pharmacoepidemiology and drug safety 2014 Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/pds.3614
ORIGINAL REPORT
Antibiotic therapy for inpatients with community-acquired pneumonia in a developing country Hieu T. Trinh1, Phuong H. Hoang1, Magnolia Cardona-Morrell2, Hai T. Nguyen1, Dinh Hoa Vu1, Phuong T. X. Dong1, Thao T. B. Cao1, Son T. Nguyen1, Van T. T. Pham1, Le Moss2, Kathryn Dinh2, Jonathan Dartnell2 and Huong T. L. Nguyen1* 1 2
Department of Clinical Pharmacy, Hanoi University of Pharmacy, Hanoi, Vietnam NPS MedicineWise (NPS), Sydney, Australia
ABSTRACT Purpose The aim of this study was to identify antibiotic prescription patterns for community-acquired pneumonia (CAP) in Vietnam. Methods Medical records for CAP adult patients admitted to 10 hospitals across the country were randomly selected from admission lists during the peak pneumonia season. CAP cases were identified from manual record reviews by clinical pharmacists. Data was collected using a standard data collection tool including patient clinical features on admission, comorbidities, microbiological culture results, and antibiotic regimens. Pneumonia severity was estimated using the CURB-65 score. Results A total of 649 medical records for adult patients (55.2% male and 52.3% urban residents, median age 68 years) met the selection criteria for CAP. Pneumonia severity was assessed as mild (64.1% of patients), moderate (23.0%), and severe (9.2%). Antibiotics were most frequently administered intravenously (93.4%) and as combination therapy (dual therapy 54.4%, monotherapy 42.5%, and triple therapy 3.1% of patients) regardless of CAP severity. Third-generation cephalosporins were used most frequently (29.3% as monotherapy and 40.4% as combination therapy). Third-generation cephalosporins were most commonly combined with penicillins and/or quinolones. Conclusions This first nationwide study provides a baseline profile of antibiotic use in the treatment of CAP. Third-generation cephalosporins were widely used for initial empirical management of CAP, often in combination with quinolones, regardless of CAP severity. The study will assist in providing an evidence base to inform new national antibiotic guidelines for CAP management and will contribute locally relevant data for the national master plan addressing antibiotic resistance and the development of educational interventions to improve CAP management. Copyright © 2014 John Wiley & Sons, Ltd. key words—community-acquired pneumonia (CAP); antibiotics; drug utilization evaluation; Vietnam; developing country; pharmacoepidemiology Received 5 September 2013; Revised 22 February 2014; Accepted 22 February 2014
INTRODUCTION Lacking of standard guidelines is one of main reasons of antibiotics misuse in treating pneumonia worldwide. According to WHO reports, there were only 70% pneumonia cases that received appropriate antibiotics in developing countries. Proportion of patient adherence in these countries was about 50%. Less than half of the countries had effective measures to increase quality use of medicines including antibiotics.1 Vietnam is a developing country in which community-acquired pneumonia (CAP) is a common *Correspondence to: Huong TL Nguyen, Department of Clinical Pharmacy, Hanoi University of Pharmacy, Hanoi, Vietnam. E-mail: [email protected]
Copyright © 2014 John Wiley & Sons, Ltd.
infectious disease that is associated with high morbidity and mortality.2 In Vietnam, respiratory diseases have high prevalence (19.09%) and are ranked as the 10th highest cause of mortality. Among respiratory diseases nationwide, pneumonia showed an overall prevalence of 409 per 100 000 population.2,3 Over the past decade, antibiotic resistance has become a serious problem in many areas of the world including Vietnam.3,4 Data from Vietnam show an alarming rate of resistance, and this rate is continuing to increase.5 In particular, the most common bacteria strains causing CAP have developed substantial resistance to commonly used antibiotics.6,7 A multi-central study conducted in 11 hospitals in Vietnam showed that isolated Streptococcus pneumoniae from
h. t. trinh et al.
infectious respiratory patients was resistant to macrolides, cefuroxime, and cefaclor with rates of 95.0%, 71.4%, and 87.5%, respectively.8 In this study, Haemophilus influenza, which is also a common pathogen in CAP, found to be highly resistant to oral antibiotics such as cotrimoxazole (82.5%), tetracycline (92.5%), and chloramphenicol (78.0%).8 These high resistance rates could be associated with the overuse of these antibiotics in a community where people could straightforwardly purchase medicines without prescriptions.9,10 Thus, the empirical treatments for CAP patients could be challenged. National and local standard treatment guidelines are important tools to tailor antibiotic choices. At the time of this study, however, the national guideline for CAP management from 2006 was not appropriate with present antibiotic resistance. Furthermore, none of hospitals have local standard treatment guidelines based on their own antibiotic resistance rate. The present study aimed to provide a cross-sectional baseline estimate of actual antibiotic prescribing practice in hospitals in Vietnam with a view to contributing evidence to support the development of new standard national antibiotic therapeutic guidelines for CAP in Vietnam. In addition, the results of this survey could assist Vietnamese authorities in determining appropriate interventions for health professionals and consumers to improve the quality of use of antibiotics. METHOD A retrospective study was undertaken to investigate antibiotic use in the treatment of CAP cases admitted to hospital in Vietnam during the identified peak season. Setting and study period A convenience sample of 10 national and provincial hospitals located in large cities across the country (North, Central, and South Vietnam) was selected from a fixed list of 16 hospitals chosen by the funding body (The Global Fund to Fight AIDS, Tuberculosis, and Malaria). These were chosen based on the availability of hospital staff to support the study data collection and a history of close collaboration with the Department of Clinical Pharmacy at Hanoi University of Pharmacy (HUP). The study period for each hospital was different depending on the peak season for respiratory illness in the region. Peak season was the period in the most recent year (2011) that included the highest number of admissions for respiratory disease determined Copyright © 2014 John Wiley & Sons, Ltd.
through examination of ICD-10 codes from hospital data. While for most hospitals (8/10), the peak season covered at least 3 months, in other hospitals (2/10), the peak season was only for 1 month. Target conditions Community-acquired pneumonia identification was based on either an admission or discharge diagnosis using ICD-10 codes J13.0-J18. Patient inclusion and exclusion criteria Eligible patients were adults aged 16 years and older, admitted for at least 24 h with a primary diagnosis of bacterial pneumonia acquired in the community (with or without underlying COPD or asthma) and were prescribed at least one antibiotic with complete information on clinical indication and drug regimen to enable analysis and clinical interpretation. Excluded patients were those who were admitted for non-bacterial pneumonia (viral, parasitic, or fungal) or did not receive antibiotics for pneumonia, hospital-acquired pneumonia, aspiration pneumonia, cases only seen in outpatient or emergency and discharged on the same day (due to absence, brevity or incompleteness of records) or who were admitted to intensive care units (due to the study’s time and resource constraints that precluded review of extensive records). Non-bacterial pneumonia was confirmed if the viral and/or parasitic test was positive. Aspiration pneumonia was confirmed if the doctor wrote “aspiration pneumonia” in the diagnosis section. Hospital-acquired pneumonia was defined as the pneumonia that occurs at least 48 h after hospitalization.11 A data collection tool to capture all variables was constructed and piloted in three hospitals before the data collection phase commenced. Then the standard data collection toolkit was used by trained pharmacists performing medical record review. Sample size calculations Sixty cases of CAP were randomly selected from each of the 10 hospitals. Sixty was estimated as a minimum number of cases required to achieve distribution of variables to approximate the Normal distribution and hence to have a representative sample of records from each hospital. This minimum number also catered for missing or incomplete items in the selected records and enabled meaningful and accurate inferences from subgroup analysis. Minimum data required for a record to be considered complete included patient Pharmacoepidemiology and Drug Safety, 2014 DOI: 10.1002/pds
ANTIBIOTICS THERAPY FOR PNEUMONIA
demographics, pneumonia diagnosis, antibiotic name, strength, dose, and route of administration. Data analysis Comparisons of demographic parameters and CAP severity used the chi-squared test to assess differences in proportions between groups. Statistical significance level was set at p = 0.05. The prevalence of antibiotic prescription was calculated as the number of patients receiving any antibiotic out of the total number of CAP patients studied. Data analysis was performed using Stata v.11.0 software program (StataCorp, College Station TX, USA). Multivariate logistic regression analysis was conducted to investigate the determinants of prescribing patterns in hospital (dichotomous outcome combined therapy) as empirical treatment. Combined therapy was defined as a therapy in which patients use dual or triple antibiotic combination. Independent variables examined were age, sex, area of residence, medical insurance status, transferred status, number of comorbidities, severity of CAP, hospital type, and ward type. As the purpose of the research group was to show the insignificant effect of less strong predictors, all variables that may have potential effect were included in the model.
RESULTS Demographic profile of patients Between January 2011 and December 2011 (depending on the peak season of pneumonia in each region), 649 medical records with CAP diagnosis from 10 hospitals of Vietnam were included in this study. The patient characteristics are presented in Table 1. CAP admitted patients were mostly elderly patients (median age 68 years, IQR 51–79), just over half of them lived in urban areas, and distribution of admissions by sex was almost equal. Clinical profile of community-acquired pneumonia cases Of 649 patients with additional clinical information available in the record, half had comorbidities. The median length of hospitalization was approximately 11 days (IQR: 8–16). Most patients were classified as having mild or moderate CAP. CURB-65 score could not be calculated for 24 patients (3.7%) during the first 2 days of admission, due to lack of information for at least one of the parameters. They were included in the overall study results but not included in some subgroup analyses.
Table 1. Demographic profile of patients with community-acquired pneumonia Demographic parameter Hospital level (n = 649) Age (year) * (n = 636) Weight (kg)* (n = 175) Sex (n = 649) Residence (n = 642) Ward (n = 640)
Comorbidity (n = 649) Type of comorbidity (n = 649)
Transferred from another hospital (n = 649) Length of hospital stay* (days) (n = 649)
Patients number (% total) National Provincial Male Female Urban Suburban Rural Respiratory Infectious diseases Geriatric General internal Others Patients without comorbidity Patients with at least 1 comorbidity 1. Other infection 2. Other acute disease 3. Hepatic disease 4. Renal disease 5. Chronic disease 6. Not documented No Yes
456 (70.3) 193 (29.7) 68 (51–79) 50.0 (45–55) 358 (55.2) 291 (44.8) 336 (52.3) 101 (15.7) 205 (32.0) 396 (61.9) 44 (6.9) 30 (4.7) 35 (5.5) 135 (21.0) 228 (35.1) 421 (64.9) 14 (2.0) 55 (7.8) 25 (3.6) 38 (5.4) 379 (54.0) 191 (27.2) 532 (86.5) 83 (13.5) 11 (8–16)
*Median (IQR)
Copyright © 2014 John Wiley & Sons, Ltd.
Pharmacoepidemiology and Drug Safety, 2014 DOI: 10.1002/pds
h. t. trinh et al.
An overview of antibiotics used is presented in Table 2. The most frequently prescribed antibiotic classes were third-generation cephalosporins (C3G) (458 courses), quinolones (276 courses), and penicillins (101 courses). A monotherapy regimen accounted for 42.4% (275) of all the antibiotic regimens. C3G were most commonly prescribed (190/275) as monotherapy or combined antibiotic therapy. In dual and triple therapy, C3G were most prescribed and were used in combination with 10 of 12 other groups of antibiotics, but most frequently quinolones. The first-generation cephalosporins, aminoglycosides, macrolides, and carbapenems were prescribed in fewer than 1% of cases. Most (93.4%) patients received antibiotics intravenously despite most cases being mild. Half (51.2%) of the patients with a provisional diagnosis of CAP had microbiological culture tests performed. The most commonly requested specimen was sputum, followed by blood, and nasopharynx swab. Out of the 47.2% with positive cultures, the two leading microorganisms were Gram-negative strains including Moraxella catarrhalis and Klebsiella pneumonia. Streptococcus species accounted for 27.4% of positive cultures (Table 3). Other notable bacteria cultures included Staphylococcus species, Pseudomonas aeruginosa and Acinetobacter baumanii, which are related to multidrug resistance and normally associated with acquired infection in hospitals.12
Table 2. Antibiotic use for initial/empiric therapy* n (%) Total antibiotics used in initial/empiric therapy One Two Three or more Antibiotics group used most commonly as monotherapy First-generation of cephalosporin Second-generation of cephalosporin Third-generation of cephalosporin Fourth-generation of cephalosporin Quinolones Macrolides Penicillins Carbapenems Most common dual and triple combination C3G + aminoglycosides C3G + quinolones C3G + macrolides Penicillin + quinolones C3G + quinolones + aminoglycosides At least one intravenous antibiotic used in initial therapy
276 (42.5) 353 (54.4) 20 (3.1) 1 (0.15) 12 (1.85) 190 (29.3) 13 (2.0) 19 (2.93) 5 (0.77) 33 (5.08) 2 (0.31) 44 (6.8) 161 (24.8) 27 (4.2) 39 (6.0) 8 (1.2) 606 (93.4)
*The percentage was calculated as number of cases over sample size (n = 649).
Copyright © 2014 John Wiley & Sons, Ltd.
Multivariate analysis After adjusting for potential confounders, the odds of prescribing combined antibiotics as empiric therapy were significantly higher in the infectious disease wards and in the respiratory wards than in the general internal medicine ward (Table 4). Although there was not much variation across most hospitals, two hospitals were significantly less likely than the referent hospital to prescribe combined therapy and one hospital was significantly more likely to do so. Patients with moderate CAP were significantly less likely than those with mild CAP (referent) to receive combined therapy. Patient’s age, sex, area of residence, insurance status, transfer status, number of comorbidities or CAP severity, did not show significant association with the use of combined antibiotic therapy on admission. DISCUSSION To the best of our knowledge, this is the first nationwide profile of the incidence of CAP and its antibiotic management in hospitals in Vietnam. This retrospective record review shows that CAP was common, microbiological tests were requested or documented half the time, and there was widespread use of intravenous antibiotics, especially third-generation cephalosporins, including in combination with quinolones, regardless of CAP severity. Given the level of severity in most cases was mild to moderate, this proportion of confirmatory testing may be reasonable as the British Thoracic Society (BTS) recommends ambulatory treatment of mild cases, hence laboratory testing may not be warranted.13 Yet, findings also indicate that there was variation in the initial management according to the ward where the patient is admitted and an unclear relationship between severity and the decision of combination therapy. This study applied the CURB-65 index to assess the severity of CAP among the sample population. 64.1% of patients had mild CAP according to the index. This could be a reflection of our inclusion criteria. According to BTS, who proposed CURB-65 classification for their treatment guideline, patients with CURB-65 scores from 0 to 1 have low mortality rates of 0.7% to 2.1%, and hospitalization is not recommended.13 In Vietnam, information from the BTS guideline has influenced the development of several Vietnamese hospital guidelines showing similar recommendations. However, the results of this study revealed that the mild CAP patients were still admitted in these hospitals, and given intravenous treatment when the BTS guideline recommend that mild and even some moderate cases be managed with oral therapy. Pharmacoepidemiology and Drug Safety, 2014 DOI: 10.1002/pds
ANTIBIOTICS THERAPY FOR PNEUMONIA Table 3. Clinical profile of community-acquired pneumonia cases Clinical characteristics
N (%)
Severity (n = 649)
Diagnostic test (n = 649)
Bacteria culture result (n = 339) Isolated bacteria (n = 182)
Gram (+)
Gram ( )
Severe Moderate Mild Insufficient data to calculate CURB score Sputum Nasopharynx swab Transtracheal aspirate Antibodies Blood Rapid PCR test Pleural fluid Not documented Positive Negative Streptococcus viridians Streptococcus pneumonia Streptococcus sp. Staphylococcus aureus Staphylococcus epidermidis Staphylococcus sp. Enterococcus sp. Moraxella catarrhalis Klebsiella pneumonia Pseudomonas aeruginosa Pseudomonas sp. Acinetobacter baumanii Acinetobacter sp. Haemophillus influenza Enterobacter sp.
60 (9.2) 149 (23.0) 416 (64.1) 24 (3.7) 255 (39.3) 24 (3.7) 1 (0.2) 1 (0.2) 58 (8.9) 9 (1.4) 10 (1.5) 317 (48.8) 160 (47.2) 179 (52.8) 17 (9.3) 11(6.0) 22(12.1) 9(4.9) 4(2.2) 2(1.1) 11(6.0) 31(17) 21(11.5) 11(6.0) 8 (4.4) 3(1.6) 3(1.6) 2(1.1) 2(1.1) 25(13.7)
Other bacteria
This raises questions on the adequacy of the resources and capacity of the primary health system in Vietnam to deal with numerous mild CAP cases and whether there are social indications for admission to higher level hospitals beyond the clinical presentation. The Vietnamese national guideline in 2006 on the management of CAP does not consider severity, but a management guideline at Bach Mai Hospital recommends different antibiotic therapy based on severity. To improve the selection of antimicrobial agents in moderate and severe CAP, attempts should be made to identify the causal pathogens through microbiological investigations because clinical presentation and radiologic features do not accurately predict the agent. While awaiting laboratory test results, initial empirical treatment for CAP should follow previously published data from surveillance of local or national resistance patterns rather than using broad spectrum antibiotics.3 BTS proposes that bacteriological tests should not be routinely done for mild CAP patients in the community. The proportion of patients with bacteriology tests in our study population was 52.2%, and there was no significant association with the severity (χ 2 = 1.062, p = 0.786). The most common diagnostic test was the sputum sample, but this can be contaminated with other Copyright © 2014 John Wiley & Sons, Ltd.
bacteria that are commonly present in the upper respiratory tract. The quantification culture can be applied to minimize misleading results; however, its high cost limits the application of this test in routine practice in Vietnam, which is reflected in the study results. In this nationwide study, there was a large variation in the antibiotics used in empirical treatment of CAP —38 active substances belonging to 13 antibiotic groups. The most common antibiotic groups were C3G (458 courses) and quinolones (276 courses). According to guidelines of the American Thoracic Society, BTS,13,14 and Bach Mai and Cho Ray Hospitals, cefotaxime or ceftriaxone should only be used to treat severe CAP. Furthermore, the C3G which have efficacy on P. aeruginosa such as ceftazidime and cefoperazone/sulbactam, should be reserved for CAP cases when P. aeruginosa is suspected. This study, however, found that while only 9.2% of CAP cases were severe, C3G were widely used for other levels of severity. The most common antibiotics were cefoperazone/ sulbactam (116 courses); cefotaxime (113 courses); ceftazidime (86 courses); and ceftriaxone (66 courses). With the widespread use of C3G, it is well known that doses of common prescriptions under minimum inhibitory concentration will facilitate the Klebsiella to Pharmacoepidemiology and Drug Safety, 2014 DOI: 10.1002/pds
h. t. trinh et al. Table 4. Univariate and multivariate analysis of the independent predictors of using combined antibiotic therapy in empirical treatments (n = 615) Univariate analysis Odds ratio (95% CI) Age (year) (n = 615)
Multivariate analysis p
Odds ratio (95% CI)
p
1.00 (0.99–1.0)
0.438
1.00 (0.98–1.01)
0.477
Sex Male (n = 342) Female (n = 273) Residence Urban (n = 321) Rural (n = 198) Semirural (n = 96)
1 0.75 (0.54–1.04)
0.077
1 0.73 (0.50–1.07)
0.113
1 0.80 (0.56–1.14) 0.84 (0.53–1.33)
0.222 0.454
1 1.11 (0.70–1.76) 1.05 (0.61–1.80)
0.657 0.860
Medical insurance Yes (n = 462) No (n = 153)
1 0.75 (0.52–1.08)
0.117
1 0.76 (0.47–1.20)
0.237
Transferred from another hospital No (n = 532) Yes (n = 83)
1 0.98 (0.62–1.56)
0.930
1 1.25 (0.68–2.30)
0.481
Number of comorbidities 0 (n = 214) 1 comorbidity (n = 230) ≥2 comorbidity (n = 171)
1 1.00 (0.69–1.45) 1.01 (0.67–1.52)
0.991 0.953
1 1.03 (0.66–1.62) 0.94 (0.56–1.57)
0.888 0.805
Severity of CAP Mild (n = 396) Moderate (n = 139) Severe (n = 37) Not determined (n = 23)
1 0.56 (0.38–0.83) 1.36 (0.75–2.44) 1.05 (0.45–2.49)
0.004 0.309 0.904
1 0.55 (0.34–0.89) 1.12 (0.54–2.29) 1.21 (0.45–3.23)
0.015 0.766 0.701
Hospital B6* (n = 74) B1 (n = 62) B2 (n = 58) B3 (n = 60) B4 (n = 60) B5 (n = 51) B7 (n = 60) B8 (n = 68) B9 (n = 56) B10 (n = 66)
1 0.27 (0.13–0.56) 1.81 (0.84–3.88) 1.45 (0.70–3.03) 0.25 (0.12–0.51) 0.29 (0.14–0.61) 0.92 (0.46–1.87) 0.61 (0.31–1.19) 1.21 (0.58–2.53) 2.13 (1.00–4.55)
0.000 0.130 0.319 0.000 0.001 0.826 0.148 0.606 0.050
1 0.21 (0.10–0.47) 1.61 (0.71–3.64) 1.00 (0.41–2.42) 0.21 (0.09–0.46) 0.78 (0.24–2.48) 2.86 (0.94–8.72) 1.63 (0.66–4.05) 2.11 (0.79–5.66) 2.46 (1.08–5.59)
0.000 0.252 0.991 0.000 0.673 0.064 0.288 0.135 0.031
Ward General internal medicine (n = 32) Respiratory (n = 385) Infectious diseases (n = 40) Geriatric (n = 30) Other wards (n = 128)
1 0.91 (0.43–1.91) 3.40 (1.10–10.47) 0.22 (0.07–0.64) 0.48 (0.22–1.07)
0.803 0.033 0.006 0.072
1 2.51 (1.00–6.29) 12.33 (3.29–46.17) 0.45 (0.10–1.89) 0.69 (0.27–1.77)
0.050 0.000 0.274 0.435
*Selected as the reference due to the largest number of community-acquired pneumonia cases remained after excluding missing data.
produce extended spectrum beta lactamase and beta-lactam-resistant Acinetobacter.15 This can also result in an increase of resistance of P. aeruginosa to C3G. Among the quinolones, the most commonly prescribed were ciprofloxacin (130 courses) and levofloxacin (102 courses), followed by moxifloxacin, ofloxacin, and pefloxacin. According to Bach Mai Hospital’s guidelines to treat CAP, respiratory quinolones (levofloxacin and moxifloxacin) are preferred, while ciprofloxacin should only be used in severe CAP or suspected P. aeruginosa or suspected Gram-negative bacilli CAP. Hence, this study showed that most quinolone use was inappropriate, especially in combination with C3G. Quinolones should be carefully restricted as they increase the risk of Copyright © 2014 John Wiley & Sons, Ltd.
production of resistant and multi-resistant bacteria.15 In addition to C3G and quinolones, antibiotics that are effective against Gram-negative bacteria include aminoglycosides, C4G, and carbapenem. In monotherapy, the four most commonly prescribed antibiotics were cefotaxime, cefoperazone/ sulbactam, ceftazidime, and amoxicillin/clavulanate. Fluoroquinolones, which are recommended for monotherapy to decrease duration of antimicrobial therapy, the length of hospital stay and development of resistance in patients with CAP,16 were not being prescribed in our target hospitals. The selection of single or multiple therapy was not related to the severity of CAP and in fact, moderate Pharmacoepidemiology and Drug Safety, 2014 DOI: 10.1002/pds
ANTIBIOTICS THERAPY FOR PNEUMONIA
cases were less likely to be treated with a combined regimen, which is contrary to the CAP management guidelines at Bach Mai Hospital. However, there is ongoing international discussion about whether the choice of single or multiple empiric therapy has an impact on the outcomes of hospitalization.17 Combination therapy has been recommended for management of severe CAP by several scientific societies in the past 6 years.13,18 Findings indicate that the most widespread combinations prescribed were C3G + quinolone, C3G + aminoglycoside, penicillin + quinolone, and C3G + macrolide. Of these, quinolones were the most common antibiotic to be used with beta-lactams. In this study, of 161 courses of C3G + quinolone, there were nearly half (72/161 courses, 44.7%) of C3G + respiratory quinolones (levofloxacin, moxifloxacin); and 89 courses (55.3%) of C3G + other quinolone such as ciprofloxacin and ofloxacin. In theory, C3G can be combined with a quinolone if there is a need to increase the efficacy against Gram-negative bacilli. However, this combination was chosen widely to treat mild CAP cases in this study (Table 2). Similarly, C3G + aminoglycosides, which is recommended for Gram-negative infection, was used for quite a large number of mild CAP patients. The antibiotic combinations being used for empirical treatment of CAP in Vietnam need to be reviewed. In general, our findings on the use of antibiotics for CAP in 10 hospitals raises questions of whether or not (i) indication for hospitalization is more than necessary; (ii) oral antibiotics are carefully considered at hospitalization before prescribing injection products; and (iii) empirical antibiotic regimens are based on bacteriological tests and resistant patterns of a specific hospital. To answer these questions, it is necessary to have antibiotic stewardship program at national-wide level and hospital level also. It is the fact that such program is not available in all 10 surveyed hospitals. When establishing the national guideline for the treatment of CAP, one should consider taking into account these issues, which could overcome some of the problems identified in this study and therefore, it is expected to solve major problems of CAP treatment in Vietnam. Strengths of this study included the random selection of patients from monthly admission lists, the nationwide coverage, the use of validated tools to determine severity, the use of a standard protocol for data extraction by trained clinical pharmacists, and the detailed documentation of treatments. The study’s main limitation was its retrospective nature based on routine clinical records, leading to incomplete evidence of the profile of possible pathogens due to a lack of diagnostic testing documentation in the hospital records. Due to the retrospective study design, Copyright © 2014 John Wiley & Sons, Ltd.
the doctor’s reason for choice of antibiotics upon admission could not be obtained. Finally, the decision on severity was made on the available notes on the clinical record, and therefore misclassification could have occurred. CONCLUSION This national profile of CAP management suggests that many cases did not warrant hospital admission, intravenous administration of antibiotics or combined antibiotic therapy. The results will support national planning on antibiotic resistance, the development of a new standard national antibiotic therapeutic guideline for CAP in Vietnam, and assist in the development of interventions for health professionals and consumers to improve the quality of use of antibiotics. Further research is needed to determine factors influencing decisions on laboratory testing, hospitalization, use of intravenous route, choice of antibiotic, and recommendation for combination therapy. CONFLICT OF INTEREST The authors declare no conflict of interest. KEY POINTS Irrational antibiotic combinations for CAP are common. • Hospitalization, choice of intravenous route, and use of combination antibiotic therapy occurred without correlation to CAP severity in Vietnamese hospitals. • Antibiotic combination highly varies among hospitals. • Further research into the factors influencing these decisions is needed.
•
ETHICS STATEMENT This study was retrospective on medical records with anonymous individual patients’ and hospitals’ information; consequently, it did not constitute human subjects research requiring review by an Institutional Review Board. ACKNOWLEDGEMENTS This project was undertaken as part of the Health System Strengthening Project—Intervention 2.1 Establishing a national pharmacovigilance system, conducted by HUP and NPS with the financial support of the Global Fund. Pharmacoepidemiology and Drug Safety, 2014 DOI: 10.1002/pds
h. t. trinh et al.
The assistance of hospital staff in retrieving medical records and that of clinical pharmacy students at HUP in entering data is also appreciated.
REFERENCES 1. WHO. The World Medicines Situation 2011 - Rational Use of Medicines (3 edn). WHO Geneva, 2011. 2. http://www.cimsi.org.vn/THONGKE.aspx?action=thongke&lang=vi. 3. World Health Organisation. World Health Organisation Global Strategy for Containment of Antimicrobial Resistance. 2001 [cited 2013 May]; Available from: http://www.who.int/csr/resources/publications/drugresist/ WHO_CDS_CSR_DRS_2001_2_EN/en/. 4. Martin SJ, Micek ST, Wood GC. Antimicrobial resistance: consideration as an adverse drug event. Crit Care Med 2010; 38(6): S155–S161. 5. Kim SH, Jong JH, Chung DR. Changing trends in antimicrobial resistance and serotypes of streptococcus pneumoniae isolates in Asian countries: an Asian Network for Surveillance of Resistant Pathogens (ANSORP) study. Antimicrob Agents Chemother 2012; 56(3): 1418–1426. 6. Albrich WC, Monnet DL, Harbarth S. Antibiotic selection pressure and resistance in streptococcus pneumoniae and streptococcus pyogenes. Emerg Infect Dis 2004; 10(3): 514–517. 7. Goossens H. Antibiotic consumption and link to resistance. Mini-review. Clin Microbiol Infect 2009; 15(Suppl 3): 12–15. 8. Van, PH, Binh PT. Antibiotic resistance rate of S. pneumoniae and H. influenzae in respiratory infection patients: a result from a multi-central study in Vietnam (SOAR) 2010–2011. Vietnamese J Pract Med 2012; 855: 6–11.
Copyright © 2014 John Wiley & Sons, Ltd.
9. Vu DH, van Rein N, Cobelens FG, Nguyen TT, Le VH, Brouwers JR. Suspected tuberculosis case detection and referral in private pharmacies in Viet Nam. Int J Tuberc Lung Dis 2012; 16(12): 1625–1629. 10. Nguyen QH, Nguyen TK, Ho D, et al. Unnecessary antibiotic use for mild acute respiratory infections during 28-day follow-up of 823 children under five in rural Vietnam. Trans R Soc Trop Med Hyg 2011; 105(11): 628–636. 11. American Thoracic Society and Infectious Diseases Society of America. Guidelines for the management of adults with hospital-acquired, ventilator-associated, and healthcareassociated pneumonia. Am J Respir Crit Care Med 2005; 171(4): 388–416. 12. Jones RN. Microbial etiologies of hospital-acquired bacterial pneumonia and ventilator-associated bacterial pneumonia. Clin Infect Dis 2010; 51(Suppl 1): S81–S87. 13. Lim WS, Baudouin SV, George RC, et al. BTS guidelines for the management of community acquired pneumonia in adults: update 2009. Thorax, 2009; 64(3): 121434. 14. Niederman MS, Mandell LA, Anzueto A, et al. Guidelines for the management of adults with community-acquired pneumonia. Diagnosis, assessment of severity, antimicrobial therapy, and prevention. Am J Respir Crit Care Med 2001; 163(7): 1730–1754. 15. Paterson DL. “Collateral damage” from cephalosporin or quinolone antibiotic therapy. Clin Infect Dis 2004; 38(Suppl 4): S341–S345. 16. Albertson TE, Dean NC, El Solh AA, Gotfried MH, Kaplan C, Niederman MS. Fluoroquinolones in the management of community-acquired pneumonia. Int J Clin Pract 2010; 64(3): 378–388. 17. Laterre PF. Monotherapy or combination therapy for hospitalized patients with community-acquired pneumonia: not yet the end of the story? Clin Infect Dis 2008; 46(10): 1510–1512. 18. 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; 1(44): S27–S72.
Pharmacoepidemiology and Drug Safety, 2014 DOI: 10.1002/pds
ORIGINAL REPORT
Antibiotic therapy for inpatients with community-acquired pneumonia in a developing country Hieu T. Trinh1, Phuong H. Hoang1, Magnolia Cardona-Morrell2, Hai T. Nguyen1, Dinh Hoa Vu1, Phuong T. X. Dong1, Thao T. B. Cao1, Son T. Nguyen1, Van T. T. Pham1, Le Moss2, Kathryn Dinh2, Jonathan Dartnell2 and Huong T. L. Nguyen1* 1 2
Department of Clinical Pharmacy, Hanoi University of Pharmacy, Hanoi, Vietnam NPS MedicineWise (NPS), Sydney, Australia
ABSTRACT Purpose The aim of this study was to identify antibiotic prescription patterns for community-acquired pneumonia (CAP) in Vietnam. Methods Medical records for CAP adult patients admitted to 10 hospitals across the country were randomly selected from admission lists during the peak pneumonia season. CAP cases were identified from manual record reviews by clinical pharmacists. Data was collected using a standard data collection tool including patient clinical features on admission, comorbidities, microbiological culture results, and antibiotic regimens. Pneumonia severity was estimated using the CURB-65 score. Results A total of 649 medical records for adult patients (55.2% male and 52.3% urban residents, median age 68 years) met the selection criteria for CAP. Pneumonia severity was assessed as mild (64.1% of patients), moderate (23.0%), and severe (9.2%). Antibiotics were most frequently administered intravenously (93.4%) and as combination therapy (dual therapy 54.4%, monotherapy 42.5%, and triple therapy 3.1% of patients) regardless of CAP severity. Third-generation cephalosporins were used most frequently (29.3% as monotherapy and 40.4% as combination therapy). Third-generation cephalosporins were most commonly combined with penicillins and/or quinolones. Conclusions This first nationwide study provides a baseline profile of antibiotic use in the treatment of CAP. Third-generation cephalosporins were widely used for initial empirical management of CAP, often in combination with quinolones, regardless of CAP severity. The study will assist in providing an evidence base to inform new national antibiotic guidelines for CAP management and will contribute locally relevant data for the national master plan addressing antibiotic resistance and the development of educational interventions to improve CAP management. Copyright © 2014 John Wiley & Sons, Ltd. key words—community-acquired pneumonia (CAP); antibiotics; drug utilization evaluation; Vietnam; developing country; pharmacoepidemiology Received 5 September 2013; Revised 22 February 2014; Accepted 22 February 2014
INTRODUCTION Lacking of standard guidelines is one of main reasons of antibiotics misuse in treating pneumonia worldwide. According to WHO reports, there were only 70% pneumonia cases that received appropriate antibiotics in developing countries. Proportion of patient adherence in these countries was about 50%. Less than half of the countries had effective measures to increase quality use of medicines including antibiotics.1 Vietnam is a developing country in which community-acquired pneumonia (CAP) is a common *Correspondence to: Huong TL Nguyen, Department of Clinical Pharmacy, Hanoi University of Pharmacy, Hanoi, Vietnam. E-mail: [email protected]
Copyright © 2014 John Wiley & Sons, Ltd.
infectious disease that is associated with high morbidity and mortality.2 In Vietnam, respiratory diseases have high prevalence (19.09%) and are ranked as the 10th highest cause of mortality. Among respiratory diseases nationwide, pneumonia showed an overall prevalence of 409 per 100 000 population.2,3 Over the past decade, antibiotic resistance has become a serious problem in many areas of the world including Vietnam.3,4 Data from Vietnam show an alarming rate of resistance, and this rate is continuing to increase.5 In particular, the most common bacteria strains causing CAP have developed substantial resistance to commonly used antibiotics.6,7 A multi-central study conducted in 11 hospitals in Vietnam showed that isolated Streptococcus pneumoniae from
h. t. trinh et al.
infectious respiratory patients was resistant to macrolides, cefuroxime, and cefaclor with rates of 95.0%, 71.4%, and 87.5%, respectively.8 In this study, Haemophilus influenza, which is also a common pathogen in CAP, found to be highly resistant to oral antibiotics such as cotrimoxazole (82.5%), tetracycline (92.5%), and chloramphenicol (78.0%).8 These high resistance rates could be associated with the overuse of these antibiotics in a community where people could straightforwardly purchase medicines without prescriptions.9,10 Thus, the empirical treatments for CAP patients could be challenged. National and local standard treatment guidelines are important tools to tailor antibiotic choices. At the time of this study, however, the national guideline for CAP management from 2006 was not appropriate with present antibiotic resistance. Furthermore, none of hospitals have local standard treatment guidelines based on their own antibiotic resistance rate. The present study aimed to provide a cross-sectional baseline estimate of actual antibiotic prescribing practice in hospitals in Vietnam with a view to contributing evidence to support the development of new standard national antibiotic therapeutic guidelines for CAP in Vietnam. In addition, the results of this survey could assist Vietnamese authorities in determining appropriate interventions for health professionals and consumers to improve the quality of use of antibiotics. METHOD A retrospective study was undertaken to investigate antibiotic use in the treatment of CAP cases admitted to hospital in Vietnam during the identified peak season. Setting and study period A convenience sample of 10 national and provincial hospitals located in large cities across the country (North, Central, and South Vietnam) was selected from a fixed list of 16 hospitals chosen by the funding body (The Global Fund to Fight AIDS, Tuberculosis, and Malaria). These were chosen based on the availability of hospital staff to support the study data collection and a history of close collaboration with the Department of Clinical Pharmacy at Hanoi University of Pharmacy (HUP). The study period for each hospital was different depending on the peak season for respiratory illness in the region. Peak season was the period in the most recent year (2011) that included the highest number of admissions for respiratory disease determined Copyright © 2014 John Wiley & Sons, Ltd.
through examination of ICD-10 codes from hospital data. While for most hospitals (8/10), the peak season covered at least 3 months, in other hospitals (2/10), the peak season was only for 1 month. Target conditions Community-acquired pneumonia identification was based on either an admission or discharge diagnosis using ICD-10 codes J13.0-J18. Patient inclusion and exclusion criteria Eligible patients were adults aged 16 years and older, admitted for at least 24 h with a primary diagnosis of bacterial pneumonia acquired in the community (with or without underlying COPD or asthma) and were prescribed at least one antibiotic with complete information on clinical indication and drug regimen to enable analysis and clinical interpretation. Excluded patients were those who were admitted for non-bacterial pneumonia (viral, parasitic, or fungal) or did not receive antibiotics for pneumonia, hospital-acquired pneumonia, aspiration pneumonia, cases only seen in outpatient or emergency and discharged on the same day (due to absence, brevity or incompleteness of records) or who were admitted to intensive care units (due to the study’s time and resource constraints that precluded review of extensive records). Non-bacterial pneumonia was confirmed if the viral and/or parasitic test was positive. Aspiration pneumonia was confirmed if the doctor wrote “aspiration pneumonia” in the diagnosis section. Hospital-acquired pneumonia was defined as the pneumonia that occurs at least 48 h after hospitalization.11 A data collection tool to capture all variables was constructed and piloted in three hospitals before the data collection phase commenced. Then the standard data collection toolkit was used by trained pharmacists performing medical record review. Sample size calculations Sixty cases of CAP were randomly selected from each of the 10 hospitals. Sixty was estimated as a minimum number of cases required to achieve distribution of variables to approximate the Normal distribution and hence to have a representative sample of records from each hospital. This minimum number also catered for missing or incomplete items in the selected records and enabled meaningful and accurate inferences from subgroup analysis. Minimum data required for a record to be considered complete included patient Pharmacoepidemiology and Drug Safety, 2014 DOI: 10.1002/pds
ANTIBIOTICS THERAPY FOR PNEUMONIA
demographics, pneumonia diagnosis, antibiotic name, strength, dose, and route of administration. Data analysis Comparisons of demographic parameters and CAP severity used the chi-squared test to assess differences in proportions between groups. Statistical significance level was set at p = 0.05. The prevalence of antibiotic prescription was calculated as the number of patients receiving any antibiotic out of the total number of CAP patients studied. Data analysis was performed using Stata v.11.0 software program (StataCorp, College Station TX, USA). Multivariate logistic regression analysis was conducted to investigate the determinants of prescribing patterns in hospital (dichotomous outcome combined therapy) as empirical treatment. Combined therapy was defined as a therapy in which patients use dual or triple antibiotic combination. Independent variables examined were age, sex, area of residence, medical insurance status, transferred status, number of comorbidities, severity of CAP, hospital type, and ward type. As the purpose of the research group was to show the insignificant effect of less strong predictors, all variables that may have potential effect were included in the model.
RESULTS Demographic profile of patients Between January 2011 and December 2011 (depending on the peak season of pneumonia in each region), 649 medical records with CAP diagnosis from 10 hospitals of Vietnam were included in this study. The patient characteristics are presented in Table 1. CAP admitted patients were mostly elderly patients (median age 68 years, IQR 51–79), just over half of them lived in urban areas, and distribution of admissions by sex was almost equal. Clinical profile of community-acquired pneumonia cases Of 649 patients with additional clinical information available in the record, half had comorbidities. The median length of hospitalization was approximately 11 days (IQR: 8–16). Most patients were classified as having mild or moderate CAP. CURB-65 score could not be calculated for 24 patients (3.7%) during the first 2 days of admission, due to lack of information for at least one of the parameters. They were included in the overall study results but not included in some subgroup analyses.
Table 1. Demographic profile of patients with community-acquired pneumonia Demographic parameter Hospital level (n = 649) Age (year) * (n = 636) Weight (kg)* (n = 175) Sex (n = 649) Residence (n = 642) Ward (n = 640)
Comorbidity (n = 649) Type of comorbidity (n = 649)
Transferred from another hospital (n = 649) Length of hospital stay* (days) (n = 649)
Patients number (% total) National Provincial Male Female Urban Suburban Rural Respiratory Infectious diseases Geriatric General internal Others Patients without comorbidity Patients with at least 1 comorbidity 1. Other infection 2. Other acute disease 3. Hepatic disease 4. Renal disease 5. Chronic disease 6. Not documented No Yes
456 (70.3) 193 (29.7) 68 (51–79) 50.0 (45–55) 358 (55.2) 291 (44.8) 336 (52.3) 101 (15.7) 205 (32.0) 396 (61.9) 44 (6.9) 30 (4.7) 35 (5.5) 135 (21.0) 228 (35.1) 421 (64.9) 14 (2.0) 55 (7.8) 25 (3.6) 38 (5.4) 379 (54.0) 191 (27.2) 532 (86.5) 83 (13.5) 11 (8–16)
*Median (IQR)
Copyright © 2014 John Wiley & Sons, Ltd.
Pharmacoepidemiology and Drug Safety, 2014 DOI: 10.1002/pds
h. t. trinh et al.
An overview of antibiotics used is presented in Table 2. The most frequently prescribed antibiotic classes were third-generation cephalosporins (C3G) (458 courses), quinolones (276 courses), and penicillins (101 courses). A monotherapy regimen accounted for 42.4% (275) of all the antibiotic regimens. C3G were most commonly prescribed (190/275) as monotherapy or combined antibiotic therapy. In dual and triple therapy, C3G were most prescribed and were used in combination with 10 of 12 other groups of antibiotics, but most frequently quinolones. The first-generation cephalosporins, aminoglycosides, macrolides, and carbapenems were prescribed in fewer than 1% of cases. Most (93.4%) patients received antibiotics intravenously despite most cases being mild. Half (51.2%) of the patients with a provisional diagnosis of CAP had microbiological culture tests performed. The most commonly requested specimen was sputum, followed by blood, and nasopharynx swab. Out of the 47.2% with positive cultures, the two leading microorganisms were Gram-negative strains including Moraxella catarrhalis and Klebsiella pneumonia. Streptococcus species accounted for 27.4% of positive cultures (Table 3). Other notable bacteria cultures included Staphylococcus species, Pseudomonas aeruginosa and Acinetobacter baumanii, which are related to multidrug resistance and normally associated with acquired infection in hospitals.12
Table 2. Antibiotic use for initial/empiric therapy* n (%) Total antibiotics used in initial/empiric therapy One Two Three or more Antibiotics group used most commonly as monotherapy First-generation of cephalosporin Second-generation of cephalosporin Third-generation of cephalosporin Fourth-generation of cephalosporin Quinolones Macrolides Penicillins Carbapenems Most common dual and triple combination C3G + aminoglycosides C3G + quinolones C3G + macrolides Penicillin + quinolones C3G + quinolones + aminoglycosides At least one intravenous antibiotic used in initial therapy
276 (42.5) 353 (54.4) 20 (3.1) 1 (0.15) 12 (1.85) 190 (29.3) 13 (2.0) 19 (2.93) 5 (0.77) 33 (5.08) 2 (0.31) 44 (6.8) 161 (24.8) 27 (4.2) 39 (6.0) 8 (1.2) 606 (93.4)
*The percentage was calculated as number of cases over sample size (n = 649).
Copyright © 2014 John Wiley & Sons, Ltd.
Multivariate analysis After adjusting for potential confounders, the odds of prescribing combined antibiotics as empiric therapy were significantly higher in the infectious disease wards and in the respiratory wards than in the general internal medicine ward (Table 4). Although there was not much variation across most hospitals, two hospitals were significantly less likely than the referent hospital to prescribe combined therapy and one hospital was significantly more likely to do so. Patients with moderate CAP were significantly less likely than those with mild CAP (referent) to receive combined therapy. Patient’s age, sex, area of residence, insurance status, transfer status, number of comorbidities or CAP severity, did not show significant association with the use of combined antibiotic therapy on admission. DISCUSSION To the best of our knowledge, this is the first nationwide profile of the incidence of CAP and its antibiotic management in hospitals in Vietnam. This retrospective record review shows that CAP was common, microbiological tests were requested or documented half the time, and there was widespread use of intravenous antibiotics, especially third-generation cephalosporins, including in combination with quinolones, regardless of CAP severity. Given the level of severity in most cases was mild to moderate, this proportion of confirmatory testing may be reasonable as the British Thoracic Society (BTS) recommends ambulatory treatment of mild cases, hence laboratory testing may not be warranted.13 Yet, findings also indicate that there was variation in the initial management according to the ward where the patient is admitted and an unclear relationship between severity and the decision of combination therapy. This study applied the CURB-65 index to assess the severity of CAP among the sample population. 64.1% of patients had mild CAP according to the index. This could be a reflection of our inclusion criteria. According to BTS, who proposed CURB-65 classification for their treatment guideline, patients with CURB-65 scores from 0 to 1 have low mortality rates of 0.7% to 2.1%, and hospitalization is not recommended.13 In Vietnam, information from the BTS guideline has influenced the development of several Vietnamese hospital guidelines showing similar recommendations. However, the results of this study revealed that the mild CAP patients were still admitted in these hospitals, and given intravenous treatment when the BTS guideline recommend that mild and even some moderate cases be managed with oral therapy. Pharmacoepidemiology and Drug Safety, 2014 DOI: 10.1002/pds
ANTIBIOTICS THERAPY FOR PNEUMONIA Table 3. Clinical profile of community-acquired pneumonia cases Clinical characteristics
N (%)
Severity (n = 649)
Diagnostic test (n = 649)
Bacteria culture result (n = 339) Isolated bacteria (n = 182)
Gram (+)
Gram ( )
Severe Moderate Mild Insufficient data to calculate CURB score Sputum Nasopharynx swab Transtracheal aspirate Antibodies Blood Rapid PCR test Pleural fluid Not documented Positive Negative Streptococcus viridians Streptococcus pneumonia Streptococcus sp. Staphylococcus aureus Staphylococcus epidermidis Staphylococcus sp. Enterococcus sp. Moraxella catarrhalis Klebsiella pneumonia Pseudomonas aeruginosa Pseudomonas sp. Acinetobacter baumanii Acinetobacter sp. Haemophillus influenza Enterobacter sp.
60 (9.2) 149 (23.0) 416 (64.1) 24 (3.7) 255 (39.3) 24 (3.7) 1 (0.2) 1 (0.2) 58 (8.9) 9 (1.4) 10 (1.5) 317 (48.8) 160 (47.2) 179 (52.8) 17 (9.3) 11(6.0) 22(12.1) 9(4.9) 4(2.2) 2(1.1) 11(6.0) 31(17) 21(11.5) 11(6.0) 8 (4.4) 3(1.6) 3(1.6) 2(1.1) 2(1.1) 25(13.7)
Other bacteria
This raises questions on the adequacy of the resources and capacity of the primary health system in Vietnam to deal with numerous mild CAP cases and whether there are social indications for admission to higher level hospitals beyond the clinical presentation. The Vietnamese national guideline in 2006 on the management of CAP does not consider severity, but a management guideline at Bach Mai Hospital recommends different antibiotic therapy based on severity. To improve the selection of antimicrobial agents in moderate and severe CAP, attempts should be made to identify the causal pathogens through microbiological investigations because clinical presentation and radiologic features do not accurately predict the agent. While awaiting laboratory test results, initial empirical treatment for CAP should follow previously published data from surveillance of local or national resistance patterns rather than using broad spectrum antibiotics.3 BTS proposes that bacteriological tests should not be routinely done for mild CAP patients in the community. The proportion of patients with bacteriology tests in our study population was 52.2%, and there was no significant association with the severity (χ 2 = 1.062, p = 0.786). The most common diagnostic test was the sputum sample, but this can be contaminated with other Copyright © 2014 John Wiley & Sons, Ltd.
bacteria that are commonly present in the upper respiratory tract. The quantification culture can be applied to minimize misleading results; however, its high cost limits the application of this test in routine practice in Vietnam, which is reflected in the study results. In this nationwide study, there was a large variation in the antibiotics used in empirical treatment of CAP —38 active substances belonging to 13 antibiotic groups. The most common antibiotic groups were C3G (458 courses) and quinolones (276 courses). According to guidelines of the American Thoracic Society, BTS,13,14 and Bach Mai and Cho Ray Hospitals, cefotaxime or ceftriaxone should only be used to treat severe CAP. Furthermore, the C3G which have efficacy on P. aeruginosa such as ceftazidime and cefoperazone/sulbactam, should be reserved for CAP cases when P. aeruginosa is suspected. This study, however, found that while only 9.2% of CAP cases were severe, C3G were widely used for other levels of severity. The most common antibiotics were cefoperazone/ sulbactam (116 courses); cefotaxime (113 courses); ceftazidime (86 courses); and ceftriaxone (66 courses). With the widespread use of C3G, it is well known that doses of common prescriptions under minimum inhibitory concentration will facilitate the Klebsiella to Pharmacoepidemiology and Drug Safety, 2014 DOI: 10.1002/pds
h. t. trinh et al. Table 4. Univariate and multivariate analysis of the independent predictors of using combined antibiotic therapy in empirical treatments (n = 615) Univariate analysis Odds ratio (95% CI) Age (year) (n = 615)
Multivariate analysis p
Odds ratio (95% CI)
p
1.00 (0.99–1.0)
0.438
1.00 (0.98–1.01)
0.477
Sex Male (n = 342) Female (n = 273) Residence Urban (n = 321) Rural (n = 198) Semirural (n = 96)
1 0.75 (0.54–1.04)
0.077
1 0.73 (0.50–1.07)
0.113
1 0.80 (0.56–1.14) 0.84 (0.53–1.33)
0.222 0.454
1 1.11 (0.70–1.76) 1.05 (0.61–1.80)
0.657 0.860
Medical insurance Yes (n = 462) No (n = 153)
1 0.75 (0.52–1.08)
0.117
1 0.76 (0.47–1.20)
0.237
Transferred from another hospital No (n = 532) Yes (n = 83)
1 0.98 (0.62–1.56)
0.930
1 1.25 (0.68–2.30)
0.481
Number of comorbidities 0 (n = 214) 1 comorbidity (n = 230) ≥2 comorbidity (n = 171)
1 1.00 (0.69–1.45) 1.01 (0.67–1.52)
0.991 0.953
1 1.03 (0.66–1.62) 0.94 (0.56–1.57)
0.888 0.805
Severity of CAP Mild (n = 396) Moderate (n = 139) Severe (n = 37) Not determined (n = 23)
1 0.56 (0.38–0.83) 1.36 (0.75–2.44) 1.05 (0.45–2.49)
0.004 0.309 0.904
1 0.55 (0.34–0.89) 1.12 (0.54–2.29) 1.21 (0.45–3.23)
0.015 0.766 0.701
Hospital B6* (n = 74) B1 (n = 62) B2 (n = 58) B3 (n = 60) B4 (n = 60) B5 (n = 51) B7 (n = 60) B8 (n = 68) B9 (n = 56) B10 (n = 66)
1 0.27 (0.13–0.56) 1.81 (0.84–3.88) 1.45 (0.70–3.03) 0.25 (0.12–0.51) 0.29 (0.14–0.61) 0.92 (0.46–1.87) 0.61 (0.31–1.19) 1.21 (0.58–2.53) 2.13 (1.00–4.55)
0.000 0.130 0.319 0.000 0.001 0.826 0.148 0.606 0.050
1 0.21 (0.10–0.47) 1.61 (0.71–3.64) 1.00 (0.41–2.42) 0.21 (0.09–0.46) 0.78 (0.24–2.48) 2.86 (0.94–8.72) 1.63 (0.66–4.05) 2.11 (0.79–5.66) 2.46 (1.08–5.59)
0.000 0.252 0.991 0.000 0.673 0.064 0.288 0.135 0.031
Ward General internal medicine (n = 32) Respiratory (n = 385) Infectious diseases (n = 40) Geriatric (n = 30) Other wards (n = 128)
1 0.91 (0.43–1.91) 3.40 (1.10–10.47) 0.22 (0.07–0.64) 0.48 (0.22–1.07)
0.803 0.033 0.006 0.072
1 2.51 (1.00–6.29) 12.33 (3.29–46.17) 0.45 (0.10–1.89) 0.69 (0.27–1.77)
0.050 0.000 0.274 0.435
*Selected as the reference due to the largest number of community-acquired pneumonia cases remained after excluding missing data.
produce extended spectrum beta lactamase and beta-lactam-resistant Acinetobacter.15 This can also result in an increase of resistance of P. aeruginosa to C3G. Among the quinolones, the most commonly prescribed were ciprofloxacin (130 courses) and levofloxacin (102 courses), followed by moxifloxacin, ofloxacin, and pefloxacin. According to Bach Mai Hospital’s guidelines to treat CAP, respiratory quinolones (levofloxacin and moxifloxacin) are preferred, while ciprofloxacin should only be used in severe CAP or suspected P. aeruginosa or suspected Gram-negative bacilli CAP. Hence, this study showed that most quinolone use was inappropriate, especially in combination with C3G. Quinolones should be carefully restricted as they increase the risk of Copyright © 2014 John Wiley & Sons, Ltd.
production of resistant and multi-resistant bacteria.15 In addition to C3G and quinolones, antibiotics that are effective against Gram-negative bacteria include aminoglycosides, C4G, and carbapenem. In monotherapy, the four most commonly prescribed antibiotics were cefotaxime, cefoperazone/ sulbactam, ceftazidime, and amoxicillin/clavulanate. Fluoroquinolones, which are recommended for monotherapy to decrease duration of antimicrobial therapy, the length of hospital stay and development of resistance in patients with CAP,16 were not being prescribed in our target hospitals. The selection of single or multiple therapy was not related to the severity of CAP and in fact, moderate Pharmacoepidemiology and Drug Safety, 2014 DOI: 10.1002/pds
ANTIBIOTICS THERAPY FOR PNEUMONIA
cases were less likely to be treated with a combined regimen, which is contrary to the CAP management guidelines at Bach Mai Hospital. However, there is ongoing international discussion about whether the choice of single or multiple empiric therapy has an impact on the outcomes of hospitalization.17 Combination therapy has been recommended for management of severe CAP by several scientific societies in the past 6 years.13,18 Findings indicate that the most widespread combinations prescribed were C3G + quinolone, C3G + aminoglycoside, penicillin + quinolone, and C3G + macrolide. Of these, quinolones were the most common antibiotic to be used with beta-lactams. In this study, of 161 courses of C3G + quinolone, there were nearly half (72/161 courses, 44.7%) of C3G + respiratory quinolones (levofloxacin, moxifloxacin); and 89 courses (55.3%) of C3G + other quinolone such as ciprofloxacin and ofloxacin. In theory, C3G can be combined with a quinolone if there is a need to increase the efficacy against Gram-negative bacilli. However, this combination was chosen widely to treat mild CAP cases in this study (Table 2). Similarly, C3G + aminoglycosides, which is recommended for Gram-negative infection, was used for quite a large number of mild CAP patients. The antibiotic combinations being used for empirical treatment of CAP in Vietnam need to be reviewed. In general, our findings on the use of antibiotics for CAP in 10 hospitals raises questions of whether or not (i) indication for hospitalization is more than necessary; (ii) oral antibiotics are carefully considered at hospitalization before prescribing injection products; and (iii) empirical antibiotic regimens are based on bacteriological tests and resistant patterns of a specific hospital. To answer these questions, it is necessary to have antibiotic stewardship program at national-wide level and hospital level also. It is the fact that such program is not available in all 10 surveyed hospitals. When establishing the national guideline for the treatment of CAP, one should consider taking into account these issues, which could overcome some of the problems identified in this study and therefore, it is expected to solve major problems of CAP treatment in Vietnam. Strengths of this study included the random selection of patients from monthly admission lists, the nationwide coverage, the use of validated tools to determine severity, the use of a standard protocol for data extraction by trained clinical pharmacists, and the detailed documentation of treatments. The study’s main limitation was its retrospective nature based on routine clinical records, leading to incomplete evidence of the profile of possible pathogens due to a lack of diagnostic testing documentation in the hospital records. Due to the retrospective study design, Copyright © 2014 John Wiley & Sons, Ltd.
the doctor’s reason for choice of antibiotics upon admission could not be obtained. Finally, the decision on severity was made on the available notes on the clinical record, and therefore misclassification could have occurred. CONCLUSION This national profile of CAP management suggests that many cases did not warrant hospital admission, intravenous administration of antibiotics or combined antibiotic therapy. The results will support national planning on antibiotic resistance, the development of a new standard national antibiotic therapeutic guideline for CAP in Vietnam, and assist in the development of interventions for health professionals and consumers to improve the quality of use of antibiotics. Further research is needed to determine factors influencing decisions on laboratory testing, hospitalization, use of intravenous route, choice of antibiotic, and recommendation for combination therapy. CONFLICT OF INTEREST The authors declare no conflict of interest. KEY POINTS Irrational antibiotic combinations for CAP are common. • Hospitalization, choice of intravenous route, and use of combination antibiotic therapy occurred without correlation to CAP severity in Vietnamese hospitals. • Antibiotic combination highly varies among hospitals. • Further research into the factors influencing these decisions is needed.
•
ETHICS STATEMENT This study was retrospective on medical records with anonymous individual patients’ and hospitals’ information; consequently, it did not constitute human subjects research requiring review by an Institutional Review Board. ACKNOWLEDGEMENTS This project was undertaken as part of the Health System Strengthening Project—Intervention 2.1 Establishing a national pharmacovigilance system, conducted by HUP and NPS with the financial support of the Global Fund. Pharmacoepidemiology and Drug Safety, 2014 DOI: 10.1002/pds
h. t. trinh et al.
The assistance of hospital staff in retrieving medical records and that of clinical pharmacy students at HUP in entering data is also appreciated.
REFERENCES 1. WHO. The World Medicines Situation 2011 - Rational Use of Medicines (3 edn). WHO Geneva, 2011. 2. http://www.cimsi.org.vn/THONGKE.aspx?action=thongke&lang=vi. 3. World Health Organisation. World Health Organisation Global Strategy for Containment of Antimicrobial Resistance. 2001 [cited 2013 May]; Available from: http://www.who.int/csr/resources/publications/drugresist/ WHO_CDS_CSR_DRS_2001_2_EN/en/. 4. Martin SJ, Micek ST, Wood GC. Antimicrobial resistance: consideration as an adverse drug event. Crit Care Med 2010; 38(6): S155–S161. 5. Kim SH, Jong JH, Chung DR. Changing trends in antimicrobial resistance and serotypes of streptococcus pneumoniae isolates in Asian countries: an Asian Network for Surveillance of Resistant Pathogens (ANSORP) study. Antimicrob Agents Chemother 2012; 56(3): 1418–1426. 6. Albrich WC, Monnet DL, Harbarth S. Antibiotic selection pressure and resistance in streptococcus pneumoniae and streptococcus pyogenes. Emerg Infect Dis 2004; 10(3): 514–517. 7. Goossens H. Antibiotic consumption and link to resistance. Mini-review. Clin Microbiol Infect 2009; 15(Suppl 3): 12–15. 8. Van, PH, Binh PT. Antibiotic resistance rate of S. pneumoniae and H. influenzae in respiratory infection patients: a result from a multi-central study in Vietnam (SOAR) 2010–2011. Vietnamese J Pract Med 2012; 855: 6–11.
Copyright © 2014 John Wiley & Sons, Ltd.
9. Vu DH, van Rein N, Cobelens FG, Nguyen TT, Le VH, Brouwers JR. Suspected tuberculosis case detection and referral in private pharmacies in Viet Nam. Int J Tuberc Lung Dis 2012; 16(12): 1625–1629. 10. Nguyen QH, Nguyen TK, Ho D, et al. Unnecessary antibiotic use for mild acute respiratory infections during 28-day follow-up of 823 children under five in rural Vietnam. Trans R Soc Trop Med Hyg 2011; 105(11): 628–636. 11. American Thoracic Society and Infectious Diseases Society of America. Guidelines for the management of adults with hospital-acquired, ventilator-associated, and healthcareassociated pneumonia. Am J Respir Crit Care Med 2005; 171(4): 388–416. 12. Jones RN. Microbial etiologies of hospital-acquired bacterial pneumonia and ventilator-associated bacterial pneumonia. Clin Infect Dis 2010; 51(Suppl 1): S81–S87. 13. Lim WS, Baudouin SV, George RC, et al. BTS guidelines for the management of community acquired pneumonia in adults: update 2009. Thorax, 2009; 64(3): 121434. 14. Niederman MS, Mandell LA, Anzueto A, et al. Guidelines for the management of adults with community-acquired pneumonia. Diagnosis, assessment of severity, antimicrobial therapy, and prevention. Am J Respir Crit Care Med 2001; 163(7): 1730–1754. 15. Paterson DL. “Collateral damage” from cephalosporin or quinolone antibiotic therapy. Clin Infect Dis 2004; 38(Suppl 4): S341–S345. 16. Albertson TE, Dean NC, El Solh AA, Gotfried MH, Kaplan C, Niederman MS. Fluoroquinolones in the management of community-acquired pneumonia. Int J Clin Pract 2010; 64(3): 378–388. 17. Laterre PF. Monotherapy or combination therapy for hospitalized patients with community-acquired pneumonia: not yet the end of the story? Clin Infect Dis 2008; 46(10): 1510–1512. 18. 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; 1(44): S27–S72.
Pharmacoepidemiology and Drug Safety, 2014 DOI: 10.1002/pds
