Annals of Clinical Microbiology, The official Journal of the Korean Society of Clinical Microbiology

6

Weeks in Review

2

Weeks to Publication
Indexed in KCI, KoreaMed, Synapse, DOAJ
Open Access, Peer Reviewed
pISSN 2288-0585 eISSN 2288-6850
Original article

Risk factor analysis of urinary tract infection by cefotaxime-resistant Escherichia coli and Klebsiella pneumoniae: a simple and effective analysis using the National Health Insurance Data Sharing Service

Jae Kwang Lee1, Yoonseon Park2, Young Ah Kim3

1Department of Research and Analysis, National Health Insurance Service Ilsan Hospital, Goyang, 2Division of Infectious Diseases Department of Internal Medicine Gil Medical Center, Gachon University College of Medicine, Incheon, 3Department of Laboratory Medicine, National Health Insurance Service Ilsan Hospital, Goyang, Korea

Corresponding to Young Ah Kim, E-mail: yakim@nhimc.or.kr; Yoonseon Park, E-mail: yoonseony@gilhospital.com

Ann Clin Microbiol 2023;26(3):59-66. https://doi.org/10.5145/ACM.2023.26.3.3
Received on 4 April 2023, Revised on 17 August 2023, Accepted on 30 August 2023, Published on 20 September 2023.
Copyright © Korean Society of Clinical Microbiology.

Abstract

Background: This study aims to analyze the risk factors for urinary tract infection (UTI) by cefotaxime-resistant Escherichia coli or Klebsiella pneumoniae, using data from the National Health Insurance Data Sharing Service. 

Methods: A retrospective case-control study was conducted to analyze the risk factors during 11 years (2010–2020). Study groups were selected based on the laboratory data of the hospital, which comprised 3,638 and 877 cases of cefotaxime-resistant E. coli and K. pneumoniae, respectively. Controls comprised 8,994 and 1,573 cases of cefotaxime-nonresistant (intermediate or susceptible) E. coli and K. pneumoniae, respectively. Clinical and socioeconomical features were obtained from the National Health Insurance service data. 

Results: In a multivariate analysis of risk factors for UTI by cefotaxime-resistant E. coli, the odds ratio (OR) of the male sex was 1.335 (95% confidence interval, 1.204–1.480), age 0–9 years was 1.794 (1.468–2.191), chronic renal disease was 1.227 (1.062–1.417), and hemodialysis was 1.685 (1.255–2.262). Moreover, the ORs of L-tube, central venous pressure catheter, and Foley catheter use were 1.204 (1.047–1.385), 1.332 (1.156–1.534), and 1.473 (1.316–1.649), respectively; the OR of previous antimicrobial use was 1.103 (1.009–1.206) and that of healthcare facility use was 1.782 (1.576–2.014). In a multivariate analysis of risk factors for UTI by cefotaxime-resistant K. pneumoniae, OR of the male sex was 1.460 (1.199–1.778), liver disease was 1.295 (1.037–1.617), and hemodialysis was 2.046 (1.263–3.315). The ORs of L-tube and Foley catheter use were 2.329 (1.861–2.915) and 1.793 (1.431–2.246), respectively, and the OR of the healthcare facility use was 1.545 (1.161–2.056). 

Conclusion: In this study, the risk factors for UTI caused by cefotaxime-resistant E. coli or K. pneumoniae were analyzed based on the data of a specific healthcare facility linked to the National Health Insurance system. We suggest that it is a simple and effective way to elucidate risk factors of infections caused by major antimicrobial-resistant pathogens.

Keywords

Risk factor, Urinary tract infection, Cefotaxime resistance, Escherichia coli, Klebsiella pneumoniae

Introduction

Urinary tract infection (UTI) refers to bacterial infection in a part of the urinary system consisting of the kidneys, ureter, bladder, and urethra [1]. UTI is a very common type of bacterial infection, and many women experience recurrent UTIs [2]. Well-known etiologic agents of UTI are Escherichia coli (most common), and other enteric gram-negative bacteria (Klebsiella, Pseudomonas, Proteus, Enterobacter, and Citrobacter spp.) [3,4]. Extended-spectrum ß-lactamase (ESBL)-producing E. coli or K. pneumoniae (most of them show the resistance to third generation cephalosporin, such as cefotaxime) are common in Korea, and resistance rates to cefotaxime were 37.1% and 24.9% for E. coli and K. pneumoniae, respectively, in 2019 [5]. 

The purpose of this study is to analyze the risk factors for UTI by cefotaxime-resistant E. coli or K. pneumoniae, using data from the National Health Insurance Data Sharing Service. The data of a specific healthcare facility can be linked to the national health insurance data with this service, which can be a simple and effective way, including not only the use of medical, but also service socio-demographic characteristics. With this methodology, entire antibacterial prescriptions can be obtained with this service, including antimicrobial use in the community.

Materials and methods

A total of 599,883 cases of urine culture over 11 years (2010-2020) were linked to the national health insurance service data in one general hospital (818 beds). Excluding repeated cases and missing data, 12,632 cases of E. coli and 2,450 cases of K. pneumoniae were included for a retrospective case-control study. Based on laboratory data of the hospital, study groups were selected. Cases comprised 3,638 cases of cefotaximeresistant E. coli and 877 cases of cefotaxime-resistant K. pneumoniae, which were isolated from urine cultures (UTI patients caused by ESBL-producers). Controls comprised 8,994 cases of cefotaxime-nonresistant (intermediate or susceptible) E. coli and 1,573 cases of cefotaxime-non-resistant K. pneumoniae, which were isolated from urine cultures (UTI patients caused by non-ESBL-producers).

Clinical and socioeconomical features were obtained from the national health insurance service data. The variables were defined as follows: underlying diseases were obtained in the inpatients during the admission period or outpatients within one month, using the major diagnostic code or four minor codes (total five codes), corresponding to malignancy (C00-D48), diabetes (E10-E14), ischemic heart disease (I20-I25), cardiovascular disease (I60-I69), chronic lower respiratory disease (J40-J47), liver disease (K70-K77), and chronic renal disease (N18).

Using health insurance claim codes, major surgery history (general anesthesia prescription), hemodialysis, anticancer treatment, and total parenteral nutrition history were confirmed. The use of a device, such as an L-tube, central venous pressure (CVP) catheter, and Foley catheter was evaluated within 90 days of pathogen isolation. Previous antimicrobial use was verified within 30 days of pathogen isolation, using J01 Anatomical Therapeutic Chemical codes, excluding local application or external preparation. The healthcare facility use comprised any healthcare facilities, including clinics before 30 days of pathogen isolation. The Charlson comorbidity index (CCI) was calculated according to the previous report [6]. 

Statistical analysis was performed with the Chi-square test and logistic regression. SAS program version 9.4 (SAS Institute, Cary, NC, USA) was used.

Results

Risk factors for cefotaxime-resistant E. coli-mediated UTI

In a univariate analysis, male sex, childhood or older age, underlying disease (malignancy, diabetes, ischemic heart disease, cardiovascular disease, chronic lower respiratory disease, liver disease, and chronic renal disease), hemodialysis, anticancer treatment, device use (L-tube, CVP catheter, and Foley catheter), previous antimicrobial use, frequent healthcare facility use, and high CCI (more than 3) were statistically significant for cefotaxime-resistant E. coli UTI (Table 1). 

In a multivariate analysis, the odds ratio (OR) of male sex was 1.335 (95% confidential interval, 1.2041.480); OR of age 0-9 years was 1.794 (1.468-2.191); OR of chronic renal disease was 1.227 (1.062-1.417); OR of hemodialysis was 1.685 (1.255-2.262); ORs of L-tube, CVP catheter, and Foley catheter use were 1.204 (1.047-1.385), 1.332 (1.156-1.534), and 1.473 (1.316-1.649), respectively; OR of previous antimicrobial use was 1.103 (1.009-1.206); OR of healthcare facility use was 1.782 (1.576-2.014) (Table 2).

Risk factors for cefotaxime-resistant K. pneumoniae-mediated UTI

In a univariate analysis, male sex, older age, underlying disease (malignancy, diabetes, ischemic heart disease, cardiovascular disease, chronic lower respiratory disease, liver disease, and chronic renal disease), hemodialysis, device use (L-tube, CVP catheter, and Foley catheter), and previous antimicrobial use, and frequent healthcare facility use were statistically significant for cefotaxime-resistant E. coli UTI (Table 3). 

In a multivariate analysis, OR of male sex was 1.460 (1.199-1.778); OR of liver disease was 1.295 (1.0371.617); OR of hemodialysis was 2.046 (1.263-3.315); OR of L-tube and Foley catheter use were 2.329 (1.861-2.915) and 1.793 (1.431-2.246), respectively; and OR of healthcare facility use was 1.545 (1.1612.056) (Table 2).

Table 1. Risk factors for cefotaxime-resistant Escherichia coli urinary tract infections: A univariate analysis.

VariableCTX-NR % (N)CTX-R % (N)P-value
All71.2 (8,994)28.8 (3,638)
SexFemale 58.2 (7,355)21.7 (2,742)< 0.0001
Male13.0 (1,639)07.1 (896)
Age (yr)0-908.6 (1,086)03.3 (416)< 0.0001
10-1900.8 (98)00.2 (20)
20-2902.7 (346)00.6 (77)
30-3903.4 (425)00.8 (106)
40-4906.2 (789)01.6 (208)
50-5909.7 (1,221)02.8 (357)
60-6909.4 (1,187)03.5 (442)
70-7915.7 (1,980)07.7 (971)
≥ 8014.7 (1,862)08.2 (1,041)
Underlying disease
MalignancyYes 35.1 (4,433)15.4 (1,946)< 0.0001
DiabetesYes 36.2 (4,576)18.0 (2,279)< 0.0001
Ischemic heart diseaseYes 25.5 (3,216)13.4 (1,699)< 0.0001
Cardiovascular diseaseYes 24.7 (3,120)13.9 (1,760)< 0.0001
Chronic LRT diseaseYes 53.0 (6,698)23.3 (2,942)< 0.0001
Liver diseaseYes 42.0 (5,303)18.9 (2,392)< 0.0001
Chronic renal diseaseYes 04.9 (625)03.8 (477)< 0.0001
Major surgeryYes 00.1 (7)00.0 (5)0.3247
HemodialysisYes 00.8 (100)00.9 (120)< 0.0001
Anticancer treatmentYes 00.7 (87)00.5 (60)0.0012
TPN useYes 00.0 (3)00.0 (3)0.2514
Device
L-tubeYes 05.6 (704)04.5 (564)< 0.0001
CVP catheterYes 05.8 (735)05.0 (629)< 0.0001
Foley catheterYes 14.9 (1,885)10.6 (1,345)< 0.0001
Previous antimicrobial useYes 41.5 (5,240)19.7 (2,483)< 0.0001
Healthcare facility useYes 25.5 (3,220)06.3 (790)< 0.0001
CCI043.3 (5,465)19.8 (2,500)< 0.0001
108.7 (1,094)02.1 (267)
205.8 (738)01.7 (212)
≥ 313.4 (1,697)05.2 (659)

Bold fonts are statistically significant. Abbreviations: CTX-NR, cefotaxime-intermediate or susceptible; CTX-R, cefotaxime-resistant; N, number; LRT, lower respiratory tract; TPN, total parenteral nutrition; CVP, central venous pressure; CCI, Charlson comorbidity index

Table 2. Risk factors of cefotaxime-resistant Escherichia coli or Klebsiella pneumoniae urinary tract infections: A multivariate analysis.

Variablecefotaxime-resistant E. colicefotaxime-resistant K. pneumoniae
OR95% CIOR95% CI
SexFemale11
Male1.3351.204-1.4801.4601.199-1.778
Age (yr)0-91.7941.468-2.1910.7220.448-1.162
10-190.8730.526-1.4491.9660.585-6.612
20-291.0140.762-1.3490.9380.372-2.364
30-391.0330.804-1.3270.9090.402-2.055
40-490.9970.820-1.2121.5180.961-2.398
50-590.9840.838-1.1550.8990.620-1.304
60-691.0180.881-1.1770.9080.651-1.267
70-791.0860.969-1.2170.9830.782-1.236
≥ 8011
Underlying disease
MalignancyNo11
Yes0.9790.895-1.0700.9130.748-1.115
DiabetesNo11
Yes0.8620.774-0.9600.8230.643-1.054
Ischemic heart diseaseNo11
Yes0.9060.822-0.9990.9660.784-1.189
Cardiovascular diseaseNo11
Yes0.7840.711-0.8650.7850.637-0.968
Chronic LRT diseaseNo11
Yes0.7430.667-0.8280.7290.575-0.925
Liver diseaseNo11
Yes0.9440.855-1.0431.2951.037-1.617
Chronic renal diseaseNo11
Yes1.2271.062-1.4171.1270.681-1.476
Major surgeryNo1
Yes1.6020.476-5.395
HemodialysisNo11
Yes1.6851.255-2.2622.0461.263-3.315
Anticancer treatmentNo11
Yes1.3220.936-1.8661.2860.691-2.393
TPN useNo1
Yes1.2740.229-7.084
Devices
L-tubeNo11
Yes1.2041.047-1.3852.3291.851-2.915
CVP catheterNo11
Yes1.3321.156-1.5341.2610.999-1.592
Foley CatheterNo11
Yes1.4731.316-1.6491.7931.431-2.246
Previous antimicrobial useNo11
Yes1.1031.009-1.2060.9790.793-1.209
Healthcare facility useNo11
Yes1.7821.576-2.0141.5451.161-2.056
CCI011
10.8340.712-0.9760.9910.660-1.487
20.9450.792-1.1260.8910.584-1.359
31.0630.926-1.2200.9250.691-1.238

Bold fonts are statistically significant. Abbreviations: OR, odds ratio; CI, confidential interval; LRT, lower respiratory tract; TPN, total parenteral nutrition; CVP, central venous pressure; CCI, Charlson comorbidity index

Table 3. Risk factors for cefotaxime-resistant Klebsiella pneumoniae urinary tract infections: A univariate analysis.

VariableCTX-NR % (N)CTX-R % (N)P-value
All64.2 (1,573)35.8 (877)
SexFemale 44.1 (1,081)22.7 (557)< 0.0001
Male20.1 (492)13.1 (320)
Age (yr)0-911.9 (292)02.2 (53)< 0.0001
10-1900.3 (8)00.2 (5)
20-2900.9 (22)00.3 (7)
30-3901.1 (28)00.4 (9)
40-4902.9 (72)01.8 (43)
50-5906.0 (148)02.6 (64)
60-6906.8 (166)03.5 (86)
70-7917.3 (423)11.7 (286)
≥ 8016.9 (414)13.2 (324)
Underlying disease
MalignancyYes 30.4 (745)20.6 (505)< 0.0001
DiabetesYes 37.2 (912)26.4 (648)< 0.0001
Ischemic heart diseaseYes 25.8 (631)19.1 (469)< 0.0001
Cardiovascular diseaseYes 27.5 (673)21.8 (534)< 0.0001
Chronic LRT diseaseYes 44.0 (1,078)28.7 (703)< 0.0001
Liver diseaseYes 38.2 (935)24.4 (597)< 0.0001
Chronic renal diseaseYes 07.8 (192)79.9 (1,957)< 0.0001
HemodialysisYes 01.5 (36)02.7 (67)< 0.0001
Anticancer treatmentYes 01.1 (26)00.9 (22)0.1429
Device
L-tubeYes 10.4 (256)15.2 (373)< 0.0001
CVP catheterYes 10.6 (260)12.9 (317)< 0.0001
Foley catheterYes 24.9 (609)24.3 (596)< 0.0001
Previous antimicrobial useYes 42.0 (1,029)26.8 (656)< 0.0001
Healthcare facility useYes 22.1 (541)07.6 (187)< 0.0001
CCI039.6 (969)23.6 (579)0.0232
105.3 (130)01.9 (46)
203.8 (92)01.8 (45)
≥ 315.6 (382)08.4 (207)

Bold fonts are statistically significant. Abbreviations: CTX-NR, cefotaxime-intermediate or susceptible; CTX-R, cefotaxime-resistant; N, number; LRT, lower respiratory tract; CVP, central venous pressure; CCI, Charlson comorbidity index

Discussion

Antimicrobial resistance and emerging ESBL infections are rising concerns in public health [7]. The production of β-lactamase enzymes is the primary mechanism by which Gram-negative bacteria resist the action of all β-lactam antibiotics, except carbapenems and cephamycins [8]. Additionally, ESBL-producing isolates have demonstrated rising rates of resistance to other classes of antibiotics, such as sulfonamides, aminoglycosides, and fluoroquinolones [8]. These multidrug-resistant organisms significantly limit the treatment options, and initial empirical therapy is often ineffective and associated with poor outcomes [9]. Defining risk factors for UTI by ESBL-producing E. coli or K. pneumoniae is very important to manage antimicrobial resistance in Korea because UTIs by these organisms are very common even in the community and they frequently progress to bloodstream infections, showing a serious prognosis [10]. The well-known risk factors for community-acquired UTI by ESBL-producing E. coli were prior use of antibiotics (OR from 2.2 to 21.4), previous hospitalization (OR from 1.7 to 3.9), and UTI history (OR from 1.3 to 3.8) after reviewing 16 previous observational studies [11]. In a multicenter study, a total of 983 patient-specific isolates were reviewed (890 were E. coli, 68 were Klebsiella species, and 25 were Proteus mirabilis) and significant risk factors for ESBL-producing Enterobacteriaceae-mediated infections were recent antibiotic use, residence in a long-term care facility, recent hospitalization, age 65 years, and male sex [12]. In this study, we found similar risk factors for UTI by cefotaxime-resistant E. coli or K. pneumoniae (ESBL-producers), such as male sex, underlying disease, hemodialysis, use of a device, previous antimicrobial use, and healthcare facility.

The National Health Insurance Data Sharing Service was used in this study, which can link the data of a specific healthcare facility to the national health insurance data. The strength of this study was that whole antibacterial prescriptions can be obtained with this service, including antimicrobial use in the community and healthcare facility use can also be identified. The risk factors for UTI caused by cefotaxime-resistant E. coli or K. pneumoniae were analyzed after the data of a specific healthcare facility were linked to the national health insurance data. Therefore, we suggest that it is a simple and effective way to elucidate risk factors of infections by major antimicrobial-resistant pathogens.

Ethics statement

The Institutional Review Board of the National Health Insurance Service Ilsan Hospital approved this study (NHIMC 2022-05-019) and waiver of informed consent was obtained.

Conflicts of interest

No potential conflicts of interest relevant to this article were reported.

Acknowledgements

This study used the National Health Insurance Data Sharing Service with permission (NHIS-2022-1-394).

Funding

This study was funded by a grant from the National Health Insurance Service Ilsan Hospital (2021-20-001).

References

1. Gupta K, Grigoryan L, Trautner B. Urinary tract infection. Ann Intern Med 2017;167:ITC49-64.

2. Jung C, Brubaker L. The etiology and management of recurrent urinary tract infections in postmenopausal women. Climacteric 2019;22:242-9.

3. Millner R, Becknell B. Urinary tract infections. Pediatr Clin North Am 2019;66:1-13.

4. Lee H, Yoon EJ, Kim D, Jeong SH, Won EJ, Shin JH, et al. Antimicrobial resistance of major clinical pathogens in South Korea, May 2016 to April 2017: first one-year report from KorGLASS. Euro Surveill 2018;23:1800047.

5. Kim D, Yoon EJ, Hong JS, Choi MH, Kim HS, Kim YR, et al. Major bloodstream infectioncausing bacterial pathogens and their antimicrobial resistance in South Korea, 2017-2019: Phase I report from Kor-GLASS. Front Microbiol 2022;12:799084.

6. Kim KH. Comparative study on three algorithms of the ICD-10 Charlson comorbidity index with myocardial infarction patients. J Prev Med Public Health 2010;43:42-9.

7. van Driel AA, Notermans DW, Meima A, Mulder M, Donker GA, Stobberingh EE, at al. Antibiotic resistance of Escherichia coli isolated from uncomplicated UTI in general practice patients over a 10-year period. Eur J Clin Microbiol Infect Dis 2019;38:2151–8.

8. Rawat D, Nair D. Extended-spectrum β-lactamases in gram negative bacteria. J Glob Infect Dis 2010;2:263–74. Clin Microbiol 2001;39:1316-8.

9. Al-Tamimi M, Abu-Raideh J, Albalawi H, Shalabi M, Saleh S. Effective oral combination treatment for extended-spectrum beta-lactamase-producing Escherichia coli. Microb Drug Resist 2019;25:1132–41.

10. Holmes CL, Anderson MT, Mobley HLT, Bachman MA. Pathogenesis of Gram-negative bacteremia. Clin Microbiol Rev 2021;34:e00234-20.

11. Larramendy S, Deglaire V, Dusollier P, Fournier JP, Caillon J, Beaudeau F, et al. Risk factors of extended-spectrum beta-lactamases-producing Escherichia coli community acquired urinary tract infections: a systematic review. Infect Drug Resist 2020;13:3945-55.

12. Ben-Ami R, Rodríguez-Baño J, Arslan H, Pitout JD, Quentin C, Calbo ES, et al. A multinational survey of risk factors for infection with extended-spectrum ß-lactamaseproducing Enterobacteriaceae in nonhospitalized patients. Clin Infect Dis 2009;49:682-90.

Table 1
Table 2
Table 3

1. Gupta K, Grigoryan L, Trautner B. Urinary tract infection. Ann Intern Med 2017;167:ITC49-64.

2. Jung C, Brubaker L. The etiology and management of recurrent urinary tract infections in postmenopausal women. Climacteric 2019;22:242-9.

3. Millner R, Becknell B. Urinary tract infections. Pediatr Clin North Am 2019;66:1-13.

4. Lee H, Yoon EJ, Kim D, Jeong SH, Won EJ, Shin JH, et al. Antimicrobial resistance of major clinical pathogens in South Korea, May 2016 to April 2017: first one-year report from KorGLASS. Euro Surveill 2018;23:1800047.

5. Kim D, Yoon EJ, Hong JS, Choi MH, Kim HS, Kim YR, et al. Major bloodstream infectioncausing bacterial pathogens and their antimicrobial resistance in South Korea, 2017-2019: Phase I report from Kor-GLASS. Front Microbiol 2022;12:799084.

6. Kim KH. Comparative study on three algorithms of the ICD-10 Charlson comorbidity index with myocardial infarction patients. J Prev Med Public Health 2010;43:42-9.

7. van Driel AA, Notermans DW, Meima A, Mulder M, Donker GA, Stobberingh EE, at al. Antibiotic resistance of Escherichia coli isolated from uncomplicated UTI in general practice patients over a 10-year period. Eur J Clin Microbiol Infect Dis 2019;38:2151–8.

8. Rawat D, Nair D. Extended-spectrum β-lactamases in gram negative bacteria. J Glob Infect Dis 2010;2:263–74. Clin Microbiol 2001;39:1316-8.

9. Al-Tamimi M, Abu-Raideh J, Albalawi H, Shalabi M, Saleh S. Effective oral combination treatment for extended-spectrum beta-lactamase-producing Escherichia coli. Microb Drug Resist 2019;25:1132–41.

10. Holmes CL, Anderson MT, Mobley HLT, Bachman MA. Pathogenesis of Gram-negative bacteremia. Clin Microbiol Rev 2021;34:e00234-20.

11. Larramendy S, Deglaire V, Dusollier P, Fournier JP, Caillon J, Beaudeau F, et al. Risk factors of extended-spectrum beta-lactamases-producing Escherichia coli community acquired urinary tract infections: a systematic review. Infect Drug Resist 2020;13:3945-55.

12. Ben-Ami R, Rodríguez-Baño J, Arslan H, Pitout JD, Quentin C, Calbo ES, et al. A multinational survey of risk factors for infection with extended-spectrum ß-lactamaseproducing Enterobacteriaceae in nonhospitalized patients. Clin Infect Dis 2009;49:682-90.