MAJOR ARTICLE
Antimicrobial Resistance among Streptococcus
pneumoniae in the United States: Have We Begun
to Turn the Corner on Resistance to Certain
Antimicrobial Classes?
Gary V. Doern, Sandra S. Richter, Ashley Miller, Norma Miller, Cassie Rice, Kristopher Heilmann
and Susan Beekmann
Department of Pathology, University of Iowa College of Medicine, Iowa City
Background. Antimicrobial resistance has emerged as a major problem in Streptococcus pneumoniae in the
United States during the past 15 years. This study was undertaken to elucidate the current scope and magnitude
of this problem in the United States and to assess resistance trends since 1994–1995.
Methods. A total of 1817 S. pneumoniae isolates obtained from patients with community-acquired respiratory
tract infections in 44 US medical centers were characterized during the winter of 2002–2003. The activity of 27
antimicrobial agents was assessed. In addition, selected isolates were examined for the presence of mutations in
the quinolone-resistance determining regions (QRDRs) of parC and gyrA that resulted in diminished fluoroquinolone activity. The results of this survey were compared with the results of 4 previous surveys conducted in a
similar manner since 1994–1995.
Results. Overall rates of resistance (defined as the rate of intermediate resistance plus the rate of resistance)
were as follows: penicillin, 34.2%; ceftriaxone, 6.9%; erythromycin, 29.5%; clindamycin, 9.4%; tetracycline, 16.2%;
and trimethoprim-sulfamethoxazole (TMP-SMX), 31.9%. No resistance was observed with vancomycin, linezolid,
or telithromycin; 22.2% of isolates were multidrug resistant; 2.3% of isolates had ciprofloxacin MICs of ⭓4.0 mg/
mL. It was estimated that 21.9% of the isolates in this national collection had mutations in the QRDRs of parC
and/or gyrA, with parC only mutations occurring most often (in 21% of all isolates). Trend analysis since 1994–
1995 indicated that rates of resistance to b-lactams, macrolides, tetracyclines, TMP-SMX, and multiple drugs have
either plateaued or have begun to decrease. Conversely, fluoroquinolone resistance among S. pneumoniae is becoming more prevalent.
Conclusion. It appears that, as fluoroquinolone resistance emerges among S. pneumoniae in the United States,
resistance to other antimicrobial classes is becoming less common.
Antimicrobial resistance among clinical isolates of
Streptococcus pneumoniae in the United States first
emerged as a significant problem during the early 1990s
[1]. Since then, the prevalence of resistance to b-lactams
has increased steadily, with rates of nonsusceptibility to
penicillin approaching 35% by 2002 [2–6]. A similar
trend has been observed in the United States for multiple non–b-lactam antimicrobial classes, including
Received 10 July 2004; accepted 3 March 2005; electronically published 7 June
2005.
Reprints or correspondence: Prof. Gary V. Doern, Medical Microbiology Div.,
Dept. of Pathology, University of Iowa College of Medicine, Iowa City, IA 52242
([email protected]).
Clinical Infectious Diseases 2005; 41:139–48
2005 by the Infectious Diseases Society of America. All rights reserved.
1058-4838/2005/4102-0001$15.00
macrolides, clindamycin, tetracyclines, trimethoprimsulfamethoxazole (TMP-SMX), and chloramphenicol.
Current rates of resistance to these agents among S.
pneumoniae have been reported to be ∼30% [2], 9%–
10% [3], 15%–17% [4], 30%–40% [5], and 12% [6],
respectively. The only class of orally administered antimicrobials that has seemingly escaped the problem of
emerging resistance among S. pneumoniae is the fluoroquinolones [2–6]. However, in view of the recent introduction in the United States of oral fluoroquinolone
therapy for the management of respiratory tract infections in adults, as well as the increasing use of agents
such as levofloxacin (a compound with borderline potency against S. pneumoniae), fluoroquinolone resistance profiles for this pathogen might be expected to
change.
Antimicrobial Resistance among S. pneumoniae in the US • CID 2005:41 (15 July) • 139
With this in mind, it was of interest to examine the results
of a recent national, multicenter surveillance program aimed
at defining current antimicrobial resistance rates among S.
pneumoniae in the United States. Since 1994–1995, we have
maintained a longitudinal, multicenter surveillance program in
the United States called the Global Respiratory Antimicrobial
Surveillance Project (GRASP). In this study, clinically significant isolates of S. pneumoniae are collected from patients in
medical centers throughout the country and are sent to a central
laboratory for characterization by means of reference standard
methods. This survey was conducted during 5 different winter
seasons: 1994–1995, 1997–1998, 1999–2000, 2001–2002, and
2002–2003. The results of the first 4 surveys have been described
elsewhere [2, 5, 6]. The results of the 2002–2003 survey are
reported herein with comparisons made to the results of surveys
conducted during the first 4 survey periods.
from their respective manufacturers as laboratory-grade powders. Isolates were tested using Mueller-Hinton broth supplemented with 3% lysed horse blood at a final volume of 100
mL/well, with a final inoculum concentration of ∼5 ⫻ 10 5 cfu/
mL. Microdilution trays were incubated at 35C in ambient air
for 22–24 h before the MICs were visually determined. The
following quality-control strains were used throughout the
study: S. pneumoniae (ATCC 49619) and 3 laboratory strains
(I-01253, 01424, and 01590) that were selected to yield onscale MIC values within both the higher and lower concentration ranges of each antibiotic tested. The MIC interpretive criteria of the NCCLS were used to calculate percentages of isolates
that were susceptible, intermediately resistant, and resistant to
each agent [8].
Characterization of mutations in the quinolone-resistance
determining regions (QRDRs) of the parC and gyrA genes was
accomplished as described elsewhere [9].
MATERIALS AND METHODS
A total of 1817 S. pneumoniae isolates were characterized in
the current study. Isolates were collected at one of 45 different
medical centers (table 1) between 1 December 2002 and 30
April 2003 and sent to the University of Iowa College of Medicine (Iowa City, IA) for further characterization. Isolates were
collected consecutively from unique patients either seen in ambulatory care settings or hospitalized for !48 h. Pneumococcal
isolates from the following specimens and media were included
in the study: middle ear fluid, sinus aspirates, CSF and other
normally sterile body fluids, blood cultures, and conjunctival
swabs. In addition, isolates from lower respiratory tract specimens (mostly of sputum and bronchoalveolar lavage) were
included, but only if the samples had been judged to be representative on the basis of Gram stain results and if the isolate
was considered to be of clinical significance by the submitting
laboratory. Isolates were shipped to the referral laboratory via
overnight courier on polyester swabs in Amies transport media
with charcoal. This transport method permitted recovery of
100% of shipped isolates. Isolates were reidentified as S. pneumoniae on the basis of Gram stain results and colony morphology, results of Optochin susceptibility tests, and sodium
deoxycholate solubility. Stock cultures of isolates were prepared
in the referral laboratory by means of the Microbank bead system
(Pro-Lab Diagnostics) and were stored indefinitely at ⫺70C.
MICs of the following 27 antimicrobial agents were determined by the broth microdilution method described by the
NCCLS [7]: penicillin, amoxicillin, amoxicillin-clavulanate,
cefaclor, cefprozil, cefuroxime axetil, ceftriaxone, cefpodoxime,
cefdinir, cefditoren, chloramphenicol, clindamycin, erythromycin, azithromycin, clarithromycin, ciprofloxacin, levofloxacin, gatifloxacin, moxifloxacin, gemifloxacin, tetracycline,
TMP-SMX, rifampin, vancomycin, linezolid, telithromycin,
and quinupristin-dalfopristin. All antibiotics were received
140 • CID 2005:41 (15 July) • Doern et al.
RESULTS
The in vitro activities of 27 antimicrobial agents against 1817
isolates of S. pneumoniae examined in this study are summarized in table 2. The overall rate of penicillin resistance (defined
as the rate of intermediate resistance plus the rate of resistance)
was 34.2% (15.7% of isolates were intermediately resistant, and
18.5% were resistant). Among resistant strains, 208 (11.4%)
had penicillin MICs of 2 mg/mL, 97 (5.3%) had MICs of 4 mg/
mL, and the remaining 31 (1.7%) had MICs of 8 mg/mL. Overall
rates of resistance to other b-lactams varied between 6.9% (for
ceftriaxone) and 33.6% (for cefaclor). In a comparison of antimicrobial MICs for individual isolates, amoxicillin, amoxicillin-clavulanate, and ceftriaxone MICs were generally the
same as MICs across the range of penicillin agents; cefuroxime,
cefpodoxime, and cefdinir MICs were 2 times higher; the cefprozil MIC was 4 times higher; and the cefaclor MIC was 32
times higher (data not shown). Cefditoren MICs were typically
2–4 times lower than penicillin MICs for individual isolates.
The overall rates of resistance to macrolides ranged from
28.7% to 29.5%. Clarithromycin MICs were consistently 2 times
lower than erythromycin MICs, which, in turn, were 2 times
lower than azithromycin MICs. Among the 536 erythromycinresistant isolates, 371 (69.2%) expressed the efflux phenotype
(erythromycin MICs of 0.5–32 mg/mL). All but 2 of these isolates
had clindamycin MICs of ⭐2 mg/mL. A total of 165 isolates
(30.8%) expressed the macrolide-lincomycin–streptogramin B
(MLSB) phenotype (all had erythromycin MICs of ⭓64 mg/mL
and clindamycin MICs of ⭓1 mg/mL). All but 3 of these isolates
had clindamycin MICs of ⭓4 mg/mL.
The overall rates of resistance to tetracycline, TMP-SMX,
and chloramphenicol were 16.2%, 31.9%, and 4.7%, respectively. Telithromycin resistance was not observed. Only a single
isolate was defined as being nonsusceptible to telithromycin
Table 1. Location of and no. of isolates contributed by participants in the Global
Respiratory Antimicrobial Surveillance Project, Winter, 2002–2003.
Region, medical center
Location
No. of
isolates
New England
Dartmouth-Hitchcock Medical Center
Lebanon, NH
Beth Israel Deaconess Medical Center
Boston, MA
24
52
Danbury Hospital
Danbury, CT
49
Lahey Clinic
Burlington, MA
27
Middle Atlantic
Columbia-Presbyterian Medical Center
New York, NY
50
Temple University Hospital
Philadelphia, PA
48
SUNY Upstate Medical University
Syracuse, NY
34
Geisinger Medical Center
Danville, PA
48
University of Rochester Medical Center
Rochester, NY
49
North Shore–LIJ Health System
Lake Success, NY
18
South Atlantic
Children’s Hospital National Medical Center
Washington, DC
5
University of North Carolina Hospital
Chapel Hill, NC
50
Dekalb General Hospital
Decatur, GA
50
Mt. Sinai Medical Center Miami Beach
Miami Beach, FL
15
Carolinas Medical Center
Charlotte, NC
49
Veterans Affairs Medical Center (VAMC) Tampa
Tampa, FL
Baptist Medical Center
Jacksonville, FL
1
50
East North Central
Cleveland Clinic Foundation
Cleveland, OH
54
Children’s Hospital of Wisconsin
Milwaukee, WI
34
Henry Ford Hospital
Detroit, MI
48
Clarian Health Methodist Hospital
Indianapolis, IN
52
Rush-Presbyterian St. Luke’s Medical Center
Chicago, IL
36
Evanston Northwestern Healthcare
Evanston, IL
27
East South Central
University of South Alabama Medical Center
Mobile, AL
50
University of Louisville Hospital
Louisville, KY
39
University of Alabama at Birmingham
Birmingham, AL
37
Mayo Clinic
Rochester, MN
49
University of Iowa Hospitals and Clinics
Iowa City, IA
42
Barnes-Jewish Hospital
St. Louis, MO
49
University of Kansas Medical Center
Kansas City, KS
51
Texas Children’s Hospital
Houston, TX
49
Parkland Hospital and Health System
Dallas, TX
37
University of Texas Health Science Center
San Antonio, TX
49
St. Francis Hospital
Tulsa, OK
51
University of New Mexico Health Sciences Center
Albuquerque, NM
46
Denver Health Medical Center
Denver, CO
50
University of Utah ARUP Laboratories
Salt Lake City, UT
34
West North Central
West South Central
Mountain
Good Samaritan Medical Center
Phoenix, AZ
48
University Medical Center of Southern Nevada
Las Vegas, NV
51
Pacific
University of California (UC)–San Diego Medical Center
San Diego, CA
UC–Los Angeles Medical Center
Los Angeles, CA
46
8
UC–San Francisco Medical Center
San Francisco, CA
44
VAMC Portland
Portland, OR
18
University of Washington Medical Center
Seattle, WA
49
UC-Irvine Medical Center
Irvine, CA
50
Table 2. In vitro activity of 27 antimicrobial agents against 1817 isolates of Streptococcus
pneumoniae collected at 44 US medical centers, Winter, 2002–2003.
MIC50,
mg/mL
MIC90,
mg/mL
Penicillin
Amoxicillin
0.03
0.03
2
2
Amoxicillin-clavulanate
Cefaclor
Cefprozil
0.03
1
0.25
2
⭓16
16
Cefuroxime axetil
Cefpodoxime
Cefdinir
Cefditoren
0.06
0.06
0.12
0.015
Antimicrobial agent(s)
Ceftriaxonea
Erythromycin
Azithromycin
Clarithromycin
Clindamycin
Telithromycin
Tetracycline
Chloramphenicol
TMP-SMX
Rifampin
Ciprofloxacin
Levofloxacin
Gatifloxacin
Moxifloxacin
Gemifloxacinb
Vancomycin
Linezolid
Quinupristin-dalfopristin
NOTE.
4
4
8
0.5
0.03
⭐0.06
1
16
⭐0.12
⭐0.06
⭐0.03
⭐0.008
⭐0.5
2
⭐0.25
0.06
1
1
0.25
0.12
16
8
0.25
0.25
⭓32
4
8
0.06
2
1
0.5
0.25
0.03
0.25
1
0.25
0.06
0.5
1
0.5
Intermediately
resistant,
% of isolates
Resistant,
% of isolates
15.7
3.2
18.5
5.3
⭐0.008 to 16
⭐0.25 to ⭓16
⭐0.03 to 64
3.7
5.6
3.5
4.6
28.0
19.9
⭐0.03
⭐0.015
⭐0.015
⭐0.004
2.3
4.1
3.3
NA
21.6
21.0
23.5
NA
⭐0.008 to 16
⭐0.06 to ⭓256
5.3
0.9
1.6
28.6
⭐0.12
⭐0.06
⭐0.03
⭐0.008
⭐0.5
⭐0.5
⭐0.25
⭐0.008
⭐0.03
⭐0.03
⭐0.015
⭐0.015
1.3
2.1
0.1
0.1
0.6
NA
6.6
0
NA
0.1
0.1
0.3
27.4
26.9
9.3
0
15.6
4.7
25.4
0.2
2.3
0.7
0.7
0.2
MIC range,
mg/mL
⭐0.008 to 8
⭐0.008 to 16
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
⭓64
32
32
4
⭓256
⭓256
⭓128
2
⭓32
⭓32
⭓16
⭓8
32
32
8
4
⭐0.004 to 2
0.12–1
0.12–2
0.25–4
0.2
NA
NA
0.3
0.2
0
0
0.06
NA, not available; TMP-SMX, trimethoprim-sulfamethoxazole.
a
The nonmeningeal MIC breakpoints of the NCCLS were used: susceptible, ⭐1 mg/mL; intermediate resistance,
2 mg/mL; and resistance, ⭓4 mg/mL [8].
b
A total of 851 isolates were tested for susceptibility to gemifloxacin.
(MIC, 2 mg/mL). No resistance to vancomycin or linezolid was
recognized.
Among the 1817 isolates characterized in this study, 404
(22.2%) were found to be multidrug resistant. Multidrug resistance was defined as intermediate resistance or resistance to
penicillin plus intermediate resistance or resistance to at least
2 of the following 4 agents: erythromycin, TMP-SMX, chloramphenicol, and tetracycline.
MIC frequency distributions for the 5 fluoroquinolones examined in this study are presented in table 3. Ciprofloxacin
and levofloxacin MICs were generally the same for individual
isolates. Gatifloxacin was consistently 4 times more active than
ciprofloxacin and levofloxacin, moxifloxacin was 8 times more
active, and gemifloxacin was 32 times more active. All 5 agents
exhibited distinct modal MIC values, as follows: ciprofloxacin
and levofloxacin, 1.0 mg/mL; gatifloxacin, 0.25 mg/mL; moxi142 • CID 2005:41 (15 July) • Doern et al.
floxacin, 0.12 mg/mL; and gemifloxacin, 0.03 mg/mL. On the
basis of current NCCLS interpretive criteria [8], the following
percentages of isolates would have been classified as intermediately resistant or resistant to levofloxacin, gatifloxacin, moxifloxacin, and gemifloxacin: 0.8%, 0.7%, 0.5%, and 0.5%, respectively. MIC breakpoints for ciprofloxacin have not been
defined by the NCCLS. However, Chen et al. [10] suggested
that a single ciprofloxacin MIC breakpoint of ⭓4 mg/mL is a
useful phenotypic means for assessing the overall activity of
fluoroquinolones against pneumococci. Among the isolates
characterized in this study, 2.3% had ciprofloxacin MICs of
⭓4.0 mg/mL.
A total of 301 isolates were examined for the presence of
mutations in the QRDRs of the genes encoding the C and E
subunits of topoisomerase IV (parC and parE) and the genes
encoding the A and B subunits of gyrase A (gyrA and gyrB).
Table 3. Fluoroquinolone MIC frequency distributions for 1817 Streptococcus pneumoniae isolates recovered at 45 US medical
centers, Winter, 2002–2003.
No. (cumulative %) of isolates, by MIC in mg/mL
Fluoroquinolone
⭐0.03
0.06
Ciprofloxacin
2 (0.1)
1 (0.2)
3 (0.3)
30 (2.0)
Levofloxacin
1 (0.1)
0
3 (0.2)
14 (1.0)
0.12
0.25
0.5
1
2
4
8
16
32
190 (12.4)
1064 (71.0)
485 (97.7)
27 (99.2)
5 (99.5)
3 (99.6)
7 (100)
317 (18.4)
1416 (96.4)
52 (99.2)
2 (99.3)
5 (99.6)
6 (99.9)
1 (100)
Gatifloxacin
3 (0.2)
9 (0.7)
190 (11.1)
1018 (67.1)
572 (98.6)
13 (99.3)
0
9 (99.8)
3 (100)
…
…
Moxifloxacin
28 (1.5)
258 (15.7)
1328 (88.8)
188 (99.2)
3 (99.3)
2 (99.5)
6 (99.8)
4 (100)
…
…
…
700 (82.3)
141 (98.8)
6 (99.5)
2 (99.8)
1 (99.9)
0
1 (100)
…
…
…
…
Gemifloxacin
a
a
Data are for 851 isolates.
These included all 42 isolates with ciprofloxacin MICs of ⭓4
mg/mL, 176 randomly selected isolates with ciprofloxacin MICs
of 2 mg/mL and 79 randomly selected isolates with ciprofloxacin
MICs of 1 mg/mL. The results are presented in table 4.
Overall rates of resistance to selected antimicrobial agents,
sorted according to geographic region, patient age, and source
of isolates, are depicted in table 5.
Because this 2002–2003 survey represented a continuation
of a survey that had been conducted 4 times previously (during
the winters of 1994–1995, 1997–1998, 1999–2000, and 2001–
2002), it was of interest to examine antibiotic resistant trends
among S. pneumoniae in the United States during the past
decade. National rates of resistance to penicillin, erythromycin,
tetracycline, TMP-SMX, and multiple drugs for the current
study and the 4 previous surveys are presented in figure 1. In
addition, the percentage of isolates with ciprofloxacin MICs of
⭓4 mg/mL during the different survey periods is presented.
With penicillin, tetracycline, and multiple drugs, resistance
rates appeared to peak in 1999–2000 and have leveled off since
then. The rate of resistance to TMP-SMX appears to be decreasing, after a peak in 1999–2000. The rates of resistance to
macrolides continue to increase. However, since 1999–2000,
the rates of increase have slowed. The relative percentage of
macrolide-resistant isolates expressing the M (efflux) phenotype
versus the high-level MLSB phenotype has remained nearly constant during the 5 survey periods (65.2%–72.2% vs. 25.3%–
32.1%). Furthermore, the macrolide MIC distributions among
strains with the M phenotype have also remained essentially
unchanged during the 5 survey periods, both when all strains
were examined collectively and when systemic pneumococcal
isolates were examined separately (unpublished data). For example, the clarithromycin MIC50 and MIC90 values for isolates
with the M phenotype were 2–4 and 8–16 mg/mL, respectively,
during all 5 survey periods, both when the entire collections
and when just systemic isolates were compared.
A conspicuously different pattern was noted for fluoroquin-
Table 4. Frequency of mutations in the quinolone-resistance determining regions for 1781 Streptococcus pneumoniae isolates
recovered at 45 US medical centers, according to ciprofloxacin MICs, Winter, 2002–2003.
Ciprofloxacin MIC
0.5 mg/mL
No. of
isolates tested/
no. recovered (%)
No. of isolates,
by mutation
None
parE
only
No. of isolates with indicated mutation, by gene
Any
parE
parC only
gyrA only
parC + gyrA
4/190 (2.1)
2
2
2
0
0
0
1 mg/mL
79/1064 (7.4)
23
35
48
Lys137-Asn (18), Gly128-Asp (2),
Ser52-Gly and Lys137-Asn (1)
0
0
2 mg/mL
176/485 (36.3)
36
109
132
Lys137-Asn (25), Asp147-Ile (1),
Gly128-Asp (1), Glu100-Ala
and Lys137-Asn (1), Ser79Phe (1), Ser79-Tyr (1)
4 mg/mL
27/27 (100)
2
10
19
Lys137-Asn (3), Asn91-Asp (2),
Asp78-Ala (1), Asp83-Gly (1),
Ser79-Phe (4), Ser79-Tyr (2),
Ser79-Tyr and Lys137-Asn (1)
8 mg/mL
5/5 (100)
0
1
4
16 mg/mL
3/3 (100)
0
0
2
0
32 mg/mL
7/7 (100)
0
0
6
0
Lys137-Asn (2)
Arg95-Cys (1)
0
Ser114-Gly (1),
Ser81-Phe (1)
Glu85-Ala (1)
0
0
Lys137-Asn + Ser81-Phe (1)
0
Ser79-Phe + Glu85-Lys (2)
Ser79-Phe + Ser81-Phe (3), Ser79Phe and Asp83-Tyr + Ser81-Phe
(1), Ser79-Phe and Lys137-Asn
+ Ser81-Phe (2), Ser79-Tyr and
Leu130-Phe + Ser81-Phe (1)
Antimicrobial Resistance among S. pneumoniae in the US • CID 2005:41 (15 July) • 143
Table 5. Geographic, demographic, and clinical characteristics associated with susceptibility rates, by antimicrobial
agent, for 1817 pneumococcal isolates recovered at 45 US medical centers, Winter, 2002–2003.
Variable
No. of
isolates
Penicillin
Erythromycin
Tetracycline
Chl
TMP-SMX
Multiple
drugs
Cpfx
Region
New England
152
15.1 and 13.8
2.6 and 20.4
0.7 and 11.2
4.6
5.3 and 18.4
13.2
5.3
Middle Atlantic
247
15.4 and 13.8
0.8 and 24.7
1.2 and 15.4
5.7
2.0 and 23.5
19.8
2.4
South Atlantic
220
16.4 and 24.1
0 and 30.9
0.5 and 10.9
2.7
5.0 and 31.4
25.5
0.9
East North Central
251
16.3 and 19.9
0.4 and 33.1
0.4 and 17.5
6.0
7.6 and 27.5
24.3
4.0
East South Central
126
14.3 and 20.6
0.8 and 40.5
0 and 15.1
5.6
7.1 and 31.0
30.2
0.8
West North Central
191
14.1 and 16.8
1.6 and 27.8
1.1 and 15.2
4.7
8.4 and 17.8
17.3
3.1
West South Central
186
18.3 and 20.4
0.5 and 31.2
0.5 and 15.1
2.2
8.6 and 30.7
26.3
1.1
Mountain
229
15.3 and 18.3
0.9 and 24.9
0 and 15.3
2.6
8.3 and 26.2
20.1
1.3
Pacific
215
15.4 and 18.6
1.4 and 26.5
0.9 and 22.8
8.4
7.4 and 21.9
24.2
1.9
0–5 years
366
18.9 and 27.9
0.8 and 38.5
1.1 and 19.7
3.3
9.0 and 36.5
32.0
0.6
6–20 years
173
17.3 and 19.1
1.2 and 28.3
0.6 and 15.0
6.9
6.9 and 26.0
24.9
3.5
21–64 years
876
14.5 and 15.0
0.8 and 24.7
0.6 and 13.1
4.5
5.4 and 21.5
18.2
2.4
⭓65 years
399
14.8 and 17.0
1.3 and 28.1
0.3 and 17.5
5.8
6.8 and 23.6
21.3
3.3
473
13.8 and 13.5
0.6 and 22.2
0.4 and 9.3
3.8
5.5 and 21.4
18.4
0.6
53
22.6 and 18.9
1.9 and 35.9
0 and 18.9
3.8
13.2 and 18.9
22.6
3.8
115
9.6 and 33.0
0.9 and 36.5
0 and 21.7
3.5
8.7 and 32.2
31.3
0.9
a
Age
Site or specimen
Blood
CSF and/or sterile body fluid
Ear
Eye
70
20.0 and 5.7
2.9 and 31.4
1.4 and 17.1
1.4
7.1 and 22.9
17.1
1.4
LRT
888
16.7 and 18.8
1.1 and 28.7
0.8 and 16.4
6.1
6.4 and 25.5
21.7
3.2
Sinus
111
18.0 and 27.9
0 and 43.2
0 and 24.3
URT
82
13.4 and 17.1
0 and 25.6
1.2 and 17.1
Other
25
12.0 and 32.0
0 and 28.0
0 and 20.0
3.6
5.4 and 41.4
36.0
4.5
0
9.8 and 19.5
20.7
2.4
0 and 36.0
28.0
0
12.0
NOTE. Data are percentages expressed as either the rate of intermediate resistance and rate of resistance or the overall rate of resistance,
unless otherwise indicated. Chl, chloramphenicol; Cpfx, ciprofloxacin; TMP-SMX, trimethoprim-sulfamethoxazole.
a
Age was unknown for 3 patients.
olones. With a ciprofloxacin MIC of ⭓4 mg/mL as an indicator
of diminished fluoroquinolone activity, a marked decrease in
fluoroquinolone activity appeared to have occurred in the
United States sometime during the 2-year period between the
winters of 1999–2000 and 2001–2002. This change was sustained through the most recent survey period, the winter of
2002–2003.
Another, perhaps more refined means for comparing pneumococci with respect to the effect of fluoroquinolones over
time is mutation analysis. On the basis of the results presented
in table 5, we estimate that 21.9% of the 2002–2003 collection
of S. pneumoniae isolates had mutations in parC (21%), gyrA
(0.3%), or both (0.6%). A similar analysis of the isolates evaluated during our 1997–1998 survey yielded an estimated QRDR
mutation rate of 4.7% [9].
DISCUSSION
The results of this study provide an objective comparison of
the activity of various antimicrobial agents against a recent
national collection of clinically significant isolates of S. pneumoniae. The overall rates of resistance to b-lactams varied considerably among different agents in this family, from a low of
144 • CID 2005:41 (15 July) • Doern et al.
6.9% for ceftriaxone to a high of 34.2% for penicillin. Variable
rates of resistance to b-lactams were primarily the result of the
different MIC interpretive criteria used to define S. pneumoniae
resistance categories for the various b-lactam agents [8]. As has
been reported elsewhere, the incremental effect of increasing
penicillin MICs was constant for all agents in the b-lactam
family across the entire range of penicillin MICs [11]. The
percentages of isolates with penicillin MICs of 2, 4, and 8 mg/
mL were 11.5%, 5.3%, and 1.7%, respectively. No isolates with
penicillin MICs of ⭓16 mg/mL were recognized.
Overall rates of resistance to macrolides (i.e., clarithromycin,
erythromycin, and azithromycin) were essentially the same
(28.7%–29.5%), despite the fact that clarithromycin MICs were
consistently 2 times lower than erythromycin MICs, which, in
turn, were 2 times lower than azithromycin MICs. Rates of
resistance to other antimicrobials varied between a high of
31.9% (for TMP-SMX) to no recognized resistance (for vancomycin, linezolid, and telithromycin). Overall, 22.2% of the
isolates examined in this study were resistant to penicillin and
at least 2 of the following agents: erythromycin, tetracycline,
TMP-SMX, and chloramphenicol. These isolates were classified
as being multidrug resistant.
Figure 1. Rates of resistance to select antimicrobial agents among respiratory tract isolates of Streptococcus pneumoniae in the United States
during 5 winters since 1994–1995. erm, ermB positive; I, intermediately resistant; mef, mefA positive; R, resistant; TMP-SMX, trimethoprimsulfamethoxazole.
Five fluoroquinolones were examined in this study. As has
been elucidated by others elsewhere [12–15], on a weight basis,
gemifloxacin was the most active, ciprofloxacin and levofloxacin the least active.
A ciprofloxacin MIC of ⭓4 mg/mL has been recommended
as a single phenotypic measure of fluoroquinolone activity that
can be used to reliably track the changing fluoroquinolone
activity profile versus S. pneumoniae [10]. This recommendation is predicated on 2 observations. First, 98%–99% of S.
pneumoniae isolates, which constitutes the modal population
of organisms with low fluoroquinolone MICs, have ciprofloxacin MICs of ⭐2 mg/mL; therefore, a ciprofloxacin MIC of ⭓4
mg/mL represents a sensitive means for detecting an upward
shift in MICs. Second, mutations in the QRDRs of parC and/
or gyrA begin to accumulate at a ciprofloxacin MIC of 4 mg/
mL [10, 16–18]. As is also true with levofloxacin, however, a
single ciprofloxacin MIC breakpoint does not necessarily reflect
the true prevalence of organisms with fluoroquinolone resistance mutations, because not infrequently, organisms with
MICs lower than proposed breakpoints have mutations, especially mutations in the QRDR of parC only [16, 18, 19].
Assessment of the frequency with which resistance mutations
occur, therefore, represents the most refined means for tracking
changes in fluoroquinolone resistance patterns.
In the current study, we estimated that 21.9% of the 1817
isolates examined had mutations in the QRDR of parC and/or
gyrA. Among these, only 10 isolates (i.e., 0.6% of the total) had
mutations in both parC and gyrA. The overall prevalence of
organisms with mutations only in gyrA was estimated to be
even lower (i.e., 0.3%). In other words, the vast majority of
Antimicrobial Resistance among S. pneumoniae in the US • CID 2005:41 (15 July) • 145
organisms with QRDR mutations (∼382 [21%] of 1817 isolates)
had mutations only in parC.
Numerous different mutations were recognized in the
QRDRs of both parC and gyrA, which raises the following
question: what is the relevance of different individual mutations
to diminished fluoroquinolone activity? Certain mutations,
such as those at the ser79 and asp83 positions of parC and the
ser81 and glu85 positions of gyrA, clearly seem to result in
diminished fluoroquinolone effect [9, 16, 20–24]. But what is
the significance of parC mutations at the lys137, gly128, asp147,
and asn91 loci and the gyrA mutations at the arg95 and ser114
positions? If these latter mutations do not contribute to fluoroquinolone “resistance,” then the overall estimated mutation
frequency in our study (i.e., 21.9%) may overstate the problem
of changing fluoroquinolone activity against S. pneumoniae in
the United States. We are left with the empirical observation
that isolates of S. pneumoniae harboring these mutations are
now recognized far more commonly than they have been in
the past in the United States and that this increased prevalence
seems to correspond to a period during which fluoroquinolones
have come to be used more often in the management of respiratory tract infections.
An analysis of resistance rates in the context of various demographic factors revealed several interesting patterns. With
the exception of the fluoroquinolones, resistance rates for other
agents were typically highest in the southern and southeastern
regions of the United States. This has been reported by others
and has seemingly been true since the problem of resistance
with S. pneumoniae first started to emerge in the United States
in the early 1990s [1–6]. With regard to fluoroquinolones, resistance was most often noted in New England and in the East
North Central and West North Central regions of the country.
Similarly, resistance to agents other than the fluoroquinolones
was generally most common among isolates of S. pneumoniae
from children ⭐5 years old and from persons with closedspaced infections, such as acute otitis media and rhinosinusitis.
By contrast, fluoroquinolone resistance was least common
among isolates from young children and from patients with
acute otitis media.
Because this survey represents a continuation of a longitudinal surveillance program that was started in 1994–1995, it
provided us with an opportunity to assess trends in the development of antimicrobial resistance among S. pneumoniae
during the past decade. Much to our surprise, such an analysis
indicated that rates of resistance to b-lactams and tetracyclines
seem to have plateaued in the United States beginning in 1999–
2000 and may even be decreasing. The same was true for the
rate of multidrug resistance among S. pneumoniae. There appears to have been a downward trend in the rate of resistance
to TMP-SMX during the first 3 years of the new millennium.
146 • CID 2005:41 (15 July) • Doern et al.
The increase in the rate of resistance to macrolides has slowed
substantially.
The only class of orally administered antimicrobials currently
relevant to the management of community-acquired respiratory
tract infections for which resistance rates seem to be increasing
markedly is the fluoroquinolones. With use of a single phenotypic definition of fluoroquinolone effect (i.e., a ciprofloxacin MIC of ⭓4 mg/mL), a conspicuous jump in resistance
rates was noted beginning in 2001–2002, with a continuation
of this trend 2 years later in 2002–2003. Fluoroquinolone resistance has been clearly documented in the most dominant
clonal groups of antimicrobial-resistant S. pneumoniae in the
United States, and there is now some evidence demonstrating
clonal expansion of these strains [21–23, 25, 26].
As noted above, the actual scope of the problem of fluoroquinolone resistance in S. pneumoniae may be better judged
by means of mutation analysis. A high percentage of isolates
in the current study (∼21.9%) were estimated to have mutations
in the QRDRs of parC and/or gyrA. This represents a mutation
rate that is substantially higher than the rate of 4.7% observed
in our 1997–1998 study [8]. The vast majority of these isolates
had mutations only in the QRDR of parC and, as result, continued to have low MICs of the more potent fluoroquinolones,
gatifloxacin, moxifloxacin, and gemifloxacin. These organisms,
however, may represent a growing reservoir of pneumococci
in nature that are 1 mutation step away from developing highlevel fluoroquinolone resistance as a result of mutations in both
parC and gyrA.
If we have really begun to turn the corner on S. pneumoniae
resistance to agents other than fluoroquinolones in this country,
the question can be asked, why? At least 4 possible factors may
be involved. Perhaps national, regional, and local campaigns
to encourage more-appropriate use of oral antibiotics in community-based practice in patients with acute respiratory tract
illnesses are beginning to pay off with a resultant decreased
burden of overall antibiotic selective pressure. There is at least
some circumstantial evidence that this may be the case [27].
Second, increasing use of pediatric pneumococcal vaccines directed at preventing infections caused by antibiotic-resistant
strains may be impacting resistance [28–33]. Third, during the
period when resistance rates among S. pneumoniae seemed to
level off if not decrease, there was a concomitant decrease in
the prevalence of influenza, perhaps as a result of either increased numbers of people being vaccinated with influenza vaccine or increased vaccine efficacy [34, 35]. Viral respiratory
tract infections, particularly influenza, result in clear predispositions to secondary bacterial respiratory tract infections. To
wit, a lower prevalence of influenza results in fewer bacterial
infections, including those caused by S. pneumoniae, which
leads in turn to less selective pressure associated with antibiotic
use.
Finally, we may have unwittingly been the beneficiary of the
relatively recent introduction of the fluoroquinolones into the
US market specifically for purposes of treating respiratory tract
infections in adults. Levofloxacin, the first fluoroquinolone
used to treat respiratory infections, was introduced into clinical
practice in the United States in 1997. It was followed in 2001–
2002 by gatifloxacin and then moxifloxacin. Levofloxacin, in
particular, has become a very popular agent in the United States
for the treatment of sinusitis, acute exacerbation of chronic
bronchitis, and community-acquired pneumonia. As its popularity has grown, there has been a proportionate decrease in
the use of nonfluoroquinolones for treating these infections. In
view of the nature of the multidrug-resistant (i.e., coresistance
to b-lactams, macrolides, tetracyclines, and/or TMP-SMX) S.
pneumoniae with stable endemicity in the United States, decreased use of any of these agents could lead to decreased rates
of resistance to all of them.
Of course, another potential consequence of increased
fluoroquinolone use in the management of respiratory tract
infections could be the emergence of fluoroquinolone resistance in S. pneumoniae. On the basis of results of the current
study, it now appears that we have entered a period of substantial change with respect to emerging fluoroquinolone resistance in S. pneumoniae in the United States. This is certainly
consistent with recent reports of the failure of fluoroquinolone therapy in US patients with pneumococcal respiratory
tract infections [36, 37].
This changing profile of activity overlaps a period during
which the use of fluoroquinolones to treat respiratory tract
infections has increased markedly. Levofloxacin, in particular,
has clearly been shown to be associated with the emergence of
S. pneumoniae resistance in individual patients [36–39]. It is
possible that this relationship is influenced by the borderline
potency of levofloxacin against S. pneumoniae.
It is obvious that the emergence of resistance in any bacterial
pathogen is a multifaceted problem. In the case of S. pneumoniae and the fluoroquinolones, it appears that we are just
now beginning to observe change, at least in the United States.
At present, given the newness and relatively modest scope of
this problem, perhaps there is still an opportunity to forestall
the development of an even bigger problem in the future. It is
hoped that the results of this study are in some way instructive
in this effort.
Acknowledgments
The authors are grateful to the following individuals for providing the
isolates of S. pneumoniae characterized in this study: Joseph D. Schwartzman (Lebanon, NH), Paola C. De Girolami (Boston, MA), Paul Iannini
(Danbury, CT), Kim Chapin (Burlington, MA), Phyllis Della-Latta (New
York, NY), Allan L. Truant (Philadelphia, PA), Deanna L. Kiska (Syracuse,
NY), Paul Bourbeau (Danville, PA), Dwight J. Hardy (Rochester, NY),
Christine C. Ginocchio (Lake Success, NY), Joseph M. Campos (Washington, DC), Peter H. Gilligan (Chapel Hill, NC), Robert Jerris (Decatur,
GA), Clarisa Suarez (Miami Beach, FL), Stephen Jenkins (Charlotte, NC),
Ann Hall (Tampa, FL), Diane C. Halstead (Jacksonville, FL), Gerri Hall
(Cleveland, OH), Susan Kehl (Milwaukee, WI), Eileen Burd (Detroit,
MI), Gerald Denys (Indianapolis, IN), Mary Hayden (Chicago, IL), Richard Thomson, Jr. (Evanston, IL), Beth Grover (Mobile, AL), James Snyder
(Louisville, KY), Ken B. Waites (Birmingham, AL), Franklin R. Cockerill
III (Rochester, MN), Joan Hoppe-Bauer (St. Louis, MO), Rebecca Horvath (Kansas City, KS), James Versalovic (Houston, TX), Paul M. Southern, Jr. (Dallas, TX), James H. Jorgenson (San Antonio, TX), Marilyn
Bartel (Tulsa, OK), Gary Overturf (Albuquerque, NM), Michael L. Wilson (Denver, CO), Karen Carroll (Salt Lake City, UT), Michael Saubolle
(Phoenix, AZ), Valerie Leslie (Las Vegas, NV), Sharon Reed (San Diego,
CA), David Bruckner (Los Angeles, CA), Rohan Nadarajah (San Francisco, CA), David Sewell (Portland, OR), Ajit Limaye (Seattle, WA), and
Ellena M. Peterson (Irvine, CA).
Financial support. Abbott Laboratories.
Potential conflicts of interest. G.V.D. currently receives research grant
support from Bayer Pharmaceutical Company and Astra-Zeneca and serves
on the speaker’s bureaus of Abbott, Astra-Zeneca, Bayer, Pfizer, Roche,
Sanofi-Aventis, and GlaxoSmithKline. All other authors: no conflicts.
References
1. Thornsberry C, Brown SD, Yee C, et al. Increasing penicillin resistance
in Streptococcus pneumoniae in the US: effect on susceptibility to oral
cephalosporins. Infect Med 1993; 93:15–24.
2. Doern GV, Heilmann KP, Huynh HK, Rhomberg PR, Coffman SL,
Brueggemann AB. Antibiotic resistance among clinical isolates of Streptococcus pneumoniae in the United States during 1999–2000, including
a comparison of resistance rates since 1994–1995. Antimicrob Agents
Chemother 2001; 45:1721–9.
3. Doern GV, Brown SD. Antimicrobial susceptibility among communityacquired respiratory tract pathogens in the USA: data from PROTEKT
US 2000–01. J Infect 2004; 48:56–65.
4. Karlowsky JA, Thornsberry C, Jones ME, Evangelista AT, Critchley IA,
Sahm DF. Factors associated with relative rates of antimicrobial resistance among Streptococcus pneumoniae in the United States: results
from the TRUST Surveillance Program (1998–2002). Clin Infect Dis
2003; 36:963–70.
5. Doern GV, Brueggemann A, Holley HP Jr, Rauch AM. Antimicrobial
resistance of Streptococcus pneumoniae recovered from outpatients in
the United States during the winter months of 1994 to 1995: results
of a 30-center national surveillance study. Antimicrob Agents Chemother 1996; 40:1208–13.
6. Doern GV, Brueggemann AB, Huynh H, Wingert E, Rhomberg P.
Antimicrobial resistance with Streptococcus pneumoniae in the United
States, 1997–98. Emerg Infect Dis 1999; 5:757–65.
7. NCCLS. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically. Approved standard M7-A6. Wayne, PA:
NCCLS, 2002.
8. NCCLS. Performance standards for antimicrobial susceptibility testing;
fourteenth information supplement. Approved standard M100-S14.
Wayne, PA: NCCLS, 2003.
9. Brueggemann AB, Coffman SL, Rhomberg P, et al. Fluoroquinolone
resistance in Streptococcus pneumoniae in the United States since
1994–1995. Antimicrob Agents Chemother 2002; 46:680–8.
10. Chen DK, McGeer A, De Azavedo JC, Low DE. Decreased susceptibility
of Streptococcus pneumoniae to fluoroquinolones in Canada. The Canadian Bacterial Surveillance Network. N Engl J Med 1999; 341:233–9.
11. Brueggemann AB, Pfaller MA, Doern GV. Use of penicillin MICs to
predict in vitro activity of other b-lactam antimicrobial agents against
Streptococcus pneumoniae. J Clin Microbiol 2001; 39:367–9.
12. Pérez-Trallero E., Garcı́a-Rey C, Martı́n-Sánchez AM, et al. Activities
of six different quinolones against clinical respiratory isolates of Streptococcus pneumonia with reduced susceptibility to ciprofloxacin in
Spain. Antimicrob Agents Chemother 2002; 46:2665–7.
13. Low DE, de Azavedo J, Weiss K, et al. Antimicrobial resistance among
Antimicrobial Resistance among S. pneumoniae in the US • CID 2005:41 (15 July) • 147
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
clinical isolates of Streptococcus pneumonia in Canada during 2000.
Antimicrob Agents Chemother 2002; 46:1295–301.
Zhanel GG, Palatnick L, Nichol KA, Bellyou T, Low DE, Hoban DJ.
Antimicrobial resistance to respiratory tract Streptococcus pneumoniae
isolates: results of the Canadian Respiratory Organism Susceptibility
Study, 1997 to 2002. Antimicrob Agents Chemother 2003; 47:1867–74.
Davies TA, Kelly LM, Pankuch GA, Credito KL, Jacobs MR, Appelbaum
PC. Antipneumococcal activities of gemifloxacin compared to those of
nine other agents. Antimicrob Agents Chemother 2000; 44:304–10.
Lim S, Bast D, McGeer A, de Azavdeo J, Low DE. Antimicrobial susceptibility breakpoints and first-step parC mutations in Streptococcus
pneumoniae: redefining fluoroquinolone resistance. Emerg Infect Dis
2003; 9:833–7.
Millichap JJ, Pestova E, Siddiqui F, Noskin GA, Peterson LR. Fluoroquinolone resistance is a poor surrogate marker for type II topoisomerase mutations in clinical isolates of Streptococcus pneumoniae. J
Clin Microbiol 2001; 39:2719–21.
Richardson DC, Bast D, McGeer A, Low DE. Evaluation of susceptibility testing to detect fluoroquinolone resistance mechanisms in Streptococcus pneumoniae. Antimicrob Agents Chemother 2001; 45:1911–4.
Low DE. Quinolone resistance among pneumococci: therapeutic and
diagnostic implications. Clin Infect Dis 2004; 38(Suppl 4):S357–62.
Morosini M-I, Loza E, del Campo R, Almaraz F, Baquero F, Cantón
R. Fluoroquinolone-resistant Streptococcus pneumoniae in Spain: activities of garenoxacin against clinical isolates including strains with
altered topoisomerases. Antimicrob Agents Chemother 2003; 47:
2692–5.
Richter SS, Heilmann KP, Beekmann SE, Miller NJ, Rice CL, Doern
GV. The molecular epidemiology of Streptococcus pneumoniae with
quinolone resistance mutations. Clin Infect Dis 2005; 40:225–35.
Canton R, Morosini M, Enright MC, Morrissey I. Worldwide incidence,
molecular epidemiology and mutations implicated in fluoroquinoloneresistant Streptococcus pneumoniae: data from the global PROTEKT
surveillance programme. J Antimicrob Agents Chemother 2003; 52:
944–52.
Peltz MWR, McGee L, Jorgensen JH, et al. Levofloxacin-resistant invasive Streptococcus pneumoniae in the United States: evidence for
clonal spread and impact of conjugate pneumococcal vaccine. The
Active Bacterial Surveillance Team. Antimicrob Chemother 2004; 48:
3491–7.
Jones ME, Sahm DF, Martin N, et al. Prevalence of gyrA, gyrB, parC
and parE mutations in clinical isolates of Streptococcus pneumoniae
with decreased susceptibilities to different fluoroquinolones and originating from worldwide surveillance studies during the 1997–1998 respiratory season. Antimicrob Agents Chemother 2002; 44:462–6.
McGee L, Goldsmith CE Klugman KP. Fluoroquinolone resistance
among clinical isolates of Streptococcus pneumoniae belonging to in-
148 • CID 2005:41 (15 July) • Doern et al.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
ternational multiresistant clones. J Antimicrob Chemother 2002; 49:
173–6.
Klugman KP. The role of clonality in the global spread of fluoroquinolone resistant bacteria. Clin Infect Dis 2003; 36:783–5.
Finkelstein JA, Stille C, Nordin J, et al. Reduction in antibiotic use
among US children, 1996–2000. Pediatrics 2003; 112:620–7.
Klugman KP. Efficacy of pneumococcal conjugate vaccines and their
effect on carriage and antimicrobial resistance. Lancet Infect Dis
2001; 1:85–91.
Dagan R, Sikuler-Cohen M, Zamir O, Janco J, Givon-Lavi N, Fraser
D. Effect of a conjugate pneumococcal vaccine on the occurrence of
respiratory infections and antibiotic use in day-care center attendees.
Pediatr Infect Dis J 2001; 20:951–8.
Klugman KP. Vaccination a novel approach to reduce antibiotic resistance. Clin Infect Dis 2004; 39:649–51.
Moore MR, Hyde TB, Hennessy TW, et al. Impact of a conjugate
vaccine on community-wide carriage of non-susceptible Streptococcus
pneumoniae in Alaska. J Infect Dis 2004; 190:2031–8.
Whitney CG, Farley MM, Hadler J, et al. Decline in invasive pneumococcal disease after introduction of protein-polysaccharide conjugate vaccine. N Engl J Med 2003; 348:1737–46.
Talbot TR, Poehling KA, Hartert TV, et al. Reduction in high rates of
antibiotic-nonsusceptible invasive pneumococcal disease in Tennessee
after introduction of the pneumococcal conjugate vaccine. Clin Infect
Dis 2004; 39:641–8.
Centers for Disease Control and Prevention. Update: influenza activity—US and worldwide, 2001–02 season, and composition of the
2002–03 influenza vaccine. MMWR Morb Mortal Wkly Rep 2002; 51:
503–6.
Centers for Disease Control and Prevention. Influenza vaccination coverage among adults aged ⭓50 years and pneumococcal vaccination
coverage among adults aged ⭓65 years—US, 2002. MMWR Morb
Mortal Wkly Rep 2003; 52:987–92.
Urban C, Rahman N, Zhao X, et al. Fluoroquinolone-resistant Streptococcus pneumoniae associated with levofloxacin therapy. J Infect Dis
2001; 184:794–8.
Anderson KB, Tan JS, File TM, DiPersio JR, Willey BM, Low DE.
Emergence of levofloxacin-resistant pneumococci in immunocompromised adults after therapy for community-acquired pneumonia.
Clin Infect Dis 2003; 37:376–81.
Ho PL, Tse WS, Tsang WT, Kwok TK, Cheng VCC, Chan RMT. Risk
factors for acquisition of levofloxacin-resistant Streptococcus pneumoniae: a case-control study. Clin Infect Dis 2001; 32:701–7.
Davidson R, Cavalcanti R, Brunton JL, et al. Resistance to levofloxacin
and failure of treatment of pneumococcal pneumonia. N Engl J Med
2002; 346:747–50.