American Journal of Emergency Medicine (2012) 30, 68-74

Original Contribution

High rates of quinolone resistance among urinary tract infections in the ED^?

Thana Khawcharoenporn MD ^a,?, Shawn Vasoo MD ^b, Edward Ward MD^c, Kamaljit Singh MD^a

^aSection of Infectious Diseases, Rush University, Chicago, IL 60612, USA ^bDepartment of Medicine, Rush University, Chicago, IL 60612, USA ^cDepartment of Emergency Medicine, Rush University, Chicago, IL 60612, USA

Received 5 August 2010; revised 25 September 2010; accepted 26 September 2010

Abstract

Objectives: The objectives of this study are to examine antibiotic resistance rates and to determine appropriate empiric oral antibiotic for patients with Urinary tract infections evaluated and discharged from the ED.

Methods: A retrospective, single-institution chart review study from August 2008 to March 2009 was conducted. Adult patients seen in the ED with UTI were identified for study inclusion from review of microbiology records. Hospitalized or asymptomatic bacteriuria cases were excluded. Health care- associated (HA)-UTI was defined as UTI with indwelling urinary catheters, health care exposure, or urologic procedures within 3 months. Prevalence of causative bacteria, antibiotic resistance rates, and risk factors for quinolone resistance were determined.

Results: There were 337 eligible patients with 83% women. The most common uropathogens among 357 bacterial isolates were Escherichia coli (71%) and Klebsiella spp. (9%). Overall levofloxacin resistance rate was 17%. Resistance rates for HA-UTIs were significantly greater than those for community-associated-UTI: levofloxacin, 38% vs 10%; trimethoprim-sulfamethoxazole, 26% vs 17%;

amoxicillin, 53% vs 45%; and amoxicillin-clavulanate, 16% vs 6%. Nitrofurantoin resistance rates were similar (9%). Independent risk factors for levofloxacin resistance were long-term medical conditions (adjusted odds ratio [aOR], 4.23; P = .001), HA-UTI (aOR, 2.56; P = .006), and prior quinolone use within 1 week (aOR, 14.90; P = .02) and within 1 to 4 weeks (aOR, 4.62; P = .04).

Conclusions: We report high rates of quinolone resistance in ED patients with UTIs at our institution. For patients with risk factors for quinolone resistance, Empiric therapy with cephalosporins or nitrofurantoin may be preferred. Urine culture and susceptibility testing should be performed to guide definitive therapy for HA-UTIs.

(C) 2012

Introduction

^? Conflict of interest: Dr K. Singh is on the Visiting Speakers Board for Wyeth and is a Scientific Advisory Board Member of Quidel Corporations.

* Corresponding author. Section of Infectious Diseases, Department of

Medicine, Rush University Medical Center, Chicago, IL 60612, USA. Tel.:

+1 312 942 5865; fax: +1 312 942 8200.

E-mail address: [email protected] (T. Khawcharoenporn).

urinary tract infection is a frequently encountered clinical condition in the ED. Most patients have community- associated (CA) lower urinary tract disease and can be safely discharged from the ED with oral antibiotics. However, there are some significant differences in UTI management in the

0735-6757/$ - see front matter (C) 2012 doi:10.1016/j.ajem.2010.09.030

ED compared with the Primary care setting [1]. Patients presenting to the ED with a complaint of UTI often have recurrent infections or failed empiric outpatient treatment, underlying comorbidities, or health care-associated (HA) infections [1]. As such, urine cultures are more commonly performed in the ED, particularly in academic center EDs, as compared with the primary care setting. In addition, empiric antibiotic selection for ED patients with UTI is often guided by hospital antibiograms, which have been shown to overestimate the frequency of drug resistance in the community [1,2]. Despite the frequency of UTI diagnosis in the ED, there are few data on appropriate management of UTIs in patients evaluated and discharged from the ED.

Previous UTI studies mostly reported on antibiotic resistance and treatment of Escherichia coli, the predominant cause of UTIs [2-7]. Increasing rates of E coli resistance to trimethoprim-sulfamethoxazole (TMP-SMX) have been reported in both ED and outpatient settings in the United States, whereas the rates of quinolone resistance appear low [1-3,7]. As a result, quinolones have become the preferred empiric therapy in many EDs, especially in areas where the prevalence of TMP-SMX resistance is 20% or more [1,2]. However, quinolone resistance has been increasingly reported, although these resistance rates vary dramatically by Geographic regions [3-7]. In addition, accurate bacterial epidemiology and prevalence of drug resistance are difficult to determine in these studies, given that urine culture was not performed in all patients. We conducted this study in response to our ED’s concern about increasingly difficult-to- treat, multidrug-resistant UTIs. Our ED physicians also cited hospital antibiogram data showing high resistance rates for E coli isolates (34% levofloxacin-resistant). Our aim was to determine the prevalence of causative bacteria, rates of antibiotic resistance, risk factors for quinolone resistance, and appropriate empiric oral antibiotics for patients with UTIs discharged from the ED.

Methods

Study design, setting, and population

All patients 12 years or older [8] who were evaluated at the Rush University Medical Center (RUMC) ED from August 1, 2008, to March 31, 2009, and had positive urine cultures were eligible for the study. Rush University Medical Center is a 613-bed tertiary academic center in Chicago, with more than

48 000 ED visits annually. Patients were retrospectively identified by daily review of urine specimens sent to the microbiology laboratory during the period from August 1, 2008, to March 31, 2009. In RUMC ED, urine cultures are performed in every patient suspected of UTI or in those with positive Urine dipstick for Leukocyte esterase and/or nitrite. Patients who had urine cultures with significant bacterial growth of 10⁴ colony-forming units per milliliter or more [9]

were included for chart review. The study was approved by the institutional review board of RUMC.

Study protocol and definitions

Chart review was performed using our electronic medical record system for ethnicity, demographics, pregnancy status, underlying medical comorbidities, relevant surgical history including urologic procedure and Obstructive uropathy, UTI history, prior antibiotic use, UTI diagnoses, treatments, causative bacteria, and antibiotic susceptibility. Health-care- associated UTI was defined as UTI in patients with long-term indwelling urinary catheters, health care exposure including hospital stay for at least 48 hours, nursing home or long-term care facility residence, regular hemodialysis clinic visits, or urologic procedures within the past 3 months [10,11]. Community-associated-UTI was defined as UTI that did not fulfill the criteria for HA-UTI. Prior use of antibiotics was categorized as use within 1 week, 1 to 4 weeks, and 1 to 3 months. The diagnosis of UTI (asymptomatic bacteriuria, cystitis, pyelonephritis, and urosepsis) was made by the treating physician, and no diagnostic parameters were established by the study investigators. We only included patients who were evaluated and discharged from the ED with the diagnosis of cystitis, pyelonephritis, and urosepsis for further analysis. Prevalence of causative bacteria, antibiotic resistance rates, and risk factors of levofloxacin resistance were determined. Identification and antibiotic susceptibilities of causative bacteria were determined using Microscan Walkaway (Siemens, Mountain View, CA). The minimum inhibitory concentration breakpoints for resistance were based on Clinical Laboratory Standards Institute interpretive criteria: levofloxacin, 8 ug/mL or more; TMP-SMX, 4/76 ug/ mL or more; nitrofurantoin, 128 ug/mL or more; gentamicin,

16 ug/mL or more; ampicillin, 32 ug/mL or more; amoxicillin-clavulanate, 32/16 ug/mL or more; cefazolin,

32 ug/mL or more; cefuroxime, 32 ug/mL or more; ceftriaxone, 64 ug/mL or more; and ertapenem, 8 ug/mL or more [12].

Statistical analysis

All analyses were performed using SPSS, version 15.0 (SPSS, Chicago, IL). The sample size required for this study to have a significant level of .05 and a power of 0.80 to detect the association between each variable and levoflox- acin resistance was 80, based on prior study data [4]. Categorical variables were compared using the Pearson ?2 or Fisher exact test, as appropriate. All P values were 2- tailed; P b .05 was considered statistically significant. Risk factors for levofloxacin resistance were identified by comparison of variables between the patients infected with levofloxacin-resistant vs susceptible bacteria. Variables that were present in more than 10% of patients at a significance level of P b .20 or that had a prior clinical significance (eg, age, diabetes mellitus, recent antibiotic use, prior UTI, and

Characteristics	All	patients (N = 337)	CA-UTI (n = 248)	HA-UTI (n = 89)	P a
Median age (IQR) (y)	38	(26-60)	34 (24-52)	60 (42-80)	b.001
Female	279	(83)	214 (86)	65 (73)	.004
Ethnicity					.02
African American	192	(57)	147 (59)	45 (51)	-
White	67	(20)	39 (16)	28 (32)	-
Hispanic	65	(19)	50 (20)	15 (17)	-
Asian	5	(2)	4 (2)	1 (1)	-
Other	8	(2)	8 (3)	0 (0)	-
Pregnancy	20	(6)	20 (8)	0 (0)	.003
Type of UTI					.29
Lower tract disease	263	(78)	190 (77)	73 (82)	-
Upper tract disease	74	(22)	58 (23)	16 (18)	-
Comorbidities
Diabetes mellitus	57	(17)	33 (13)	24 (27)	.003
Renal transplant	10	(3)	7 (3)	3 (3)	.79
Long-term medical conditions b	192	(57)	115 (46)	77 (87)	b.001
Obstructive uropathy	36	(11)	18 (7)	18 (20)	.001
Prior UTI	114	(34)	66 (27)	48 (54)	b.001

Prior antibiotic use within 3 mo	89 (26)	36 (15)	53 (60)	b.001
Quinolones	40 (12)	10 (4)	30 (34)
Cephalosporins	25 (7)	8 (3)	17 (19)
Penicillins	21 (6)	7 (3)	14 (16)
TMP-SMX	14 (4)	5 (2)	9 (10)
Nitrofurantoin	13 (4)	6 (2)	7 (8)

ethnicity) [2,4-6] were entered into logistic regression models. Significant variables that were thought to be covariates were grouped, and only 1 variable from each group was chosen for model entry. The model’s overall robustness was confirmed by Hosmer-Lemeshow goodness- of-fit statistic. Adjusted odd ratios (aORs) and 95%

Table 1 Characteristics of the 337 study patients stratified by type of UTI

Data are in numbers (%), unless otherwise indicated. IQR indicates interquartile range.

^a Compared between CA-UTI and HA-UTI.

^b Any medical conditions that require regular follow-up and treatment, and that had been present for at least 6 months before positive urine culture.

Table 2 Distribution of 357 unique bacteria isolates from all study patients stratified by type of acquisition

confidence intervals (CIs) were calculated to determine independent risk factors of levofloxacin resistance.

Results

A total of 676 patients with positive urine cultures were identified; 156 patients who were hospitalized and 183 who

Table 3 Antibiotic resistance rates among 357 unique bacterial isolates from all study patients

Bacteria	CA-UTI (n = 260)	HA-UTI (n = 97)	P
Gram-negative E coli Klebsiella spp. Proteus spp. Citrobacter spp. Enterobacter spp. Serratia spp. P aeruginosa Gram-positive Enterococcus spp. Staphylococcus aureus Staphylococcus saprophyticus	245 (94) 199 (77) 23 (9) 14 (5) 6 (2) 2 (1) 1 (0.4) 0 (0) 15 (6) 7 (3) 6 (2) 2 (1)	85 (88) 54 (56) 9 (9) 4 (4) 5 (5) 3 (3) 0 (0) 10 (10) 12 (12) 10 (10) 2 (2) 0 (0)	.04 b.001 .90 .79 .17 .13 1.00 b.001 .04 .003 1.00 1.00
Data are in numbers (%).

Antibiotics	All (N = 357)	CA-UTI (n = 260)	HA-UTI (n = 97)	P a
Levofloxacin	17%	10%	38%	b.001
TMP-SMX	20%	17%	26%	.01
Nitrofurantoin	9%	9%	9%	1.00
Gentamicin	8%	7%	11%	.17
Ampicillin	47%	45%	53%	.03
Amoxicillin-clavulanate	8%	6%	16%	b.001
Cefazolin	10%	7%	18%	b.001
Cefuroxime	5%	2%	11%	b.001
Ceftriaxone	5%	2%	11%	b.001
Ertapenem	0.3%	0%	1%	.23
a Compared between CA-UTI and HA-UTI.

Antibiotics	All (N =	253)	CA-UTI (n = 199)	HA-UTI (n = 54)	P a
Levofloxacin	16%	10%		39%	b.001
TMP-SMX	25%	21%		39%	.007
Nitrofurantoin	4%	4%		4%	1.00
Gentamicin	10%	9%		15%	.22
Ampicillin	45%	41%		61%	.008
Amoxicillin-clavulanate	7%	5%		15%	.007
Cefazolin	9%	7%		19%	.007
Cefuroxime	4%	2%		13%	.002
Ceftriaxone	3%	2%		7%	.04
Ertapenem	0%	0%		0%	1.00
a Compared between CA-UTI and HA-UTI.

had asymptomatic bacteriuria were excluded. The study population consisted of 337 patients with symptomatic UTIs who were discharged from the ED with oral antibiotic therapy. Demographics, comorbidities, UTI, and antibiotic history are summarized in Table 1. The median age was 38 years (range, 14-96 years), and most patients were women (83%) and African American (57%). Long-term medical conditions were present in 192 patients (57%); the most common conditions reported were hypertension (34%), dyslipidemia (17%), and diabetes mellitus (16%). Thirty-six patients (11%) had obstructive uropathy, including urinary tract calculous diseases, cervical cancer, bladder cancer, vaginal cancer, and ureteropelvic junction obstruction. Prior antibiotic use within the last 3 months was reported by 89 (26%) of 337 patients, and quinolones were the most commonly used antibiotic (12%). Lower UTIs were diag- nosed in 263 patients (78%).

Table 4 Antibiotic resistance rates among 253 unique E coli

isolates

Table 5 Risk factors associated with levofloxacin resistance among all 337 study patients

Two hundred forty-eight (74%) of 337 patients had CA- UTIs (Table 1). Patients with HA-UTI were older and more likely to be male, white, have long-term medical conditions, obstructive uropathy, prior UTIs, and antibiotic exposure within 3 months (P b .05). There were 357 nonduplicate bacterial isolates recovered from 337 patients. The types and distribution of bacteria in patients with CA-UTI and HA-UTI are summarized in Table 2. Most uropathogens were gram- negative bacteria (330 [92%]). E coli (253 [71%]), Kleb-

siella spp. (31 [9%]), and Proteus spp. (18 [5%]) were the 3 most common uropathogens. E coli was significantly more common in patients with CA-UTIs than those with HA-UTIs (77% vs 56%; P b .001), whereas Enterococcus spp. and Pseudomonas aeruginosa were more common in patients with HA-UTI (P b .05).

Rates of antibiotic resistance for all uropathogens and E coli are summarized in Tables 3 and 4, respectively. The overall resistance rates for TMP-SMX and levofloxacin were 20% and 17%, respectively. Antibiotic resistance rates for HA-UTIs were significantly greater than CA-UTIs for most tested antibiotics including levofloxacin, 38% vs 10% (Table 3). Among a subset of patients with E coli UTI (253 isolates), antibiotic resistance rates were significantly greater in patients with HA-UTI including levofloxacin, 39% vs 10% (Table 4).

We analyzed the risk factors associated with levofloxacin resistance by comparing patients with levofloxacin-resistant bacteria (60 [17%]) vs susceptible bacterial isolates (297 [83%]) (Table 5). By univariate analysis, 50 years or older, male gender, diabetes mellitus, long-term medical comor- bidities, prior UTI, HA-UTI, obstructive uropathy, and prior use of any antibiotics or quinolones within 4 weeks were associated with levofloxacin resistance. In the multivariate logistic regression analysis, independent risk factors for

Risk factors	Univariate analysis			Multivariate analysis
	OR (95% CI)	P		aOR (95% CI)	P

Age, >=50 y	3.10 (1.76-5.47)	b.001	-	-
Male	3.49 (1.88-6.48)	b.001	-	-
Diabetes mellitus	2.09 (1.10-3.97)	.02	-	-
Long-term medical conditions a	7.34 (3.24-16.65)	b.001	4.23 (1.77-10.11)	.001
Prior UTI	3.14 (1.79-5.50)	b.001	-	-
HA infection b	4.84 (2.71-8.62)	b.001	2.56 (1.31-5.02)	.006
Upper tract infection	1.57 (0.76-3.24)	.23	-	-
Prior any antibiotic use within:
N1-4 wk	5.92 (3.05-11.48)	b.001	1.63 (0.61-4.34)	.33
1 wk	8.56 (2.70-27.17)	b.001	3.27 (0.89-11.98)	.07
Prior quinolone use within:
N1-4 wk	13.87 (4.69-41.08)	b.001	4.62 (1.09-19.61)	.04
1 wk	29.85 (2.69-327.17)	.01	14.90 (2.09-105.44)	.02
a Any medical conditions that require regular follow-up and treatment and that had been present for at least 6 months before positive urine culture. b Defined as an infection in the patients with chronic indwelling urinary catheters, having health care exposure including short- and long-term care hospital stay for more than 48 hours, being nursing home residents, having regular hemodialysis clinic visit, or undergoing urologic procedures within the past 3 months.

Risk factors	Univariate analysis			Multivariate analysis
	OR (95% CI)	P		aOR (95% CI)	P

Age, >=50 y	2.04 (0.88-4.73)	.09	1.39 (0.62-3.75)	.52
Male	1.77 (0.62-5.07)	.28	-	-
Diabetes mellitus	2.07 (0.77-5.58)	.14	-	-
Long-term medical conditions a	5.09 (1.85-14.02)	.001	4.57 (1.36-15.38)	.01
Prior UTI	1.37 (0.56-3.32)	.49	-	-
Upper tract infection	1.23 (0.44-3.43)	.69	-	-
Prior any antibiotic use within:
N1-4 wk	3.86 (1.13-13.19)	.02	0.85 (0.07-9.92)	.89
1 wk	31.77 (3.17-318.50)	b.001	22.83 (1.64-317.51)	.02
Prior quinolone use within:
N1-4 wk	20.26 (1.77-232.08)	.001	54.46 (1.57-894.17)	.03
1 wk	28.71 (2.28-355.65)	.01	18.80 (2.36-147.91)	.03
a Any medical conditions that require regular follow-up and treatment and that had been present for at least 6 months before positive urine culture.

levofloxacin resistance were long-term medical conditions (aOR, 4.23; 95% CI, 1.77-10.11; P = .001), HA-UTI (aOR,

Table 6 Risk factors associated with levofloxacin resistance in 248 patients with CA-UTIs

2.56; 95% CI, 1.31-5.02; P = .006), and prior fluoroquino-

lone use within 1 week (aOR, 14.90; 95% CI, 2.09-105.44;

P = .02) and within 1 to 4 weeks (aOR, 4.62; 95% CI, 1.09- 19.61; P = .04). The classification of UTI, upper vs lower, was not significantly associated with levofloxacin resistance. Subset analysis of patients with CA-UTI identified the same independent predictors for levofloxacin resistance (Table 6). All study patients were prescribed oral antibiotics: levofloxacin (145 [43%]), nitrofurantoin (67 [20%]),

TMP-SMX (27 [8%]), and third-generation cephalosporins (23 [7%]). Based on in vitro antibiotic susceptibility testing,

44 patients (13%) received Empiric treatment with an antibiotic that tested nonsusceptible. Patients with HA-UTIs were significantly more likely to receive a nonsusceptible

antibiotic compared with those with CA-UTIs (22% vs 10%; P = .005).

Discussion

Our study highlights the importance of monitoring local, up-to-date antibiotic resistance rates and determining risk factors for antibiotic-resistant UTI. Given that there are data supporting an association between in vitro drug resistance and clinical failure in patients with UTIs [13,14], appropri- ate empiric antibiotic selection is crucial. This is especially true in the ED where Empiric antibiotics are prescribed before urine culture results are available and patient follow- up and monitoring of response are more difficult. In the ED, UTIs are frequently classified as complicated (lower or

Fig. 1 Proposed algorithm of empiric antibiotic therapy in patients with UTIs discharged from the ED. ^aDefined as UTIs in the patients with chronic indwelling urinary catheters, having health care exposure including short- and long-term care hospital stay for more than 48 hours, being nursing home residents, having regular hemodialysis clinic visit, or undergoing urologic procedures within the past 3 months. ^bAny medical conditions that require regular follow-up and treatment and that had been present for at least 6 months before positive urine culture.

upper tract infections) and uncomplicated infections in an attempt to improve antibiotic selection [1]. In addition, results of our study suggest that empiric antibiotic selection should also take into consideration whether the infection is a CA-UTI vs HA-UTI, given the predictable differences in causative pathogens and antibiotic resistance between these 2 populations (Fig. 1).

Many studies on antibiotic resistance rates for UTIs only examine the susceptibility patterns for E coli, which is the dominant pathogen. However, the rates of UTIs due to E coli range from 58% to 89%, depending on the proportion of patients with HA-UTIs included in each study [2,3]. We found that E coli was the causative pathogen in 77% of CA- UTIs and only 56% of HA-UTIs seen in the ED. This suggests that use of an antibiogram for E coli alone may be inaccurate for guiding empiric antibiotic selection. We found high rates of TMP-SMX resistance, 20% among patients with UTIs, which is consistent with results from previous studies [2,3,15-17]. More worrisome was the high rate of levoflox- acin resistance (17%) in patients evaluated and discharged from the ED. The rate of levofloxacin resistance was significantly higher in patients with HA-UTIs, 38% vs 10% for CA-UTI. In the multivariate analysis of risk factors for levofloxacin resistance, we found that HA-UTI, long-term underlying medical conditions, and prior fluoroquinolone use within 4 weeks were independent risk factors for levofloxacin resistance. Although other characteristics, including age more than 50 years, the presence of complicated UTIs, and recurrent UTIs, have been reported as risk factors for levofloxacin resistance [4,5,7,17-19]; they did not reach statistical significance in the multivariate analysis.

Although high rates of quinolone resistance have been reported in studies from Europe [18,20], our study shows a significant increase in levofloxacin resistance among urinary isolates (including E coli) in a US setting. These findings are in contrast with a recent multicenter study of pyelonephritis in EDs in the United States, which reported overall quinolone resistance rates less than 3% [2]. Factors that may contribute to this disparity include the difference in the study period (2008-2009 vs 2000-2004), study inclusion criteria, and definitions of CA-UTI. Our study results are not unexpected, given recent reports of declining antibiotic susceptibility of Enterobacteriaceae in the United States and worldwide [4,21]. In addition, recent guidelines from the Infectious Diseases Society of America caution that quinolone-resistant E coli have become common in some communities, and as such, quinolones should not be used empirically unless hospital surveys indicate greater than 90% susceptibility [22]. The strength of our study included our ability to evaluate overall susceptibility data of bacteria causing UTIs in the ED setting including separation of patients into HA-UTI and CA-UTI. Results of our study support the cautious use of quinolones as empiric therapy for CA-UTIs only in communities with documented low rates of resistance and in patients with no risk factors for such resistance. Oral antibiotic alternatives for patients with CA-UTI who have

risk factors for quinolone resistance include the oral cephalosporins or nitrofurantoin. Of note, many microbiol- ogy laboratories do not offer antibiotic susceptibility testing of oral cephalosporins [12], and although nitrofurantoin demonstrates low resistance against E coli, it has poor activity against non-E coli uropathogens and is not recommended for upper UTIs or in patients with moderate- to-severe Renal impairment [1,3,4]. Our results also indicate that quinolones are no longer appropriate empiric agents for patients with HA-UTI in institutions with high quinolone resistance rates. Empiric oral antibiotic therapy for HA-UTI in such settings is even more problematic because many of these infections are due to non-E coli bacteria and are frequently multidrug resistant. We propose an algorithm for empiric antibiotic therapy in the ED based on our study findings (Fig. 1). At our institution, we recommend that patients with HA-UTIs should be hospitalized and treated with parenteral broad-spectrum antibiotics. Subsequent antibiotic regimen should be guided by the results of urine culture and susceptibilities. Nitrofurantoin and oral third- generation cephalosporin may be empiric alternatives in patients with lower or uncomplicated HA-UTIs.

Limitations

Some limitations of our study include the fact that data were obtained from chart review, which relies on the completeness of history, physical examination, and labora- tory data recorded by the ED physician. We attempted to minimize information and misclassification biases inherent in chart review by using a standardized data-gathering form. Second, our findings were from a single medical center and may not be applicable to other settings with different local susceptibility patterns and patient populations. Lastly, clinical and microbiological outcomes with regard to antibiotic resistance and appropriate empiric antibiotic therapy were not evaluated. However, regularly updated surveillance of local microbial prevalence and resistance patterns are recommended for use to guide the empiric therapy.

Conclusions

We report high rates of quinolone resistance among UTIs in our ED setting. Quinolones can only be recommended for empiric therapy for CA-UTIs in patients without risk factors for quinolone resistance. For patients with long-term medical conditions, HA-UTIs, and prior quinolone use within 4 weeks, empiric oral cephalosporins, or nitrofurantoin may be considered. Patients with HA-UTIs should be differentiated from those with CA-UTIs, and their definitive therapy should be guided by results of urine culture and susceptibilities. Antibiograms that are derived from locally conducted studies in the ED are desirable to better determine optimal empiric antibiotic selection.

References

1. Abrahamian FM, Moran GJ, Talan DA. Urinary tract infections in the emergency department. Infect Dis Clin North Am 2008;22:73-87.
2. Talan DA, Krishnadasan A, Abrahamian FM, et al. Prevalence and risk factor analysis of trimethoprim-sulfamethoxazole- and fluoroquino- lone-resistant Escherichia coli infection among emergency department patients with pyelonephritis. Clin Infect Dis 2008;47:1150-8.
3. Zhanel GG, Hisanaga TL, Laing NM, et al. Antibiotic resistance in outpatient urinary isolates: final results from the North American Urinary Tract Infection Collaborative Alliance (NAUTICA). Int J Antimicrob Agents 2005;26:380-8.
4. Johnson L, Sabel A, Burman WJ, et al. Emergence of fluoroquinolone resistance in outpatient urinary Escherichia coli isolates. Am J Med 2008;121:876-84.
5. Colodner R, Kometiani I, Chazan B, et al. Risk factors for community- acquired urinary tract infection due to quinolone-resistant E. coli. Infection 2008;36:41-5.
6. Karlowsky JA, Hoban DJ, Decorby MR, et al. Fluoroquinolone- resistant urinary isolates of Escherichia coli from outpatients are frequently multidrug resistant: results from the North American Urinary Tract Infection Collaborative Alliance-Quinolone Resistance study. Antimicrob Agents Chemother 2006;50:2251-4.
7. Olson RP, Harrell LJ, Kaye KS. Antibiotic resistance in urinary isolates of Escherichia coli from college women with urinary tract infections. Antimicrob Agents Chemother 2009;53:1285-6.
8. Warren JW, Abrutyn E, Hebel JR, et al. Guidelines for Antimicrobial treatment of uncomplicated acute bacterial cystitis and acute pyelonephritis in women. Infectious Diseases Society of America (IDSA). Clin Infect Dis 1999;29:745-58.
9. Rubin RH, Beam Jr TR, Stamm WE. An approach to evaluating antibacterial agents in the treatment of urinary tract infection. Clin Infect Dis 1992;14:S246-251.
10. Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control 2008;36:309-32.
11. American Thoracic Society and Infectious Diseases Society of America. Guidelines for the management of adults with hospital- acquired, ventilator-associated, and healthcare-associated pneumonia. Am J Respir Crit Care Med 2005;171:388-416.
12. Clinical Laboratory Standards Institue. Performance standards for antimicrobial susceptibility testing: 17th informational supplement.

CLSI document M100-S17. Wayne, PA: Clinical and Laboratory Standards Institute; 2007.

1. Raz R, Chazan B, Kennes Y, et al. Empiric use of trimethoprim- sulfamethoxazole (TMP-SMX) in the treatment of women with Uncomplicated urinary tract infections, in a Geographical area with a high prevalence of TMP-SMX-resistant uropathogens. Clin Infect Dis 2002;34:1165-9.
2. Talan DA, Stamm WE, Hooton TM, et al. Comparison of ciprofloxacin (7 days) and trimethoprim-sulfamethoxazole (14 days) for acute uncomplicated pyelonephritis in women: a randomized trial. JAMA 2000;283:1583-90.
3. Kahlmeter G. ECO.SENS: a n international survey of the antimicrobial susceptibility of pathogens from uncomplicated urinary tract infec- tions: the ECO.SENS Project. J Antimicrob Chemother 2003;51: 69-76.
4. Colgan R, Johnson JR, Kuskowski M, et al. Risk factors for trimethoprim-sulfamethoxazole resistance in patients with acute uncomplicated cystitis. Antimicrob Agents Chemother 2008;52: 846-51.
5. Smith SP, Manges AR, Riley LW. temporal changes in the prevalence of community-acquired antimicrobial-resistant urinary tract infection affected by Escherichia coli clonal group composition. Clin Infect Dis 2008;46:689-95.
6. Arslan H, Azap OK, Ergonul O, et al. Urinary Tract Infection Study Group. Risk factors for ciprofloxacin resistance among Escherichia coli strains isolated from community-acquired urinary tract infections in Turkey. J Antimicrob Chemother 2005;56:914-8.
7. Killgore KM, March KL, Guglielmo BJ. Risk factors for community- acquired ciprofloxacin-resistant Escherichia coli urinary tract infec- tion. Ann Pharmacother 2004;38:1148-52.
8. Cagnacci S, Gualco L, Debbia E, et al. European emergence of ciprofloxacin-resistant Escherichia coli clonal groups O25:H4-ST 131 and O15:K52:H1 causing community-acquired uncomplicated cystitis. J Clin Microbiol 2008;46:2605-12.
9. Hoban DJ, Bouchillon SK, Hawser SP, et al. Susceptibility of gram- negative pathogens isolated from patients with complicated intra- Abdominal infections in the United States, 2007-2008: results of the Study for Monitoring Antimicrobial resistance Trends (SMART). Antimicrob Agents Chemother 2010;54:3031-4.
10. Solomkin JS, Mazuski JE, Bradley JS, et al. Diagnosis and management of complicated intra-abdominal infection in adults and children: guidelines by the Surgical Infection Society and the Infectious Diseases Society of America. Clin Infect Dis 2010;50: 133-64.