dent. Department of Medicine, Queen Elizabeth II Hospital, Hertfordshire, United Kingdom Background: British Thoracic Society (BTS) guidelines recommend the CURB 65 score (new Confusion < 9/10, Urea >7, Respiratory rate > 30, Blood pressure < 90 systolic or 60 diastolic, age > 65) to assess the severity and treatment of community acquired pneumonia. Accordingly, oral antibiotics should always be prescribed when the CURB 65 score is 0-1 and intravenous antibiotics when the score is 3 or above. Methods We retrospectively reviewed 244 consecutive admissions for suspected pneumonia during 2005 at the QE II Hospital, Hertforshire. Only the case notes of those subsequently diagnosed with pneumonia were included. We calculated the CURB 65 score in all patients with a diagnosis of pneumonia. Subsequent prescription of oral or intravenous antibiotics was noted. Results: Of the 244 admissions with suspected pneumonia 106 were actually diagnosed and treated as pneumonia. In only 2 cases was the CURB 65 score recorded. 16 patients fulfilled the criteria for intravenous antibiotics (ie a CURB 65 score of 3 or above), and all of these were treated appropriately. However, of the 52 patients who fit the criteria for non-severe pneumonia (CURB 65 score of 0 or 1) 28 (54%) were given intravenous antibiotics contrary to current BTS guidelines. Conclusion: Inappropriate and unnecessary use of intravenous antibiotics is expensive and promotes antibiotic resistance, as well as hospital acquired infections including Clostridium difficile and Methicillin Resistant Staphylococcus aureus (MRSA). Despite BTS guidelines, intravenous antibiotics are being inappropriately prescribed which may contribute to the morbidity and mortality of inpatients in UK hospitals.

P2091 Promising prophylactic means and methods for prevention of community-acquired pneumonia and respiratory infections among servicemen in Russia Konstantin D. Jogolev 1 , Sergey D. Zhogolev 2 , Pavel I. Ogarkov 2 , Nikolay N. Bespalov 3 , Boris S. Sukhanov 2 . 1 Immunology Department, Military Medical Academy, St Petersburg, Russia; 2 Epidemiology Department, Military Medical Academy, St Petersburg, Russia; 3 Medical Equipment Department, Mordovsky State University, Saransk, Russia Efficiency of some means and methods used to prevent respiratory infections was studied in summer period of 2005 in a military training center of the North-Western region of Russia. The investigation included three groups of recruits (May call up) aged 18-20 (70 persons in each group) resided in similar living and service conditions. The first group was given ascorbic acid (0.3 g) and Dibazol (0.04 g t.i.d.) for first 10 days after the center arrival. The second group was located in the sleeping accommodation equipped with a special ceiling radiator “Effluvion-1.3” based on A.L.Chizhevsky method. This device being an air ionizator emits negative oxygen aeroions through the distribution electrode network mounted on the ceiling. The aeroions have a beneficial effect on the human body. The third group served as a control. Five-month follow-up showed that respiratory infection incidence in the first group was 2.2 times lower than in the control group whereas angina incidence was 3.3 times lower and acute sinusitis incidence – 4.9 times lower. In the second group where servicemen breathed with ionized air at nights incidence of respiratory infectious was 1.8 times lower (angina incidence – 2.4 times lower and acute sinusitis – 4.8 times lower) than in the control group. Thus, both means proved to be efficient for prophylaxis of respiratory infections, angina and acute sinusitis.

P2092 Value of bacteraemia as an unfavourable prognostic factor for community-acquired pneumonia F. Sanz, J. Blanquer, L. Briones, R. Blanquer, E. Chiner, C. Aguar. NAC-CV Group, Valencia, Spain Aims: The aim of this work is to identify etiologic agents and risk factors for

350s Abstract printing supported by Nonin Medical, Inc. Visit Nonin Medical, Inc. at stand C09

Poster Discussion

Room C1e - 08:30-10:30 M ONDAY, S EPTEMBER 4 TH 2006

bacteraemia in community-acquired pneumonia (B-CAP) and their influence on clinical outcome. Methods: An epidemiologic, prospective, multicenter study was performed in which 1,314 CAP cases were analyzed. Two blood cultures, urinary antigen detection, and serology for atypical bacterias and viruses were performed in order to obtain an etiological diagnosis. Results: Among 1,037 (78.9%) blood cultures, 77 (7.4%) bacteraemic pneumonias were identified: Streptococcus pneumoniae: 46 cases (59.7%), Streptococcus spp: 14 cases (18.2%), coagulase-positive staphylococci: 6 cases (7.8%), Escherichia coli: 6 cases (7.8%), P. aeruginosa: 3 cases (3.9%), other: in 6 blood cultures (7.8%). Alcoholism was the risk factor associated with B-CAP (OR 2.54; 95% CI 1.43-4.52; p=0.002), whereas concordant prior antibiotic therapy was a protective factor for B-CAP (OR 0.31; 95%CI 0.13-0.73; p=0.007). Bacteraemia was associated with a higher hospital stay, more clinical severity (PSI) and a worse outcome than non bacteraemic pneumonia, but ICU admission was the only significant factor associated with bacteraemia (OR 3,97; 95% CI 1,71-9,25; p=0,001). The mortality rate was higher in B-CAP than in non-bacteraemic pneumonias (14.5% vs 5.1%; p=0.001). Conclusions: 1. Streptococcus pneumoniae was the main etiologic agent of bacteraemic pneumonia. 2. Early antibiotic treatment acts as a protective factor for bacteraemia when it is concordant. 3. In our experience, alcoholism is the only risk factor for bacteraemia. 4. Bacteraemic pneumonias are associated with a worse outcome, with higher ICU admission and mortality than non-bacteraemic pneumonias.

P2093 Recognising poor prognosis community-acquired pneumonia in young adults with simple clinical measurements Oxana V. Fessenko. Pulmonology, State Postgraduate Medical Institute, Moscow, Russia Introduction: Although severe community-acquired pneumonia (CAP) and death occur most frequently in the elderly and those with chronic underlying diseases, young patient also die. Death from pneumonia in this group of patients might be considered to be most likely to be preventable. Nowadays there isn’t any robust criteria to estimate risk of mortality from CAP in young patients. Previous studies show that all developed scoring systems (ATS, BTS, APACHE, PSI, etc) allow to estimate pneumonia severity in adult patients but their predictive value is limited for population of young people. It can be explained that all of these systems use patient’s age as one of principal risk factor. Aim: To identify risk factors of mortality from severe CAP in young people. Methods: We reviewed data from 267 non-immunocompromissed young patients (all male) with CAP admitted to ICU. Mean age was 22±4. 39 patients died. StatSoft Software (version 5.0, 1997) was used for calculations. Results: We analyzed 23 factors (temperature, blood pressure, respiratory rate, multilobar extension, blood count, etc) measured at patient’s admission. Multivariate logistic regression revealed that decreased WBC count (≤6000 cells/ml) (OR 9,6, 95% CI 4,4–20,9), dyspnea at rest (OR 7, 95% CI 3,4–14,8) and low level of consciousness (OR 6,6, 95% CI 3,1–14,2) are three independent predictors of mortality. Mortality risk for CAP is 91% in young patients with these 3 symptoms presence. Conclusion: Although our model requires further validation it may be useful as simple and fast method for identifying young patient with poor prognosis CAP.

P2094 Validating CURB-65 score among inpatients with community acquired pneumonia in Pakistan Faisal F. Zuberi, Ahmed S. Haque, Javaid A. Khan, Muhammad Islam. Pulmonology & Critical Care, Aga Khan University, Karachi, Sindh, Pakistan Background: British Thoracic Society recommends CURB-65 score; Confusion, Blood Urea Nitrogen (BUN) ≥20mg/dl, Respiratory Rate (RR) ≥30/min, Blood pressure (systolic<90/diastolic ≤60mmHg) and Age≥ 65yrs to categorize patients with community acquired pneumonia (CAP) into six strata (range 0-5) of increasing risk of mortality. Objective: To validate CURB-65 score based risk of mortality in our population. Methods: Records of all adult inpatients over 1 year with CAP at a 500-bed teaching hospital were analyzed. They were grouped into Low (CURB-65score:01), Intermediate (CURB-65score:2) and High (CURB-65score:3-5) mortality risk groups with outcome as discharged alive or expired in hospital. Results: 268 patients were studied (mean age 61 yrs, 55.2% male). Total observed mortality was 34 (12.7%). BUN (OR 7.4; 95% CI 2.8-19.7) was the strongest, Confusion (OR 3.7; 95% CI 1.7-8.3) intermediate and RR (OR 2.4; 95% CI 1.1-5.0) the weakest variable predicting mortality on univariate analysis. Low BP and Age were not statistically significant. Only BUN and Confusion were independently associated with risk of mortality on multivariate analysis. The observed mortality in our population was 6(4.5%) in the Low mortality group, 13(17.1%) in the Intermediate and 15(25.0%) in the High group. This was higher than CURB-65 predicted risk of 1.5%, 9.2% and 22% respectively. Conclusion: CURB-65 score for CAP underestimated the risk of mortality in particular for the Low and Intermediate risk groups in our population. Amongst CURB-65 variables only BUN and Confusion were independently associated with

risk of mortality. Further research is needed to identify risk factors in CAP predicting severity and mortality for our population.

P2095 Risk factors for community-acquired pneumonia in hospitalized adult patients, a case- control study Safaa M. Wafy, Aliae A. Mohamed. Chest Department, Assiut University Hospital, Assiut, Egypt Although CAP remains a major cause of hospitalization in developed countries, few studies on risk factors have been performed. The purpose was to identify risk factors for pneumonia diagnosed in our community, using a case control study. All 104 CAP patients admitted to chest department in Assiut University Hospital- Egypt, during 18 months were compared with 300 controls adults. Results: males were 71.2% with age mean ± SD 44.77± 17.37, and females were 28.8 with age mean ± SD 45.77± 11.59. CAP was common among males 37.4% versus 14.6%for females. It significantly affected patients aged more than 50 years (30.1% versus 22.8%). The incidence of CAP was higher in workers, farmers and employers (56.4%, 35.4% and32.3% respectively) and among smokers (46.1%) versus (19.9%) none smokers. Significant risk factors in univariate analysis included renal, heart disease, hepatic, smoking, systemic diseases, diabetes and steroid therapy. Multiple logistic regression analysis showed statistically significant influence of heart failure (OR = 2.57; CI: 1.1-5.99; p = 0.029) and chronic renal failure (OR = 3.35; CI: 1.06-10.54; p = 0.039) on high risk.Another independent risk factors were smoking (OR=3.56; 95% CI: 1.65-8.42, p=<0.001), chronic bronchitis (9.24, 1.48 - 57.74, p = 0.000) and asthma (3.20, 1.98-7.15.p=<0.001). Hepatic (6.02, 1.08-33.3, p=<0.001), systemic diseases (11.57, 6.68-20.04, p=<0.001) corticosteroid therapy (OR 3.52, 95% CI 1.99 – 6.24, p =<0.001) and diabetes (OR 2.40, 95% CI 1.12 – 5.14, p =<0.001) were validated as additional independent risk factors. These data suggest that cigarette smoking is the main avoidable risk factor for community-acquired pneumonia in adults.

P2096 The post-tuberculosis lung changes as risk factor of the community-acquired pneumonia delayed course Tamara V. Rubanik 1 , Natalia L. Shaporova 2 , Vasily I. Trofimov 2 . 1 Pulmonology Department, City Clinico-Diagnostic Center, Saint-Petersburg, Russia; 2 Hospital Therapy, Pavlov’s State Medical University, Saint-Petersburg, Russia To investigate the role of post-tuberculosis lung changes (PTLC) as risk factor of the community-acquired pneumonia (CAP) delayed course we examined 60 patients with CAP delayed course and PTLC as main group and 102 ones with CAP delayed course and without PTLC as the control, with average age 53,5 ± 1,1 years. There were 68,3% of small and 31,7% of large PTLC forms in the main group. The ptlc presence significantly influenced the CAP severity (p<0,001). The haemoptysis as the clinical symptom of CAP took place more frequent in the patients with PTLC (p<0,05).And smoking patients demonstrated haemoptysis 3 times more often to be compared with none smoking ones. The prolongation of the cough as the clinical symptom of CAP was significantly longer in the main group (43,8 ± 1,0 and 34,4 ± 0,8, p<0,001). We still observed cough after the reversion of XR-changes in the patients with PTLC, while in the control group cough disappeared before the infiltration. The average prolongation of CAP in patients of main group was approximately equal 42,3 ± 0,9 days, in control - 35,9 ± 0,6 days, so the prolongation of CAP was significantly longer in the patients with PTLC (p<0,001). The antibiotic treatment courses quantity was higher in the patients with PTLC (p<0,05). The bronchoscopy investigation revealed the PTLC in all the patients of main and in 38% patients of control group. We can conclude that the endoscopy signs of PTLC are more frequent then the roentgenological ones. So our investigation showed that PTLC are the important risk factor of delayed CAP course, which influence the clinical features, prolongation and treatment peculiarities of disease.

P2097 Incidence of pneumonia in patients after stroke Zoran S. Arsovski 1 , Dejan V. Dokic 1 , Anita A. Arsovska 2 , Ante M. Popovski 2 . 1 Scientific Department, Clinic of Pulmonology, Skopje, Macedonia; 2 ICU, Clinic of Neurology, Skopje, Macedonia The purpose of this study was to evaluate the incidence of pneumonia in patients after stroke. We have made a retrospective analysis of 400 patients (211 male, 189 female) with stroke, confirmed with computer tomography (CT) of the brain, average age 63, which developed pneumonia in ICU. The incidence of pneumonia was analyzed in two different groups of patients, who were classified according to CT findings after stroke. First group-352 patients (88%) with involvement of a.carotis interna system. The second group-48 patients (12%) with involvement of the vertebrobasilar system. Criteria for diagnosis of pneumonia were: new pneumonic infiltration seen on a chest X-ray, physical chest findings and one or more of the following symptoms: temperature higher then 37,5 C, dyspnea and cough. Pneumonia was found in 45 patients (12,78%) in the first group and 22 patients (45,83%) in the second group. Aspiration pneumonia was mainly cause of pneumonia, 82% were developed due to post stroke swallowing disorders. 18%

351s Abstract printing supported by Nonin Medical, Inc. Visit Nonin Medical, Inc. at stand C09

Poster Discussion

Room C1e - 08:30-10:30 M ONDAY, S EPTEMBER 4 TH 2006

of patients developed nosocomial pneumonia. Pefloxacinum (Abactal) 400 mg bid for 10 days was the therapy of choice in 68,75% of the cases. We can conclude that patients with stroke which affected the vertebrobasilar system should be subjected to developed pneumonia (p<0,05) owing to frequent aspiration. Early recognition of the swallowing disorders could reduce the incidence of pneumonia in patients with stroke.

P2098 Predictive factors of in-hospital mortality of patients with community-acquired pneumonia Sandra Saleiro, Vitor Braz, António Oliveira e Silva, Sofia Pereira, Margarida Freitas Silva. Pneumology, Hospital São João, Porto, Portugal; Internal Medicine, Hospital São João, Porto, Portugal Community-acquired pneumonia (CAP) is a frequent cause of hospital admission and mortality worldwide.To evaluate predictive factors of in-hospital mortality in patients admitted with CAP, a retrospective study of patients admitted to a Medicine ward, between January and December 2003, was carried out. Data related to demographics, comorbilities, clinical and radiological presentation were collected through medical records review. Statistical analysis was made using Epi Info and Stata. A p value < 0.05 was considered statistically significant. The study included 333 patients (51.4% male; mean age 72.8 ± 16.4 years). In-hospital mortality was 17.7%. In univariate analysis, malignant disease [OR: 2.39 (1.08-5.25)], cerebrovascular disease [OR: 1.84 (1.00-3.38)], chronic hepatic disease [OR: 3.53 (0.92-13.15)] and current treatment with oral steroids [OR: 2.99 (0.91-9.53)] were associated with in-hospital mortality, as was impaired conscious level [OR: 2.78 (1.48-5.24)] and bilateral pulmonary involvement in chest radiograph [OR: 2.86 (1.42-5.74)]. Chronic obstructive pulmonary disease (COPD) was negatively associated with in-hospital mortality [OR: 0.22 (0.05-0.79)]. In multivariate analysis, current treatment with oral steroids [OR: 3.35 (1.02-11.0)], impaired conscious level [OR: 2.14 (1.04-4.42)], bilateral pulmonary involvement [OR: 2.89 (1.41-5.89)] and COPD [OR: 0.24 (0.07-0.85)] remained as independent variables associated with in-hospital mortality. Conclusion: This study identified the presence of current treatment with oral steroids, impaired conscious level and pulmonary bilateral involvement in chest radiograph as predictive factors of in-hospital mortality of patients with CAP.

P2099 Community acquired pneumonia in elderly: clinical aspects and prognostic factors Ivanka Djordjevic, Tatjana Pejcic, Slavica Golubovic, Milan Radovic, Tatjana Radjenovic-Petkovic, Dragana Dacic. Department for Nonspecific Lung Diseases, Clinic for Lung Diseases, Nis, Serbia, Serbia & Montenegro The aim of this study were to determine the clinical and epidemiological characteristics of community acquired pneumonia (CAP) in elderly, to identify prognostic factors and to established a predictive model for mortality of CAP. Patients (pts) ≥65 years old with CAP admitted in the clinic over the last year were included in study. Multivariate analysis was used to identified prognostic factors from variables present on admission, from which a discrimination rule was constructed to predict mortality. Among total of 172 pts with CAP, there were 75 elderly pts. (43,6%). Clinical pictures lasted 21 days on average and was atypical in 15 pts (20%). The main clinical feature were cough in 54 pts (72%). Most pts 48 (66%) had some kind of accompained or undrelying disease. Microbiological diagnosis was made in 9 pts (12%). There were 6 (8%) deaths. The prognostic factors in multivariate analysis on admission were bilateral radiographic infiltrates, hyperazotaemia, absence of fever, tachypnoea, alteration in mental status and shock. The dicriminating rule to predict mortality comprising three or more of these factors was 56% sensitivity with specificity of 88% and an overall accuracy of 84%. CAP in elderly is associated with a high degree of mortality.The discriminating rule incorporating the prognostic factors identified is powerful predictor of bad course of CAP.

4.68; 95%CI 2.17-8.60, p<0.0001), mechanical ventilation (O.r. 6.64; 95%CI 1.64-27.37, p=0.0063), ICU admission (O.r. 3.33; 95% CI 1.43-7.75, p=0.0063), ischemic heart disease (O.r. 7.56; 95%CI 2.58-22.10, p<0.0001), the absence of microbiological analyses (O.r. 50.00; 95%CI 9.15-273.10), p<0.0001), the use of narcotic analgesics (O.r. 8.67; 95%CI 2.2-34.36, p=0.0019), the average day infusion volume more than 2000ml (O.r. 6.00; 95%CI 1.47-24.46, p=0.017), temperature by admission less than 37.7 C without antipyretics (O.r.4.65; 95%CI 2.96-21.87, p=0.003). Conclusion: while there are some common predictors of in-hospital mortality, mortality is affected by the use of narcotic analgesics, high infusion volume, the absence of microbiological analyses and temperature less than 37.7 C.

P2101 Atypical bacterial and viral ethiology of community-acquired pneumonia in Chile: a preliminary report Mauricio H. Ruiz, Maria A. Martinez, Vivian R. Luchsinger, Enna M. Zunino, Lucia R. Aguad, Pamela L. Arce, Mauricio A. Lopez, Luis F. Avendano. Enfermedades Respiratorias, Hospital Clinico Universidad Chile, Santiago, Chile; Microbiologia ICBM, Universidad de Chile, Santiago, Chile; Virologia ICBM, Universidad de Chile, Santiago, Chile; Infeccioso, Hospital Lucio Cordova, Santiago, Chile; Infeccioso, Hospital Lucio Cordova, Santiago, Chile; Medicina, Hospital Clinico Universidad Chile, Santiago, Chile; Medicina, Hospital Clinico Universidad Chile, Santiago, Chile; Virologia ICBM, Universidad de Chile, Santiago, Chile Community-acquired pneumonia (CAP) is a worldwide leading cause of mortality. S. pneumoniae seems to be the principal pathogen; the etiology remains usually unclear. Objectives. To determine the etiology of CAP in ambulatory and hospital settings in Chile and clinical features associated to viral and atypical bacteria etiology. Methods: Persons ≥ 18 years old with radiographic confirmed CAP were enrolled in 2005 in Santiago. Clinical features were registered. Laboratory study included: viral culture and immunofluorescent antigen detection for RSV, adenovirus, influenza and parainfluenza viruses; urinaryS. pneumoniae antigen detection; sputum culture for conventional bacteria and Legionella; polymerase chain reaction for RSV, metapneumovirus, Mycoplasma pneumoniae, Chlamydia pneumoniae and Legionella pneumophila; paired sera test for antibodies to RSV, M pneumoniae, C. pneumoniae. Results: The average age of the137 cases studied (66 female) was 57 ± 20 years. 15% were outpatient, 58% hospitalized and 23% had severe CAP. Pathogens were identified in 53% of cases: S. pneumoniae in 20%, atypical bacteria (M. pneumoniae or Ch. Pneumoniae) in 28%, viruses in 12%, mixed infections in 11%. Co-infecction were significantly associated to ambulatory cases (p<0,01). No relationship was found between clinical severity and etiology. Atypical bacterias were detected more frequently in spring and summer (p<0,01), while viral and mixed infections were common in winter and fall (p<0,01 and <0,05 respectively). Conclusion: Viral and atypical bacteria are agents frequently associated to CAP, with seasonal distribution and without severity relationship. Supported by FONDECYT 1050734.

P2100 Factors influencing in-hospital mortality in community-acquired pneumonia (CAP) Natalia B. Lazareva, Lubov S. Dolgenkova, Anton A. Igonin, Vladimir V. Arkhipov, Alla N. Tsoi. Clinical Pharmacology, Pulmonology, Sechenov Moscow Medical Academy, Moscow, Russia Aim: to determinate the factors that predict in-hospital mortality among patients require hospitalization for the treatment of CAP. Methods: a retrospective observational study of the patients who were admitted to Moscow city clinical hospital in 2005. Results: a total of 150 patients were enrolled in the study, 16 of whom died (10.66%). The mortality among the patients with severe CAP (according to BTS, 2002) was 48%. The median age was 63 (17-85) years. The population was 59% male. The non-survived patients had 30(15-57) points by SAPS II vs. 25 (6-57), p<0.05. The initial temperature (36.7±0.89 vs.38.3±0.89, p=0.0037) and maximal temperature (37.6±1.19 vs.38.7±0.77, p=0.0014) were lower in the group of the non-survived patients in comparison with the survived patient. The following factors were associated with increased mortality: two-sided infiltration (Odds ratio

352s Abstract printing supported by Nonin Medical, Inc. Visit Nonin Medical, Inc. at stand C09