Infections were recorded in nearly three out of every 10 adult hospitalizations for liver transplantation in a nationwide U.S. analysis. Compared with hospitalizations without recorded infections, those with infections were associated with higher odds of death and complications, longer stays and higher adjusted costs. The findings describe an association; they do not show that infections caused the worse outcomes.
Sahiljot Singh Bhupal and colleagues analyzed 50,630 adult liver-transplant hospitalizations in the National Inpatient Sample, a database of U.S. hospital discharges. The retrospective study covered 2016 through 2022 and was published in the peer-reviewed journal PLOS ONE. Infections were recorded in 15,130 hospitalizations, or 29.9% of the total.
The most frequently reported infection types were pneumonia (8.5%), urinary tract infections (8.3%) and cholangitis (7.5%). The database grouped cases by whether an infection was recorded during the transplant hospitalization. It could not establish whether those infections began before, around or after the transplant operation.
After adjusting for measured differences between the groups, the researchers found that infection-coded hospitalizations had higher odds of in-hospital death (adjusted odds ratio 1.47) and total complications (1.65). They also had higher odds of discharge to a non-home setting (1.89). These ratios compare odds, not the probability that a particular person will experience an outcome.
The analysis also estimated a substantial difference in hospital use. Adjusted predicted stays averaged 30.1 days for hospitalizations with infections, compared with 16.2 days without them. Adjusted predicted costs were $235,953 and $156,336, respectively; the study expressed costs in 2022 U.S. dollars. These figures point to a greater hospital burden alongside the worse outcomes, though the analysis cannot determine how much of the difference was due to infection itself.
The nationwide scope offers a broad picture of infection diagnoses and outcomes during transplant-related admissions. The results may help frame questions about infection prevention and care in this population, but they do not test a prevention strategy or show that any particular intervention would improve outcomes.
The study’s records did not include microbiological details, so the researchers could not identify the specific organisms or antimicrobial resistance patterns in these cases. The database also captures coded diagnoses for the full admission, rather than the timing or duration of each infection. It does not include infections treated outside the hospital or those occurring after discharge. As with other observational analyses, coding errors and unmeasured differences between groups may affect the results, and adjustment cannot rule out residual confounding.
The findings therefore show that recorded infections and worse hospital outcomes occurred together in this national sample, while leaving open when infections arose and how much they contributed to those outcomes. The authors say prospective studies with microbiological and biochemical data are needed to clarify infection timing and refine risk assessment.