- Email: [email protected]
Analysis of 5 years of bacteraemias: Importance of stratification of microbial susceptibilities by source of patients
]ournal of Infection (1997) 35, 17-23
Analysis of 5 Years of Bacteraemias: Importance of Stratification of Microbial Susceptibilities by Source of Patients A. M. Yinnon .1, Y. Schlesinger 1, D. Gabbay 1 and B. Rudensky 2 ilnfectious Diseases Unit and 2Clinical Microbiology Laboratory, Shaare Zedek Medical Center, Jerusalem, Israel Many factors need to be considered when selecting empiric antimicrobial treatment for infections; foremost are the principal pathogens causing the diagnosed infection and their antimicrobial susceptibility patterns. These susceptibifities are location specific. This study analyses blood cultures of a 5-year period (1990-94) at a 550 bed community hospital and stratifies antimicrobial susceptibilities by source of patients. Data included: date of culture, patient location, number of positive bottles with the same organism over a period of 2 weeks and results of susceptibility testing. Positive cultures from patients in the Emergency Department were deemed to reflect communityacquired strains; positive cultures from patients in the Intensive Care Unit were considered nosocomial organisms. During the study period 52 055 blood cultures were drawn; 5652 (11%) from 2742 patients grew at least one organism, excluding skin contaminants. Organisms cultured most frequently were: Enterobacteriaceae: 1162 patients (42%); Staphylococcus aureus: 442 (16%); Enterococcus: 429 (16%); and Pseudomonas: 196 (7%). Antimicrobial susceptibility percentages of Enterobacteriaceae from Emergency Room patients (1l = 370) were significantly greater to all tested antimicrobials than from ICU patients (n = 161) (P<0.O01). Overall, 143 isolates of S. aureus from 442 patients (32%) were methicillin resistant (MR); stratification by department revealed a range of 20/142 (14%) MR in community acquired strains to 49/67 (73%) from ICU patients (P
Introduction Patients with moderate to severe infectious processes are usually treated with antimicrobial agents before results of relevant cultures are available. Isolation of pathogens and susceptibility testing m a y subsequently lead to changes in antimicrobial management.~ The appropriate empiric treatment, chosen w h e n the patient is most severely ill, appears to be of critical importance for survival and well-being. 1'2 M a n y factors have to be considered w h e n selecting empiric treatment; foremost are the principal pathogens causing the diagnosed infection and their antimicrobial susceptibility patterns. 3 These latter factors are very m u c h location-specific. It is therefore important that physicians have access to microbiological data pertaining to the patient population they serve. Computers are essential tools in the generation of these data. We describe the results of computerization of blood culture data and the importance of stratification * Address correspondence to: Amos M. Yinnon, MD, Infectious Diseases Unit, Department of Medicine, Shaare Zedek Medical Center. P.O. Box 3235, Jerusalem 91031, Israel. This work was presented in part at the 35th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco, CA, September, 1995. Accepted for publication 13 October 1996. 0163-4453/97/040017 + 07 $12.00/0
of organisms and their susceptibility patterns by source of patients.
Methods This study was undertaken in Shaare Zedek Medical Center, Jerusalem's second largest hospital. The centre, a 5 50 bed facility, is university affiliated and encompasses all the major departments, including medical and surgical intensive care units, cardiac surgery and haematologyoncology. The microbiology laboratory receives a n n u a l l y between 10 0 0 0 and 11 0 0 0 blood culture sets for diagnostic work-up. A set consists of an aerobic and anaerobic bottle, cultured by using the BACTEC 9 2 4 0 (Becton Dickenson Diagnostic I n s t r u m e n t Systems, Sparks, MD, U.S.A.). Blood culture bottles were routinely incubated for up to 5 days unless prolonged incubation was clinically indicated. Terminal subcultures were not routinely performed unless clinically indicated. Isolates were identified using the API system (Biomerieux, Marcy l'Etoile, France). Instrument-generated positive bottles were Gram-stained or acridine-stained, and plated, with subsequent sensitivity testing by the Kirby-Bauer method. 4 Minimal inhibitory concentrations were not routinely © 1997 The British Society for the Study of Infection
18
A . M . Yinnon e t al. 14
14
12
12
lO
i0
s
8
6i
~ @
[]
4 z
r~
f
4
2 403
o
9
1990
450
518
565
692
1991
1992 Year
1993
1994
o
Figure 1. Yield of blood cultures. (m) Total number of blood cultures drawn (in thousands); ([Z) percentages of patients with positive blood cultures out of all drawn cultures; (O) absolute number of patients with positive cultures.
performed unless clinically indicated and then performed with the E-test (PDM epsilometer; AB Biodisk, 8olna, Sweden). 5 Data from blood cultures from 1990 onwards were entered in a computer program. We used the QuattroPro software package. ~ A spreadsheet was prepared incorporating all data deemed relevant, including: date of culture, the patient's identification number, department, isolated organisms, number of a patient's positive bottles with the same organism over a period of up to 2 weeks, and results of sensitivity testing to up to 18 antimicrobial agents. For analysis of antimicrobial susceptibilities an organism isolated from any patient was considered only once, even though the same organism could have been grown from more than one culture. Growth of Staphylococcus epidermidis, diphtheroids, and alpha haemolytic streptococci in blood cultures is often secondary to skin contamination or sampling error and these organisms were therefore not entered. Results were analysed using the QuattroPro package, requiring moderate proficiency only. Statistical calculations were performed using Student's t-test and Chi-squared analysis. 7
Results During the 1990-94 study period 52 055 blood cultures were drawn; 5652 (11%) cultures from 2742 patients grew at least one organism, excluding skin contaminants (Fig. 1). The mean number of positive cultures per patient with at least one positive culture was 2.1___0.5 (mean _+s.D.) with a range of 1-21. A breakdown of
patients with bacteraemia by department of origin is presented in Fig. 2: Numbers represent patients rather than culture sets or bottles in this figure and all subsequent tables and analyses. The three departments generating the greatest number of positive cultures were: (1) the Emergency Department with 785/2742 (28.6%) cultures; (2) the Intensive Care Unit with 413/2742 (15.1%); and (3) the Medical Departments with 321/ 2742 (11.7%). All surgical departments constituted 307/ 2742 (11.2%) of positive blood cultures. Table I lists the isolated organisms. The five most frequently isolated organisms were: E. coli (480 patients, 17.5%), S. aureus (442 patients, 16.1%), Klebsiella pneumoniae (266 patients, 9.7%), Pseudomonas aeruginosa (196 patients, 7.1%), and Enterococcus (194 patients, 7.1%). The relative frequency of candidaemia remained stable at about 4.3% of positive cultures over the 5 year study period. A total of 105 anaerobes were isolated (3.8% of 2742 patients with bacteraemia), including: Bacteroides fragilis (54 cases), Clostridium (37), Peptostreptococcus (six), Veillonella (five), Fusobacterium (two) and Peptococcus (one). Table II shows antimicrobial susceptibilities of Enterobacteriaceae. No major changes in susceptibility percentages occurred over time, with two notable exceptions. Firstly, sensitivity of Enterobacteriaceae to ciprofloxacin decreased from 94% to 81% within 1 year of introduction (1990) (P
19
Analysis of 5 Years of Bacteraemias Dialysis (10.3%) cy D e p a r t m e n t (28.6%)
Other, < 2% eael Neonatal ICU (6.5~ Paediatries (7.0%)
G e r i a t r i c s (8.6%1 s i r e C a r e U n i t (15.1%) Surg~j
....... M e d i e i n e (11.7%) Figure 2. Bacteraemia by department (n = 2742).
Table I. List of isolated orgamsms. Organism
1990
1991
1992
1993
1994
Total
Enterobacteriaceae* Staphylococcus aureus Pseudomonas aeruginosa Enterococcus spp. Streptococcus pneumoniae Candida spp. Acinetobacter spp. Bacteroides spp. Streptococcus Group A Haemophilus influenzae CIostridiwn perfringens Streptococcus Group B Brucella melitensis Neisseria meningitidis Other o r g a n i s m s t
157 67 28 22 25 17 13 11 10 5 7 5 4 1 31
181 93 37 26 26 24 9 7 6 7 5 4 1 1 23
276 78 51 33 31 24 11 17 8 10 4 5 1 2 24
260 93 45 54 23 20 22 7 17 9 8 12 4 4 44
295 111 35 59 32 30 22 12 10 8 13 11 5 3 46
1169 442 196 194 137 115 77 54 51 39 37 37 15 11 168
Total
403
450
575
622
692
2742
* Most frequently isolated Enterobacteriaceae: Escherichia coli: 4 8 0 (41%), KlebsieIla pneumoniae: 266 (23%), Enterobacter spp.: 125 (11%), Proteus mirabilis: 83 (7%), Salmonella spp.: 51 (4%), Citrobacter spp.: 39 (3%), Klebsiella oxytoca: 35 (3%), Serratia marcescens: 31 (3%), others: 59 (5%). ]- Organisms isolated fewer t h a n four times annually.
Table II. Antimicrobial susceptibilities of Enterobacteriaceae (% sensitive), 1 9 9 0 - 9 4 . Antimicrobial agents
1990 (n=157)
1991 (n=181)
1992 (n=276)
1993 (n=260)
1994 (n=295)
Ampicillin Amikacin Cefazolin Chloramphenicol Ciprofloxacin Ceftriaxone Cefuroxime Co-trimoxazole Gentamicin Imipenem Mezlocilltn Tobramicin
32 94 41 70 94 90 70 64 87 100 67 86
35 92 41 60 81 79 70 57 80 100 49 79
30 95 47 70 84 81 66 71 84 100 58 87
30 91 51 64 80 84 70 62 73 100 52 78
34 99 51 68 80 88 74 64 86 100 59 84
A. M. Yinnon et al.
20
Table IlL Antimicrobial sensitivities of selected Gram-negative organisms (% sensitive). Antimicrobial
Ampicillin Amikacin Cefazolin Chloramphenicol Ciprofloxacin Ceftriaxone Cefuroxime Co-trimoxazole Gentamicin Imipenem Mezlocillin Tobramycin
Escherichia coli
Klebsiella pneumoniae
Pseudomonas Enterobacter aeruginosa spp.
Proteus mirabilis
(n = 480)
(n = 266)
(n= 196)
(n= 125)
(n= 82)
50 98 62 80 90 95 91 68 96 i 00 51 95
0 85 36 43 66 58 43 45 51 100 35 60
ND 91 ND ND 77 ND ND ND 82 94 79 83
22 96 3 72 89 71 49 79 90 100 72 91
37 100 46 47 71 99 92 44 74 100 46 91
ND = Not Determined.
Table IV. Antimicrobial susceptibilities of Staphylococcus aureus and Enterococcus spp. (in %), 1990-94. Year
1990 1991 1992 1993 1994
Staphylococcusa u r e u s
Enterococcus spp.
N
% methicillin resistant*
n
% ampicillin resistant1-
67 93 78 90 111
27 29 27 40 37
22 26 33 54 59
14 10 19 17 21
* Percentage of isolates that were methicillin resistant. Percentage of isolates that were ampicillin resistant; no vancomycinresistance was encountered.
Discussion This study found that 11% of d r a w n blood cultures yielded organisms. S. epidermidis, alpha haemolytic streptococci and diphtheroids, three frequent c o n t a m i n a n t s of blood cultures, were excluded from this study; some of these could have involved true bacteraemias. On the other hand, some of the organisms included in this analysis could have been contaminants. The percentage of positive blood cultures is reported to range from 8.4% to 17.6%, 8-1° and of clinically significant blood cultures from 6.7% to 9% 11 with a m e a n rate of 7%. 12 Accordingly, we believe t h a t the yield of blood cultures in our laboratory reflects current trends. During the 1 9 9 0 - 9 4 study period, the yearly n u m b e r of positive cultures increased a n n u a l l y between 8.2% and 28% as compared to the previous year; this reflects the growing n u m b e r of patients being treated by our medical center rather t h a n increasing sensitivity of blood culture isolation techniques, as s h o w n by Fig. 1. The organisms most frequently isolated were, in order
of decreasing frequency, E. cell (17%), S. aureus (16.1%), K pneumoniae (9.7%), P. aeruginosa (7.1%) and Enterococcus spp. (7.1%). This list is similar to those reported in other large series) °' 11.13 Candida spp. constituted 4.3% of the positive cultures, and anaerobes 3.8%. A l t h o u g h the overall n u m b e r of positive cultures increased during the 5-year study period, no major changes in relative incidence of the principal organisms occurred (Table I). The same is the case with antimicrobial susceptibility patterns; this constency in susceptibility percentages allowed pooling and analysis of data from the entire study period. Microbiology laboratories in m a n y m o d e r n hospitals distribute a yearly or twice yearly leaflet with the most frequently isolated organisms and the susceptibilities to antimicrobial agents employed in their institutions. These tables are intended to facilitate the rational choice of optimal empiric antimicrobial therapy of patients with infectious diseases. The results from our study indicate that susceptibility data have to be stratified in several ways in order to be meaningful and useful. Firstly, our data have clearly demonstrated the major differences in antimicrobial sensitivities between the c o m m u n i t y acquired and nosocomial strains. Susceptibilities of microorganisms to antimicrobial agents can be viewed as a continuum, with, at the one extreme, the most sensitive bacteria from the c o m m u n i t y and, at the other extreme, the frequently multidrug resistant organisms from the Intensive Care Unit. Reports from microbiology laboratories have to contain stratified susceptibility results according to the source of patients rather t h a n pooled data in order to be useful to physicians w h e n selecting empiric antibiotic treatment for individual patients. Obviously, the larger the n u m b e r of isolates involved in true infections within a particular location, the more reliable the antimicrobial susceptibilities' report and the more predict-ive
Analysis of 5 Years of Bacteraemias
21
100 9080 7060 h~
5040 30 20 10 0
Amp Cef Cxm Mez Chl Cot Caz Tob Ctx G e n Cip Amk Imp Antimicrobial agents
Figure 3. Stratified antimicrobiaI susceptibilities of Enterobacteriaceae. (D) ICU patients (n= 161); ( I ) Emergency Room patients (n=370). Stratification of susceptibilities is by source of patients. Antimicrobial abbreviations: Amp = ampicillin; Cef= cefazolin; Cxm=cefuroxime; Mez = mezlocyllin; Chl = chloramphenicol; Cot = co-trimoxazole (trimethoprim-sulfamethoxazole); Caz = ceftazidime;Tob = tobramycin: Ctx = ceftriaxone; Gen=gentamicin; Cip=ciprofloxacin; Amk=amikacin; Imp=imipenem.
Table V. Antimicrobial sensitivities of Enterobacteriaceae; stratification by source of patients (% fully sensitive). Anl:imicrobial
Emergency Department (n=370)
All isolates (n=1169)
Intensive Care Unit (n=161)
Ampicillin Amikacin Ceftazidime Cefazolin Chloramphenicol Ciprofloxacin Ceftriaxone Cefuroxime Cotrimoxazole Gentamicin Iinipenem Mezlocillin Tobramycin
46 100 87 62 75 90 96 86 68 90 100 59 93
32 94 61 46 66 84 84 70 64 82 100 57 83
14 85 58 21 47 65 60 31 55 63 100 44 60
of a n i n d i v i d u a l p a t i e n t ' s isolate. A department-specific list of m i c r o b i a l isolates a n d sensitivities w o u l d be the u l t i m a t e objective; however, t h e r e is quite some mobility of patients a n d p e r s o n n e l w i t h i n hospitals, so too rigid c o m p a r t m e n t a l i z a t i o n of o r g a n i s m s a n d their susceptibilities w o u l d n e i t h e r be possible n o r t r u l y reflect the microbiological e n v i r o n m e n t w i t h i n hospitals.
It should be n o t e d t h a t the n a t u r e of infections m a y be quite different in v a r i o u s settings; (a) acute admissions will largely be infected w i t h c o m m u n i t y - a c q u i r e d flora w i t h a small a d m i x t u r e of o r g a n i s m s exported from hospital; (b) infections in Intensive Care Unit patients m a y n o t h a v e been acquired in the Intensive Care Unit - t h e y could be c o m m u n i t y - a c q u i r e d or acquired in the wards, or in the Intensive Care Unit; a n d (c) such infections are g e n e r a l l y e n d o g e n o u s , their p a t h o g e n s being derived from flora reflecting either the c o m m u n i t y , or the w a r d population or Intensive Care Unit population, d e p e n d i n g on h o w long the p a t i e n t h a s b e e n in e a c h setting. In o t h e r words, the flora in e a c h setting is n o t h o m o g e n e o u s l y derived. Nevertheless, our d a t a h a v e clearly d e m o n s t r a t e d the overall clinical usefulness of d i s t i n g u i s h i n g b e t w e e n the a n t i m i c r o b i a l sensitivities of the c o m m u n i t y - a c q u i r e d a n d n o s o c o m i a l strains. Secondly, we believe it to be a d v a n t a g e o u s to physicians to possess susceptibility d a t a on categories of o r g a n i s m s r a t h e r t h a n for p a r t i c u l a r o r g a n i s m s only. For example, a p a t i e n t being a d m i t t e d to hospital b e c a u s e of urosepsis requires empiric t r e a t m e n t of Enterobacteriaceae, the g r o u p of o r g a n i s m s u s u a l l y involved in this p a r t i c u l a r infection. Available d a t a on susceptibility p a t t e r n s of individual m i c r o - o r g a n i s m s m a y be confusing as it re-
22
A . M . Yinnon et al, Dialysis m Paediatrics m Emergency Department
l0
27
~
105
Neonatal ICU
23
Medicine All Isolates
54
~
442
Surgery
23
Geriatrics
30
Intensive Care Unit
67 0
r 10
I 20
I 30
I 40
I 50
I 60
I 70
80
% methicillin r e s i s t a n c e (n) Figure 4. Percentage methicillin resistance of Staphylococcus aureus.
quires specific knowledge concerning all principal organisms causing the particular infection. The identity of the causative organism is unknown at the time of selection of empiric antimicrobial therapy; therefore, it would be more logical to possess information on susceptibility patterns of categories of causative organisms. Subsequent isolation of the organism and determination of antimicrobial sensitivities will allow adaptation of treatment of the individual patient. Figure 3 and Table V present data concerning Enterobacteriaceae that are relevant to the clinician. Staphylococcus aureus frequently produces both community-acquired and nosocomial infections. An analysis concerning methicillin resistance in isolates of S. aureus showed that in infections originating in the community, exemplified by ambulatory dialysis patients, fewer than 10% manifest methicillin resistance. On the other hand, strains of S. aureus causing infections in the Intensive Care Unit were methicilliri resistant in more than 70% of instances, reflecting worldwide trends. TM 1s Methicillin resistance occurred in 31% of 439 patients with staphylococcal bacteraemia. These data indicate the importance of the individual patient's location when selecting empiric treatment for suspected staphylococcal infections. In summary, the two extremes of community-acquired organisms, and nosocomial (Intensive Care Unit)acquired organisms, with their respective susceptibility patterns, give a useful framework that provides physicians
with a reliable tool to choose appropriate antimicrobial therapy, As hospitals harbour different organisms with different antimicrobial sensitivities, each hospital needs to generate its own data. This study has demonstrated that stratification of susceptibilities according to the source of the patients is necessary in order to be meaningful and useful.
Acknowledgement We wish to express our appreciation to the staff of the microbiology laboratory, who with their dedication and professionalism have assisted in the care of countless patients.
References 1 Moellering RC. Principles of anti-infective therapy. In: Mandell GL, Douglas RG Jr, DoIin R, eds. Principles and Practice of Infectious Diseases. 4th ed, 1995; 199-212. 2 Joshi N, Milfred D. The use and misuse of new antibiotics. Arch Intern Med 1995; 155: 569-577. 3 Reese RE, Betts RF. Antibiotic use. In: Reese RE, Betts RF, eds. A Practical Approach to Infectious Diseases, Little, Brown & Co, 3rd ed, 1991: 821-1007. 4 Bauer AM, Kirby WMM, Sherris JC, Turck M. Antibiotic susceptibility testing by a standard, simple disk method. Am J Clin Pathol 1966; 45: 493-496. 5 Murray PR. Manual of Clinical Microbiology. ASM Press, 6th ed, 1995; 1336. 6 QuattroPro. Borland International Co, Inc. Version 4.0, User's Guide, 1992.
Analysis of 5 Years of Bacteraemias 7 Glantz SA. Primer of Biostatistics. McGraw-Hill, Inc., 3rd ed, 1992. 8 Roberts FJ. A review of positive blood cultures: identification and source of microorganisms and patterns of sensitivity to antibiotics. Rev Infect Dis 1980; 2: 329-339. 9 Nitzan Y, Arielly H, Maayan MC, Rozenszajn A. Gram negative bacteria isolated from blood cultures in a general hospital. Mixrobiologiea 1994; 17: 111-122. 0 Chamberland S, L'Ecuyer J, Lessard C, Bernier M, Provencher P, Bergeron MG. Antibiotic susceptibility profiles of 941 Gram-negative bacteria isolated from septicemic patients throughout Canada. The 7~anadian Study Group. Clin Infect Dis 1992; 15: 615-628. 1 Williams GT, Houang ET, Shaw EJ, Tabaqchali S. Bacteraemia in a London teaching hospital 1966-75. Lancet 1976; ii: 1291-1293.
23
I2 Brown DFI, Perry SF. Methods used in the United Kingdom for the culture of micro-organisms from blood. I Clin Pathol 1992; 45: 468-474. 13 Pittet D, Wenzel RP. Nosocomial bloodstream infections. Secular trends in rates, mortality, and contribution to total hospital deaths. Arch Intern Med 1995; 155: 1177-1184. 14 Mulligan ME, Murray-Leisure KA, Ribner BS et aI. Methicillinresistant Staphylococcus aureus: a consensus review of the microbiology, pathogenesis, and epidemioIogy with implications for prevention and management. Am l Med 1993; 94: 313-328. 15 BoyceJM. Methicillin-resistantStaphylococcusaureusinhospitalsand long-term care facilities: microbiology, epidemiology, and preventive measures. Infect Control Hosp Epidemiol 1992; 13: 725-737.
Analysis of 5 Years of Bacteraemias: Importance of Stratification of Microbial Susceptibilities by Source of Patients A. M. Yinnon .1, Y. Schlesinger 1, D. Gabbay 1 and B. Rudensky 2 ilnfectious Diseases Unit and 2Clinical Microbiology Laboratory, Shaare Zedek Medical Center, Jerusalem, Israel Many factors need to be considered when selecting empiric antimicrobial treatment for infections; foremost are the principal pathogens causing the diagnosed infection and their antimicrobial susceptibility patterns. These susceptibifities are location specific. This study analyses blood cultures of a 5-year period (1990-94) at a 550 bed community hospital and stratifies antimicrobial susceptibilities by source of patients. Data included: date of culture, patient location, number of positive bottles with the same organism over a period of 2 weeks and results of susceptibility testing. Positive cultures from patients in the Emergency Department were deemed to reflect communityacquired strains; positive cultures from patients in the Intensive Care Unit were considered nosocomial organisms. During the study period 52 055 blood cultures were drawn; 5652 (11%) from 2742 patients grew at least one organism, excluding skin contaminants. Organisms cultured most frequently were: Enterobacteriaceae: 1162 patients (42%); Staphylococcus aureus: 442 (16%); Enterococcus: 429 (16%); and Pseudomonas: 196 (7%). Antimicrobial susceptibility percentages of Enterobacteriaceae from Emergency Room patients (1l = 370) were significantly greater to all tested antimicrobials than from ICU patients (n = 161) (P<0.O01). Overall, 143 isolates of S. aureus from 442 patients (32%) were methicillin resistant (MR); stratification by department revealed a range of 20/142 (14%) MR in community acquired strains to 49/67 (73%) from ICU patients (P
Introduction Patients with moderate to severe infectious processes are usually treated with antimicrobial agents before results of relevant cultures are available. Isolation of pathogens and susceptibility testing m a y subsequently lead to changes in antimicrobial management.~ The appropriate empiric treatment, chosen w h e n the patient is most severely ill, appears to be of critical importance for survival and well-being. 1'2 M a n y factors have to be considered w h e n selecting empiric treatment; foremost are the principal pathogens causing the diagnosed infection and their antimicrobial susceptibility patterns. 3 These latter factors are very m u c h location-specific. It is therefore important that physicians have access to microbiological data pertaining to the patient population they serve. Computers are essential tools in the generation of these data. We describe the results of computerization of blood culture data and the importance of stratification * Address correspondence to: Amos M. Yinnon, MD, Infectious Diseases Unit, Department of Medicine, Shaare Zedek Medical Center. P.O. Box 3235, Jerusalem 91031, Israel. This work was presented in part at the 35th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco, CA, September, 1995. Accepted for publication 13 October 1996. 0163-4453/97/040017 + 07 $12.00/0
of organisms and their susceptibility patterns by source of patients.
Methods This study was undertaken in Shaare Zedek Medical Center, Jerusalem's second largest hospital. The centre, a 5 50 bed facility, is university affiliated and encompasses all the major departments, including medical and surgical intensive care units, cardiac surgery and haematologyoncology. The microbiology laboratory receives a n n u a l l y between 10 0 0 0 and 11 0 0 0 blood culture sets for diagnostic work-up. A set consists of an aerobic and anaerobic bottle, cultured by using the BACTEC 9 2 4 0 (Becton Dickenson Diagnostic I n s t r u m e n t Systems, Sparks, MD, U.S.A.). Blood culture bottles were routinely incubated for up to 5 days unless prolonged incubation was clinically indicated. Terminal subcultures were not routinely performed unless clinically indicated. Isolates were identified using the API system (Biomerieux, Marcy l'Etoile, France). Instrument-generated positive bottles were Gram-stained or acridine-stained, and plated, with subsequent sensitivity testing by the Kirby-Bauer method. 4 Minimal inhibitory concentrations were not routinely © 1997 The British Society for the Study of Infection
18
A . M . Yinnon e t al. 14
14
12
12
lO
i0
s
8
6i
~ @
[]
4 z
r~
f
4
2 403
o
9
1990
450
518
565
692
1991
1992 Year
1993
1994
o
Figure 1. Yield of blood cultures. (m) Total number of blood cultures drawn (in thousands); ([Z) percentages of patients with positive blood cultures out of all drawn cultures; (O) absolute number of patients with positive cultures.
performed unless clinically indicated and then performed with the E-test (PDM epsilometer; AB Biodisk, 8olna, Sweden). 5 Data from blood cultures from 1990 onwards were entered in a computer program. We used the QuattroPro software package. ~ A spreadsheet was prepared incorporating all data deemed relevant, including: date of culture, the patient's identification number, department, isolated organisms, number of a patient's positive bottles with the same organism over a period of up to 2 weeks, and results of sensitivity testing to up to 18 antimicrobial agents. For analysis of antimicrobial susceptibilities an organism isolated from any patient was considered only once, even though the same organism could have been grown from more than one culture. Growth of Staphylococcus epidermidis, diphtheroids, and alpha haemolytic streptococci in blood cultures is often secondary to skin contamination or sampling error and these organisms were therefore not entered. Results were analysed using the QuattroPro package, requiring moderate proficiency only. Statistical calculations were performed using Student's t-test and Chi-squared analysis. 7
Results During the 1990-94 study period 52 055 blood cultures were drawn; 5652 (11%) cultures from 2742 patients grew at least one organism, excluding skin contaminants (Fig. 1). The mean number of positive cultures per patient with at least one positive culture was 2.1___0.5 (mean _+s.D.) with a range of 1-21. A breakdown of
patients with bacteraemia by department of origin is presented in Fig. 2: Numbers represent patients rather than culture sets or bottles in this figure and all subsequent tables and analyses. The three departments generating the greatest number of positive cultures were: (1) the Emergency Department with 785/2742 (28.6%) cultures; (2) the Intensive Care Unit with 413/2742 (15.1%); and (3) the Medical Departments with 321/ 2742 (11.7%). All surgical departments constituted 307/ 2742 (11.2%) of positive blood cultures. Table I lists the isolated organisms. The five most frequently isolated organisms were: E. coli (480 patients, 17.5%), S. aureus (442 patients, 16.1%), Klebsiella pneumoniae (266 patients, 9.7%), Pseudomonas aeruginosa (196 patients, 7.1%), and Enterococcus (194 patients, 7.1%). The relative frequency of candidaemia remained stable at about 4.3% of positive cultures over the 5 year study period. A total of 105 anaerobes were isolated (3.8% of 2742 patients with bacteraemia), including: Bacteroides fragilis (54 cases), Clostridium (37), Peptostreptococcus (six), Veillonella (five), Fusobacterium (two) and Peptococcus (one). Table II shows antimicrobial susceptibilities of Enterobacteriaceae. No major changes in susceptibility percentages occurred over time, with two notable exceptions. Firstly, sensitivity of Enterobacteriaceae to ciprofloxacin decreased from 94% to 81% within 1 year of introduction (1990) (P
19
Analysis of 5 Years of Bacteraemias Dialysis (10.3%) cy D e p a r t m e n t (28.6%)
Other, < 2% eael Neonatal ICU (6.5~ Paediatries (7.0%)
G e r i a t r i c s (8.6%1 s i r e C a r e U n i t (15.1%) Surg~j
....... M e d i e i n e (11.7%) Figure 2. Bacteraemia by department (n = 2742).
Table I. List of isolated orgamsms. Organism
1990
1991
1992
1993
1994
Total
Enterobacteriaceae* Staphylococcus aureus Pseudomonas aeruginosa Enterococcus spp. Streptococcus pneumoniae Candida spp. Acinetobacter spp. Bacteroides spp. Streptococcus Group A Haemophilus influenzae CIostridiwn perfringens Streptococcus Group B Brucella melitensis Neisseria meningitidis Other o r g a n i s m s t
157 67 28 22 25 17 13 11 10 5 7 5 4 1 31
181 93 37 26 26 24 9 7 6 7 5 4 1 1 23
276 78 51 33 31 24 11 17 8 10 4 5 1 2 24
260 93 45 54 23 20 22 7 17 9 8 12 4 4 44
295 111 35 59 32 30 22 12 10 8 13 11 5 3 46
1169 442 196 194 137 115 77 54 51 39 37 37 15 11 168
Total
403
450
575
622
692
2742
* Most frequently isolated Enterobacteriaceae: Escherichia coli: 4 8 0 (41%), KlebsieIla pneumoniae: 266 (23%), Enterobacter spp.: 125 (11%), Proteus mirabilis: 83 (7%), Salmonella spp.: 51 (4%), Citrobacter spp.: 39 (3%), Klebsiella oxytoca: 35 (3%), Serratia marcescens: 31 (3%), others: 59 (5%). ]- Organisms isolated fewer t h a n four times annually.
Table II. Antimicrobial susceptibilities of Enterobacteriaceae (% sensitive), 1 9 9 0 - 9 4 . Antimicrobial agents
1990 (n=157)
1991 (n=181)
1992 (n=276)
1993 (n=260)
1994 (n=295)
Ampicillin Amikacin Cefazolin Chloramphenicol Ciprofloxacin Ceftriaxone Cefuroxime Co-trimoxazole Gentamicin Imipenem Mezlocilltn Tobramicin
32 94 41 70 94 90 70 64 87 100 67 86
35 92 41 60 81 79 70 57 80 100 49 79
30 95 47 70 84 81 66 71 84 100 58 87
30 91 51 64 80 84 70 62 73 100 52 78
34 99 51 68 80 88 74 64 86 100 59 84
A. M. Yinnon et al.
20
Table IlL Antimicrobial sensitivities of selected Gram-negative organisms (% sensitive). Antimicrobial
Ampicillin Amikacin Cefazolin Chloramphenicol Ciprofloxacin Ceftriaxone Cefuroxime Co-trimoxazole Gentamicin Imipenem Mezlocillin Tobramycin
Escherichia coli
Klebsiella pneumoniae
Pseudomonas Enterobacter aeruginosa spp.
Proteus mirabilis
(n = 480)
(n = 266)
(n= 196)
(n= 125)
(n= 82)
50 98 62 80 90 95 91 68 96 i 00 51 95
0 85 36 43 66 58 43 45 51 100 35 60
ND 91 ND ND 77 ND ND ND 82 94 79 83
22 96 3 72 89 71 49 79 90 100 72 91
37 100 46 47 71 99 92 44 74 100 46 91
ND = Not Determined.
Table IV. Antimicrobial susceptibilities of Staphylococcus aureus and Enterococcus spp. (in %), 1990-94. Year
1990 1991 1992 1993 1994
Staphylococcusa u r e u s
Enterococcus spp.
N
% methicillin resistant*
n
% ampicillin resistant1-
67 93 78 90 111
27 29 27 40 37
22 26 33 54 59
14 10 19 17 21
* Percentage of isolates that were methicillin resistant. Percentage of isolates that were ampicillin resistant; no vancomycinresistance was encountered.
Discussion This study found that 11% of d r a w n blood cultures yielded organisms. S. epidermidis, alpha haemolytic streptococci and diphtheroids, three frequent c o n t a m i n a n t s of blood cultures, were excluded from this study; some of these could have involved true bacteraemias. On the other hand, some of the organisms included in this analysis could have been contaminants. The percentage of positive blood cultures is reported to range from 8.4% to 17.6%, 8-1° and of clinically significant blood cultures from 6.7% to 9% 11 with a m e a n rate of 7%. 12 Accordingly, we believe t h a t the yield of blood cultures in our laboratory reflects current trends. During the 1 9 9 0 - 9 4 study period, the yearly n u m b e r of positive cultures increased a n n u a l l y between 8.2% and 28% as compared to the previous year; this reflects the growing n u m b e r of patients being treated by our medical center rather t h a n increasing sensitivity of blood culture isolation techniques, as s h o w n by Fig. 1. The organisms most frequently isolated were, in order
of decreasing frequency, E. cell (17%), S. aureus (16.1%), K pneumoniae (9.7%), P. aeruginosa (7.1%) and Enterococcus spp. (7.1%). This list is similar to those reported in other large series) °' 11.13 Candida spp. constituted 4.3% of the positive cultures, and anaerobes 3.8%. A l t h o u g h the overall n u m b e r of positive cultures increased during the 5-year study period, no major changes in relative incidence of the principal organisms occurred (Table I). The same is the case with antimicrobial susceptibility patterns; this constency in susceptibility percentages allowed pooling and analysis of data from the entire study period. Microbiology laboratories in m a n y m o d e r n hospitals distribute a yearly or twice yearly leaflet with the most frequently isolated organisms and the susceptibilities to antimicrobial agents employed in their institutions. These tables are intended to facilitate the rational choice of optimal empiric antimicrobial therapy of patients with infectious diseases. The results from our study indicate that susceptibility data have to be stratified in several ways in order to be meaningful and useful. Firstly, our data have clearly demonstrated the major differences in antimicrobial sensitivities between the c o m m u n i t y acquired and nosocomial strains. Susceptibilities of microorganisms to antimicrobial agents can be viewed as a continuum, with, at the one extreme, the most sensitive bacteria from the c o m m u n i t y and, at the other extreme, the frequently multidrug resistant organisms from the Intensive Care Unit. Reports from microbiology laboratories have to contain stratified susceptibility results according to the source of patients rather t h a n pooled data in order to be useful to physicians w h e n selecting empiric antibiotic treatment for individual patients. Obviously, the larger the n u m b e r of isolates involved in true infections within a particular location, the more reliable the antimicrobial susceptibilities' report and the more predict-ive
Analysis of 5 Years of Bacteraemias
21
100 9080 7060 h~
5040 30 20 10 0
Amp Cef Cxm Mez Chl Cot Caz Tob Ctx G e n Cip Amk Imp Antimicrobial agents
Figure 3. Stratified antimicrobiaI susceptibilities of Enterobacteriaceae. (D) ICU patients (n= 161); ( I ) Emergency Room patients (n=370). Stratification of susceptibilities is by source of patients. Antimicrobial abbreviations: Amp = ampicillin; Cef= cefazolin; Cxm=cefuroxime; Mez = mezlocyllin; Chl = chloramphenicol; Cot = co-trimoxazole (trimethoprim-sulfamethoxazole); Caz = ceftazidime;Tob = tobramycin: Ctx = ceftriaxone; Gen=gentamicin; Cip=ciprofloxacin; Amk=amikacin; Imp=imipenem.
Table V. Antimicrobial sensitivities of Enterobacteriaceae; stratification by source of patients (% fully sensitive). Anl:imicrobial
Emergency Department (n=370)
All isolates (n=1169)
Intensive Care Unit (n=161)
Ampicillin Amikacin Ceftazidime Cefazolin Chloramphenicol Ciprofloxacin Ceftriaxone Cefuroxime Cotrimoxazole Gentamicin Iinipenem Mezlocillin Tobramycin
46 100 87 62 75 90 96 86 68 90 100 59 93
32 94 61 46 66 84 84 70 64 82 100 57 83
14 85 58 21 47 65 60 31 55 63 100 44 60
of a n i n d i v i d u a l p a t i e n t ' s isolate. A department-specific list of m i c r o b i a l isolates a n d sensitivities w o u l d be the u l t i m a t e objective; however, t h e r e is quite some mobility of patients a n d p e r s o n n e l w i t h i n hospitals, so too rigid c o m p a r t m e n t a l i z a t i o n of o r g a n i s m s a n d their susceptibilities w o u l d n e i t h e r be possible n o r t r u l y reflect the microbiological e n v i r o n m e n t w i t h i n hospitals.
It should be n o t e d t h a t the n a t u r e of infections m a y be quite different in v a r i o u s settings; (a) acute admissions will largely be infected w i t h c o m m u n i t y - a c q u i r e d flora w i t h a small a d m i x t u r e of o r g a n i s m s exported from hospital; (b) infections in Intensive Care Unit patients m a y n o t h a v e been acquired in the Intensive Care Unit - t h e y could be c o m m u n i t y - a c q u i r e d or acquired in the wards, or in the Intensive Care Unit; a n d (c) such infections are g e n e r a l l y e n d o g e n o u s , their p a t h o g e n s being derived from flora reflecting either the c o m m u n i t y , or the w a r d population or Intensive Care Unit population, d e p e n d i n g on h o w long the p a t i e n t h a s b e e n in e a c h setting. In o t h e r words, the flora in e a c h setting is n o t h o m o g e n e o u s l y derived. Nevertheless, our d a t a h a v e clearly d e m o n s t r a t e d the overall clinical usefulness of d i s t i n g u i s h i n g b e t w e e n the a n t i m i c r o b i a l sensitivities of the c o m m u n i t y - a c q u i r e d a n d n o s o c o m i a l strains. Secondly, we believe it to be a d v a n t a g e o u s to physicians to possess susceptibility d a t a on categories of o r g a n i s m s r a t h e r t h a n for p a r t i c u l a r o r g a n i s m s only. For example, a p a t i e n t being a d m i t t e d to hospital b e c a u s e of urosepsis requires empiric t r e a t m e n t of Enterobacteriaceae, the g r o u p of o r g a n i s m s u s u a l l y involved in this p a r t i c u l a r infection. Available d a t a on susceptibility p a t t e r n s of individual m i c r o - o r g a n i s m s m a y be confusing as it re-
22
A . M . Yinnon et al, Dialysis m Paediatrics m Emergency Department
l0
27
~
105
Neonatal ICU
23
Medicine All Isolates
54
~
442
Surgery
23
Geriatrics
30
Intensive Care Unit
67 0
r 10
I 20
I 30
I 40
I 50
I 60
I 70
80
% methicillin r e s i s t a n c e (n) Figure 4. Percentage methicillin resistance of Staphylococcus aureus.
quires specific knowledge concerning all principal organisms causing the particular infection. The identity of the causative organism is unknown at the time of selection of empiric antimicrobial therapy; therefore, it would be more logical to possess information on susceptibility patterns of categories of causative organisms. Subsequent isolation of the organism and determination of antimicrobial sensitivities will allow adaptation of treatment of the individual patient. Figure 3 and Table V present data concerning Enterobacteriaceae that are relevant to the clinician. Staphylococcus aureus frequently produces both community-acquired and nosocomial infections. An analysis concerning methicillin resistance in isolates of S. aureus showed that in infections originating in the community, exemplified by ambulatory dialysis patients, fewer than 10% manifest methicillin resistance. On the other hand, strains of S. aureus causing infections in the Intensive Care Unit were methicilliri resistant in more than 70% of instances, reflecting worldwide trends. TM 1s Methicillin resistance occurred in 31% of 439 patients with staphylococcal bacteraemia. These data indicate the importance of the individual patient's location when selecting empiric treatment for suspected staphylococcal infections. In summary, the two extremes of community-acquired organisms, and nosocomial (Intensive Care Unit)acquired organisms, with their respective susceptibility patterns, give a useful framework that provides physicians
with a reliable tool to choose appropriate antimicrobial therapy, As hospitals harbour different organisms with different antimicrobial sensitivities, each hospital needs to generate its own data. This study has demonstrated that stratification of susceptibilities according to the source of the patients is necessary in order to be meaningful and useful.
Acknowledgement We wish to express our appreciation to the staff of the microbiology laboratory, who with their dedication and professionalism have assisted in the care of countless patients.
References 1 Moellering RC. Principles of anti-infective therapy. In: Mandell GL, Douglas RG Jr, DoIin R, eds. Principles and Practice of Infectious Diseases. 4th ed, 1995; 199-212. 2 Joshi N, Milfred D. The use and misuse of new antibiotics. Arch Intern Med 1995; 155: 569-577. 3 Reese RE, Betts RF. Antibiotic use. In: Reese RE, Betts RF, eds. A Practical Approach to Infectious Diseases, Little, Brown & Co, 3rd ed, 1991: 821-1007. 4 Bauer AM, Kirby WMM, Sherris JC, Turck M. Antibiotic susceptibility testing by a standard, simple disk method. Am J Clin Pathol 1966; 45: 493-496. 5 Murray PR. Manual of Clinical Microbiology. ASM Press, 6th ed, 1995; 1336. 6 QuattroPro. Borland International Co, Inc. Version 4.0, User's Guide, 1992.
Analysis of 5 Years of Bacteraemias 7 Glantz SA. Primer of Biostatistics. McGraw-Hill, Inc., 3rd ed, 1992. 8 Roberts FJ. A review of positive blood cultures: identification and source of microorganisms and patterns of sensitivity to antibiotics. Rev Infect Dis 1980; 2: 329-339. 9 Nitzan Y, Arielly H, Maayan MC, Rozenszajn A. Gram negative bacteria isolated from blood cultures in a general hospital. Mixrobiologiea 1994; 17: 111-122. 0 Chamberland S, L'Ecuyer J, Lessard C, Bernier M, Provencher P, Bergeron MG. Antibiotic susceptibility profiles of 941 Gram-negative bacteria isolated from septicemic patients throughout Canada. The 7~anadian Study Group. Clin Infect Dis 1992; 15: 615-628. 1 Williams GT, Houang ET, Shaw EJ, Tabaqchali S. Bacteraemia in a London teaching hospital 1966-75. Lancet 1976; ii: 1291-1293.
23
I2 Brown DFI, Perry SF. Methods used in the United Kingdom for the culture of micro-organisms from blood. I Clin Pathol 1992; 45: 468-474. 13 Pittet D, Wenzel RP. Nosocomial bloodstream infections. Secular trends in rates, mortality, and contribution to total hospital deaths. Arch Intern Med 1995; 155: 1177-1184. 14 Mulligan ME, Murray-Leisure KA, Ribner BS et aI. Methicillinresistant Staphylococcus aureus: a consensus review of the microbiology, pathogenesis, and epidemioIogy with implications for prevention and management. Am l Med 1993; 94: 313-328. 15 BoyceJM. Methicillin-resistantStaphylococcusaureusinhospitalsand long-term care facilities: microbiology, epidemiology, and preventive measures. Infect Control Hosp Epidemiol 1992; 13: 725-737.