Multi-Institutional Predictors of Antibiotic Resistance in Patients Presenting to the Emergency Department with Urosepsis Secondary to Ureteral Obstruction

Abstract

Introduction and Objective:

Patients presenting with a urinary tract infection with kidney or ureteral stones is a urologic emergency often achieve early clinical stability but remain hospitalized while awaiting results from urine antibiotic sensitivity analyses. We aimed to identify clinical predictors of antibiotic resistance in patients who underwent urgent urinary tract decompression for sepsis and obstructive urolithiasis to facilitate early discharge on empiric oral antibiotics.

Methods:

Patients who underwent emergent urinary tract decompression for sepsis and an obstructing ureteral stone from 2014 to 2018 at two academic medical institutions were identified. Emergent stent placement was performed and patients were treated with broad-spectrum intravenous antibiotics. We assessed the association between clinical parameters at the time of presentation and resistance to at least one antibiotic from urine culture using the Wilcoxon test and Fisher exact test for continuous and categorical variables, respectively. Multivariate logistic regression was then performed using all significant variables from univariate analysis.

Results:

Out of 134 patients, 84 patients (62.7%) had urine cultures resistant to at least one antibiotic. On univariate analysis, patients with resistant cultures were significantly more likely to have had previous ureteroscopy, require postoperative intensive care unit-level care, have bacteremia, and a longer length of stay. In multivariate analysis using significant variables from univariate analysis, only previous ureteroscopy was significantly associated with antibiotic resistance with an increased odds of 6.95 (p = 0.011).

Conclusions:

In this study, we show that a history of ureteroscopy is significantly associated with antibiotic resistance in both univariate and multivariate analyses. Our findings suggest that patients with history of ureteroscopy should await urine culture results, while those without a history of ureteroscopy may be discharged early on empiric oral antibiotics. However, future studies are necessary to determine the effectiveness of this predictor.

Introduction

Urolithiasis is a prevalent condition that affects ∼8% of the U.S. population¹ and can cause secondary unilateral or bilateral renal obstruction. The presence of signs of a urinary tract infection (UTI) and/or anuria in patients with kidney or ureteral stones is a urologic emergency. Urgent decompression, with either a ureteral stent or percutaneous nephrostomy^2,3 along with adequate assessment of culture data and adherence to an appropriate therapy, is often necessary to prevent further complications in infected hydronephrosis.

According to the American Urological Association (AUA) guidelines, definitive stone removal should be delayed until the infection is cleared following a completed course of antimicrobial therapy.⁴ These patients are often hospitalized for hemodynamic support and given broad-spectrum antibiotics. However, a subset of patients achieve early clinical stability but remain hospitalized while awaiting results from urine antibiotic sensitivity analyses. If validated in future prospective studies, these patients with antibiotic-sensitive cultures may be discharged on empiric therapy. Thus, preliminary analysis of factors associated with antibiotic sensitivity is imperative and may help meet an unmet clinical need.

With increasing antibiotic resistance patterns in nosocomial infections, it is important to establish local strategies to reduce the risk for antibiotic resistance, such as rationalization of the empiric use of antibiotics and limiting antibiotic prophylaxis to those patients with predetermined risk factors along with the prevention and management of infectious complications in kidney stone disease. The aim of this study was to identify patterns and clinical predictors of antibiotic resistance in patients who underwent urgent urinary tract decompression for sepsis and obstructive urolithiasis.

Materials and Methods

Patient cohort

Patients who underwent emergent urinary tract decompression for sepsis and an obstructing ureteral stone from 2014 to 2018 at two academic medical institutions in New York, NY (Mount Sinai Hospital), and Providence, RI (Brown University), were identified. All patients satisfied the standard criteria for sepsis, defined as systemic inflammatory response syndrome, fulfilling at least two of the following criteria: body temperature >38.0°C or <36.0°C, heart rate >90 beats per minute, respiratory rate >20 breaths per minute or arterial CO₂ tension <32 mm Hg, white blood cell count >12,000/mm³ or <4000/mm³, or immature neutrophils >10%⁵ along with clinical suspicion of UTI.

Retrospective data collection

As per established guidelines,⁶ emergent stenting was performed and patients were treated with broad-spectrum intravenous antibiotics. Chart review was performed for all identified patients. Clinical parameters, including surgical history, stone analysis, length of stay, blood and urine cultures, and length of intensive care unit (ICU) admission, were obtained from the electronic medical record. Complete blood count and vitals at the time of admission were also obtained. Results from urine cultures were analyzed for antibiotic sensitivities. We defined a culture as resistant if it was resistant to at least one antimicrobial. Pan-sensitive cultures included all other cultures (i.e., sensitive to all antimicrobials). This study was approved by the Institutional Review Board (HS14-00879).

Statistical analysis

Univariate comparison of clinical and demographic variables was performed between patients with pan-sensitive and resistant urine cultures. Patterns of antibiotic resistance between the two institutions included in the study were also compared. Statistical significance was assessed using Wilcoxon test and Fisher exact test for continuous and categorical variables, respectively.

Multivariate analysis using binomial logistic regression was then performed using all variables significantly associated with resistant cultures in univariate analysis.

Results

Demographics

One hundred thirty-four patients who met the inclusion criteria were identified. Eighty-four patients (62.7%) had urine cultures with antibiotic resistance. We found no significant differences in age, sex, body mass index, history of diabetes, paraplegia, stone size, serum white blood cell count, temperature at admission, systolic blood pressure at admission, and heart rate at admission between patients with resistant and sensitive urine cultures (Table 1).

Table 1.

Univariate Analysis of Predictors of Resistant Urine Culture

	Pan-sensitive (n = 50)	Resistant (n = 84)	p
Age, years	59.9	58.9	0.763
Gender, male, %	22	27.1	0.516
BMI, kg/m²	31.2	31.8	0.874
Diabetes, %	22.4	22.4	0.962
Paraplegia, %	2	4.7	0.425
Stone size, mm	6.7	7.6	0.249
WBC, serum	15	15.9	0.403
Temperature, F	99.3	100.4	0.489
Systolic blood pressure, mm Hg	113	113	0.782
Heart rate, bpm	107	109	0.498
Previous urologic surgery, %	22 (44)	44.7 (53)	0.008^*
Previous ureteroscopy, %	8 (16)	38.9 (46)	>0.001
Previous PCNL, %	4 (8)	7.1 (8.4)	0.471
ICU admission postoperatively, %	12 (24)	27.1 (32)	0.039^*
Length of stay, mean, days	3.4 (6.8)	5.4 (6.4)	0.026^*
Positive blood culture, %	24 (48)	48.2 (57)	0.005^*
History of UTI, %	21 (42)	28 (33)	0.291

Pan-sensitive indicates sensitivity to all antimicrobials, while resistant indicates resistance to at least one antimicrobial. Significance was calculated using Wilcoxon test for continuous variables and Fisher exact test for categorical variables.

Indicates p < 0.05.

BMI = body mass index; ICU = intensive care unit; PCNL = percutaneous nephrolithotomy; UTI = urinary tract infection; WBC = white blood cell.

Predictors of urine culture resistance

We then sought to determine whether clinical parameters taken from the time of admission were associated with urine culture resistance. Table 1 presents univariate statistics for variable-wise comparison between antibiotic-sensitive and antibiotic-resistant patients. Patients with resistant cultures were more likely to have had previous urologic surgery (44.7% vs 22.0%, p = 0.008). The most notable difference was in patients who had had previous ureteroscopy (38.9% vs 8.0%, p = 0.0002). Those with resistant cultures were more likely to require postoperative ICU-level care (27.1% vs 12.0%, p = 0.039), have bacteremia (48.2% vs 24.0%, p = 0.005), and a longer length of stay (5.4 vs 3.4 days, p = 0.026). Comparing urine culture antibiotic resistance patterns between institutions (Table 2), patients from Mount Sinai had a significantly greater rate of ampicillin/sulbactam resistance (60.0% vs 12.7%; p < 0.001) and piperacillin/tazobactam resistance (20.0% vs 3.1%; p < 0.001) compared with those at Brown University.

Table 2.

Patterns of Antibiotic Resistance by Institution

Antibiotic	Brown University (n = 102)	Mount Sinai Hospital (n = 32)	p
Ampicillin/sulbactam	12.30%	60.00%	0.0002^*
Aztreonam	7.70%	10.00%	0.746
Cephalexin	13.80%	30.00%	0.212
Ciprofloxacin	35.40%	45.00%	0.619
Gentamicin	16.90%	30.00%	0.33
Levofloxacin	29.20%	45.00%	0.302
Meropenem	1.50%	5.00%	0.378
Piperacillin/tazobactam	3.10%	20.00%	0.009^*
Trimethoprim/sulfamethoxazole	29.50%	45.00%	0.302

Indicates p < 0.05.

We then identified variables that were independently predictive of antibiotic-resistant cultures using multivariate logistic regression. In multivariate logistic analysis using significant variables from univariate analysis, only previous ureteroscopy was significantly associated with antibiotic resistance with an increased odds of 6.95 (p = 0.011; Table 3).

Table 3.

Multivariate Analysis of Significant Predictors from Univariate Model

	Odds ratio (95% confidence interval)	p
Prior urologic surgery	0.82 (0.26–2.6)	0.737
Prior ureteroscopy	6.95 (1.65–33.98)	0.011^*
ICU admission postoperatively	0.91 (0.31–2.74)	0.859
Length of stay	1.13 (0.99–1.34)	0.091
Positive blood culture	1.37 (0.62–3.05)	0.436

All variables reaching significantly associated (p < 0.05) with antibiotic resistance in univariate analysis were included in a multivariate logistic regression model.

Indicates p < 0.05.

Organisms identified by urine culture

We explored the distribution of organisms found on urine culture at both Mount Sinai and Brown University. In both populations, Escherichia coli was the most common organism (Table 4). However, the proportion of E. coli-positive urine cultures varied by institution, with 65% at Mount Sinai and 46% at Brown University. We also observed a larger range of organisms found in the patient population at Brown University (12), compared with Mount Sinai (7). Acinetobacter and Corynebacterium were detected only at Mount Sinai, whereas Staph-, Enterobacter-, Candida-, Group B Strep-, Citrobacter-, Achromobacter-, and Providencia-positive urine cultures were observed only at Brown University (Table 4).

Table 4.

Organisms Identified in Urine Cultures in Mount Sinai and Brown Cohorts

Organism	Mount Sinai	Brown	Total
Escherichia coli	20	49	69
Proteus	6	9	15
Klebsiella	1	13	14
Pseudomonas	1	2	3
Enterococcus	1	8	9
Acinetobacter	1	0	1
Corynebacterium	1	0	1
Staph	0	9	9
Enterobacter	0	5	5
Candida	0	7	7
Group B Strep	0	1	1
Citrobacter	0	2	2
Achromobacter	0	1	1
Providencia	0	1	1
Total	31	107	138

Multiple organisms grew in a subset of individuals.

Discussion

Approximately 20% to 30% of sepsis cases result from a UTI and commonly involve an obstructive disease of the urinary tract, such as calculi, tumors, or stenosis.⁷ Septic patients with concomitant unitary tract obstruction undergo emergent urinary tract decompression and are treated with antibiotic therapy. Following decompression, patients are hospitalized while awaiting results from urinary cultures to determine an appropriate antibiotic regimen upon discharge.

To our knowledge, no previous work has addressed factors predictive of resistant urine culture in urosepsis patients. In this work, we identified clinical factors predictive of pan-sensitive urine cultures, potentially allowing for early discharge of these patients on broad-spectrum oral antibiotics.

Patterns of antibiotic sensitivity in our cohort matched with previous surveillance reports.⁸ A 2017 report⁸ from the European Centre for Disease Prevention and Control found wide variation in resistance rates within Europe by geographical region, with lower resistance in the North and greater in the South and East. Rates also vary considerably by institution, indicating the need for institution-specific treatment regimens. Similarly, in our study, we found significant differences in rates of amoxicillin/sulbactam and piperacillin/tazobactam resistance between the two institutions included. These differences in antibiotic resistance patterns may be attributed, in part, to differences in local microbiome colonization due to differences in diet and geography, and population resistance patterns.⁹

In univariate analysis, patients with resistant cultures had a significantly greater rate of previous urologic surgery, specifically ureteroscopy, and were much more likely to require postoperative ICU admission. It is possible that patients were prescribed an antibiotic during the perioperative period at the time of prior urologic surgery that led to the development of antibiotic resistance. However, a history of UTI was not significantly associated with antibiotic resistance, suggesting that if patients were treated with antibiotics for a previous UTI, this did not drive current resistance. In multivariate analysis, only a history of ureteroscopy was significantly associated with resistant culture. This raises the possibility that community-acquired infections treated with antibiotics may be less likely to create resistance patterns than prophylactic or therapeutic antibiotics given in a hospital setting, even in the absence of an infection.

In a recent study, previous intravenous antibiotic administration was associated with antibiotic resistance in patients presenting with sepsis.¹⁰ In addition, patients with prior exposure to the health care system have an increased risk for colonization by antibiotic resistance bacteria¹¹ and may contribute to resistant urine cultures. However, many patients are unaware of specific antibiotics they have previously received and medical records from outside institutions are frequently unavailable. Moreover, patients in our study had community-acquired infections for which knowledge of previous antibiotic administration is necessary for appropriate empiric therapy. When antibiotic administration history is unknown, clinicians must rely on broad-spectrum antibiotics, which increase the risk of emergence of resistant bacterial strains.¹² For these patients, history of hospitalizations or surgeries may serve as surrogate markers for prior antibiotic exposure and guide treatment plans.

Our findings highlight the important role that the history of ureteroscopy plays in driving antibiotic resistance. In our study, the history of ureteroscopy may be a surrogate marker reflective of prior antibiotic administration; however, this must be validated in future studies. When knowledge of prior antibiotic treatment is unavailable, clinicians treating patients presenting with urosepsis secondary to obstruction may use surgical history to guide the treatment plan. For example, a history of ureteroscopy may warrant suspicion of antibiotic-resistant culture. Our findings suggest that these patients should remain hospitalized and await results of urine cultures due to increased risk of antibiotic-resistant cultures. Patients without a history of ureteroscopy are at lower risk for antibiotic-resistant infections and may be considered for early discharge. In addition, our study shows the great variations in infective agents present by institution. Thus, although our results may inform clinician judgment, specific infective agents prevalent at individual institutions must be considered in determining the appropriate choice of antibiotic therapy.

Since our study is limited by its retrospective nature, a prospective trial evaluating the benefits of early discharge without urine cultures is necessary. Our study is also limited by the lack of information on historical surgeries since patients were frequently treated at other institutions. Thus, we were not able to ascertain information on postoperative management for prior urologic surgeries or time from past surgery. In addition, we did not find a relationship between time of prior urological surgery and the development of antibiotic resistance in our series, but our study was limited by the lack of information on historical surgeries since some patients may have been treated at other institutions, for which we did not have medical record access. For patients without a history of ureteroscopy, alternative diagnostic modalities, such as novel rapid molecular diagnostic tests, may also have clinical utility. Rapid molecular assays have been applied to determine antibiotic sensitivity patterns for patients with bacteremia and pneumonia^13,14 and may be applied in the setting of urosepsis.

Conclusion

In this study, we identified predictors of antibiotic resistance in urine cultures using demographic and clinical variables available at the time of presentation to the emergency department with ureteral obstruction and urosepsis. We show that a history of ureteroscopy is significantly associated with antibiotic resistance in both univariate and multivariate analyses. Our findings may help guide clinical decision-making for patients presenting with urosepsis secondary to obstruction. Patients with a history of ureteroscopy should await urine culture results, while those without a history of ureteroscopy may be discharged early on empiric oral antibiotics, however, future studies are necessary to determine the effectiveness of this predictor.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was provided for this article.

Abbreviations Used

References

Scales

, Smith

, Hanley

, Saigal

. Prevalence of kidney stones in the United States. Eur Urol, 2012; 62:160–165.

Uppot

. Emergent nephrostomy tube placement for acute urinary obstruction. Tech Vasc Interv Radiol, 2009; 12:154–161.

Preminger

, Tiselius

, Assimos

, et al. 2007 Guideline for the management of ureteral calculi. J Urol, 2007; 178:2418–2434.

Assimos

, Krambeck

, Miller

, et al. Surgical management of stones: American Urological Association/Endourological Society Guideline, Part I. J Urol, 2016; 196:1153–1160.

Singer

, Deutschman

, Seymour

, et al. The third international consensus definitions for sepsis and septic shock (sepsis-3). JAMA, 2016; 315:801–810.

Pearle

, Goldfarb

, Assimos

, et al

Medical management of kidney stones

: AUA guideline. J Urol, 2014; 192:316–324.

Brun-Buisson

. The epidemiology of the systemic inflammatory response. Intensive Care Med, 2000; 26(Suppl 1):S64–S74.

European Centre for Disease Prevention and Control. Surveillance of antimicrobial resistance in Europe: Annual report of the European Antimicrobial Resistance Surveillance Network (EARS-Net). ECDC, Stockholm, Sweden, 2018.

Gupta

, Paul

, Dutta

. Geography, ethnicity or subsistence-specific variations in human microbiome composition and diversity. Front Microbiol, 2017; 8:1162.

10.

Micek

, Hampton

, Kollef

. Risk factors and outcomes for ineffective empiric treatment of sepsis caused by gram-negative pathogens: Stratification by onset of infection. Antimicrob Agents Chemother, 2017; 62:e01577-17.

11.

Shorr

, Zilberberg

, Reichley

, Kan

, Hoban

, Hoffman

, Micek

, Kollef

. Validation of a clinical score for assessing the risk of resistant pathogens in patients with pneumonia presenting to the emergency department. Clin Infect Dis, 2012; 54:193–198.

12.

Kollef

, Micek

. Rational use of antibiotics in the ICU balancing stewardship and clinical outcomes. JAMA, 2014; 312:1403–1404.

13.

Timbrook

, Morton

, Mcconeghy

, Caffrey

, Mylonakis

, LaPlante

. The effect of molecular rapid diagnostic testing on clinical outcomes in bloodstream infections: A systematic review and meta-analysis. Clin Infect Dis, 2017; 64:15–23.

14.

Parta

, Goebel

, Thomas

, Matloobi

, Stager

, Musher

. Impact of an assay that enables rapid determination of staphylococcus species and their drug susceptibility on the treatment of patients with positive blood culture results. Infect Control Hosp Epidemiol, 2010; 31:1043–1048.