Investigators identified 50,630 adult hospitalizations for liver transplantation in the National Inpatient Sample from 2016 to 2022. Roughly 29.9% of those admissions carried at least one documented infection. The three most common infection types were pneumonia (8.5%), urinary tract infection (UTI, 8.3%), and cholangitis (7.5%). Bacteremia, Clostridioides difficile infection, and skin and soft tissue infections were less frequent, each occurring in approximately 2–3% of admissions annually. The cohort demonstrated a dose–response pattern: mortality rose as the number of distinct infection types increased, from 2.2% with no infection to 42.9% when four infections were recorded.
The infected cohort skewed middle-aged and male. Most hospitalizations with infections occurred in patients aged 45–65, and a majority identified as White. Private insurance represented the largest payer group among both infected and non-infected patients. Comorbidity burden—measured by the modified Charlson comorbidity index (mCCI)—was high across the cohort; however, a smaller proportion of infected admissions had mCCI ≥3 than non-infected admissions. The study team reported that certain underlying liver disease etiologies clustered with infection: alcohol-associated liver disease (ALD) and metabolic dysfunction-associated steatohepatitis (MASH) were common in both groups, while cholestatic disease appeared more frequently in infected admissions. Admissions with infections had higher prevalences of ascites, hepatic encephalopathy, hepatorenal syndrome, and variceal bleeding.
Data indicated that hospitalizations with infections required more organ-support interventions. The infected group had higher use of vasopressors, mechanical ventilation, and renal replacement therapy than the non-infected group. Correspondingly, admissions with infections carried a higher burden of complications: total complication prevalence was 71.2% with infections versus 56.0% without. Specific increases were seen in neurological, pulmonary, cardiovascular, and gastrointestinal complications. Transplant-related complications were also more frequent in the infected cohort. Genitourinary and endocrine complications did not differ meaningfully between groups.
The authors used multivariable logistic regression and survey-weighted generalized linear models, adjusting for demographics, hospital characteristics, mCCI, liver disease etiology, hepatic decompensations, and year. After adjustment, infections remained associated with worse outcomes. The adjusted odds ratio (aOR) for in-hospital all-cause mortality was 1.47 (95% CI 1.14–1.90). The aOR for total complications was 1.65 (95% CI 1.49–1.82), and the aOR for discharge to a non-home setting was 1.89 (95% CI 1.70–2.10). Adjusted predicted length of stay increased from 16.2 days without infections to 30.1 days with infections—an adjusted mean difference of 13.9 days. Adjusted predicted hospitalization costs rose from $156,336 to $235,953 in the presence of infections, with an adjusted mean difference of $79,618. These resource-use differences paralleled the higher intervention rates and complication burden described above.
The prevalence of most infection types remained relatively stable across the study period. UTI showed a downward trend from 2016 to 2022, while hospital-acquired pneumonia demonstrated year-to-year fluctuation with notable increases in some intervals. Bacteremia, C. difficile infection, and skin and soft tissue infections remained low and relatively constant. The study team presented these trends graphically and treated year as a categorical covariate in adjusted models to account for temporal variation.
The authors acknowledged the study’s reliance on administrative coding; infections and complications were identified using ICD-10-CM codes, and absence of a code was treated as absence of the condition. The NIS provides hospitalization-level (not patient-level) data and lacks granular clinical detail such as microbiology results, antimicrobial therapy timing, perioperative variables, donor factors, or outpatient follow-up. The source did not report specifics on timing of infection relative to transplant procedure, microbial etiology beyond broad categories, or antimicrobial management strategies. These gaps limit causal inference and prevent assessment of preventable versus unavoidable infections. Future work should clarify which infections are amenable to targeted prevention, the roles of perioperative screening and prophylaxis, and how infection timing influences post-transplant trajectories.