Clinical characteristics and outcomes

A total of 224 CRAb isolates were identified at our institution from 2019-Q1 2026 (Figure S1). Of these, 13.39% (30/224) underwent analysis and review. All patients included in the analysis had at least one comorbidity, as listed in Table 1, with diabetes, chronic respiratory failure, and chronic kidney disease being the most common. Three patients were immunocompromised per CDC/ACIP definitions, including one with CD4 count <200 from HIV infection, one with chronic steroid use (>20 mg daily prednisone for >14 days), and one with history of splenectomy. Most patients presented to the hospital from a long-term care facility (LTCF) (67%) with 63% of the same patient population discharging to an LTCF.

Table 1.

Clinical and epidemiologic characteristics of patients with carbapenemresistant Acinetobacter baumannii

Table 1. Long description

A table with 30 rows and 25 columns detailing the clinical and epidemiologic characteristics of patients with carbapenem-resistant Acinetobacter baumannii. The table includes data on sex, age, comorbidities, presence of invasive devices, antibiotic exposure, history of specific colonizations, COVID-19 infection status, and location immediately prior to admission. Key columns include male sex, female sex, age median, comorbidities such as diabetes, chronic respiratory failure, and chronic kidney disease, presence of invasive devices like ventilators and tracheostomy tubes, antibiotic exposure prior to specimen collection, history of various colonizations, COVID-19 infection status, and location prior to admission. Notable trends include a high prevalence of diabetes, chronic respiratory failure, and chronic kidney disease among the patients, as well as a significant number of patients with a history of antibiotic exposure and presence of invasive devices.

^a Includes head and neck cancer, bladder cancer, myelodysplastic syndrome, prostate cancer, and hepatocellular cancer. None of these patients were receiving immunosuppressive therapy at the time of their CRAb infection.

^b Defined as antibiotics received within 90-days prior to collection of culture growing CRAb isolate documented in chart. There are four cases listed as unknown due to lack of documents or details including antibiotic administration prior to collection.

^c Co-infection defined as additional organisms isolated from same culture as identified CRAb and treated accordingly. Multidrug resistant defined as resistant to >3 antimicrobial drug classes.

Isolates were collected from a variety of sources including respiratory (18), blood (5), wounds (4), intra-abdominal (2), and urine (1) (Table S1). Eighteen of thirty (60%) patients received CRAb-specific antimicrobial therapy; the remainder were considered colonized or cured with source control alone. The most prescribed combination therapy was sulbactam-durlobactam with meropenem (3/18 (17%)), while cefiderocol 4/18 (24%) was the most commonly prescribed monotherapy (figure S2). Of the 30 patients, six (20%) died during their hospitalization.

Genetic diversity, resistome, and phylogenetics

We analyzed thirty CRAb isolates. Nine different sequencing types (STs), based on Oxford groups, were identified, while two were novel STs (Figure 1). STs were not consistent with and any epidemiological indications such as LTCFs, year, or community. Interestingly, we found several highly genetically related isolates across different facilities and communities throughout the six-year study period. However, despite the genetic diversity among clinical isolates, there was a high degree of homogeneity in their AMR gene profiles, with 90% co-expressing bla _OXA-23, bla _OXA-66, and bla _ADC-25, 86.67% arm(A), and 70% tet(B). Additionally, SNP analysis revealed the isolates from two ICU patients in adjacent rooms differed by 0 SNPs, confirming nosocomial transmission (Figure S3). Two other patients admitted at the same ICU also had genetically related isolates that differed by 25 SNPs. Although their hospital stays were more than 30 days apart, both patients were from the same SNF, suggesting possible intra-SNF transmission.

Figure 1.

Diversely isolated A. baumannii complex samples display genotypic and phenotypic antimicrobial resistance patterns. Feature frequency profile via Jensen-Shannon divergences phylogenetic tree constructed from K-mer alignment. ST: sequence type; NST: new sequencing type; N/A: not applicable; year: year of isolation. Prior residence includes long-term acute care (LTAC), skilled nursing facility (SNF), outside hospital (OSH), home derived (HD), donor derived (DD). Antibiotic susceptibility results were from micro-broth dilutions and the presence of resistance genes from WGS results of each isolate are displayed.

Correlation between genotypic and phenotypic antimicrobial resistance profiles

Presence of specific AMR genes including bla _OXA-23, bla _NDM, and tet(B) were highly predictive of phenotypic susceptibility profiles. The combination of bla _OXA-23, bla _OXA-66, and bla _ADC-25 was the predominant β-lactamase, with all isolates harboring this combination (100%, 27/27) demonstrating resistance to all tested cephalosporins, carbapenems, and ampicillin/sulbactam when susceptibility results were available. Additionally, bla _NDM-1 was identified in four isolates (1 in 2023, 1 in 2025 and 2 in 2026), all of which (100% 4/4) demonstrated resistance to all cephalosporins, carbapenems, and ampicillin/sulbactam antibiotics, however all four isolates also contained bla _OXA-23, bla _OXA-66, and bla _ADC-25. The presence of non- bla _OXA-23 genes (bla _OXA-64 and bla _OXA-223) showed less of a resistance phenotype with either imipenem or ampicillin/sulbactam susceptibility. Tet(B), when present, was highly predictive (90.5%, 19/21) of minocycline resistance by AST.