Antibiograms created by aggregating hospital-wide susceptibility data from diverse patients can be misleading. To demonstrate the utility of age- and location-stratified antibiograms, we compared stratified antibiograms for three common bacterial pathogens, E. coli, S. aureus, and S. pneumoniae. We created stratified antibiograms based on patient age (<18 years, 18–64 years, >/=65 years), and inpatient or outpatient location using all 2009 E. coli and S. aureus, and all 2008–2009 S. pneumoniae isolates submitted to our clinical microbiology laboratory. We compared susceptibility rates among cumulative and stratified antibiograms using descriptive statistics.
For E. coli and S. aureus, the institution-wide antibiogram overestimated resistance in pediatic isolates and underestimated resistance in isolates from the elderly. For E. coli, pediatric isolates were less susceptible to ampicillin and ampicillin-sulbactam and more susceptible to gentamicin and ciprofloxacin compared to adult isolates (p < 0.05 for all), and isolates from patients >65 years were least susceptible to ciprofloxacin (71%). For S. aureus, susceptibility to oxacillin, clindamycin, and levofloxacin was highest among children and decreased with increasing age (p < .001 for all). For S. pneumoniae, pediatric isolates were less susceptible than adult isolates to all agents except penicillin (IV breakpoint). Within children there were significant differences in susceptibility of inpatient and outpatient isolates of E. coli but not of S. aureus or S. pneumoniae.
Stratified antibiograms reveal age - associated differences in susceptibility of E. coli, S. aureus, and S. pneumoniae that are obscured by hospital-wide antibiograms. Age-stratified antibiograms have potential to influence antibiotic selection.
Keywords:Antibiogram; S. aureus; E. coli; S. pneumoniae; Age-stratified
Surveillance of local antimicrobial resistance is an integral part of antimicrobial stewardship. In many institutions, antimicrobial resistance rates are reported using hospital-wide, cumulative antibiograms. Cumulative antibiograms that aggregate data across a hospital can obscure differences among patient populations (e.g. children, adults, or long-term care facility residents), hospital units (e.g. outpatient, ward, nursery, intensive care unit), or anatomic sites (e.g. blood, respiratory, urine). Accordingly, the Clinical and Laboratory Standards Institute (CLSI) has published guidelines recommending stratification of susceptibility data by patient location (e.g. specific ward or long term care facility), clinical service, specimen type, or patient population (e.g. cystic fibrosis patients) (Clinical and Laboratory Standards Institute 2006; Hindler & Stelling 2007).
Despite publication of these guidelines and studies demonstrating the utility of stratified antibiograms (Binkley et al. 2006; Kaufman et al. 1998; Kuster et al. 2008), there continues to be great variability in antibiogram creation and reporting (Xu et al. 2012). Among 65 academic centers in the US, two thirds combined outpatient and inpatient isolates in their antibiograms, and half did not create unit-specific antibiograms (Lautenbach & Nachamkin 2006). It is not clear how many institutions create age group-specific antibiograms. A recent study reported that 6 of 10 pediatric hospitals within larger adult hospitals had cumulative antibiograms without stratification by age (Boggan et al. 2012).
To demonstrate the feasibility and utility of creating stratified antibiograms at our institution, we constructed and compared age and location -stratified antibiograms for three common bacterial pathogens, E. coli, S. aureus, and S. pneumoniae.
Mayo Clinic, in Rochester, MN, is a tertiary care referral center for patients of all ages. Its children’s hospital is not a free-standing facility, but rather a hospital within the adult hospital. Mayo Clinic creates an institution- wide antibiogram that aggregates susceptibility data of isolates from inpatients and outpatients of all ages and specimen types. We created stratified antibiograms based on patient age (<18 years, 18–64 years, >/=65 years), and location of specimen collection (inpatient [IP] or outpatient [OP]) using all 2009 E. coli and S. aureus, and all 2008–2009 S. pneumoniae isolates. We excluded isolates that were not tested for antibiotic susceptibilities and included one isolate per patient per specimen source. Specimen sources were blood or respiratory (all pathogens), urine (E. coli and S. aureus), and soft tissue and bone (S. aureus). Susceptibility was defined using the 2009 CLSI breakpoints. Intermediate susceptibility was considered nonsusceptible. X2 or Fischer’s exact tests were used to compare susceptibility rates. P value < 0.05 was considered statistically significant. This study was approved by the Mayo Clinic institutional review board.
Comparison of the cumulative and age-stratified antibiograms is shown in Table 1. For E. coli, pediatric isolates were less susceptible to ampicillin and ampicillin-sulbactam and more susceptible to gentamicin and ciprofloxacin compared to adult isolates (p < 0.05 for all). E. coli isolates from patients >65 years were less susceptible to ciprofloxacin than were isolates from younger patients (p < .001). There were no differences in susceptibility to trimethoprim-sulfamethoxazole (TMP-SMX) or nitrofurantoin by age.
Table 1. Susceptibility of E. coli, S. aureus , and S. pneumoniae isolates by patient age, Mayo Clinic Rochester, MN
For S. aureus, susceptibility to oxacillin, clindamycin, and levofloxacin was highest among children and decreased with increasing age (p < .001 for all). TMP-SMX susceptibility did not change with age. For methicillin-susceptible S. aureus (MSSA), clindamycin susceptibility did not vary significantly by age group (84% in children, 79% in adults 18–64 years, and 81% in >65 years). However, for methicillin-resistant S. aureus (MRSA), clindamycin susceptibility decreased dramatically with age (76% in children, 43% in adults 18–64 years, 20% in >65 years). For S. pneumoniae, pediatric isolates were less susceptible than adult isolates to penicillin (oral breakpoint), ceftriaxone (CNS breakpoint), tetracyclines, macrolides, and TMP-SMX. There were no significant differences in S. pneumoniae resistance between the adult age groups.
Location-stratified antibiograms for E. coli and S. aureus are shown in Table 2. The institution-wide antibiogram stratified by location, but not age, did not reveal clinically significant differences in susceptibility between IP and OP isolates (data not shown). However, when stratified by both location and age, IP E. coli isolates from children and adults 18–64 years were significantly less susceptible to ampicillin, cefazolin, ceftriaxone, gentamicin, and ciprofloxacin than were OP E. coli isolates (Table 2). In those >65 years, there were no clinically significant differences between IP and OP E. coli susceptibility.
Table 2. Susceptibility of E. coli, S. aureus , and S. pneumoniae isolates by patient age and location, Mayo Clinic Rochester, MN
In contrast, for S. aureus there were no differences in susceptibility between IP and OP pediatric isolates, and small but statistically significant differences between IP and OP adult isolates. For S. pneumoniae, there were no significant differences in susceptibility when isolates were stratified by age and location (data not shown).
At our institution age-stratified antibiograms for E. coli, S. aureus, and S. pneumoniae were significantly different from the institution-wide, cumulative antibiogram. Within age groups, susceptibility varied by IP and OP location at time of specimen collection for E. coli but not for S. aureus or S. pneumoniae.
We observed, as in previous studies, that E. coli and S. aureus isolates from children were the least drug resistant, while those from patients > 65 years were the most drug resistant (Boggan et al. 2012; David et al. 2006). However, this trend was obscured by the institution-wide antibiogram which reported average values, thus overestimating resistance in pediatric isolates and underestimating resistance in isolates from the elderly. A high prevalence of E. coli and S. aureus drug resistance in the elderly is well-documented and most likely reflects greater comorbidities, hospitalizations, and antimicrobial exposure among older patients (David et al. 2006; Swami et al. 2012). In contrast, for S. pneumoniae, pediatric isolates were more drug resistant than adult isolates. The greater drug resistance among pediatric S.pneumoniae isolates is probably due to the high utilization of penicillins in children and the likelihood that the S. pneumoniae isolates used to create the antibiogram reflected complicated or refractory infections, since S. pneumoniae is not routinely cultured in uncomplicated otitis media or pneumonia.
Clinician reliance on institution-wide antibiograms that do not accurately reflect susceptibility rates in certain patient groups might lead to inappropriate empiric antibiotic prescribing. Overestimating resistance in pediatric S. aureus or E. coli could lead to prescribing of unnecessarily broad antibiotics in children which, in turn, can lead to increasingly drug resistant pathogens and C. difficile infections. For example, we have observed that local providers avoid clindamycin for treatment of pediatric skin and soft tissue infections because the cumulative antibiogram reports high rates of clindamycin resistance. An age-stratified analysis revealed that in our geographic region, clindamycin susceptibility among pediatric S. aureus isolates is above 80% and this drug remains an important therapeutic option for treatment of skin and soft tissue infections, including MRSA, in children but not in older adults. Similarly, providers in our community have followed national trends (Copp & Hersh 2011) by increasingly prescribing third generation cephalosporins for empiric treatment of pediatric urinary tract infections. Prescribing of fewer broad - spectrum agents could be encouraged by creation and dissemination of a pediatric antibiogram demonstrating E. coli with high susceptibility to narrow - spectrum cephalosporins. Boggan et al. recently reported that effective antibiotic prescribing among pediatricians, as measured through responses to clinical vignettes, improved when pediatric-specific antibiograms were provided (Boggan et al. 2012).
For elderly patients, availability of age-specific antibiograms that do not underestimate drug resistance is also likely to enhance appropriate antibiotic selection, which in turn, can optimize outcomes (Ibrahim et al. 2000). We recently determined that in our county, a quarter of elderly patients with fluoroquinolone-resistant E. coli infections received ineffective empiric therapy with a fluoroquinolone and had persistent or recurrent infections, likely due to lack of awareness among providers about local resistance rates in this age group. Furthermore, because residents of long-term care facilities are at high risk for colonization or infection with multidrug-resistant organisms, some have advocated creation of antibiograms specific for long-term care facility residents (Philippe et al. 2011). However, because many long-term care facilities lack resources to create their own antibiograms, a more practical option may be for providers to rely on hospital-based age-stratified antibiograms, as we have created.
Within age groups, we noted clinically significant differences in susceptibility of IP and OP isolates of E. coli but not of S. aureus or S. pneumoniae. E. coli OP isolates were generally more drug susceptible than IP isolates, especially among children. In contrast, for S. aureus, differences in susceptibilities between IP and OP isolates from children as well as adults were not statistically significant. These findings suggest that age- and location- stratification of E. coli, the most common pathogen isolated in children and in urinary tract infections in general, might be a valuable tool to guide empiric antibiotic selection for management of pediatric urinary tract infections.
This study has several limitations. Since it was laboratory-based, we did not have associated clinical history and could not determine if cultures represented infection or colonization, or were community-associated vs. healthcare-associated isolates. Although we eliminated duplicate isolates, our sample was also likely biased toward patients with complicated or refractory infections and prior antibiotic exposure, since such patients have cultures sent more frequently than patients with uncomplicated infections. Thus, the isolates in our collection may be more drug-resistant than those in the general population. Lastly, our susceptibility data reflect local epidemiology and may not be generalizable to other geographic regions or institutions. Despite these limitations, we demonstrate that creation of age and location-stratified antibiograms is feasible and valuable.
In conclusion, stratified antibiograms reveal age - associated differences in susceptibility of E. coli, S. aureus, and S. pneumoniae that are obscured by hospital-wide antibiograms. Further stratification of E. coli isolates by both age and IP or OP location may also be useful to clinicians who manage pediatric urinary tract infections. Although the proportion of institutions that create stratified vs. cumulative antibiograms is not clear, we believe that more facilities should create age - stratified antibiograms especially if they serve diverse patient groups (i.e. are not free-standing children’s hospitals or long-term care facilities). More research is needed to determine if improved antibiograms can be valuable stewardship tools that facilitate appropriate empiric antibiotic selection and enhance surveillance of antimicrobial resistance trends (Hebert et al. 2012; McGregor et al. 2009).
Previously presented in abstract form at the Infectious Disease Society of America meeting, 2011, Boston, MA.
All authors report no conflicts of interest relevant to this article.
SS performed data analysis and drafted the manuscript. RB developed research question and assisted with data analysis and manuscript preparation. Both authors read and approved the final manuscript.
We thank Peggy Kohner and Dr. Robin Patel, Mayo Clinic, for providing us with susceptibility data from the clinical microbiology laboratory at Mayo Clinic.
This study was funded in part by the T. Denny Sanford Pediatric Collaborative Research Fund.
Binkley S, Fishman NO, LaRosa LA, et al. (2006) Comparison of unit-specific and hospital-wide antibiograms: potential implications for selection of empirical antimicrobial therapy. Infect Control Hosp Epidemiol 27(7):682-687 PubMed Abstract | Publisher Full Text
Copp HLSD, Hersh AL (2011) National ambulatory antibiotic prescribing patterns for pediatric urinary tract infection, 1998–2007. Pediatrics 127:1027-1033 PubMed Abstract | Publisher Full Text | PubMed Central Full Text
David M, Crawford S, Boyle-Vavra S, Hostetler M, Kim D, Daum R (2006) Contrasting pediatric and adult methicillin-resistant Staphylococcus aureus isolates. Emerg Infect Dis 12:631-637 PubMed Abstract | Publisher Full Text | PubMed Central Full Text
Hebert C, Ridgway J, Vekkhter B, Brown E, Weber S, Robicsek A (2012) Demonstration of a weighted-incidence syndromic combination antibiogram: An empiric prescribing decision aid. Infect Control Hosp Epidemiol 33:381-388 PubMed Abstract | Publisher Full Text
Hindler JF, Stelling J (2007) Analysis and presentation of cumulative antibiograms: a new consensus guideline from the Clinical and Laboratory Standards Institute. Clin Infect Dis 44(6):867-873 PubMed Abstract | Publisher Full Text
Ibrahim E, Sherman S, Ward S, VJ F, MH K. (2000) The influence of inadequate antimicrobial treatment of bloodstream infections on patient outcomes in the ICU setting. Chest 118:146-155 PubMed Abstract | Publisher Full Text
Kaufman D, Haas CE, Edinger R, Hollick G (1998) Antibiotic susceptibility in the surgical intensive care unit compared with the hospital-wide antibiogram. Arch Surg 133(10):1041-1045 PubMed Abstract | Publisher Full Text
Kuster SP, Ruef C, Zbinden R, et al. (2008) Stratification of cumulative antibiograms in hospitals for hospital unit, specimen type, isolate sequence and duration of hospital stay. J Antimicrob Chemother 62(6):1451-1461 PubMed Abstract | Publisher Full Text
Lautenbach E, Nachamkin I (2006) Analysis and presentation of cumulative antimicrobial susceptibility data (antibiograms): Substantial variability across medical centers in the United States. Infect Control Hosp Epidemiol 27:409-412 PubMed Abstract | Publisher Full Text
McGregor JC, Dumyati G, Casiano-Colon AE, Chang PJ, Klevens RM (2009) Usefulness of antibiogram surveillance for methicillin-resistant Staphylococcus aureus in outpatient pediatric populations. Diagn Microbiol Infect Dis 64(1):70-75 PubMed Abstract | Publisher Full Text
Swami S, Liesinger J, Shah N, Baddour L, Banerjee R (2012) Incidence of antibiotic-resistant Escherichia coli bacteriuria according to age and location of onset: a population-based study from Olmsted County, Minnesota. Mayo Clin Proc 87:753-759 PubMed Abstract | Publisher Full Text | PubMed Central Full Text
Xu R, Polk R, Stencel L, et al. (2012) Antibiogram compliance in University HealthSystem Consortium participating hospitals with Clinical and Laboratory Standards Institute guidelines. Am J Health Syst Pharm 69:598-604 PubMed Abstract | Publisher Full Text