Skip to main content
NCBI home page
Journal of Antimicrobial Chemotherapy logoLink to Journal of Antimicrobial Chemotherapy
. 2023 Nov 14;79(1):82–95. doi: 10.1093/jac/dkad351

Incidence of acute kidney injury (AKI) and its impact on patient outcomes among adult hospitalized patients with carbapenem-resistant Gram-negative infections who received targeted treatment with a newer β-lactam or β-lactam/β-lactamase inhibitor-, polymyxin- or aminoglycoside-containing regimen

Thomas P Lodise 1,✉, Emre Yucel 2, Engels N Obi 3, Alexandre H Watanabe 4, Brian H Nathanson 5
PMCID: PMC10761276  PMID: 37962080

Abstract

Background

Limited comparative data exist on acute kidney injury (AKI) risk and AKI-associated outcomes in hospitalized patients with carbapenem-resistant Gram-negative infections (CR-GNIs) treated with a newer β-lactam/β-lactam-β-lactamase inhibitor (BL/BL-BLI)-, polymyxin (PB)- or aminoglycoside (AG)-containing regimen. This study quantified the risk of AKI and AKI-related outcomes among patients with CR-GNIs treated with a newer BL/BL-BLI-, PB- or AG-containing regimen.

Methods

A multicentre, retrospective, observational study was performed (2016–20). The study included adult hospitalized patients with (i) baseline estimated glomerular filtration rates ≥30 mL/min/1.73 m2; (ii) CR-GN pneumonia, complicated urinary tract infection or bloodstream infection; and (iii) receipt of newer BL/BL-BLI, PG or AG within 7 days of index CR-GN culture for ≥3 days. Outcomes included AKI, in-hospital mortality and hospital costs.

Results

The study included 750 patients and most (48%) received a newer BL/BL-BLI. The median (IQR) treatment duration was 8 (5–11), 5 (4–8) and 7 (4–8) days in the newer BL/BL-BLI group, AG group and PB group, respectively. The PB group had the highest adjusted AKI incidence (95% CI) (PB: 25.1% (15.6%–34.6%) versus AG: 8.9% (5.7%–12.2%) versus newer BL/BL-BLI: 11.9% (8.1%–15.7%); P = 0.001). Patients with AKI had significantly higher in-hospital mortality (AKI: 18.5% versus ‘No AKI’: 5.6%; P = 0.001) and mean hospital costs (AKI: $49 192 versus ‘No AKI’: $38,763; P = 0.043).

Conclusions

The AKI incidence was highest among PB patients and patients with AKI had worse outcomes. Healthcare systems should consider minimizing the use of antibiotics that augment AKI risk as a measure to improve outcomes in patients with CR-GNIs.

Introduction

Despite notable advances in patient care, it is estimated that approximately 1 in 25 hospitalized patients will develop a healthcare-associated infection and over half of infections will be resistant to a first-line treatment option.1–4 While there has been a reduction in antibiotic-resistant Gram-positive infections in recent years,4 the prevalence of carbapenem-resistant Gram-negative infections (CR-GNIs) has increased dramatically over the past decade among hospitalized patients worldwide.5,6 Patients with CR-GNIs have extended hospital stays, higher mortality rates and greater healthcare costs compared with patients with serious carbapenem-susceptible GNIs.7–13 As a testament to the importance of CR-GNIs, the WHO identified carbapenem-resistant Acinetobacter baumannii, Pseudomonas aeruginosa and Enterobacterales as critical priority antibiotic-resistant pathogens that pose the greatest threat to human health.14,15

Several factors contribute to the increased morbidity and mortality among patients with CR-GNIs.16–21 Patients with CR-GNIs are often critically ill and have multiple comorbidities.16–21 One common comorbid condition among patients with CR-GNIs is renal disease.22–31 Studies show that hospitalized patients with impaired renal function are at an increased risk of death and other healthcare-related problems relative to those without impaired renal function.22–34 Additionally, data indicate that hospitalized patients who develop even modest cases of acute kidney injury (AKI), regardless of the cause, are at greater risk of unfavourable outcomes.22–31,35 This is a critically important consideration for patients with CR-GNIs since potential agents include the polymyxins (PB) and aminoglycosides (AG), antibiotics with well-described renal adverse event profiles.36–46 Although expert guidance documents have relegated AG and PB as last-line agents due their toxicity profiles,47–49 they are still frequently used in many patients with CR-GNIs.16,50

While the deleterious consequences of AKI are well described across most therapeutic domains,22–31,35,40,43,46 few comparative real-world evidence studies have compared the cumulative incidence of AKI between commonly used antibiotics among adult, hospitalized patients with serious CR-GNIs, and assessed the effects of treatment-associated AKI (overall and by baseline renal function status) on outcomes of patients with CR-GNIs.39,41,46 Given these literature gaps, the objectives of this study were to quantify the incidence of AKI51 and outcomes associated with AKI, overall and by presence of baseline renal impairment on index treatment day, among hospitalized adult patients with CR-GNIs who received targeted treatment with a newer β-lactam or β-lactam/β-lactamase inhibitor (newer BL/BL-BLI)-, PB- or AG-containing regimen.

Patients and methods

Study design and population

A retrospective, multicentre observational study of adult hospitalized patients in the PINC AI™ Healthcare Data (PHD),52 between 2016 and 2020 was performed (Appendix A, available as Supplementary data at JAC Online). Patients from the PINC AI Healthcare Database were included in this study if they met the following criteria: (i) age ≥ 18 years; (ii) hospital discharge between 2016 and 2020; (iii) diagnosis for pneumonia (PNA), complicated urinary tract infection (cUTIs), or bloodstream infection (BSI) (Appendix B); (iv) presence of a CR-GN pathogen(s) other than carbapenem-resistant A. baumannii or Stenotrophomonas maltophilia on a clinical culture site consistent with infection diagnosis (index CR-GN culture day); (v) receipt of any antibiotic within −2 days to +3 days of the index CR-GN culture (Appendix C); (vi) receipt of a newer BL/BL-BLI- (i.e. ceftolozane/tazobactam, ceftazidime/avibactam, imipenem/relebactam, cefiderocol or meropenem/vaborbactam), AG (i.e. tobramycin, gentamicin or amikacin)- or PB (i.e. polymyxin B or colistin)-containing regimen within ≤7 days of the index CR-GN culture collection day (index treatment day); (vii) administration of a newer BL/BL-BLI -, AG- or PB-containing regimen for ≥3 days; (viii) serum creatinine (SCR) data within ± 2 days of admission; (ix) estimated glomerular filtration rate (eGFR) ≥ 30 mL/min/1.73 m2 between −2 days to +2 days of the index treatment day; (x) ≥1 SCR value(s) from +2 days post index treatment day through to +3 days after treatment discontinuation; (xi) no receipt of any renal replacement therapy (RRT) from −3 months to +2 days of the index treatment day; and (xii) no diagnosis of cystic fibrosis or moderate to severe bronchiectasis.

For patients who received a newer BL/BL-BLI, AG or PB within ≤ 7 days of the index CR-GN culture, the first treatment received for ≥3 days defined the treatment group. Patients were excluded if they received multiple treatments of interest (e.g. both AG and newer BL/BL-BLI) on the same index day for ≥3 days or if carbapenem-resistant A. baumannii or S. maltophilia was identified on the index CR-GN culture. We excluded patients with a carbapenem-resistant A. baumannii or S. maltophilia given the limited microbiological activity of approved newer BL/BL-BLI agents during the study period against these pathogens. Finally, only the first encounter was considered among patients with ≥1 hospital admissions that met the study criteria during the study period. Because this study utilized an already existing Health Insurance Portability and Accountability Act (HIPAA)-compliant fully de-identified data, it was exempt from Institutional Review Board (IRB) review.52

Baseline data covariates

Hospital-level variables included region, population served, teaching status and hospital size. Patient-level variables included information on demographics, medical history, hospitalization course, infection characteristics, renal function and medications received. Patient demographics included age, sex, race, primary payer and admission source. Medical history included hospitalization within ≤ 6 months of index admission and Charlson comorbidity index (CCI) (overall score and individual conditions).53 Data collected during the hospital course included hospital length of hospitalization prior to the index CR-GN culture collection day, residence in an ICU on the index CR-GN culture day, and receipt of mechanical ventilation (MV) on the index CR-GN culture day. Infection characteristics included CR-GN pathogen(s) on the index culture and infection type(s). Renal function was assessed using available SCR and eGFR values [calculated with the modification of diet in renal disease (MDRD) equation54] recorded from admission through to 3 days after newer BL/BL-BLI, AG or PB discontinuation. Antibiotic(s) received between admission and the index treatment day, duration of newer BL/BL-BLI, AG or PB treatment, other antibiotics received from the index treatment allocation day through to 3 days after newer BL/BL-BLI, AG or PB discontinuation, and renal-toxic medications (Appendix D) received between admission and the end of treatment with a newer BL/BL-BLI, AG or PB were included to capture patient medication use.

Outcomes

The AKI outcomes were treatment-associated AKI and receipt of any RRT. Treatment-associated AKI incidence proportions (i.e. cumulative incidence) were based on the risk, injury, and failure components of the risk of renal dysfunction, injury to kidney, failure or loss of kidney function, and end-stage kidney disease (RIFLE) criteria51 and it was assessed from Day 2 of newer BL/BL-BLI, AG or PB treatment through to 3 days after newer BL/BL-BLI, AG or PB discontinuation. The following RIFLE classifications were evaluated: (i) RIFLE risk (≥1.5 times increase in index treatment day SCR); (ii) RIFLE injury (≥2 times increase in index treatment day SCR); and (iii) RIFLE failure (≥3 times increase in index treatment day SCR or SCR ≥ 4.0 mg/dL).51 Receipt of any RRT was assessed from the index treatment day to hospital discharge. Clinical outcomes included 30 day mortality, in-hospital mortality, in-hospital mortality or discharge to hospice, and discharge destination (home versus other). Healthcare resource utilization (HRU) outcomes evaluated were hospital length of stay (LOS) from the index treatment day to hospital discharge, hospital costs from the index treatment day to hospital discharge, and 30 day hospital readmissions among survivors.

Statistical methods

Two sets of analyses were performed. We first compared baseline covariates and AKI outcomes between the three treatment groups. As part of the treatment–AKI outcome analyses, we evaluated the association between AKI outcomes and treatment by eGFR at the initiation of treatment (≥60 versus <60 mL/min/1.73 m2). Restricted analyses were also performed in the newer BL/BL-BLI subset to assess the associations between (i) AKI outcomes and treatment by receipt of AG/PB for ≥ 2 days during newer BL/BL-BLI treatment and (ii) AKI outcomes and each newer BL/BL-BLI received. Second, we compared baseline variables and clinical/HRU outcomes between patients who experienced AKI, defined by RIFLE, relative to those who did not.

We conducted unadjusted statistical inferences for each set of analyses. The Student’s t-test was used to compare means between two groups (e.g. AKI versus ‘No AKI’), a one-way ANOVA was used to compare means between more than two groups (e.g. the three treatment groups) and the Kruskal–Wallis test was used to compare continuous distributions non-parametrically between two or more groups. The chi-squared test was used to compare frequencies by groups unless a cell count was <5, wherein the Fisher exact test was used. Kaplan–Meier survival curves were used to compare the time to AKI between treatment groups. We compared survivor functions of the three treatment groups with the log-rank test.

Multivariable regression models were then developed to examine the associations between (i) treatment and occurrence of AKI (risk component of RIFLE criteria) and (ii) occurrence of AKI and clinical/HRU outcomes.51 Logistic regression was used to examine binary outcomes. Generalized linear models with a logarithmic link and a gamma distribution were used to examine continuous outcomes due to the skewed distributions of these outcomes. Inverse probability of weighting (IPW)55,56 was then used to evaluate the associations between the occurrence of AKI (risk component of RIFLE criteria51) and the clinical and economic outcomes, while adjusting for potential baseline confounding variables. For the propensity score model, clinically plausible confounders were selected a priori for adjustment (Appendix E).57 When constructing the propensity model, collinearity was assessed between potential predictors using correlation statistics and the variance inflation factor. As a sensitivity analysis, we also developed multivariable regression models to examine the association between the occurrence of AKI (risk component of RIFLE criteria)51 and the clinical and economic outcomes while adjusting for clinically plausible confounders (Appendix E). All analyses were done using Stata/MP 17.0 for Windows (StataCorp LLC, College Station, TX, USA). P values of <0.05 were considered statistically significant.

Results

During the study period, 750 patients met the study criteria (Table S1). Among the 750 patients, 48% received a newer BL/BL-BLI, 39% received an AG and 19% received a PB. The median (IQR) treatment duration was 8 days (5–11) in the newer BL/BL-BLI group, 5 (4–8) days in the AG group and 7 (4–8) days in the PB group. In the newer BL/BL-BLI group, most patients either received ceftolozane/tazobactam (60.6%) or ceftazidime/avibactam (35.2%). Tobramycin (60.4%), followed by gentamicin (25.3%) were the most frequently administered AG, while 96.8% of patients in the PB group received colistin. The median (IQR) and mean (SD) number of days with SCR values per patient Day 2 of newer BL/BL-BLI, AG or PB treatment through to 3 days after newer BL/BL-BLI, AG or PB discontinuation were 5 (2–8) and 6.4 (6.5) mg/dL, respectively. Table 1 presents comparisons of baseline characteristics between treatment groups. All hospital-level variables were significantly different between treatment groups. For patient-level variables, treatment groups were significantly different with regard to mean age, sex, certain comorbid conditions, CR-GN pathogen, infection type, eGFR at admission and on the index treatment day, other antibiotics received prior to the index treatment day and during treatment, and concurrent receipt of other nephrotoxins. Among the baseline variables that were significantly different between treatment groups, infection type and concurrent treatment with a tetracycline-like antibiotic (i.e. receipt of doxycycline, eravacycline, minocycline, omadacycline, tetracycline and/or tigecycline) were the only ones that were significantly associated with AKI (defined by RIFLE risk criteria).

Table 1.

Baseline stratified by treatment group and presence or absence of AKI defined by risk in RIFLE criteria51

Newer BL/BL-BLI AG PB P value No AKI AKI P value
N = 363 N = 293 N = 94 N = 659 N = 91
Census region
 Midwest 96 (26.5) 40 (13.7) 9 (9.6) 131 (19.9) 14 (15.4)
 Northeast 52 (14.3) 21 (7.2) 19 (20.2) 81 (12.3) 11 (12.1)
 South 212 (58.4) 227 (77.5) 66 (70.2) <0.001 440 (66.8) 65 (71.4) 0.747
 West 3 (0.8) 5 (1.7) 0 (0) 7 (1.1) 1 (1.1)
Number of beds
 <300 86 (23.7) 118 (40.3) 25 (26.7) 203 (30.1) 26 (28.6)
 300 to 499 117 (32.2) 92 (31.4) 40 (42.6) <0.001 212 (32.2) 37 (40.7) 0.255
 500+ 160 (44.1) 83 (28.3) 29 (30.9) 244 (37.0) 28 (30.8)
Teaching hospital 203 (55.9) 84 (28.7) 51 (54.3) <0.001 302 (45.8) 36 (39.6) 0.312
Urban location of hospital 331 (91.2) 234 (79.9) 76 (80.9) <0.001 566 (85.9) 75 (82.4) 0.379
Mean age, years (SD) 63.3 (15.4) 60.2 (16.9) 56.6 (18.3) <0.001 61.3 (16.6) 60.7 (16.0) 0.753
Sex: male 243 (66.7) 170 (58.0) 65 (69.1) 0.031 421 (63.9) 57 (62.6) 0.817
Race
 White 263 (72.4) 221 (75.4) 66 (70.2) 489 (74.2) 61 (67.0)
 Black 68 (18.7) 46 (15.7) 14 (14.9) 0.330 117 (17.8) 11 (12.1) 0.001
 Other 26 (7.2) 24 (8.2) 11 (11.7) 43 (6.5) 18 (19.8)
 Unknown 6 (1.7) 2 (0.7) 3 (3.2) 10 (1.5) 1 (1.1)
Admission source
 Non-healthcare facility (including home) 264 (72.7) 217 (74.1) 68 (72.3) 481 (73.0) 68 (74.7)
 Clinic 16 (4.4) 19 (6.5) 1 (1.1) 0.369 35 (5.3) 1 (1.1)
 Transfer from SNF, ICF 34 (9.4) 22 (7.5) 7 (7.5) 58 (8.8) 5 (5.5) 0.115
 Transfer from another non-acute care facility or other/unknown 49 (13.5) 35 (12.0) 18 (19.2) 85 (12.9) 17 (18.7)
Insurance
 Medicare 244 (67.2) 172 (58.7) 61 (65.0) 414 (62.8) 63 (69.2)
 Medicaid 60 (16.5) 63 (21.5) 21 (22.3) 129 (19.6) 15 (16.5)
 Managed care 42 (11.6) 33 (11.3) 6 (6.4) 0.067 72 (10.9) 9 (9.9) 0.876
 Commercial/Worker’s Comp/Self-Pay 11 (3.0) 14 (4.8) 6 (6.4) 28 (4.3) 3 (3.3)
 Other 6 (1.7) 11 (3.8) 0 (0) 16 (2.4) 17 (18.7)
Hospitalization in 6 months prior to index CR-GN admission 141 (34.9) 99 (32.0) 33 (34.7) 0.393 245 (37.2) 28 (30.8) 0.234
CCI score
 Mean (SD) 3.2 (2.3) 2.9 (2.1) 2.8 (2.1) 0.219 2 (1–4) 3 (2–5) 0.004
 Median (IQR) 3 (2–4) 2 (1–4) 2 (1–4) 0.338 2.9 (2.2) 3.6 (2.3) 0.006
Charlson comorbidities
 Acute myocardial infarction 51 (14.1) 30 (10.2) 14 (14.9) 0.271 77 (11.7) 18 (19.8) 0.030
 Congestive heart failure 118 (32.5) 93 (31.7) 17 (18.1) 0.021 194 (29.4) 34 (37.4) 0.123
 Peripheral vascular disease 41 (11.3) 23 (7.9) 11 (11.7) 0.289 58 (8.8) 17 (18.7) 0.003
 Cerebrovascular disease 45 (12.4) 40 (13.7) 11 (11.7) 0.842 84 (12.8) 12 (13.2) 0.906
 Dementia 46 (12.7) 36 (12.3) 8 (8.5) 0.532 78 (11.8) 12 (13.2) 0.710
 COPD 118 (32.5) 144 (49.2) 43 (45.7) <0.001 258 (39.2) 47 (51.7) 0.023
 Rheumatoid disease 12 (3.3) 8 (2.7) 3 (3.2) 0.914 20 (3.0) 3 (3.3) 0.752
 Peptic ulcer disease 10 (2.8) 7 (2.4) 1 (1.1) 0.754 16 (2.4) 2 (2.2) 1.000
 Mild liver disease 14 (3.9) 22 (7.5) 6 (6.4) 0.122 36 (5.5) 6 (6.6) 0.660
 Diabetes 94 (25.9) 61 (20.8) 27 (28.7) 0.180 157 (23.8) 25 (27.5) 0.447
 Diabetes with complications 67 (18.5) 42 (14.3) 6 (6.4) 0.013 99 (15.0) 17 (18.7) 0.344
 Hemiplegia or paraplegia 70 (19.3) 48 (16.4) 19 (20.2) 0.552 118 (17.9) 19 (20.9) 0.491
 Renal disease 83 (22.9) 45 (15.4) 14 (14.9) 0.029 119 (18.1) 23 (25.3) 0.100
 Cancer 22 (6.1) 20 (6.8) 6 (6.4) 0.924 41 (6.2) 7 (7.7) 0.591
 Moderate/severe liver disease 5 (1.4) 6 (2.1) 3 (3.2) 0.416 12 (1.8) 2 (2.2) 0.683
 Metastatic cancer 16 (4.4) 10 (3.4) 3 (3.2) 0.803 27 (4.1) 2 (2.2) 0.563
 AIDS 2 (0.6) 2 (0.7) 1 (1.1) 0.687 4 (0.6) 1 (1.1) 0.477
LOS prior to index CR-GN culture, days
 Mean (SD) 5.8 (10.7) 6.7 (13.0) 6.7 (10.6) 0.501 5.8 (10.5) 9.6 (17.6) 0.003
 Median (IQR) 1 (1–5) 1 (1–6) 2 (1–8) 0.109 1 (1–5) 2 (1–10) 0.001
Residence in ICU on index CR-GN culture day 146 (44.2) 112 (38.2) 45 (47.9) 0.252 256 (38.9) 47 (51.7) 0.020
MV on index CR-GN culture day 123 (33.9) 89 (20.4) 41 (43.6) 0.061 213 (32.3) 40 (44.0) 0.028
Carbapenem-resistant organisms on index culture
 Escherichia coli 12 (33.3) 9 (3.1) 3 (3.2) 1.000 19 (2.9) 5 (5.5) 0.185
 Enterobacter aerogenes 0 (0) 0 (0) 0 (0) 1.000 0 (0) 0 (0) 1.000
 Enterobacter cloacae 11 (3.0) 12 (4.1) 1 (1.1) 0.378 23 (3.5) 1 (1.1) 0.344
 Klebsiella pneumoniae 73 (20.1) 25 (8.5) 27 (28.7) <0.001 108 (16.4) 17 (18.7) 0.582
 Klebsiella oxytoca 2 (0.6) 1 (0.3) 1 (1.1) 0.608 3 (0.5) 1 (1.1) 0.405
 Morganella morganii 0 (0) 0 (0) 0 (0) 1.000 0 (0) 0 (0) 1.000
 Proteus mirabilis 1 (0.3) 4 (1.4) 1 (1.1) 0.227 6 (0.9) 0 (0) 1.000
 P. aeruginosa 251 (69.2) 228 (77.8) 59 (62.8) 0.006 477 (72.4) 61 (67.0) 0.288
 Serratia marcescens 3 (0.8) 3 (1.0) 0 (0) 1.000 6 (0.9) 0 (0) 1.000
Other Gram negative 17 (4.7) 15 (5.1) 6 (6.4) 0.798 31 (4.7) 7 (7.7) 0.223
Number of CR-GN organisms in the index culture
 1 356 (98.1) 289 (98.6) 90 (95.7) 645 (97.9) 90 (98.9)
 2 7 (1.9) 4 (1.4) 4 (4.3) 0.210 14 (2.1) 1 (1.1) 1.000
 3+ 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)
Infection type
 HABP/VABP 214 (53.0) 203 (65.5) 75 (76.5) <0.001 418 (63.4) 74 (81.3) 0.001
 cUTI 146 (36.1) 102 (32.9) 22 (22.5) 0.008 246 (37.3) 24 (26.4) 0.041
 BSI 80 (19.8) 34 (11.0) 21 (21.4) <0.001 122 (18.5) 13 (14.3) 0.325
Number of CR-GN infection types
 1 290 (72.8) 249 (80.3) 70 (71.4) 537 (81.5) 72 (79.1)
 2 69 (17.1) 42 (13.6) 24 (24.5) 0.109 117 (17.8) 18 (19.8) 0.587
 3 4 (1.0) 2 (0.7) 0 (0) 5 (0.8) 1 (1.1)
eGFR at admission, mL/min/1.73 m2
 Mean (SD) 68.8 (38.9) 70.5 (31.1) 87.6 (67.1) <0.001 72.5 (41.3) 67.2 (30.9) 0.237
 Median (IQR) 60.0 (51.8–81.0) 60.0 (60.0–85.0) 60.0 (60.0–110.0) 0.005 60 (57–88) 60 (53–81) 0.215
SCR at admission, mg/dL
 Mean (SD) 1.1 (0.6) 0.9 (0.6) 1.0 (0.7) 0.067 1.0 (0.6) 1.0 (0.6) 0.872
 Median (IQR) 0.9 (0.6–1.3) 0.8 (0.6–1.1) 0.8 (0.6–1.2) 0.115 0.8 (0.6–1.2) 0.8 (0.6–1.3) 1.000
eGFR on index treatment day
 Mean (SD) 75.1 (34.3) 78.4 (31.5) 106.9 (112.1) <0.001 80.9 (53.2) 75.9 (33.8) 0.383
 Median (IQR) 60.0 (59.5–90.0) 60.0 (60.0–94.0) 60.0 (60.0–113.0) 0.003 60 (60–94) 60 (60–90) 0.500
SCR on index treatment day, mg/dL
 Mean (SD) 0.8 (0.4) 0.7 (0.3) 0.7 (0.4) <0.001 0.8 (0.4) 0.7 (0.4) 0.059
 Median (IQR) 0.8 (0.5–1.0) 0.7 (0.5–0.9) 0.6 (0.4–0.9) 0.003 0.7 (0.5–1.0) 0.6 (0.4–0.9) 0.026
Antibiotics received between admission and index treatment day
 Aminoglycosidea 18 (5.0) 72 (24.6) 11 (11.7) <0.001 88 (13.4) 13 (14.3) 0.807
 β-Lactam (older)b 335 (92.3%) 273 (93.2) 91 (96.8%) 0.300 611 (92.7) 88 (96.7) 0.157
 Fluoroquinolone 62 (17.1) 62 (21.2) 16 (17.0) 0373 122 (18.5) 18 (19.8) 0.771
 Vancomycin 233 (64.2) 172 (58.7) 58 (61.7) 0.356 405 (61.5) 58 (63.7) 0.675
 Daptomycin 10 (2.8) 7 (2.4) 6 (6.4) 0.132 18 (2.7) 5 (5.5) 0.152
 Other glycopeptide/glycopeptide-like agentsc 0 (0) 0 (0) 0 (0) 1.000 0 (0) 0 (0) 1.000
 Macrolided 26 (7.2) 34 (11.6) 12 (12.8) 0.085 66 (10.0) 6 (6.6) 0.299
 Oxazolidone 24 (6.6) 5 (1.7) 6 (6.4) 0.009 31 (4.7) 4 (4.4) 1.000
 Polymyxin 2 (0.6) 4 (1.4) 17 (18.1)e <0.001 18 (2.7) 5 (5.5) 0.184
 Rifamycin 0 (0) 0 (0) 0 (0) 1.000 0 (0) 0 (0) 1.000
 Sulpha-like agentsf 9 (2.5) 10 (3.4) 2 (2.1) 0.741 19 (2.9) 2 (2.2) 1.000
 Tetracycline-like agentsg 22 (6.1) 21 (7.2) 7 (7.5) 0.809 40 (6.1) 10 (11.0) 0.078
Number of antibiotics received between admission and index treatment day
 0 13 (3.6) 15 (5.1) 4 (4.3) 30 (4.6) 2 (2.2)
 1 228 (62.8) 142 (48.5) 43 (45.7) 364 (55.2) 49 (53.9)
 2 97 (26.7) 100 (34.1) 33 (35.1) 0.012 198 (30.1) 32 (35.2) 0.595
 3 19 (5.2) 28 (9.6) 11 (11.7) 53 (8.0) 5 (5.5)
  ≥4 6 (1.7) 8 (2.7) 3 (3.2) 14 (2.1) 3 (3.3)
Receipt of any of the other treatment allocations of interest from index treatment day through to 3 days after newer BL/BL-BLI, AG or PB discontinuation
 Newer BL/BL-BLI NA 38 (13.0) 18 (22.8) NA 46 (7.0) 10 (11.0) 0.173
 AG 70 (19.3) NA 11 (11.7) NA 71 (10.8) 10 (11.0) 0.951
 PB 19 (5.2) 18 (6.1) NA NA 29 (4.4) 8 (8.8) 0.070
Other antibiotics received from index treatment day through to 3 days after newer BL/BL-BLI, AG or PB discontinuation
 β-Lactam (older)b 268 (73.8) 242 (82.6) 79 (84.0) 0.009 520 (78.9) 69 (75.8) 0.502
 Fluoroquinolone 41 (11.3) 66 (22.5) 12 (12.8) <0.001 106 (16.1) 13 (14.3) 0.660
 Vancomycin 163 (44.9) 126 (43.0) 39 (41.5) 0.795 279 (42.3) 49 (53.8) 0.038
 Daptomycin 22 (6.1) 15 (5.1) 6 (6.4) 0.840 39 (5.9) 4 (4.4) 0.809
 Other glycopeptide/glycopeptide-like agentsc 1 (0.3) 0 (0) 0 (0) 1.000 0 (0) 1 (1.1) 0.121
 Macrolided 14 (3.9) 17 (5.8) 6 (6.4) 0.409 33 (5.0) 4 (4.4) 1.000
 Oxazolidone 28 (7.7) 16 (5.5) 5 (5.3) 0.448 41 (6.2) 8 (8.8) 0.352
 Rifamycin 0 (0) 0 (0) 0 (0) 1.000 0 (0) 0 (0) 1.000
 Sulpha-like agentsf 13 (3.6) 13 (4.4) 3 (3.2) 0.833 28 (4.3) 1 (1.1) 0.240
 Tetracycline-like agentsg 32 (8.9) 24 (8.2) 18 (19.2) 0.005 58 (8.8) 16 (17.6) 0.008
Receipt of medications known to cause renal toxicity between admission and end of treatment with a newer BL/BL-BLI, AG or PB 331 (92.1) 293 (100) 80 (85.1) <0.001 616 (93.5) 88 (96.7) 0.229
Receipt of IV contrast media from index treatment day through to 3 days after newer BL/BL-BLI, AG or PB discontinuation 71 (19.6) 32 (10.9) 16 (17.0) 0.010 102 (15.5) 17 (18.7) 0.43
Open in a new tab

SNF, skilled nursing facility; ICF, intermediate care facility. All values are given as n (%) unless otherwise stated.

aAminoglycosides included gentamicin, tobramycin and amikacin.

bOlder β-lactams included amoxicillin/clavulanate, amoxicillin, ampicillin, ampicillin/sulbactam, aztreonam, bacampicillin, carbenicillin, cefaclor, cefadroxil, cefamandole, cefazolin, cefdinir, cefditoren pivoxil, cefepime, cefixime, cefonicid, cefoperazone, cefotaxime, cefotetan, cefoxitin, cefpodoxime, cefprozil, ceftaroline, ceftibuten, ceftizoxime, ceftriaxone, cefalexin, cefapirin, dicloxacillin, doripenem, ertapenem, imipenem, loracarbef, meropenem, mezlocillin, penicillin, piperacillin, piperacillin/tazobactam, ticarcillin/clavulanate and ticarcillin.

cOther glycopeptide/glycopeptide-like agents included dalbavancin, oritavancin and telavancin.

dMacrolides included azithromycin, clarithromycin, erythromycin/sulfisoxazole and quinupristin/dalfopristin.

eMean time from start of PB from admission to index PB treatment day was 1.2 days (SD = 2.8); only two patients in the PB group received a PB for more than 1 day before the index day.

fSulpha-like drugs included sulfamethoxazole, sulfamethoxazole/trimethoprim, trimethoprim and sulfisoxazole.

gTetracycline-like drugs included doxycycline, eravacycline, minocycline, omadacycline and tigecycline.

Unadjusted treatment–AKI outcome comparisons are shown in Table 2. In the overall bivariate analyses, significant associations were noted between treatment and AKI by RIFLE risk (PB: 23.4% versus AG: 10.6% versus newer BL/BL-BLI: 10.5%; P = 0.002) and AKI by RIFLE injury (PB: 10.6% versus AG: 3.1% versus newer BL/BL-BLI: 5.2%; P = 0.014). No significant differences between treatment and RIFLE failure and receipt of RRT were observed. Similar associations were observed between treatment and AKI outcomes within each eGFR category (Table 2). For all three treatment groups, patients who experienced AKI (RIFLE risk) had longer treatment durations relative to those who did not experience AKI (Figure S1). For patients in the newer BL/BL-BLI group who received PB for ≥2 days (n = 15), 5 (33%) developed AKI (RIFLE risk) versus 9.5% (33 out of 38) who received neither PB nor AG concurrently (Table S2). No differences in AKI outcomes were observed within the newer BL/BL-BLI treatments (Table S3). Eighteen patients (22.8%) in the PB group received a newer BL/BL-BLI. Among those who received a newer BL/BL-BLI, the incidence of AKI was 16.7% (3/18). Among those who did not receive a newer BL/BL-BLI, the incidence of AKI was 25.0% (19/76). The P value comparing the two AKI incidence proportions was P = 0.550.

Table 2.

Unadjusted comparisons of AKI per RIFLE criteria51 between treatment groups

Newer BL/BL-BLI (N = 363) AG (N = 293) PB (N = 94) P value
Overall (N = 750)
 RIFLE risk 38 (10.47) 31 (10.58) 22 (23.40) 0.002
 RIFLE injury 19 (5.23) 9 (3.07) 10 (10.64) 0.014
 RIFLE failure 5 (1.38) 3 (1.02) 4 (4.26) 0.090
 Receipt of RRT 6 (1.65) 4 (1.37) 2 (2.13) 0.845
eGFR < 60 mL/min/1.73 m2 on index treatment day (n = 160) n = 96 n = 56 n = 18
 RIFLE risk 9 (9.38) 7 (15.22) 3 (16.67) 0.472
 RIFLE injury 3 (3.13) 1 (2.17) 1 (5.56) 0.633
 RIFLE failure 0 (0.00) 0 (0.00) 1 (5.56) 0.113
 Receipt of any RRT 1 (1.04) 1 (2.17) 2 (11.11) 0.053
eGFR ≥ 60 mL/min/1.73 m2 on index treatment day (n = 590) n = 267 n = 247 n = 76
 RIFLE risk 29 (10.86) 24 (9.72) 19 (25.00) 0.001
 RIFLE injury 16 (5.99) 8 (3.24) 9 (11.84) 0.016
 RIFLE failure 5 (1.87) 3 (1.21) 3 (3.95) 0.262
 Receipt of any RRT 5 (1.87) 3 (1.21) 0 (0.00) 0.693
Open in a new tab

All values are given as n (%). The following RIFLE classifications were evaluated: (1) RIFLE risk (increase in index treatment day SCR by ≥1.5 times); (2) RIFLE injury (increase in index treatment day SCR by ≥2 times); and (3) RIFLE failure (increase in index treatment day SCR by ≥3 times, or SCR ≥ 4.0 mg/dL).51

Comparison of the Kaplan–Meier time-to-AKI (RIFLE risk) curves between treatment groups (overall and by eGFR category on the index treatment day) are shown in Figure 1. In the Kaplan–Meier analyses (Figure 1), significant differences in time to AKI (RIFLE risk) between treatment groups were observed overall and within each eGFR category (P < 0.001 for each plot). Results of the adjusted treatment–AKI (RIFLE risk) logistic regression analyses are shown in Figure 2. Consistent with the unadjusted analyses, patients in the PB group had the highest adjusted AKI incidence (95% CIs) of RIFLE risk overall: PB: 25.1% (15.6%–34.6%) versus AG: 8.9% (5.7%–12.2%) versus newer BL/BL-BLI: 11.9% (8.1%–15.7%); P = 0.001.

Figure 1.

Figure 1.

Open in a new tab

Time to treatment-associated AKI defined by risk in the RIFLE criteria51 for (a) all patients (n = 750), (b) patients with eGFR < 60 mL/min/1.73 m2 on the index treatment day (n = 160) and (c) patients with eGFR > 60 mL/min/1.73 m2 on the index treatment day (n = 590). Patients in Kaplan–Meier plots were censored on AKI day or last day of treatment, whichever came first.

Figure 2.

Figure 2.

Open in a new tab

Adjusted probability of treatment-associated AKI defined by risk in the RIFLE criteria51 stratified by eGFR value on the index treatment day.

Unadjusted and IPW comparisons of clinical and HRU outcomes between patients who experienced AKI (RIFLE risk) relative to those who did not experience AKI are shown in Table 3. Patients with AKI had higher incidences of mortality, and increased hospital LOS and hospital costs compared with patients who did not experience AKI (RIFLE risk). No differences were noted by AKI status with regard to discharge destination (home versus other) or 30 day readmissions. Results of the multivariable regression analyses that examined the association between AKI (risk component of RIFLE criteria) and the study outcomes were consistent with the IPW comparisons (Table S4).

Table 3.

Unadjusted and Inverse probability of weighting (IPW)-adjusted comparisons of clinical and healthcare resource utilization outcomes between patients who experienced treatment-associated AKI defined by risk in the RIFLE criteria51 relative to patients who did not experience treatment-associated AKI

Unadjusted comparisons IPW-adjusted comparisons
No treatment-associated AKI (n = 659) Treatment-associated AKI (n = 91) P value No treatment-associated AKI (n = 659) Treatment-associated AKI (n = 91) P value
30 Day mortality, % (95% CI) 4.4 (3.0–6.3) 17.6 (10.4–27.0) <0.001 4.5 (2.9–6.1) 16.7 (9.4–24.1) 0.001
In-hospital mortality, % (95% CI) 5.3 (3.7–7.3) 20.9 (13.1–30.7) <0.001 5.6 (3.8–7.3) 18.5 (10.8–26.2) 0.001
In-hospital mortality or discharge to hospice, % (95% CI) 11.2 (8.9–13.9) 30.8 (21.5–41.3) <0.001 11.7 (9.3–14.2) 26.7 (17.9–35.4) 0.001
Discharge destination among all patients: home versus other, % (95% CI) 32.9 (29.3–36.7) 26.4 (17.7–36.7) 0.209 32.0 (28.5–35.5) 29.5 (18.9–40.0) 0.651
Mean hospital LOS (days) from index treatment day to hospital discharge, % (95% CI)a 14.3 (13.1–15.5) 25.8 (11.4–40.3) <0.001 14.8 (13.4–16.2) 18.5 (15.2–21.8) 0.032
Mean hospital costs from index treatment day to hospital discharge (95% CI)a $37 027 ($32 531–$41 523) $68 046 ($37 154–$98 937) <0.001 $38 763 ($33 959–$43 566) $49 192 ($39 697–$58 686) 0.043
30 Day readmission among survivors, % (95% CI) 24.7 (21.3–28.3) 26.4 (16.7–38.1) 0.751 24.7 (21.3–28.0) 29.8 (18.4–41.2) 0.396
Open in a new tab

Baseline covariates included as predictor variables included US census region, number of hospital beds, type of hospital (teaching versus non-teaching), age, sex, race (white versus other), admission source (home/community versus other), history of hospitalization in the 6 months prior to index culture, baseline eGFR category, eGFR on index treatment day, CCI score, hospital LOS prior to the index CR-GN culture, residence in ICU on the day of the index CR-GN culture, receipt of MV on the day of the index CR-GN culture, CR-GN pathogen on the day of index culture (P. aeruginosa versus Klebsiella sp. versus other), infection type (HABP/VABP versus cUTI versus BSI), the treatment groups (newer BL/BL-BLI, AG or PB), antibiotics received between admission and index treatment day (aminoglycoside versus β-lactam versus vancomycin versus other) and receipt of other treatment allocation agents (newer BL/BL-BLI, AG or PB) from index β-lactam, aminoglycoside or polymyxin treatment through to 3 days after β-lactam, aminoglycoside or polymyxin discontinuation.

aIndex treatment day was first day of treatment with a newer BL/BL-BLI-, AG- or PB-containing regimen.

Discussion

In this study of adult hospital patients with CR-GNIs at high risk for AKI due to their underlying disease severity, comorbidities and concomitant receipt of other potentially nephrotoxic medications,16–21 the adjusted AKI incidence51 was considerably higher among patients who received a PB-containing regimen versus patients who received a newer BL/BL-BLI- or AG-containing regimen. The overall PB-associated AKI incidence proportion of 23.4% observed in this study was comparable or lower to the AKI proportions reported in other studies that assessed the incidence of nephrotoxicity among patients who received a PB intravenously.36,39–41,43–46,58–63 In a systematic review of the estimated incidences of nephrotoxicity in patients treated with systemic PB, the observed nephrotoxicity incidence proportion among studies (103 studies: 21 451 patients) using the RIFLE criteria to define AKI was 38.8% (95% CI: 35.8%–42.0%).41 Additionally, fewer PB patients in this study developed severe AKI relative to other studies.36,41

Several factors potentially contributed to the lower incidence of PB-associated AKI observed here. This study only included patients with eGFR ≥ 30 mL/min/1.73 m2 at admission and was limited to the index CR-GNI episode during hospitalization. We also excluded patients who recently received any RRT or had an CR-GNI due to carbapenem-resistant A. baumannii or S. maltophilia. Consequently, few patients in the PB group (23.4%) had eGFR ≤ 60 mL/min/1.73 m2, the average age of PB patients was 56.6 years, and fewer than half of patients in the PB group were in the ICU on the index CR-GNI culture day. In contrast, most other studies that evaluated the association between PB treatment and AKI included older, more acutely ill patients with worse baseline renal function.36,39–41,43–46,58–63 Patients received, on average, 7 days of PB-based therapy in this study, and data indicate that PB-associated AKI increases as a function of treatment duration.62,63 Finally, the observed associations between PB and AKI relative to non-PB-based therapies on the relative scale were nearly identical between this study (RR: 2.16; 95% CI: 1.31–3.55; OR: 2.71; 95% CI: 1.36–5.40) and a recent large-scale meta-analysis (OR: 2.23; 95% CI: 1.58–3.15).41 This finding further suggests that baseline and treatment duration differences likely accounted for the lower PB AKI incidence observed in this study relative to others.64

Another notable finding in this study was the comparable AKI incidences between patients who received a newer BL/BL-BLI- or AG-containing regimen. AG-associated AKI in this study was found to be 6%–25% lower than that observed in other recent studies that assessed the association between AG treatment and AKI among patients with CR-GNIs.39,58,65 In the pathogen-specific Combating Antibiotic-Resistant Enterobacteriaceae (CARE) trial, 16.7% of patients who received plazomicin had AKI, defined as a ≥0.5 mg/dL increase in SCR.58 Similarly, Pogue et al.39 reported that 23% of patients who received an AG-containing regimen for their drug-resistant P. aeruginosa infection experienced AKI, defined by RIFLE criteria, relative to 6% of those who received ceftolozane/tazobactam. It is likely similar factors (e.g. younger population, low proportion of patients with baseline eGFR < 60 mL/min/1.73 m2 or residence in ICU on the index CR-GNI culture collection date) that accounted for the lower observed PB-associated AKI incidence in this study also contributed to the lower AG-associated AKI incidence proportion. The median duration of AG use in this study was 5 days and available clinical data indicate >5–7 days of AG therapy are needed to cause AKI.66–68 As shown in Figure 1, there was an accelerated increase in AKI among patients receiving AG for >7 days relative to those receiving a newer BL/BL-BLI-containing regimen.

Although β-lactams can cause AKI,69 we anticipate the adjusted AKI incidence of 11.9% (95% CI: 8.1%–15.7%) in the newer BL/BL-BLI group was most likely due to their underlying risk factors for AKI16–21 versus a true drug-associated adverse event.70,71 Notably, the median treatment duration in the newer β-lactam group was 8 days and most AKI events occurred after 10 days of therapy (Figure S1). This finding suggests the observed AKI incidence in the newer BL/BL-BLI group was more likely related to their underlying disease severity and comorbidities that augmented their risk of experiencing an AKI over time.70,71 Furthermore, nearly all (92%) patients in the BL/BL-BLI group received other potentially nephrotoxic medications (Table 1) and we speculate this contributed to the observed adjusted AKI incidence in the BL/BL-BLI group. Of note, the factors that likely contributed to the observed AKI cumulative incidence in the BL/BL-BLI group were also responsible, in part, for the observed AKI incidences in the PB and AG groups.16–21 As such, the proportion of the observed treatment-associated AKI that was actually attributable to either BL/BL-BLI, AG or PB (i.e. attributable AKI risk percentage) in this study was likely less than the observed AKI incidence.

The findings from the AKI–outcomes analyses have important implications for antimicrobial stewardship. Although stewardship encompasses a range of coordinated activities to promote the prudent use of antimicrobials, one of its most critical functions is to ensure patients with serious infections receive effective therapies, in a timely fashion, that maximize clinical response with minimal adverse events.72,73 Consistent with other studies,35,40,43–46,74,75 we observed that patients with CR-GNIs who experience AKI have a 2- to 3-fold increase in mortality, 4–5 day longer hospital stay and additional hospital costs in excess of $11 000 USD. From a clinical perspective, it is important to recognize that SCR is a crude biomarker that only increases after a substantial amount of kidney injury has already occurred.76 Because of renal reserve, it is estimated that patients may lose up to 50% of nephrons before SCR increases and AKI cannot be averted by close SCR monitoring.76,77 Therefore, clinicians and antimicrobial stewards should minimize the use of agents associated with an increased risk of AKI whenever possible as a measure to optimize patient-centric outcomes. From a clinical pathway/formulary perspective, antibiotics that augment the risk of AKI should be designated as non-preferred or second-line agents in patients with CR-GNIs and their use should only be considered in clinical situations when the benefits outweigh the risks (e.g. patient failed a first-line agent that has a lower potential to cause AKI).

Several limitations should be noted when interpreting the findings of this study. First are those inherent to this study design, which include study selection bias, unmeasured confounding, and confounding by indication. Study design restrictions, stratified analyses and multivariable methods (IPW, multivariable regression modelling) were used to minimize the influence of the potential systematic biases.78 For example, our study design produced treatment groups that were largely comparable at baseline. Infection type and concurrent treatment with a tetracycline-like antibiotic were the only baseline variables that were significantly different between treatment groups [i.e. hospital-acquired bacterial pneumonia/ventilator-associated bacterial pneumonia (HABP/VABP) and concurrent tetracycline use were more pronounced in PB patients) and significantly associated with AKI. The persistence of findings in the multivariable analyses after accounting for these and other baseline variables suggests confounding likely had a minimal impact on the observed exposure–outcome findings. However, these methods cannot account for unmeasured confounders between groups and some degree of caution should be exercised when interpreting the results.

Data on antibiotic dosing and drug concentrations were not available in the PHD. As such, we were unable to assess the potential impact of drug exposure or dosing regimens received on the observed treatment-associated AKI cumulative proportions. Patients with infections due to carbapenem-resistant A. baumannii or S. maltophilia were excluded since most of the newer agents apart from cefiderocol have reliable in vitro activity against these CR-GNs. As such, it is unknown how applicable these findings are to other populations that were not included in the study. Third, we were unable to assess the effect of time to receipt of a microbiologically active agent because susceptibility data were not available in many patients and the impact of delayed appropriate therapy on the observed outcomes could not be determined as part of this study. Fourth, potential misclassification may be present from using diagnosis codes, microbiological culture and treatment data to define patients with HABP/VABP and cUTI due to CR-GNI. In addition, some HABP/VABP cases may have been missed as pneumonia and are often undercoded in hospitalized ICU patients.79–81 Since there are no specific codes for cUTI, a composite case definition was utilized based on a previous study.82,83 However, the codes used to define the study cohort have been previously validated to have high positive predictive values.84–88

In conclusion, there are two critical components of optimal antimicrobial chemotherapy. First, therapy should be efficacious; second, therapy should be non-toxic. A potentially serious adverse effect associated with some of the agents used for patients with CR-GNIs, namely the PB, is AKI.36–46 Although caution should be exercised when interpreting the findings due to the observational nature of the study, we observed that patients who received PB had a significantly higher incidence of AKI. Furthermore, patients with CR-GNIs who experienced treatment-associated AKI defined by the RIFLE criteria had substantial increases in morbidity, mortality and healthcare resource utilization, with each AKI event resulting in excess hospital costs of ≥11 000 USD. As data indicate AKI cannot be averted by close SCR monitoring,76,77 clinicians should minimize use of agents associated with an increased risk of AKI whenever possible.

Supplementary Material

dkad351_Supplementary_Data
Click here for additional data file. (39.2KB, docx)

Contributor Information

Thomas P Lodise, Albany College of Pharmacy and Health Sciences, Department of Pharmacy Practice, 106 New Scotland Avenue, Albany, NY, USA.

Emre Yucel, Merck & Co., Inc., 2025 E Scott Ave, Rahway, NJ, USA.

Engels N Obi, Merck & Co., Inc., 2025 E Scott Ave, Rahway, NJ, USA.

Alexandre H Watanabe, Merck & Co., Inc., 2025 E Scott Ave, Rahway, NJ, USA.

Brian H Nathanson, OptiStatim LLC, 25 Willow Circle, Longmeadow, MA, USA.

Funding

Funding was provided for this project by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA.

Transparency declarations

T.P.L.: AbbVie (Consultant), BioFire Diagnostics (Grant/Research Support), Cidara (Advisor/Consultant), Entasis (Grant/Research Support), Ferring (Advisor/Consultant/Speaker), Genentech (Consultant), ICPD (Consultant), Johnson & Johnson (Consultant), Melinta (Advisor/Consultant), Merck (Advisor/Consultant, Grant/Research Support), Paratek (Advisor/Consultant), Roche (Consultant), Shionogi (Advisor/Consultant/Speaker), Spero (Advisor/Consultant), Wockhardt (Grant/Research Support) and Venatorx (Advisor/Consultant). B.H.N.: Merck (Consultant), EviMed (Consultant). E.Y., E.N.O. and A.H.W. are employees of Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA.

Author contributions

T.P.L. and B.H.N. led the development of the research question, study design, implementation of the study protocol, analysis and interpretation of data, and drafting the report, along with E.Y.., E.N.O. and A.H.W. All authors provided critical reviews and final approval of the manuscript.

Supplementary data

Figure S1, Tables S1 to S4 and Appendices A to E are available as Supplementary data at JAC Online.

References

  • 1. Haque M, Sartelli M, McKimm Jet al. Health care-associated infections—an overview. Infect Drug Resist 2018; 11: 2321–33. 10.2147/IDR.S177247 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Centers for Disease Control and Prevention (U.S.) . CDC Winnable battles final report. 2016. https://stacks.cdc.gov/view/cdc/43072
  • 3. Weiner LM, Webb AK, Limbago Bet al. Antimicrobial-resistant pathogens associated with healthcare-associated infections: summary of data reported to the National Healthcare Safety Network at the Centers for Disease Control and Prevention, 2011–2014. Infect Control Hosp Epidemiol 2016; 37: 1288–301. 10.1017/ice.2016.174 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. CDC . Antibiotic Resistance Threats in the United States, 2019.2019. https://www.cdc.gov/drugresistance/pdf/threats-report/2019-ar-threats-report-508.pdf.
  • 5. CDC . Antibiotic Resistance Threats in the United States, 2013. 2013. http://www.cdc.gov/drugresistance/threat-report-2013/pdf/ar-threats-2013-508.pdf.
  • 6. Tacconelli E, Carrara E, Savoldi Aet al. Discovery, research, and development of new antibiotics: the WHO priority list of antibiotic-resistant bacteria and tuberculosis. Lancet Infect Dis 2018; 18: 318–27. 10.1016/S1473-3099(17)30753-3 [DOI] [PubMed] [Google Scholar]
  • 7. Falagas ME, Tansarli GS, Karageorgopoulos DEet al. Deaths attributable to carbapenem-resistant Enterobacteriaceae infections. Emerg Infect Dis 2014; 20: 1170–5. 10.3201/eid2007.121004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Bogan C, Kaye KS, Chopra Tet al. Outcomes of carbapenem-resistant Enterobacteriaceae isolation: matched analysis. Am J Infect Control 2014; 42: 612–20. 10.1016/j.ajic.2014.02.013 [DOI] [PubMed] [Google Scholar]
  • 9. Zilberberg MD, Nathanson BH, Sulham Ket al. Carbapenem resistance, inappropriate empiric treatment and outcomes among patients hospitalized with Enterobacteriaceae urinary tract infection, pneumonia and sepsis. BMC Infect Dis 2017; 17: 279. 10.1186/s12879-017-2383-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Lodise TP, Berger A, Altincatal Aet al. Antimicrobial resistance or delayed appropriate therapy—does one influence outcomes more than the other among patients with serious infections due to carbapenem-resistant versus carbapenem-susceptible Enterobacteriaceae? Open Forum Infect Dis 2019; 6: ofz194. 10.1093/ofid/ofz194 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Bonine NG, Berger A, Altincatal Aet al. Impact of delayed appropriate antibiotic therapy on patient outcomes by antibiotic resistance status from serious Gram-negative bacterial infections. Am J Med Sci 2019; 357: 103–10. 10.1016/j.amjms.2018.11.009 [DOI] [PubMed] [Google Scholar]
  • 12. Martin A, Fahrbach K, Zhao Qet al. Association between carbapenem resistance and mortality among adult, hospitalized patients with serious infections due to Enterobacteriaceae: results of a systematic literature review and meta-analysis. Open Forum Infect Dis 2018; 5: ofy150. 10.1093/ofid/ofy150 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Martirosov DM, Lodise TP. Emerging trends in epidemiology and management of infections caused by carbapenem-resistant Enterobacteriaceae. Diagn Microbiol Infect Dis 2016; 85: 266–75. 10.1016/j.diagmicrobio.2015.10.008 [DOI] [PubMed] [Google Scholar]
  • 14. WHO . Global Priority List of Antibiotic-Resistant Bacteria to Guide Research, Discovery, and Development of New Antibiotics. https://remed.org/wp-content/uploads/2017/03/lobal-priority-list-of-antibiotic-resistant-bacteria-2017.pdf
  • 15. Palacios-Baena ZR, Giannella M, Manissero Det al. Risk factors for carbapenem-resistant Gram-negative bacterial infections: a systematic review. Clin Microbiol Infect 2021; 27: 228–35. 10.1016/j.cmi.2020.10.016 [DOI] [PubMed] [Google Scholar]
  • 16. Kadri SS, Adjemian J, Lai YLet al. Difficult-to-treat resistance in Gram-negative bacteremia at 173 US hospitals: retrospective cohort analysis of prevalence, predictors, and outcome of resistance to all first-line agents. Clin Infect Dis 2018; 67: 1803–14. 10.1093/cid/ciy378 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Lakbar I, Medam S, Ronfle Ret al. Association between mortality and highly antimicrobial-resistant bacteria in intensive care unit-acquired pneumonia. Sci Rep 2021; 11: 16497. 10.1038/s41598-021-95852-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Tabah A, Buetti N, Staiquly Qet al. Epidemiology and outcomes of hospital-acquired bloodstream infections in intensive care unit patients: the EUROBACT-2 international cohort study. Intensive Care Med 2023; 49: 178–90. 10.1007/s00134-022-06944-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Tabah A, Koulenti D, Laupland Ket al. Characteristics and determinants of outcome of hospital-acquired bloodstream infections in intensive care units: the EUROBACT International Cohort Study. Intensive Care Med 2012; 38: 1930–45. 10.1007/s00134-012-2695-9 [DOI] [PubMed] [Google Scholar]
  • 20. Vincent JL, Rello J, Marshall Jet al. International study of the prevalence and outcomes of infection in intensive care units. JAMA 2009; 302: 2323–9. 10.1001/jama.2009.1754 [DOI] [PubMed] [Google Scholar]
  • 21. Vincent JL, Sakr Y, Singer Met al. Prevalence and outcomes of infection among patients in intensive care units in 2017. JAMA 2020; 323: 1478–87. 10.1001/jama.2020.2717 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Abdalrahim MS, Khalil AA, Alramly Met al. Pre-existing chronic kidney disease and acute kidney injury among critically ill patients. Heart Lung 2020; 49: 626–9. 10.1016/j.hrtlng.2020.04.013 [DOI] [PubMed] [Google Scholar]
  • 23. Bagshaw SM, George C, Dinu Iet al. A multi-centre evaluation of the RIFLE criteria for early acute kidney injury in critically ill patients. Nephrol Dial Transplant 2008; 23: 1203–10. 10.1093/ndt/gfm744 [DOI] [PubMed] [Google Scholar]
  • 24. Bouchard J, Acharya A, Cerda Jet al. A prospective international multicenter study of AKI in the intensive care unit. Clin J Am Soc Nephrol 2015; 10: 1324–31. 10.2215/CJN.04360514 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Hong D, Ren Q, Zhang Jet al. A new criteria for acute on preexisting kidney dysfunction in critically ill patients. Ren Fail 2023; 45: 2173498. 10.1080/0886022X.2023.2173498 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Hoste EA, Bagshaw SM, Bellomo Ret al. Epidemiology of acute kidney injury in critically ill patients: the multinational AKI-EPI study. Intensive Care Med 2015; 41: 1411–23. 10.1007/s00134-015-3934-7 [DOI] [PubMed] [Google Scholar]
  • 27. Hoste EAJ, Kellum JA, Selby NMet al. Global epidemiology and outcomes of acute kidney injury. Nat Rev Nephrol 2018; 14: 607–25. 10.1038/s41581-018-0052-0 [DOI] [PubMed] [Google Scholar]
  • 28. Kellum JA, Prowle JR. Paradigms of acute kidney injury in the intensive care setting. Nat Rev Nephrol 2018; 14: 217–30. 10.1038/nrneph.2017.184 [DOI] [PubMed] [Google Scholar]
  • 29. Lebiedz P, Knickel L, Engelbertz Cet al. Impact of preexisting chronic kidney disease on acute and long-term outcome of critically ill patients on a medical intensive care unit. J Nephrol 2014; 27: 73–80. 10.1007/s40620-013-0016-1 [DOI] [PubMed] [Google Scholar]
  • 30. Pickkers P, Darmon M, Hoste Eet al. Acute kidney injury in the critically ill: an updated review on pathophysiology and management. Intensive Care Med 2021; 47: 835–50. 10.1007/s00134-021-06454-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Singbartl K, Kellum JA. AKI In the ICU: definition, epidemiology, risk stratification, and outcomes. Kidney Int 2012; 81: 819–25. 10.1038/ki.2011.339 [DOI] [PubMed] [Google Scholar]
  • 32. Arulkumaran N, Annear NM, Singer M. Patients with end-stage renal disease admitted to the intensive care unit: systematic review. Br J Anaesth 2013; 110: 13–20. 10.1093/bja/aes401 [DOI] [PubMed] [Google Scholar]
  • 33. CDC . Chronic Kidney Disease in the United States, 2019.2019. https://www.cdc.gov/kidneydisease/pdf/2019_National-Chronic-Kidney-Disease-Fact-Sheet.pdf.
  • 34. National Kidney Foundation . K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis 2002; 39: S1–266. https://www.kidney.org/sites/default/files/docs/ckd_evaluation_classification_stratification.pdf [PubMed] [Google Scholar]
  • 35. Chertow GM, Burdick E, Honour Met al. Acute kidney injury, mortality, length of stay, and costs in hospitalized patients. J Am Soc Nephrol 2005; 16: 3365–70. 10.1681/ASN.2004090740 [DOI] [PubMed] [Google Scholar]
  • 36. Sisay M, Hagos B, Edessa Det al. Polymyxin-induced nephrotoxicity and its predictors: a systematic review and meta-analysis of studies conducted using RIFLE criteria of acute kidney injury. Pharmacol Res 2021; 163: 105328. 10.1016/j.phrs.2020.105328 [DOI] [PubMed] [Google Scholar]
  • 37. Nation RL, Rigatto MHP, Falci DRet al. Polymyxin acute kidney injury: dosing and other strategies to reduce toxicity. Antibiotics (Basel) 2019; 8: 24. 10.3390/antibiotics8010024 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. Bland CM, Pai MP, Lodise TP. Reappraisal of contemporary pharmacokinetic and pharmacodynamic principles for informing aminoglycoside dosing. Pharmacotherapy 2018; 38: 1229–38. 10.1002/phar.2193 [DOI] [PubMed] [Google Scholar]
  • 39. Pogue JM, Kaye KS, Veve MPet al. Ceftolozane/tazobactam versus polymyxin or aminoglycoside-based regimens for the treatment of drug-resistant Pseudomonas aeruginosa. Clin Infect Dis 2020; 71: 304–10. 10.1093/cid/ciz816 [DOI] [PubMed] [Google Scholar]
  • 40. Rigatto MH, Behle TF, Falci DRet al. Risk factors for acute kidney injury (AKI) in patients treated with polymyxin B and influence of AKI on mortality: a multicentre prospective cohort study. J Antimicrob Chemother 2015; 70: 1552–7. 10.1093/jac/dku561 [DOI] [PubMed] [Google Scholar]
  • 41. Wagenlehner F, Lucenteforte E, Pea Fet al. Systematic review on estimated rates of nephrotoxicity and neurotoxicity in patients treated with polymyxins. Clin Microbiol Infect 2021; 27: 671–86. 10.1016/j.cmi.2020.12.009 [DOI] [PubMed] [Google Scholar]
  • 42. Wagenlehner FME, Cloutier DJ, Komirenko ASet al. Once-daily plazomicin for complicated urinary tract infections. N Engl J Med 2019; 380: 729–40. 10.1056/NEJMoa1801467 [DOI] [PubMed] [Google Scholar]
  • 43. Baradaran S, Black DJ, Keyloun KRet al. The impact of acute kidney injury on the risk of mortality and health care utilization among patients treated with polymyxins for severe gram-negative infections. Open Forum Infect Dis 2018; 5: ofy191. 10.1093/ofid/ofy191 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44. Durante-Mangoni E, Andini R, Signoriello Set al. Acute kidney injury during colistin therapy: a prospective study in patients with extensively-drug resistant Acinetobacter baumannii infections. Clin Microbiol Infect 2016; 22: 984–9. 10.1016/j.cmi.2016.08.004 [DOI] [PubMed] [Google Scholar]
  • 45. Miano TA, Lautenbach E, Wilson FPet al. Attributable risk and time course of colistin-associated acute kidney injury. Clin J Am Soc Nephrol 2018; 13: 542–50. 10.2215/CJN.06980717 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46. Doremus C, Marcella SW, Cai Bet al. Utilization of colistin versus β-lactam and β-lactamase inhibitor agents in relation to acute kidney injury in patients with severe Gram-negative infections. Infect Dis Ther 2022; 11: 187–99. 10.1007/s40121-021-00556-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47. Tamma PD, Aitken SL, Bonomo RAet al. Infectious Diseases Society of America 2022 guidance on the treatment of extended-spectrum β-lactamase producing Enterobacterales (ESBL-E), carbapenem-resistant Enterobacterales (CRE), and Pseudomonas aeruginosa with difficult-to-treat resistance (DTR-P. aeruginosa). Clin Infect Dis 2022; 75: 187–212. 10.1093/cid/ciac268 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48. Tamma PD, Aitken SL, Bonomo RAet al. Infectious Diseases Society of America guidance on the treatment of AmpC β-lactamase-producing Enterobacterales, carbapenem-resistant Acinetobacter baumannii, and Stenotrophomonas maltophilia infections. Clin Infect Dis 2022; 74: 2089–114. 10.1093/cid/ciab1013 [DOI] [PubMed] [Google Scholar]
  • 49. Paul M, Carrara E, Retamar Pet al. European Society of Clinical Microbiology and Infectious Diseases (ESCMID) guidelines for the treatment of infections caused by multidrug-resistant Gram-negative bacilli (endorsed by European Society of Intensive Care Medicine). Clin Microbiol Infect 2022; 28: 521–47. 10.1016/j.cmi.2021.11.025 [DOI] [PubMed] [Google Scholar]
  • 50. Strich JR, Warner S, Lai YLet al. Needs assessment for novel Gram-negative antibiotics in US hospitals: a retrospective cohort study. Lancet Infect Dis 2020; 20: 1172–81. 10.1016/S1473-3099(20)30153-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51. Bellomo R, Ronco C, Kellum JAet al. Acute renal failure—definition, outcome measures, animal models, fluid therapy and information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care 2004; 8: R204–12. 10.1186/cc2872 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52. PINC. White Paper. PINC AI™ Healthcare Database: Data That Informs and Performs. July 2023. https://offers.pinc-ai.com/PINC-AI-Healthcare-Database-White-Paper-LP.html
  • 53. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992; 45: 613–9. 10.1016/0895-4356(92)90133-8 [DOI] [PubMed] [Google Scholar]
  • 54. Levey AS, Coresh J, Greene Tet al. Using standardized serum creatinine values in the Modification of Diet in Renal Disease Study equation for estimating glomerular filtration rate. Ann Intern Med 2006; 145: 247–54. 10.7326/0003-4819-145-4-200608150-00004 [DOI] [PubMed] [Google Scholar]
  • 55. Cattaneo MD. Efficient semiparametric estimation of multi-valued treatment effects under ignorability. J Econom 2010; 155: 138–54. 10.1016/j.jeconom.2009.09.023 [DOI] [Google Scholar]
  • 56. Chesnaye NC, Stel VS, Tripepi Get al. An introduction to inverse probability of treatment weighting in observational research. Clin Kidney J 2022; 15: 14–20. 10.1093/ckj/sfab158 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57. Harrell FE. Regression Modeling Strategies: With Applications to Linear Models, Logistic Regression, and Survival Analysis. Springer, 2001. [Google Scholar]
  • 58. McKinnell JA, Dwyer JP, Talbot GHet al. Plazomicin for infections caused by carbapenem-resistant Enterobacteriaceae. N Engl J Med 2019; 380: 791–3. 10.1056/NEJMc1807634 [DOI] [PubMed] [Google Scholar]
  • 59. Motsch J, de Oliveira C M, Stus Vet al. RESTORE-IMI 1: a multicenter, randomized, double-blind trial comparing efficacy and safety of imipenem/relebactam versus colistin plus imipenem in patients with imipenem-nonsusceptible bacterial infections. Clin Infect Dis 2020; 70: 1799–808. 10.1093/cid/ciz530 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60. Wunderink RG, Giamarellos-Bourboulis EJ, Rahav Get al. Effect and safety of meropenem-vaborbactam versus best-available therapy in patients with carbapenem-resistant Enterobacteriaceae infections: the TANGO II randomized clinical trial. Infect Dis Ther 2018; 7: 439–55. 10.1007/s40121-018-0214-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61. van Duin D, Lok JJ, Earley Met al. Colistin versus ceftazidime-avibactam in the treatment of infections due to carbapenem-resistant Enterobacteriaceae. Clin Infect Dis 2018; 66: 163–71. 10.1093/cid/cix783 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62. Phe K, Lee Y, McDaneld PMet al. In vitro assessment and multicenter cohort study of comparative nephrotoxicity rates associated with colistimethate versus polymyxin B therapy. Antimicrob Agents Chemother 2014; 58: 2740–6. 10.1128/AAC.02476-13 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 63. Sorli L, Luque S, Grau Set al. Trough colistin plasma level is an independent risk factor for nephrotoxicity: a prospective observational cohort study. BMC Infect Dis 2013; 13: 380. 10.1186/1471-2334-13-380 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64. Rothman KJ, Greenland S. Causation and causal inference in epidemiology. Am J Public Health 2005; 95Suppl 1: S144–50. 10.2105/AJPH.2004.059204 [DOI] [PubMed] [Google Scholar]
  • 65. Shields RK, Clancy CJ, Press EGet al. Aminoglycosides for treatment of bacteremia due to carbapenem-resistant Klebsiella pneumoniae. Antimicrob Agents Chemother 2016; 60: 3187–92. 10.1128/AAC.02638-15 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66. Drusano GL, Ambrose PG, Bhavnani SMet al. Back to the future: using aminoglycosides again and how to dose them optimally. Clin Infect Dis 2007; 45: 753–60. 10.1086/520991 [DOI] [PubMed] [Google Scholar]
  • 67. Rybak MJ, Abate BJ, Kang SLet al. Prospective evaluation of the effect of an aminoglycoside dosing regimen on rates of observed nephrotoxicity and ototoxicity. Antimicrob Agents Chemother 1999; 43: 1549–55. 10.1128/AAC.43.7.1549 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68. Nicolau DP, Freeman CD, Belliveau PPet al. Experience with a once-daily aminoglycoside program administered to 2,184 adult patients. Antimicrob Agents Chemother 1995; 39: 650–5. 10.1128/AAC.39.3.650 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 69. Tune BM. Nephrotoxicity of beta-lactam antibiotics: mechanisms and strategies for prevention. Pediatr Nephrol 1997; 11: 768–72. 10.1007/s004670050386 [DOI] [PubMed] [Google Scholar]
  • 70. Jiang YJ, Xi XM, Jia HMet al. Risk factors, clinical features and outcome of new-onset acute kidney injury among critically ill patients: a database analysis based on prospective cohort study. BMC Nephrol 2021; 22: 289. 10.1186/s12882-021-02503-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 71. Garzotto F, Piccinni P, Cruz Det al. RIFLE-based data collection/management system applied to a prospective cohort multicenter Italian study on the epidemiology of acute kidney injury in the intensive care unit. Blood Purif 2011; 31: 159–71. 10.1159/000322161 [DOI] [PubMed] [Google Scholar]
  • 72. Barlam TF, Cosgrove SE, Abbo LMet al. Implementing an antibiotic stewardship program: guidelines by the Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America. Clin Infect Dis 2016; 62: e51–77. 10.1093/cid/ciw118 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 73. Dellit TH, Owens RC, McGowan JE Jret al. Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship. Clin Infect Dis 2007; 44: 159–77. 10.1086/510393 [DOI] [PubMed] [Google Scholar]
  • 74. Doyle JF, Forni LG. Acute kidney injury: short-term and long-term effects. Crit Care 2016; 20: 188. 10.1186/s13054-016-1353-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 75. Arrayasillapatorn N, Promsen P, Kritmetapak Ket al. Colistin-induced acute kidney injury and the effect on survival in patients with multidrug-resistant Gram-negative infections: significance of drug doses adjusted to ideal body weight. Int J Nephrol 2021; 2021: 7795096. 10.1155/2021/7795096 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 76. Waikar SS, Bonventre JV. Creatinine kinetics and the definition of acute kidney injury. J Am Soc Nephrol 2009; 20: 672–9. 10.1681/ASN.2008070669 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 77. Ronco C, Rosner MH. Acute kidney injury and residual renal function. Crit Care 2012; 16: 144. 10.1186/cc11426 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 78. Curtis LH, Hammill BG, Eisenstein ELet al. Using inverse probability-weighted estimators in comparative effectiveness analyses with observational databases. Med Care 2007; 45: S103–7. 10.1097/MLR.0b013e31806518ac [DOI] [PubMed] [Google Scholar]
  • 79. Lagu T, Lindenauer P. Can we make performance measures more resilient to the effects of coding? Crit Care Med 2015; 43: 1141–3. 10.1097/CCM.0000000000000897 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 80. Sjoding MW, Iwashyna TJ, Dimick JBet al. Gaming hospital-level pneumonia 30-day mortality and readmission measures by legitimate changes to diagnostic coding. Crit Care Med 2015; 43: 989–95. 10.1097/CCM.0000000000000862 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 81. Rothberg MB, Pekow PS, Priya Aet al. Variation in diagnostic coding of patients with pneumonia and its association with hospital risk-standardized mortality rates: a cross-sectional analysis. Ann Intern Med 2014; 160: 380–8. 10.7326/M13-1419 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 82. Lodise T, Ye MJ, Zhao Q. Prevalence of invasive infections due to carbapenem-resistant Enterobacteriaceae among adult patients in US hospitals. Antimicrob Agents Chemother 2017; 61: 6. 10.1128/AAC.00228-17 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 83. Carreno JJ, Tam IM, Meyers JLet al. Longitudinal, nationwide, cohort study to assess incidence, outcomes, and costs associated with complicated urinary tract infection. Open Forum Infect Dis 2019; 6: ofz446. 10.1093/ofid/ofz446 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 84. Huang X, Peterson S, Lavergne Ret al. Predicting the cost of health care services: a comparison of case-mix systems and comorbidity indices that use administrative data. Med Care 2020; 58: 114–9. 10.1097/MLR.0000000000001247 [DOI] [PubMed] [Google Scholar]
  • 85. Sharabiani MT, Aylin P, Bottle A. Systematic review of comorbidity indices for administrative data. Med Care 2012; 50: 1109–18. 10.1097/MLR.0b013e31825f64d0 [DOI] [PubMed] [Google Scholar]
  • 86. Wiese AD, Griffin MR, Stein CMet al. Validation of discharge diagnosis codes to identify serious infections among middle age and older adults. BMJ Open 2018; 8: e020857. 10.1136/bmjopen-2017-020857 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 87. Barber C, Lacaille D, Fortin PR. Systematic review of validation studies of the use of administrative data to identify serious infections. Arthritis Care Res (Hoboken) 2013; 65: 1343–57. 10.1002/acr.21959 [DOI] [PubMed] [Google Scholar]
  • 88. Schneeweiss S, Robicsek A, Scranton Ret al. Veteran’s affairs hospital discharge databases coded serious bacterial infections accurately. J Clin Epidemiol 2007; 60: 397–409. 10.1016/j.jclinepi.2006.07.011 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

dkad351_Supplementary_Data
Click here for additional data file. (39.2KB, docx)

Articles from Journal of Antimicrobial Chemotherapy are provided here courtesy of Oxford University Press

RESOURCES