Antimicrobial Susceptibility Trends Observed in Urinary Pathogens Obtained From New York State

Abstract International guidelines recommend using local susceptibility data to direct empiric therapy for acute uncomplicated cystitis. We evaluated outpatient urinary isolate susceptibility trends in New York State. Nitrofurantoin had the lowest resistance prevalence whereas trimethoprim-sulfamethoxazole and fluoroquinolones had higher prevalences. This study highlights the need for local outpatient antimicrobial stewardship programs.

International guidelines recommend using local susceptibility data to direct empiric therapy for acute uncomplicated cystitis. We evaluated outpatient urinary isolate susceptibility trends in New York State. Nitrofurantoin had the lowest resistance prevalence whereas trimethoprim-sulfamethoxazole and fluoroquinolones had higher prevalences. This study highlights the need for local outpatient antimicrobial stewardship programs.
Acute uncomplicated cystitis is a prevalent outpatient condition [1,2]. It is estimated that there are approximately 6-8 million annual visits to a physician or clinic by patients for treatment of an acute uncomplicated cystitis event [3,4]. Treatment guidelines for acute uncomplicated cystitis in premenopausal, nonpregnant women are well defined in international guidelines [2]. In these international guidelines, empiric treatment recommendations are provided, but these empiric recommendations are accompanied by the caveat that empiric antibiotic selection should be guided by local organism susceptibility data and patient-specific risk factors.
Although treatment selection for acute uncomplicated cystitis was straightforward in the past, management of these conditions has been complicated by reports of rising antibiotic resistance rates for several key urinary pathogens. It is important to note that much of the data that describe rising rates of resistance for uncomplicated cystitis pathogens were derived from hospitalized inpatients with urinary tract infections. Data that describe outpatient resistance rates are consistent with published inpatient reports but are limited [5]. It is also unclear if resistance rates among acute uncomplicated cystitis pathogens are applicable to all age groups, and to both men and women. Given these gaps in the literature, this study sought to describe the prevalence and resistance patterns of urinary pathogens in New York State in the outpatient setting. The intent was to use these data to help inform treatment decisions for patients who present with uncomplicated cystitis in the outpatient setting and to assess the appropriateness of empiric treatment recommendations found in national guidelines for New York State.

METHODS
A retrospective analysis was conducted on all urine cultures received from outpatient settings (defined as physician offices or outpatient clinics), with antimicrobial susceptibility testing performed from January 1, 2016, to December 31, 2016, at a major clinical microbiology reference laboratory (Quest Diagnostics Laboratory, Teterboro, NJ). As antimicrobial testing is not routinely recommended for Streptococcus agalactiae and Staphylococcus saprophyticus, these organisms were not included in the analysis. Data from 17 New York State counties were included in the overall sample (Table 1). Urine cultures demonstrating 1 or 2 bacterial isolates at >10 5 colony-forming units/mL were included (only the first reported isolate from dual infections was included). In accordance with standards for summarizing antimicrobial susceptibilities for antibiograms, each species required a minimum of 30 isolates for inclusion in antimicrobial susceptibility estimates.
Urine pathogens were isolated from bi-plates of Trypticase Soy Agar with 5% Sheep Blood (TSA II) and MacConkey II Agar, followed by identification and automated drug susceptibility testing (Vitek-2). Antimicrobials tested included ampicillin, ceftazidime, cefazolin, ciprofloxacin, nitrofurantoin, piperacillin-tazobactam, trimethoprim-sulfamethoxazole, tetracycline, oxacillin, and vancomycin, as appropriate. Minimum inhibitory concentration (MIC) susceptibility interpretations were derived from CLSI M100 S-25 [6]. Antibiotic sensitivity percentages are reported as number of susceptible isolates divided by number of isolates tested. Susceptibility was reported overall and by sex and age (children: ≤17 years; adults: 18-64 years; and older adults: ≥65 years). Chi-square tests or Fisher exact tests were used to compare the overall prevalence of bacterial resistance 2  within and between agents tested and between age and sex groups.

RESULTS
A total of 78 078 urine culture susceptibility reports were included ( . Of all the isolates tested for nitrofurantoin sensitivity (n = 73 191), 90.4% were susceptible. Nitrofurantoin resistance was more commonly noted in males as compared with females (10.6% vs 9.1%, P < .001) and older adults as compared with all other ages combined (12.3% vs 8.1%, P < .001). High overall rates of susceptibility were reported for isolates tested for cefazolin sensitivity (90.4%). However, due to the high prevalence of Enterococcus spp. (which is intrinsically resistant to cefazolin), cefazolin only has activity against 68.2% of the isolates. Only 77.2% of all isolates tested (n = 67 711) were trimethoprim-sulfamethoxazole susceptible. Resistance to trimethoprim-sulfamethoxazole was more prevalent in men than women (26.3% vs 22.7%, P < .001) and in older adults than all other age groups combined (25.1% vs 22.1%, P < .001). Of the isolates tested for ciprofloxacin sensitivity (n = 68 709), 80.2% were susceptible. Resistance to ciprofloxacin was more frequent in males than females (35.0% vs 17.3%, P < .001) and in older adults than in all other ages combined (30. % vs 14.0%, P < .001). Additional data on pathogen-specific sensitivity overall and by sex and age groups are presented in Table 1. County-specific data are provided in the Supplementary Data.

DISCUSSION
The current Infectious Disease Society of America (IDSA) guidelines for acute uncomplicated cystitis recommend nitrofurantoin, trimethoprim-sulfamethoxazole, or fosfomycin as firstline agents for treatment [2]. Overall, we found that bacterial isolates in the outpatient setting in New York State were 90% sensitive in vitro to nitrofurantoin. The high probability of in vitro activity with nitrofurantoin is likely due to the high prevalence and susceptibility rates for E. coli. Although the IDSA guidelines do not define a resistance prevalence threshold for assessing the appropriateness of nitrofurantoin for empiric use in treating acute uncomplicated cystitis, if we apply the trimethoprim-sulfamethoxazole threshold (20%, per the IDSA guidelines), our data support the utilization of nitrofurantoin in the New York State outpatient setting for this condition. In contrast, trimethoprim-sulfamethoxazole appears to have more limited utility as an empiric treatment regimen for acute uncomplicated cystitis as the overall prevalence of trimethoprim-sulfamethoxazole resistance exceeded 20%. Unfortunately, fosfomycin susceptibility data were unavailable in this data set as testing was not routinely performed. Given the limited activity of trimethoprim-sulfamethoxazole and ciprofloxacin for acute uncomplicated cystitis, it would be prudent for laboratories to consider testing of fosfomycin in urinary isolates, especially for Escherichia coli and Enterococcus faecalis. In species other than Escherichia coli and Enteroccocus faecalis, the utility and validity of susceptibility results for fosfomycin have yet to be determined [6]. Importantly, recent data show that 1 day of fosfomycin is inferior to 5 days of nitrofurantoin for acute uncomplicated cystitis [7]. As such, fosfomycin should be used with caution when treating acute uncomplicated cystitis.
Susceptibility results from this study were also unfavorable to the empiric use of fluoroquinolones and β-lactam antibiotics. The overall prevalence of ciprofloxacin resistance was 19.8%, exceeding the IDSA-recommended 10% resistance threshold for empiric use. In light of the relatively high prevalence of fluoroquinolone resistance as well as growing concerns about fluoroquinolone-associated disability, our study supports the recent Food and Drug Administration recommendation to avoid empiric fluoroquinolone use unless no other alternative agents are available [7]. We also examined cefazolin as a surrogate for cephalexin susceptibility. When only Gram-negative bacterial species were isolated, cefazolin appeared to be an appropriate empiric agent. However, the utility of cefazolin as an empiric agent is less than favorable due to the high prevalence of Enterococcus, an organism that is intrinsically resistant to cefazolin. Given that β-lactam medications require longer duration of treatment as compared with other therapies and are associated with lower efficacy [8], our data suggest that β-lactam drugs should be considered as empiric agents only when potential benefits outweigh risks.
These data also indicate that the overall antimicrobial susceptibility percentages of antimicrobials are significantly lower for individuals in the ≥65 age group compared with all other ages combined and among males of all ages compared with females. These findings highlight the need to not rely exclusively on overall susceptibility results and to consider sex, age, and prior urine culture results when selecting an agent for a given patient. As susceptibility rates varied by age and sex, subsequent adjustment of therapy based on individualized culture and susceptibility reports should be performed in an effort to promote use of narrow-spectrum antibiotics where possible.
There are several caveats to be considered with respect to these findings. We did not include coagulase-negative staphylococci in this analysis due to its questionable pathogenicity. However, our conclusions would be affected minimally by the inclusion of coagulase-negative staphylococci because coagulase-negative staphylococci only accounted for 2% of isolates. Another potential limitation of this study is lack of data on clinical presentation of the patients. We did not determine whether these were symptomatic urinary tract infections or asymptomatic bacteriuria. Last, this study may overrepresent antimicrobial resistance rates as practitioners may only send urine cultures for patients with recurrent infections or treatment failure.
Although these data are from New York State, there are several generalizable aspects of this study. First, this project highlights the framework for creating a regional antibiogram. Regional antibiograms are important tools that could be used by health departments and other regional authorities to help influence prescribing and/or track antimicrobial resistance for key pathogens. Second, New York State (especially Kings, Queens, New York, Bronx, and Richmond counties, the 5 New York City counties) is unique in that it has an incredibly high population density [9]. Along with antimicrobial utilization, population density has been shown to be associated with antimicrobial resistance prevalence [10]. Hence, it stands to reason that New York State can be thought of as a regional bellwether for antimicrobial resistance.
In summary, we conducted a 1-year retrospective analysis of outpatient urine isolates collected from patients in New York State. Data indicated that nitrofurantoin has retained activity against many of the urinary pathogens since the publication of the 2011 IDSA guidelines and has a high prevalence of in vitro activity. In contrast, we found that the overall prevalence of resistance in bacterial urinary isolates exceeded the predefined IDSA empiric therapy thresholds of 20% for trimethoprim-sulfamethoxazole and 10% for fluoroquinolones, with resistance rates being highest in patients ≥65 years. These data highlight the need in outpatient settings for (1) urinary culture specimen submission, (2) antimicrobial stewardship efforts, and (3) assembly of local susceptibility data to guide empiric therapy in acute uncomplicated cystitis. Finally, these data also support the desperate need for new oral antimicrobials to treat acute uncomplicated cystitis in the outpatient setting.

Supplementary Data
Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.