ABSTRACT
The objectives were to analyze treatment, clinical outcomes, and predictors of mortality in hospitalized patients with Acinetobacter baumannii infection. This was a retrospective cohort study of inpatients with A. baumannii cultures and treatment from 2010 to 2019. Patients who died during admission were compared to those who survived, to identify predictors of inpatient mortality, using multivariable unconditional logistic regression models. We identified 4,599 inpatients with A. baumannii infection; 13.6% died during admission. Fluoroquinolones (26.8%), piperacillin-tazobactam (24%), and carbapenems (15.6%) were used for treatment. Tigecycline (3%) and polymyxins (3.7%) were not used often. Predictors of inpatient mortality included current acute respiratory failure (adjusted odds ratio [aOR] 3.94), shock (aOR 3.05), and acute renal failure (aOR 2.01); blood (aOR 1.94) and respiratory (aOR 1.64) infectious source; multidrug-resistant A. baumannii (MDRAB) infection (aOR 1.66); liver disease (aOR 2.15); and inadequate initial treatment (aOR 1.30). Inpatient mortality was higher in those with MDRAB versus non-MDRAB (aOR 1.61) and in those with CRAB versus non-CRAB infection (aOR 1.68). Length of stay >10 days was higher among those with MDRAB versus non-MDRAB (aOR 1.25) and in those with CRAB versus non-CRAB infection (aOR 1.31). In our national cohort of inpatients with A. baumannii infection, clinical outcomes were worse among those with MDRAB and/or CRAB infection. Predictors of inpatient mortality included several current conditions associated with severity, infectious source, underlying illness, and inappropriate treatment. Our study may assist health care providers in the early identification of admitted patients with A. baumannii infection who are at higher risk of death.
KEYWORDS: Acinetobacter baumannii, antibiotic treatment, mortality, outcomes, predictors
INTRODUCTION
Acinetobacter baumannii, primarily a nosocomial pathogen, is one of the most antibiotic resistant pathogens in clinical medicine (1). Estimates of mortality rates among patients with A. baumannii infections have ranged from 26.0% to 55.7%, with estimated attributable mortality rates between 8.4% and 36.5% (2). While bacteremia and pneumonia are the most severe infections caused by A. baumannii, this organism can cause a variety of other serious infections including urinary tract infections, skin and soft tissue infections, wound infections, osteomyelitis, and meningitis (1). A. baumannii infections have become increasingly difficult to treat due to the emergence of multidrug-resistant A. baumannii (MDRAB) and carbapenem-resistant A. baumannii (CRAB) strains (3). The high prevalence of MDRAB and/or CRAB strains presents a real challenge for clinical treatment, as resistance is associated with inappropriate initial therapy and worse outcomes for patients with A. baumannii infection, including higher mortality (3, 4).
In addition to resistance, identification of other risk factors for mortality could be helpful in early identification of patients at higher risk of death and improving clinical outcomes associated with A. baumannii infections. Previous studies have already identified several risk factors for mortality among hospitalized patients with A. baumannii infection, such as intensive care unit (ICU) stay, older age, renal failure, and septic shock. However, these studies were limited to patients with single-system infections, generally bacteremia or pneumonia, and were small, older, single center studies, which did not control for differences in antibiotic treatment (5–7). As such, the aims of this work were to identify predictors of inpatient mortality while controlling for antibiotic treatment and to evaluate the impact of resistance on clinical outcomes in a national cohort of hospitalized patients with A. baumannii infection.
RESULTS
This study identified a cohort of 4,599 hospitalized patients treated for an A. baumannii infection. The infectious source of their A. baumannii infection was most commonly urine (28.5%, n = 1,313), followed by skin and soft tissue (25.4%, n = 1,166) and respiratory (24.9%, n = 1,146) sources. The overall inpatient mortality rate was 13.6% (n = 626). Demographics and clinical characteristics are presented in Table 1. Those who died during admission were significantly older (mean age 70.9 [± 11.2] versus 66.7 [±12.1] years) and more likely to be treated by intensive care (43.5% versus 19.4%) than those who survived. Those who died were significantly more likely to have MDRAB (65.3% versus 34.1%) and/or CRAB (45.5% versus 21.9%) infections than those who survived, and less likely to have coinfections with other organisms (56.2% versus 65.5%).
TABLE 1.
Baseline demographics and clinical characteristics among hospitalized patients with Acinetobacter baumannii infectiona
| Demographics and clinical characteristics | All hospitalized patients with A. baumannii infection n = 4,599 | Inpatient mortality n = 626 (13.6) | Inpatient survival n = 3,973 (86.4) | P |
|---|---|---|---|---|
| Age, yrs, mean (SD) | 67.3 (12.1) | 70.9 (11.2) | 66.7 (12.1) | <0.001 |
| Male | 4,484 (97.5) | 616 (98.4) | 3,868 (97.4) | 0.120 |
| White | 3,008 (65.4) | 406 (64.9) | 2,602 (65.5) | 0.756 |
| Hispanic or Latino | 448 (9.7) | 111 (17.7) | 337 (8.5) | <0.001 |
| Married | 1,811 (39.4) | 285 (45.5) | 1,526 (38.4) | 0.001 |
| Community admission | 1,851 (40.2) | 182 (29.1) | 1,669 (42) | <0.001 |
| Intensive care unit treatment specialty | 1,041 (22.6) | 272 (43.5) | 769 (19.4) | <0.001 |
| Prior antibiotic and healthcare exposures, 30 days prior to admission | ||||
| Antibiotic exposure | 1,761 (38.3) | 219 (35) | 1,542 (38.8) | 0.067 |
| Hospital exposure | 962 (20.9) | 141 (22.5) | 821 (20.7) | 0.288 |
| Intensive care unit exposure | 244 (5.3) | 43 (6.9) | 201 (5.1) | 0.060 |
| Nursing home exposure | 100 (2.2) | 16 (2.6) | 84 (2.1) | 0.481 |
| Multidrug-resistant Acinetobacter baumannii (MDRAB) infectionb | 1,762 (38.3) | 409 (65.3) | 1,353 (34.1) | <0.001 |
| Carbapenem-resistant Acinetobacter baumannii (CRAB) infectionc | 1,155 (25.1) | 285 (45.5) | 870 (21.9) | <0.001 |
| Resistance phenotype of index Acinetobacter baumannii isolate | ||||
| Aminoglycosided | 1,440 (31.3%) | 345 (55.1) | 1095 (27.6) | <0.001 |
| Ampicillin/sulbactam | 867 (18.9%) | 219 (35.0) | 648 (16.3) | <0.001 |
| Anti-pseudomonal penicillins + β-lactamase inhibitorse | 1,268 (27.6%) | 243 (38.8) | 1,025 (25.8) | <0.001 |
| Extended-spectrum cephalosporinsf | 3,252 (70.7%) | 505 (80.7) | 2,747 (69.1) | <0.001 |
| Fluoroquinolonesg | 1,988 (43.2%) | 437 (69.8) | 1,551 (39.0) | <0.001 |
| Sulfamethoxazole/trimethoprim | 1,462 (31.8%) | 314 (50.2) | 1,148 (28.9) | <0.001 |
| Culture characteristics | ||||
| Time to culture from admission, days, median (interquartile range) | 1 (0–7) | 8 (1–27) | 1 (0–5) | <0.001 |
| Time to culture report completion, days, median (interquartile range) | 4 (3–5) | 4 (3–5) | 4 (3–5) | 0.003 |
| Infectious sourceh | ||||
| Blood | 534 (11.6) | 99 (15.8) | 435 (10.9) | <0.001 |
| Bone and joint | 204 (4.4) | 20 (3.2) | 184 (4.6) | 0.105 |
| Respiratory | 1,146 (24.9) | 339 (54.2) | 807 (20.3) | <0.001 |
| Skin and tissue | 1,166 (25.4) | 71 (11.3) | 1,095 (27.6) | <0.001 |
| Urine | 1,313 (28.5) | 87 (13.9) | 1,226 (30.9) | <0.001 |
| Other | 338 (7.3) | 38 (6.1) | 300 (7.6) | 0.187 |
| Coinfectioni | 2,953 (64.2) | 352 (56.2) | 2,601 (65.5) | <0.001 |
| Citrobacter | 81 (1.8) | 15 (2.4) | 66 (1.7) | 0.194 |
| Enterobacter | 246 (5.3) | 23 (3.7) | 223 (5.6) | 0.045 |
| Enterococcus | 857 (18.6) | 88 (14.1) | 769 (19.4) | 0.002 |
| Escherichia coli | 466 (10.1) | 57 (9.1) | 409 (10.3) | 0.360 |
| Klebsiella | 455 (9.9) | 79 (12.6) | 376 (9.5) | 0.014 |
| Morganella morganii | 68 (1.5) | 8 (1.3) | 60 (1.5) | 0.655 |
| Proteus mirabilis | 340 (7.4) | 47 (7.5) | 293 (7.4) | 0.906 |
| Pseudomonas aeruginosa | 554 (12) | 87 (13.9) | 467 (11.8) | 0.126 |
| Serratia marcescens | 60 (1.3) | 15 (2.4) | 45 (1.1) | 0.010 |
| Staphylococcus aureus | 959 (20.9) | 109 (17.4) | 850 (21.4) | 0.023 |
| Streptococcus pneumoniae | 10 (0.2) | 0 (0) | 10 (0.3) | 0.209 |
| Other | 761 (16.5) | 76 (12.1) | 685 (17.2) | 0.001 |
Data are n (%), unless otherwise indicated. Categorical variables were compared using chi-square or Fisher’s exact tests where appropriate, means were compared using t tests, and medians were compared using nonparametric Wilcoxon tests.
Multidrug-resistant Acinetobacter baumannii (MDRAB) infection was defined as infection due to an MDR A. baumannii strain compared to infection due to a non-MDR A. baumannii strain, due to either susceptibility or absence of susceptibility testing.
Carbapenem-resistant Acinetobacter baumannii (CRAB) infection was defined as infection due to a carbapenem-resistant A. baumannii strain compared to infection due to a noncarbapenem-susceptible A. baumannii strain, due to either susceptibility or absence of susceptibility testing.
Aminoglycosides (amikacin, gentamicin, tobramycin).
Anti-pseudomonal penicillins + β-lactamase inhibitors (piperacillin-tazobactam, clavulanate/ticarcillin).
Extended-spectrum cephalosporins (cefepime, ceftazidime, cefotaxime, ceftriaxone).
Fluoroquinolones (ciprofloxacin, levofloxacin).
Counts and percentages are not mutually exclusive as patients may have had multiple positive culture sites positive for Acinetobacter baumannii.
Coinfections were assessed from three days prior through the day of Acinetobacter baumannii culture collection. Counts and percentages are not mutually exclusive as patients may have had multiple positive organisms in the same culture site, or multiple positive culture sites with multiple organisms.
Current acute conditions and medical history are shown in Table 2. Those who died during admission were significantly more likely to have current acute respiratory failure (75.7% versus 23.6%), shock (44.2% versus 7.6%), and acute renal failure (60.1% versus 29.2%) than those who survived the admission. Those who died were also significantly more likely to have a medical history of liver disease (9.1% versus 6.7%) and cancer or malignancy (29.7% versus 26.2%) than those who survived. The median Charlson score in those who died was 4 (interquartile [IQR] range 2–6) compared to 3 (IQR 1–5) in those that survived.
TABLE 2.
Current conditions and medical history among hospitalized patients with Acinetobacter baumannii infectiona
| Current conditions and medical history | All hospitalized patients with A. baumannii infection n = 4,599 | Inpatient mortality n = 626 (13.6) | Inpatient survival n = 3,973 (86.4) | P |
|---|---|---|---|---|
| Current conditions | ||||
| Anemia | 1,711 (37.2) | 283 (45.2) | 1,428 (35.9) | <0.001 |
| Acute renal failure | 1,535 (33.4) | 376 (60.1) | 1,159 (29.2) | <0.001 |
| Acute respiratory failure | 1,411 (30.7) | 474 (75.7) | 937 (23.6) | <0.001 |
| Adverse effects of medical care | 575 (12.5) | 61 (9.7) | 514 (12.9) | 0.025 |
| Bacterial infection, unspecified site | 2,064 (44.9) | 215 (34.3) | 1,849 (46.5) | <0.001 |
| Complication of surgical or medical procedure | 993 (21.6) | 167 (26.7) | 826 (20.8) | <0.001 |
| Complication of device, implant, or graft | 656 (14.3) | 96 (15.3) | 560 (14.1) | 0.410 |
| Fever | 420 (9.1) | 62 (9.9) | 358 (9) | 0.471 |
| Osteomyelitis and infective arthritis | 888 (19.3) | 88 (14.1) | 800 (20.1) | <0.001 |
| Pneumonia | 1,443 (31.4) | 387 (61.8) | 1,056 (26.6) | <0.001 |
| Septicemia | 1,828 (39.7) | 452 (72.2) | 1,376 (34.6) | <0.001 |
| Shock | 580 (12.6) | 277 (44.2) | 303 (7.6) | <0.001 |
| Skin and subcutaneous infection | 1,092 (23.7) | 91 (14.5) | 1,001 (25.2) | <0.001 |
| Urinary tract infection | 1,940 (42.2) | 263 (42) | 1,677 (42.2) | 0.926 |
| Wound | 366 (8.0) | 42 (6.7) | 324 (8.2) | 0.214 |
| Medical history | ||||
| Atherosclerosis | 1,377 (29.9) | 205 (32.7) | 1,172 (29.5) | 0.099 |
| Cancer or malignancy | 1,225 (26.6) | 186 (29.7) | 1,039 (26.2) | 0.061 |
| Cerebrovascular disease | 531 (11.5) | 86 (13.7) | 445 (11.2) | 0.065 |
| Chronic kidney disease | 1,145 (24.9) | 199 (31.8) | 946 (23.8) | <0.001 |
| Chronic obstructive pulmonary disease | 1,483 (32.2) | 232 (37.1) | 1,251 (31.5) | 0.006 |
| Congestive heart failure | 1,057 (23.0) | 186 (29.7) | 871 (21.9) | <0.001 |
| Diabetes mellitus | 2,111 (45.9) | 285 (45.5) | 1,826 (46) | 0.840 |
| Hypertension | 3,328 (72.4) | 463 (74) | 2,865 (72.1) | 0.336 |
| Immunocompromised | 43 (0.9) | 9 (1.4) | 34 (0.9) | 0.160 |
| Liver disease | 322 (7) | 57 (9.1) | 265 (6.7) | 0.027 |
| Transplant | 39 (0.8) | 10 (1.6) | 29 (0.7) | 0.028 |
| Charlson score, median (interquartile range) | 3 (2–6) | 4 (2–6) | 3 (1–5) | <0.001 |
| Elixhauser score, median (interquartile range) | 5 (3–7) | 6 (3–7) | 5 (3–7) | <0.001 |
Data are n (%), unless otherwise indicated. Categorical variables were compared using chi-square or Fisher’s exact tests where appropriate, and medians were compared using non-parametric Wilcoxon tests. Current conditions, including infections and other acute events, were identified using the International Classification of Diseases, Ninth or Tenth Revision (ICD-9 or -10) diagnosis and procedure codes during the admission. Medical history was identified using ICD-9/10 codes in the year prior to the admission.
Antibiotic treatment exposures for the study cohort are shown in Table 3. Fluoroquinolones (26.8%) and piperacillin-tazobactam (24%) were the most used antibiotics. Carbapenems were used in 15.6% of patients. Most carbapenem usage was as monotherapy (74.0%, n = 530/716), as was fluroquinolone (63.6%, 783/1,232) and piperacillin-tazobactam usage (77.0%, n = 849/1,102). Combination therapy was used in 18.4% of patients. Only 3% of patients were treated with tigecycline (53.3%, n = 73/137 as combination therapy) and 3.7% with a polymyxin (55.0%, n = 93/169 as combination therapy). Those who died during were significantly more likely to be treated with tigecycline (7.2% versus 2.3%), a carbapenem (32.9% versus 12.8%), or a polymyxin (12.9% versus 2.2%) than those who survived. Inadequate initial treatment was significantly more common in those who died compared to those who survived (59.9% versus 46.5%).
TABLE 3.
Antibiotic exposures among hospitalized patients with Acinetobacter baumannii infectiona
| Antibiotic treatments | All hospitalized patients with an A. baumannii infection n = 4,599 | Inpatient mortality n = 626 (13.6) | Inpatient survival n = 3,973 (86.4) | P |
|---|---|---|---|---|
| Inadequate initial treatmentb | 2,221 (48.3) | 375 (59.9) | 1,846 (46.5) | <0.001 |
| Amikacin | 56 (1.2) | 11 (1.8) | 45 (1.1) | 0.185 |
| Ampicillin/sulbactam | 380 (8.3) | 44 (7) | 336 (8.5) | 0.228 |
| Cefepime | 462 (10) | 77 (12.3) | 385 (9.7) | 0.044 |
| Ceftazidime | 104 (2.3) | 21 (3.4) | 83 (2.1) | 0.048 |
| Ceftriaxone | 418 (9.1) | 23 (3.7) | 395 (9.9) | <0.001 |
| Ciprofloxacin | 814 (17.7) | 44 (7) | 770 (19.4) | <0.001 |
| Colistin | 87 (1.9) | 37 (5.9) | 50 (1.3) | <0.001 |
| Doripenem | 28 (0.6) | 8 (1.3) | 20 (0.5) | 0.021 |
| Doxycycline | 163 (3.5) | 4 (0.6) | 159 (4) | <0.001 |
| Gentamicin | 78 (1.7) | 15 (2.4) | 63 (1.6) | 0.144 |
| Imipenem | 345 (7.5) | 115 (18.4) | 230 (5.8) | <0.001 |
| Levofloxacin | 431 (9.4) | 30 (4.8) | 401 (10.1) | <0.001 |
| Meropenem | 346 (7.5) | 84 (13.4) | 262 (6.6) | <0.001 |
| Minocycline | 72 (1.6) | 3 (0.5) | 69 (1.7) | 0.019 |
| Piperacillin-tazobactam | 1,102 (24) | 155 (24.8) | 947 (23.8) | 0.615 |
| Polymyxin B | 84 (1.8) | 46 (7.3) | 38 (1) | <0.001 |
| Sulfamethoxazole/trimethoprim | 368 (8) | 18 (2.9) | 350 (8.8) | <0.001 |
| Tigecycline | 137 (3) | 45 (7.2) | 92 (2.3) | <0.001 |
| Tobramycin | 52 (1.1) | 12 (1.9) | 40 (1) | 0.045 |
| Antibiotic classes | ||||
| Aminoglycosidec | 186 (4) | 38 (6.1) | 148 (3.7) | 0.006 |
| Carbapenemsd | 716 (15.6) | 206 (32.9) | 510 (12.8) | <0.001 |
| Extended-spectrum cephalosporinse | 980 (21.3) | 120 (19.2) | 860 (21.6) | 0.160 |
| Fluoroquinolonesf | 1232 (26.8) | 73 (11.7) | 1,159 (29.2) | <0.001 |
| Anti-pseudomonal penicillins + β-lactamase inhibitorsg | 1,106 (24) | 156 (24.9) | 950 (23.9) | 0.583 |
| Polymyxinsh | 169 (3.7) | 81 (12.9) | 88 (2.2) | <0.001 |
| Tetracyclinesi | 235 (5.1) | 7 (1.1) | 228 (5.7) | <0.001 |
Data are n (%). Categorical variables were compared using chi-square or Fisher’s exact tests where appropriate. Assessed antibiotic treatment with activity against A. baumannii, which included amikacin, ampicillin/sulbactam, cefepime, cefotaxime, ceftazidime, ceftriaxone, ciprofloxacin, clavulanate/ticarcillin, colistin, doripenem, doxycycline, gentamicin, imipenem, levofloxacin, meropenem, minocycline, piperacillin-tazobactam, polymyxin B, sulfamethoxazole/trimethoprim, tetracycline, tigecycline and tobramycin.
Inadequate initial treatment was assessed from culture collection of the A. baumannii isolate until the 4th day after culture and defined as lack of receipt of at least one antibiotic with susceptibility.
Aminoglycosides (amikacin, gentamicin, tobramycin).
Carbapenems (imipenem, meropenem, doripenem).
Extended-spectrum cephalosporins (cefepime, ceftazidime, cefotaxime, ceftriaxone).
Fluoroquinolones (ciprofloxacin, levofloxacin).
Anti-pseudomonal penicillins + β-lactamase inhibitors (piperacillin-tazobactam, clavulanate/ticarcillin).
Polymyxins (colistin, polymyxin B).
Tetracyclines (tetracycline, minocycline, doxycycline).
Several (n = 12) predictors of inpatient mortality (Table 4), controlling for antibiotic treatment, time to culture from admission, year of treatment, and factors associated with survival, were identified, including presence of current acute respiratory failure (aOR 3.94; 95% confidence intervals [CI], 3.07–5.05), shock (aOR 3.05; 95% CI, 2.36–3.94), and acute renal failure (aOR 2.01; 95% CI, 1.62–2.49); blood (aOR 1.94; 95% CI, 1.40–2.69) and respiratory (aOR 1.64; 95% CI, 1.29–2.09) infectious source; MDRAB infection (aOR 1.66; 95% CI, 1.30–2.12); medical history of liver disease (aOR 2.15, 95% CI, 1.46–3.17) and cancer or malignancy (aOR 1.40; 95% CI, 1.11–1.76); and inadequate initial treatment (aOR 1.30; 95% CI, 1.04–1.64). Results of subgroup analyses can be found in the supplemental material. Current acute respiratory failure, shock, acute renal failure, and MDRAB infection remained significant predictors of inpatient mortality in all subgroups assessed (Table S1 in the supplemental material).
Table 4.
Independent predictors of mortality among hospitalized patients with Acinetobacter baumannii infectiona
| Predictor | Adjusted odds ratio | Lower 95% confidence limit | Upper 95% confidence limit |
|---|---|---|---|
| Current conditions | |||
| Acute respiratory failure | 3.94 | 3.07 | 5.05 |
| Shock | 3.05 | 2.36 | 3.94 |
| Acute renal failure | 2.01 | 1.62 | 2.49 |
| Septicemia | 1.62 | 1.27 | 2.08 |
| Anemia | 1.33 | 1.08 | 1.65 |
| Inadequate initial treatmentb | 1.30 | 1.04 | 1.64 |
| Infectious source | |||
| Blood source | 1.94 | 1.40 | 2.69 |
| Respiratory source | 1.64 | 1.29 | 2.09 |
| Multidrug-resistant Acinetobacter baumannii (MDRAB) infectionc | 1.66 | 1.30 | 2.12 |
| Medical history | |||
| Liver disease | 2.15 | 1.46 | 3.17 |
| Cancer or malignancy | 1.40 | 1.11 | 1.76 |
| Age, yrs | 1.04 | 1.03 | 1.05 |
The adjusted odds ratios are estimated from multivariable analysis of the data. The final multivariable unconditional logistic regression model included all predictive variables listed in the table above (odds ratio >1) and controlled for the following variables: treatment with tigecycline, tetracyclines, polymyxins, fluroquinolones, and/or carbapenems, time to culture from admission, osteomyelitis and infective arthritis, bacterial infection of an unspecified site, and year of treatment.
Inadequate initial treatment was assessed from culture collection of the A. baumannii isolate until the 4th day after culture and defined as lack of receipt of at least one antibiotic with susceptibility.
Multidrug-resistant Acinetobacter baumannii (MDRAB) infection was defined as infection due to an MDR A. baumannii strain compared to infection due to a non-MDR A. baumannii strain, due to either susceptibility or absence of susceptibility testing.
Clinical outcomes can be found in Table 5. Most clinical outcomes were significantly worse for those with MDRAB or CRAB infection, including inpatient mortality (MDRAB 23.1% versus non-MDRAB 7.7%; CRAB 24.7% versus non-CRAB 8.5%) and length of stay greater than 10 days (MDRAB 68.7% versus non-MDRAB 42.5%; CRAB 71.7% versus non-CRAB 44.8%). Inpatient mortality was significantly higher in those with MDRAB versus non-MDRAB (aOR 1.61; 95% CI 1.26–2.06) and in those with CRAB versus non-CRAB infection (aOR 1.68; 95% CI 1.31–2.17), and the odds of a longer length of stay (>10 days) were significantly higher among those with MDRAB versus non-MDRAB (aOR 1.25; 95% CI 1.03–1.52) and in those with CRAB versus non-CRAB infection (aOR 1.31; 95% CI 1.04–1.65). Results of subgroup analyses were generally similar to the overall cohort (Table S2). However, the odds of a longer length of stay (>10 days) were significantly lower among those with MDRAB versus non-MDRAB (aOR 0.37; 95% CI 0.19–0.73) and among those with CRAB versus non-CRAB infection (aOR 0.37; 95% CI 0.16–0.86) in the subgroup of patients with blood source infections and not significantly different for MDRAB or CRAB infections in the subgroup of patients with respiratory source infections.
TABLE 5.
Clinical outcomes among hospitalized patients with Acinetobacter baumannii infection overall and stratified by resistancea
| Clinical outcomes | All hospitalized patients with an A. baumannii infection (n = 4,599) | MDRAB infectionb (n = 1,762) | Non-MDRAB infection (n = 2,802) | Adjusted odds ratio, (95% confidence interval) | CRAB infectionc (n = 1,155) | Non-CRAB infection (n = 2,482) | Adjusted odds ratio, (95% confidence interval) |
|---|---|---|---|---|---|---|---|
| Inpatient mortality | 626 (13.6%) | 409 (23.1%) | 216 (7.7%) | 1.61 (1.26–2.06) d | 285 (24.7%) | 210 (8.5%) | 1.68 (1.31–2.17) e |
| Mortality within 30 days of culture | 676 (14.7%) | 391 (22.2%) | 282 (10.1%) | 1.71 (1.40–2.10) f | 270 (23.4%) | 266 (10.7%) | 1.62 (1.29–2.04) g |
| Reinfection within 30 days of dischargeh | 157/3,973 (4.0%) | 100/1,353 (7.4%) | 55/2,586 (2.1%) | 2.66 (1.88–3.77) i | 68/870 (7.8%) | 61/2,272 (2.7%) | 2.31 (1.61–3.33) j |
| Length of hospital stay greater than 10 days | 2,411 (52.4%) | 1,210 (68.7%) | 1,192 (42.5%) | 1.25 (1.03–1.52) k | 821 (71.1%) | 1,112 (44.8%) | 1.31 (1.04–1.65) l |
| Readmission within 30 days of dischargeh | 935/3,973 (23.5%) | 341/1,353 (25.2%) | 582/2,586 (22.5%) | 1.02 (0.87–1.19)m | 215/870 (24.7%) | 528/2,272 (23.2%) | 0.95 (0.78 1.14)n |
Data are n (%) or adjusted odds ratio (95% confidence interval). Categorical variables were compared using chi-square or Fisher’s exact tests where appropriate, means were compared using t tests, and medians were compared using non-parametric Wilcoxon tests. Bolded indicates P value <0.05 for comparison of resistant and non-resistant phenotypes. The adjusted odds ratios are estimated from multivariable analysis of the data.
Multidrug-resistant Acinetobacter baumannii (MDRAB) infection was defined as A. baumannii infection due to an isolate that demonstrated non-susceptibility (intermediate or resistant) to at least 1 drug in at least 3 antibiotic classes (extended-spectrum cephalosporins, fluoroquinolones, aminoglycosides, carbapenems, piperacillin-tazobactam, and ampicillin/sulbactam). Non-MDRAB infection was defined as infection due to an isolate that demonstrated susceptibility to at least 1 drug in at least 3 antibiotic classes. Data not available in 35 patients to define MDRAB or non-MDRAB, as two or fewer antibiotic classes were tested for susceptibility.
Carbapenem-resistant Acinetobacter baumannii (CRAB) infection was defined as A. baumannii infection due to an isolate that demonstrated nonsusceptibility to at least 1 carbapenem (imipenem, meropenem, or doripenem). Non-CRAB infection was defined as infection due to an isolate that demonstrated susceptibility to at least 1 carbapenem. Data not available in 962 patients to define CRAB or non-CRAB, as no carbapenems were tested for susceptibility.
Adjusted for the current conditions (acute renal failure, acute respiratory failure, anemia, osteomyelitis and infective arthritis, septicemia, shock), medical history (cancer or malignancy, liver disease), infectious source (blood, respiratory), age, inadequate initial treatment, antibiotic treatment (carbapenems, fluoroquinolones, polymyxins, tigecycline), time to culture from admission, and year of treatment.
Adjusted for current conditions (acute renal failure, acute respiratory failure, anemia, bacterial infection unspecified site, complication of surgical or medical procedure, septicemia, shock), medical history (cancer or malignancy, liver disease), infectious source (blood, respiratory), age, antibiotic treatment (carbapenems, fluoroquinolones, polymyxins, tetracyclines, tigecycline), time to culture from admission, and year of treatment.
Adjusted for current conditions (acute renal failure, acute respiratory failure, anemia, bacterial infection unspecified site, fever, pneumonia, osteomyelitis and infective arthritis, septicemia, shock), medical history (cancer or malignancy, chronic obstructive pulmonary disease, liver disease), infectious source (blood, respiratory), age, antibiotic treatment (carbapenems, tigecycline), and year of treatment.
Adjusted for current conditions (anemia, acute renal failure, acute respiratory failure, fever, osteomyelitis and infective arthritis, septicemia, shock, wound), medical history (cancer or malignancy, liver disease), infectious source (blood, respiratory), age, Charlson score, antibiotic treatment (carbapenems, polymyxins, tigecycline), and year of treatment.
Only measured among patients who were discharged alive.
Adjusted for current conditions (acute respiratory failure), infectious source (blood), antibiotics in the prior 30 days, and year of treatment.
Adjusted for current conditions (acute respiratory failure), infectious source (blood), antibiotics in the prior 30 days, nursing home stay in the prior 30 days, and year of treatment.
Adjusted for current conditions (anemia; acute renal failure; acute respiratory failure; complication of device, implant, or graft; pneumonia; osteomyelitis and infective arthritis; septicemia), medical history (cancer or malignancy), infectious source (respiratory), age, inadequate initial treatment, antibiotic treatment (doxycycline, fluoroquinolones, piperacillin-tazobactam, polymyxins), intensive care unit treatment specialty, time to culture from admission, and year of treatment.
Adjusted for current conditions (anemia; acute renal failure; acute respiratory failure; fever; complication of device, implant, or graft; pneumonia; osteomyelitis and infective arthritis; septicemia; wound), Charlson score, infectious source (respiratory), age, inadequate initial treatment, antibiotic treatment (doxycycline, fluoroquinolones, piperacillin-tazobactam, polymyxins), time to culture from admission, and year of treatment.
Adjusted for current condition (anemia), Charlson score, Elixhauser score, antibiotics in the prior 30 days, time to culture from admission, and year of treatment.
Adjusted for current conditions (anemia, septicemia), Charlson score, antibiotics in the prior 30 days, and year of treatment.
DISCUSSION
To our knowledge, this is the first multicenter study to identify predictors of inpatient mortality associated with all types of A. baumannii infections. This study demonstrated poor outcomes for hospitalized patients with A. baumannii infection, with significantly worse outcomes among those with MDRAB and/or CRAB infection.
The inpatient mortality rate we observed (13.6%) was similar to several previous studies that have described inpatient mortality rates of 12.7% and 17.6% (3, 8). Similar to previous work, we identified several predictors of mortality related to the severity of infection, including septicemia, shock, acute respiratory failure, and acute renal failure (7, 9). These conditions are all signs of severe infection and resultant multiple organ dysfunction that are often present in critically ill patients and contribute to poor outcomes including mortality (9). We also identified that current anemia was an independent predictor of mortality, in contrast to previous findings (10). As patients with anemia have low hemoglobin levels and may have disrupted transportation of oxygen to organ systems leading to hypoxia, anemia may exacerbate multiple organ dysfunction seen in critically ill patients, thus contributing to excess mortality (11). Additionally, in a previous study of 175 hospitals in the United States, anemia was more common among patients with MDRAB versus non-MDRAB pneumonia and sepsis (50.6% versus 38.5%, P < 0.001), and hospital mortality was higher in those with MDRAB (12). Finally, our results may be related to the need for red blood cell transfusion in patients with severe anemia. Previous work demonstrated that red blood cell transfusion was a strong independent predictor of in-hospital mortality among patients with MDRAB ventilator associated pneumonia (13).
We also found that that underlying medical conditions (liver disease and cancer) and increasing age were predictors of mortality, which is consistent with previous findings (9, 14–16). Previously, among two separate cohorts of hospitalized patients with A. baumannii bacteremia (n = 188 and n = 122), malignancy was predictive of mortality (16, 17). A meta-analysis of 19 observational studies found that liver disease was an important factor associated with mortality in patients infected with CRAB (9).
Blood source and respiratory sources of infection were also predictive of mortality in our cohort. These results are expected as pneumonia and bacteremia are generally associated with more serious A. baumannii disease and worse patient outcomes than other infection types (1). Previous work among hospitalized patients with A. baumannii also demonstrated that blood source of A. baumannii infection was an independent predictor of mortality (OR, 4.64; 95% CI, 1.26–17.06) (10). Bacteremia is an important and common cause of death. Previous work has demonstrated that among patients with ventilator associated pneumonia (VAP) A. baumannii, the mortality rate was higher in patients with VAP and bacteremia compared to those with nonbacteremic VAP (32.4% versus 9.6%, P < 0.005) (18). Pneumonia is also an important cause of death. Previously, recovery of A. baumannii from the respiratory tract was identified as a major risk factor related to mortality among patients in the ICU with nosocomial A. baumannii infections (19). Among 338 patients with nosocomial A. baumannii bacteremia, bacteremia occurring after severe pneumonia was an independent risk factor mortality (14).
As with several previous studies, we found that MDRAB infection was an independent risk factor for mortality. Infection with MDRAB often leads to high treatment failure and worse outcomes than susceptible A. baumannii infection. A previous retrospective, matched cohort study in 2 hospitals found that after controlling for severity of illness and underlying disease, MDRAB was independently associated with increased hospital and intensive care unit length of stay compared to susceptible MDRAB (3). The association between MDRAB infection and worse clinical outcomes is likely related to increased probability for inappropriate initial antibiotic therapy, delay to active antibiotics, and resultant increased severity of disease and increased risk for treatment failure, rather than enhanced virulence of the organism (15, 20–22). When controlling for treatment and severity, MDRAB was still an independent predictor of mortality in our study. In prior work, those with MDRAB were 5-fold more likely to receive inappropriate empirical treatment than those with non-MDRAB pneumonia or sepsis, and inappropriate empirical treatment nearly doubled in-hospital mortality (12).
As expected, clinical outcomes were significantly worse among those with MDRAB and CRAB compared to susceptible A. baumannii infections. Similarly, previous work has demonstrated that resistant A. baumannii infections are associated with higher mortality and morbidity, including increased length of hospital stay, readmission rates, and reinfection rates (3). All-cause readmissions and readmissions during which MDR organisms are isolated are common among patients with infections due to MDR pathogens (23). As with MDRAB, poor outcomes among those with CRAB are likely related to inappropriate initial antibiotic therapy. A systematic review of observational studies that included over 2,500 patients with CRAB and susceptible A. baumannii infection found that CRAB was associated with a greater risk of mortality (OR 2.22; 95% CI 1.66–2.98) and those with CRAB were more likely to receive inappropriate initial antibiotic therapy (4).
There are limitations inherent in the present work. Clinical signs and symptoms of infection were not assessed, and thus some of the A. baumannii cultures we captured may have represented colonization rather than true infection. However, only patients in which antibiotics with activity against A. baumannii were used were included, and patients had to be treated for a minimum duration of at least 2 days. Additionally, several infection diagnoses for which A. baumannii is known to be the cause were common during the admission, including bacterial infection of an unspecified site (which includes bacteremia) 44.9%, pneumonia 31.9%, and urinary tract infection 42.2%. Moreover, results were similar among the subgroups of patients with blood source and respiratory source infections. To capture the full spectrum of antibiotic treatment among hospitalized patients with A. baumannii infection, patients who had positive cultures for other organisms (coinfections) were not excluded; thus, some of the antibiotic treatment captured may have been targeting other non-A. baumannii organisms and/or infections. Reassuringly, results were similar among the subgroup of patients with monomicrobial A. baumannii infections. For clinical outcomes, only readmissions and reinfections that were treated within the VA health care system were captured. Additionally, all outcomes assessed were all cause and not necessarily A. baumannii infection related. In our predictive analysis, we considered several risk factors for poor outcomes in patients with A. baumannii infection that have been previously described; however, there may be other known and unknown risk factors that were not included in our study. The generalizability of this study may be limited to the patients admitted to VA hospitals.
In our national cohort of hospitalized patients with A. baumannii infection, 13.6% of patients died during admission and clinical outcomes were worse for those with MDRAB or CRAB infection. While most of the predictors we identified have been previously identified (separately or together with other identified predictors) in various other studies of patients with A. baumannii infection, mostly bacteremia and pneumonia, these studies did not control for differences in antibiotic treatments. Additionally, our study is the first to identify this full list of predictors of inpatient morality among hospitalized patients with all types of A. baumannii infection and the first to identify anemia as an independent predictor. Controlling for antibiotic treatment, the predictors of inpatient morality we identified included current conditions, infectious source, underlying illness, and MDRAB infection. Our study may assist health care providers in the early identification of admitted patients with A. baumannii infection who are at higher risk of death.
MATERIALS AND METHODS
The national VA data sets used for this work included inpatient admissions, inpatient and outpatient care, diagnoses, procedures, vital status, microbiology, and pharmacy.
The retrospective cohort study selected a cohort of adult inpatients (age ≥18 years) admitted to VA hospitals nationally with positive A. baumannii cultures from any infectious source (blood, urine, respiratory, skin and tissue, and other) collected between January 1, 2010 and April 30, 2019. Further inclusion criteria consisted of antibiotic treatment with activity against A. baumannii within 2 days of culture collection, treatment duration of 2 days or more, index culture and susceptibility report completion time 10 days or less, and the first qualifying admission meeting these criteria during the study period. In the case of multiple positive A. baumannii cultures during the admission, the first culture was considered the index culture.
Clinical characteristics of our cohort were evaluated, including demographics, current conditions and medical history, and prior antibiotic and health care exposures. Current conditions, including infections and other acute events, were identified using the International Classification of Diseases, Ninth or Tenth Revision (ICD-9 or -10) diagnosis and procedure codes during the admission. Medical history was identified using ICD-9/10 codes in the year prior to the admission. Coinfections were defined as cultures positive for other organisms assessed from 3 days prior through the day of A. baumannii collection. Antibiotic treatments were evaluated by antibiotic agent and class. Inadequate initial treatment was defined as lack of receipt of at least one treatment in which the index A. baumannii isolate demonstrated susceptibility to within 4 days after culture collection.
MDRAB infection was defined as A. baumannii infection due to an isolate that demonstrated nonsusceptibility (intermediate or resistant) to at least 1 drug in at least 3 antibiotic classes (extended-spectrum cephalosporins, fluoroquinolones, aminoglycosides, carbapenems, piperacillin-tazobactam, and ampicillin/sulbactam) (24). Isolates not meeting the definition of MDRAB, due to either susceptibility or absence of susceptibility testing against one of more antibiotic classes, were defined as non-MDRAB. CRAB infection was defined as A. baumannii infection due to an isolate that demonstrated nonsusceptibility to at least 1 carbapenem (imipenem, meropenem, or doripenem) (24). Isolates not meeting the definition of CRAB, due to either susceptibility or absence of susceptibility testing against carbapenems, were defined as non-CRAB.
Clinical outcomes evaluated included all-cause inpatient mortality during the admission and 30-day mortality from culture collection, length of stay greater than 10 days, reinfection defined as a subsequent positive A. baumannii culture within 30 days of discharge, and readmission to a VA hospital within 30 days of discharge.
Clinical characteristics and antibiotic treatment were compared between patients who died during the admission and those who survived. Categorical variables were compared using chi-square or Fisher’s exact tests where appropriate, means were compared using t tests, and medians were compared using nonparametric Wilcoxon tests. Backwards, manual, stepwise unconditional logistic regression (initial selection P value <0.1, retained in model P value <0.05) was used to identify characteristics that were predictive of inpatient mortality while controlling for confounding, which included imbalances in antibiotic treatment, time to culture collection, and year of treatment between survivors and nonsurvivors, and other factors protective against mortality (25). Potential predictors were selected a priori and were based on previously described predictors of poor outcomes in patients with A. baumannii infection (4, 10). Adjusted odds ratios (aORs) and 95% confidence intervals (CI) were calculated for independent predictors of inpatient mortality (aORs > 1.0). Absence of collinearity between the variables in the final model was assessed from tolerance and variance inflation (25).
We also calculated aORs and 95% CIs for resistance (MDRAB versus non-MDRAB, and CRAB versus non-CRAB) and each of the aforementioned clinical outcomes. Confounders significantly associated with both resistance and clinical outcomes were controlled for in the adjusted models (backwards, automatic, stepwise unconditional logistic regression, initial selection P value <0.1, retained in model P value <0.05). For our analysis assessing the impact of resistance on clinical outcomes, we excluded patients without MDR or carbapenem susceptibility results. We conducted subgroup analyses by culture source (for blood cultures and respiratory cultures), and monomicrobial Acinetobacter baumannii cultures.
Data availability.
The deidentified datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request and approval of the Providence Veterans Affairs Medical Center Institutional Review Board.
ACKNOWLEDGMENTS
The views expressed are those of the authors and do not necessarily reflect the position or policy of the United States Department of Veterans Affairs. This material is based upon work supported, in part, by the Office of Research and Development, Department of Veterans Affairs.
The study was approved by the Institutional Review Board and the Research and Development Committee of the Providence Veterans Affairs Medical Center prior to initiation with a waiver of the informed consent process. All methods were carried out in accordance with relevant guidelines and regulations.
Author contributions. Conception and design of the study: A.R.C., H.A., K.L. Data generation: H.A., A.R.C., V.L. Analysis and interpretation of the data: H.A., A.R.C., K.L. Preparation or critical revision of the manuscript: H.A., A.R.C., K.L.
Haley Appaneal has received research funding from Shionogi Kerry LaPlante has received research funding or acted as a scientific advisor for Merck, Parateck, Pfizer, Spero, and Shionogi. Aisling Caffrey has received research funding from Gilead, Merck, Pfizer, and Shionogi, and has received speaking honoraria from Merck.
This work was funded, in part, by Shionogi, Inc.
Footnotes
Supplemental material is available online only.
REFERENCES
- 1.Alsan M, Klompas M. 2010. An emerging and important pathogen. J Clin Outcomes Manag 17:363–369. [PMC free article] [PubMed] [Google Scholar]
- 2.Falagas ME, Rafailidis PI. 2007. Attributable mortality of Acinetobacter baumannii: no longer a controversial issue. Crit Care 11:134. doi: 10.1186/cc5911. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Sunenshine RH, Wright MO, Maragakis LL, Harris AD, Song X, Hebden J, Cosgrove SE, Anderson A, Carnell J, Jernigan DB, Kleinbaum DG, Perl TM, Standiford HC, Srinivasan A. 2007. Multidrug-resistant Acinetobacter infection mortality rate and length of hospitalization. Emerg Infect Dis 13:97–103. doi: 10.3201/eid1301.060716. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Lemos EV, de la Hoz FP, Einarson TR, McGhan WF, Quevedo E, Castañeda C, Kawai K. 2014. Carbapenem resistance and mortality in patients with Acinetobacter baumannii infection: systematic review and meta-analysis. Clin Microbiol Infect 20:416–423. doi: 10.1111/1469-0691.12363. [DOI] [PubMed] [Google Scholar]
- 5.Metan G, Sariguzel F, Sumerkan B. 2009. Factors influencing survival in patients with multi-drug-resistant Acinetobacter bacteraemia. Eur J Intern Med 20:540–544. doi: 10.1016/j.ejim.2009.05.005. [DOI] [PubMed] [Google Scholar]
- 6.Dizbay M, Tunccan OG, Sezer BE, Hizel K. 2010. Nosocomial imipenem-resistant Acinetobacter baumannii infections: epidemiology and risk factors. Scand J Infect Dis 42:741–746. doi: 10.3109/00365548.2010.489568. [DOI] [PubMed] [Google Scholar]
- 7.Katsaragakis S, Markogiannakis H, Samara E, Pachylaki N, Theodoraki EM, Xanthaki A, Toutouza M, Toutouzas KG, Theodorou D. 2010. Predictors of mortality of Acinetobacter baumannii infections: a 2-year prospective study in a Greek surgical intensive care unit. Am J Infect Control 38:631–635. doi: 10.1016/j.ajic.2010.01.009. [DOI] [PubMed] [Google Scholar]
- 8.Yang S, Sun J, Wu X, Zhang L. 2018. Determinants of mortality in patients with nosocomial Acinetobacter baumannii bacteremia in southwest China: a five-year case-control study. Can J Infect Dis Med Microbiol 2018:3150965. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Du X, Xu X, Yao J, Deng K, Chen S, Shen Z, Yang L, Feng G. 2019. Predictors of mortality in patients infected with carbapenem-resistant Acinetobacter baumannii: a systematic review and meta-analysis. Am J Infect Control 47:1140–1145. doi: 10.1016/j.ajic.2019.03.003. [DOI] [PubMed] [Google Scholar]
- 10.Zhang H, Zhao Y, Zheng Y, Kong Q, Lv N, Liu Y, Zhao D, Li J, Ye Y. 2020. Development and validation of a model for predicting the risk of death in patients with Acinetobacter baumannii infection: a retrospective study. Infect Drug Resist 13:2761–2772. doi: 10.2147/IDR.S253143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Hemauer SJ, Kingeter AJ, Han X, Shotwell MS, Pandharipande PP, Weavind LM. 2017. Daily lowest hemoglobin and risk of organ dysfunctions in critically ill patients. Crit Care Med 45:e479–e84. doi: 10.1097/CCM.0000000000002288. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Zilberberg MD, Nathanson BH, Sulham K, Fan W, Shorr AF. 2016. Multidrug resistance, inappropriate empiric therapy, and hospital mortality in Acinetobacter baumannii pneumonia and sepsis. Crit Care 20:221. doi: 10.1186/s13054-016-1392-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Čiginskienė A, Dambrauskienė A, Rello J, Adukauskienė D. 2019. Ventilator-associated pneumonia due to drug-resistant Acinetobacter baumannii: risk factors and mortality relation with resistance profiles, and independent predictors of in-hospital mortality. Medicina 55:49. doi: 10.3390/medicina55020049. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Zhou H, Yao Y, Zhu B, Ren D, Yang Q, Fu Y, Yu Y, Zhou J. 2019. Risk factors for acquisition and mortality of multidrug-resistant Acinetobacter baumannii bacteremia: a retrospective study from a Chinese hospital. Medicine (Baltimore) 98:e14937. doi: 10.1097/MD.0000000000014937. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Liu Q, Li W, Du X, Zhong T, Tang Y, Feng Y, Tao C, Xie Y. 2015. Risk and prognostic factors for multidrug-resistant Acinetobacter baumannii complex bacteremia: a retrospective study in a tertiary hospital of West China. PLoS One 10:e0130701. doi: 10.1371/journal.pone.0130701. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Gu Z, Han Y, Meng T, Zhao S, Zhao X, Gao C, Huang W. 2016. Risk factors and clinical outcomes for patients with Acinetobacter baumannii bacteremia. Medicine (Baltimore) 95:e2943. doi: 10.1097/MD.0000000000002943. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Park SY, Choo JW, Kwon SH, Yu SN, Lee EJ, Kim TH, Choo EJ, Jeon MH. 2013. Risk factors for mortality in patients with Acinetobacter baumannii bacteremia. Infect Chemother 45:325–330. doi: 10.3947/ic.2013.45.3.325. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Brotfain E, Borer A, Koyfman L, Saidel-Odes L, Frenkel A, Gruenbaum SE, Rosenzweig V, Zlotnik A, Klein M. 2017. Multidrug resistance Acinetobacter bacteremia secondary to ventilator-associated pneumonia: risk factors and outcome. J Intensive Care Med 32:528–534. doi: 10.1177/0885066616632193. [DOI] [PubMed] [Google Scholar]
- 19.Karabay O, Yahyaoğlu M, Oğütlü A, Sandıkçı O, Tuna N, Ceylan S. 2012. Factors associated with mortality in Acinetobacter baumannii infected intensive care unit patients. Mikrobiyol Bul 46:335–337. [PubMed] [Google Scholar]
- 20.Gordon NC, Wareham DW. 2010. Multidrug-resistant Acinetobacter baumannii: mechanisms of virulence and resistance. Int J Antimicrob Agents 35:219–226. doi: 10.1016/j.ijantimicag.2009.10.024. [DOI] [PubMed] [Google Scholar]
- 21.Zaragoza R, Artero A, Camarena JJ, Sancho S, González R, Nogueira JM. 2003. The influence of inadequate empirical antimicrobial treatment on patients with bloodstream infections in an intensive care unit. Clin Microbiol Infect 9:412–418. doi: 10.1046/j.1469-0691.2003.00656.x. [DOI] [PubMed] [Google Scholar]
- 22.Falagas ME, Kasiakou SK, Rafailidis PI, Zouglakis G, Morfou P. 2006. Comparison of mortality of patients with Acinetobacter baumannii bacteraemia receiving appropriate and inappropriate empirical therapy. J Antimicrob Chemother 57:1251–1254. doi: 10.1093/jac/dkl130. [DOI] [PubMed] [Google Scholar]
- 23.Burnham JP, Kwon JH, Olsen MA, Babcock HM, Kollef MH. 2018. Readmissions with multidrug-resistant infection in patients with prior multidrug resistant infection. Infect Control Hosp Epidemiol 39:12–19. doi: 10.1017/ice.2017.254. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Centers for Disease Control and Prevention. January 2020, revision date. Antimicrobial resistant phenotype definitions 2020. https://www.cdc.gov/nhsn/pdfs/ps-analysis-resources/aur/ar-phenotype-definitions-508.pdf.
- 25.Hosmer DW, Lemeshow S. 2000. Applied logistic regression, 2nd ed. Wiley, New York, NY. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental Tables S1 and S2. Download aac.01975-21-s0001.pdf, PDF file, 0.05 MB (49.1KB, pdf)
Data Availability Statement
The deidentified datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request and approval of the Providence Veterans Affairs Medical Center Institutional Review Board.