Impact of Pre-Transplant Frailty on Early Outcomes Following Liver Transplantation: A Propensity-Matched Multicenter Cohort Study
1
Department of Internal Medicine, Lahey Hospital and Medical Center, Burlington, MA 01805, USA
2
Department of Medicine, Royal College of Surgeons in Ireland, Medical University of Bahrain, Adliya P.O. Box 15503, Bahrain
3
Department of Gastroenterology and Hepatology, Salmaniya Medical Complex, Manama P.O. Box 12, Bahrain
*
Author to whom correspondence should be addressed.
J. Clin. Med. 2026, 15(11), 4003; https://doi.org/10.3390/jcm15114003
Submission received: 22 April 2026
/
Revised: 10 May 2026
/
Accepted: 21 May 2026
/
Published: 22 May 2026
(This article belongs to the Section Gastroenterology & Hepatopancreatobiliary Medicine)
Abstract
Background/Objectives: Frailty is a validated predictor of waitlist mortality and perioperative risk in liver transplant candidates, but its association with early post-transplant outcomes in large real-world cohorts remains incompletely characterized. This study evaluated the association between administratively defined pre-transplant frailty and early clinical outcomes following liver transplantation. Methods: We conducted a retrospective cohort study using the TriNetX US Collaborative Research Network. Adults undergoing first-time isolated liver transplantation through February 2026 were included. Frailty was identified using ICD-10-CM codes for frailty, sarcopenia, cachexia, weakness, abnormal gait/mobility, or reduced mobility documented within 12 months before transplantation; patients coded only for nonspecific weakness were excluded from the frailty cohort. Patients underwent 1:1 propensity score matching using 18 baseline covariates, including demographics, comorbidities, laboratory values, albumin, and MELD-Na. The primary outcome was all-cause mortality at 7, 30, and 90 days. Secondary outcomes included acute kidney injury, prolonged mechanical ventilation, vasopressor requirement/hemodynamic instability, renal replacement therapy, ICU and hospital length of stay, and 90-day readmission. Sensitivity analyses used a restrictive ≥ 2-code frailty definition and substituted MELD 3.0 for MELD-Na in the propensity model. Results: Among 4860 eligible recipients, 742 had administratively defined frailty and 4118 did not. After matching, 730 patients remained in each group with well-balanced covariates. Administratively defined frailty was associated with higher mortality at 7, 30, and 90 days, with numerically smaller relative risks at later time points. It was also associated with higher risks of acute kidney injury, prolonged mechanical ventilation, vasopressor requirement/hemodynamic instability, renal replacement therapy, longer ICU and hospital stays, and 90-day readmission. Findings were directionally consistent in both sensitivity analyses. Etiology-stratified analyses were exploratory and showed no statistically significant heterogeneity across liver disease etiologies. Conclusions: In this large propensity-matched multicenter cohort, administratively defined pre-transplant frailty was associated with worse early outcomes after liver transplantation. Because frailty and several outcomes were identified using structured EHR and administrative data, findings should be interpreted as associative and hypothesis-generating. Prospective studies using validated frailty instruments and granular donor, intraoperative, and center-level variables are needed to confirm these findings.
1. Introduction
Frailty is a multidimensional syndrome of diminished physiologic reserve that confers vulnerability to adverse outcomes following physiologic stressors [1,2]. In patients with cirrhosis and end-stage liver disease, frailty has been consistently associated with waitlist mortality, hospitalization, and impaired functional recovery [1,2]. The Liver Frailty Index (LFI), a validated performance-based tool incorporating grip strength, chair stands, and balance, has emerged as the reference standard for frailty assessment in this population and is endorsed by the 2025 AASLD/AST Practice Guideline on Liver Transplantation for pre-transplant evaluation [2,3,4].
The landmark multicenter FrAILT study (n = 1166 recipients across 8 US centers) demonstrated that pre-transplant frailty (LFI ≥ 4.5) was independently associated with post-transplant mortality (adjusted HR 2.13; 95% CI: 1.39–3.26), prolonged hospitalization (OR 2.00), prolonged ICU stay (OR 1.56), and nonhome discharge (OR 2.50) [5]. More recently, the Liver Transplant Comorbidity Index (LTCI) integrated frailty with coronary artery disease, hepatocellular carcinoma, renal dysfunction, and diabetes to predict 3-year post-transplant mortality, further establishing frailty as a central component of composite risk stratification [6]. Importantly, liver transplantation has been shown to provide net survival benefit at all levels of frailty, underscoring that the clinical goal is not to exclude frail candidates but to optimize perioperative care and risk stratification [4,7].
Despite these advances, several knowledge gaps persist. First, the FrAILT study evaluated frailty using the Liver Frailty Index but did not report granular early post-operative outcomes such as AKI, RRT, vasopressor requirement, or prolonged mechanical ventilation [5]. Second, the temporal dynamics of administratively defined pre-transplant frailty-associated risk—whether associations are strongest immediately postoperatively and numerically attenuate over time—have not been systematically characterized. Third, whether administratively defined pre-transplant frailty-outcome associations differ by underlying liver disease etiology remains unexplored. Fourth, emerging evidence on prehabilitation programs suggests that frailty may be modifiable before transplantation, with demonstrated improvements in LFI and 6 min walk test performance, but identifying which early outcomes are most associated with frailty is essential to guide prehabilitation targets [8,9].
To address these gaps, we conducted a large multicenter retrospective cohort study using the TriNetX US Collaborative Research Network [10,11]. This study aimed to evaluate the association between administratively defined pre-transplant frailty and a comprehensive spectrum of early post-transplant outcomes, describe temporal patterns in these associations, and explore etiology-specific differences in frailty-outcome associations.
2. Methods
2.1. Data Source and Study Design
We performed a retrospective cohort study using the TriNetX US Collaborative Research Network, a federated real-world electronic health record database and analytics platform provided by TriNetX, LLC (Cambridge, MA, USA). Available data include demographics, diagnoses (ICD-10-CM), procedures (CPT/ICD-10-PCS), medications, laboratory values, and vital signs. This study did not require institutional review board approval because the data were de-identified and HIPAA-compliant. The study was conducted in accordance with the STROBE guidelines for observational research.
Because TriNetX is a federated EHR network, patient-level chart review, detailed donor information, intraoperative variables, ventilator flow-sheet data, urine output, and center-specific transplant practices are not uniformly available. Therefore, exposures and outcomes were defined using structured diagnosis, procedure, medication, laboratory, and encounter data.
2.2. Study Population and Cohort Definitions
We identified all adult patients (≥18 years) who underwent first-time, isolated liver transplantation using ICD-10-PCS codes (0FY00Z0, 0FY00Z1, 0FY00Z2) and CPT code 47135. The index date was defined as the date of the transplant procedure.
Patients were classified into two cohorts based on the presence or absence of documented pre-transplant frailty within 12 months before the index transplant date:
Frailty cohort: Patients with at least one of the following ICD-10-CM codes docu-mented within 12 months before transplantation: R54 (age-associated physical debili-ty/frailty), M62.84 (sarcopenia), M62.81 (muscle weakness, generalized), R64 (cachexia), R63.4 (abnormal weight loss), R26.89 (other abnormalities of gait and mobility), Z74.09 (other reduced mobility), or R53.1 (weakness). To improve specificity, patients coded only with R53.1 (weakness) without any other qualifying code were excluded from the frailty cohort. This coding strategy was intended to identify an administratively defined, claims-based frailty phenotype rather than directly measured physiologic frailty. A single qualifying frailty-related ICD-10-CM code within 12 months before transplantation was sufficient for primary cohort assignment. This approach differs from performance-based frailty instruments such as the Liver Frailty Index, which directly assesses grip strength, chair stands, and balance.
Non-frailty cohort: Patients without any of the above codes documented within 12 months before transplantation.
Exclusion criteria (applied uniformly to both cohorts): age <18 years; incomplete demographic data (missing age, sex, or race); multiorgan transplantation (simultaneous liver-kidney, liver-heart, or liver-lung); retransplantation; acute/subacute hepatic failure due to trauma, acetaminophen overdose, or other acute toxic etiologies (ICD-10-CM K71.1x, T39.1x); and insufficient baseline data for propensity matching (defined as missing ≥3 of the 18 matching covariates).
2.3. Baseline Characteristics and Covariates
Eighteen baseline covariates were collected for propensity score matching and multivariable adjustment, including demographic variables (age, sex, and race/ethnicity [White, Black, Hispanic]); comorbidities documented within 12 months before transplant (type 2 diabetes mellitus [E11.x], hypertension [I10–I16], chronic kidney disease [N18.x], hepatic encephalopathy [K72.x, G93.4x], ascites [R18.x], hepatocellular carcinoma [C22.0], and pre-transplant dialysis [Z99.2]); and laboratory values measured within 90 days before transplant (total bilirubin, INR, creatinine, sodium, and albumin). MELD-Na was also included as a composite severity variable and was calculated from bilirubin, INR, creatinine, and sodium using the standard OPTN formula. MELD-Na was retained as the primary composite marker of liver disease severity because it remains widely used in transplant outcomes literature and allows direct comparison with prior landmark studies, including the FrAILT cohort and the Liver Transplant Comorbidity Index [5,6].
2.4. Study Endpoints
Primary outcome: All-cause mortality at 7, 30, and 90 days after liver transplantation.
Secondary outcomes:
- Acute kidney injury (AKI): Acute kidney injury (AKI) was defined using ICD-10-CM N17.x codes within the specified postoperative follow-up window. KDIGO-based AKI staging was not used because granular urine output data and sufficiently time-stamped serial creatinine values are not uniformly available across all participating healthcare organizations in the TriNetX federated platform [12]. Therefore, this definition likely captures clinically recognized AKI, particularly moderate-to-severe events, rather than all creatinine-defined AKI episodes.
- Prolonged mechanical ventilation: Prolonged mechanical ventilation was defined as >48 h of continuous mechanical ventilation using ICD-10-PCS code 5A1955Z or CPT codes 94002/94003 with duration >48 h. Ventilator flow-sheet data, ventilator settings, and exact extubation timing were not uniformly available in TriNetX; therefore, procedure-based definitions were used to ensure reproducibility across participating healthcare organizations.
- Vasopressor requirement or hemodynamic instability: This was defined by administration of norepinephrine, vasopressin, epinephrine, or phenylephrine, or ICD-10-CM coding for shock (R57.x), within the follow-up window. Because vasopressor dose, duration, and indication were not uniformly available, this outcome was intended to capture clinically significant postoperative hemodynamic instability rather than quantify vasopressor intensity.
- Renal replacement therapy (RRT): ICD-10-PCS 5A1D70Z/5A1D80Z or CPT 90935/90937/90945/90947.
- ICU length of stay (days).
- Hospital length of stay (days).
- 90-day all-cause hospital readmission.
2.5. Statistical Analysis
Baseline characteristics were compared using chi-square tests for categorical variables and Student’s t-tests or Wilcoxon rank-sum tests for continuous variables, as appropriate. To account for baseline differences between cohorts, we performed 1:1 propensity score matching (PSM) using the nearest-neighbor algorithm without replacement with a caliper width of 0.01 standard deviations of the logit of the propensity score. All 18 covariates listed above were included in the propensity score model. Covariate balance after matching was assessed using standardized mean differences (SMD), with values <0.10 considered indicative of adequate balance.
For each binary outcome, we calculated relative risks (RR) and risk differences (RD) with 95% confidence intervals (CI) at predefined follow-up intervals of 7, 30, and 90 days. For continuous outcomes, mean differences with 95% CI were calculated. To address time-to-event relationships, we performed Cox proportional hazards regression analysis in the propensity-matched cohort without additional covariate adjustment, as all measured baseline covariates were well balanced after matching. Hazard ratios (HR) with 95% CI were calculated for all primary and secondary outcomes. The proportional hazards assumption was assessed using Schoenfeld residuals.
Sensitivity and subgroup analyses included:
- Restricted analysis: Excluding recipients transplanted from the ICU or requiring pre-transplant mechanical ventilation or vasopressors, with new PSM performed within this subgroup.
- Restrictive frailty-definition sensitivity analysis: To improve specificity and address potential misclassification from single-code ascertainment, we repeated the primary analysis using a more restrictive, administratively defined frailty phenotype requiring at least two qualifying frailty-related ICD-10-CM codes within 12 months before transplantation. New PSM was performed within this sensitivity cohort.
- MELD 3.0 sensitivity analysis: To assess whether findings were robust to the use of a contemporary liver disease severity metric, we repeated PSM after substituting MELD 3.0 for MELD-Na in the propensity score model.
- Etiology-stratified analysis: Outcomes were examined within subgroups defined by primary liver disease etiology, including alcohol-associated liver disease, MASLD, viral hepatitis, and autoimmune/cholestatic liver disease, to assess consistency of frailty-outcome associations. These subgroup analyses were considered exploratory and were not powered for definitive subgroup-specific conclusions. Interaction p-values were used to assess statistical heterogeneity across etiologic subgroups.
Statistical significance was defined as a two-sided p-value <0.05. All analyses were conducted within the TriNetX analytics platform.
3. Results
3.1. Study Population
A total of 5996 adult patients with liver transplantation procedure codes were identified in the TriNetX US Collaborative Research Network. After sequential application of exclusion criteria, 4860 patients comprised the final eligible cohort: 742 (15.3%) with administratively defined pre-transplant frailty and 4118 (84.7%) without. After 1:1 PSM, 730 patients remained in each group. The patient selection process is illustrated in Figure 1.
3.2. Baseline Characteristics
Before PSM, the administratively defined frailty cohort was significantly older, more likely to be female, and had higher prevalences of type 2 diabetes, hypertension, CKD, hepatic encephalopathy, ascites, and pre-transplant dialysis, as well as higher MELD-Na scores and worse laboratory parameters. After PSM, all 18 covariates were well balanced (all SMD < 0.10) (Table 1).
3.3. Primary and Secondary Outcomes
At 7 days, all-cause mortality was significantly higher in the administratively defined frailty cohort (RR 2.571; 95% CI: 1.102–5.998; p = 0.021). This association persisted at 30 days (RR 1.958; 95% CI: 1.224–3.133; p = 0.004) and 90 days (RR 1.646; 95% CI: 1.182–2.293; p = 0.003). Frailty was significantly associated with increased risk of AKI, prolonged mechanical ventilation, vasopressor requirement, and RRT at all assessed time points, with numerically larger relative risks at earlier postoperative time points. Ninety-day readmission was also significantly higher in frail recipients (Table 2).
3.4. Length of Stay and Continuous Outcomes
Frailty was associated with significantly longer hospital and ICU stays, higher peak creatinine, greater creatinine rise from baseline, and longer ventilator duration among ventilated patients (Table 3).
3.5. Cox Proportional Hazards Model
Cox regression confirmed significantly higher hazards for all primary and secondary outcomes in frail recipients (Table 4). The proportional hazards assumption was met for all models (Schoenfeld residual global test p > 0.05 for all outcomes).
3.6. Sensitivity Analyses
3.6.1. Restricted Acuity Analysis
To address residual confounding by pre-transplant illness acuity, analyses were repeated in a restricted subgroup excluding recipients transplanted from the ICU or requiring pre-transplant mechanical ventilation or vasopressors. After a new PSM within this subgroup (n = 540 per group), all covariates remained balanced (all SMD < 0.10). Associations persisted but were attenuated compared with the full cohort, suggesting that administratively defined frailty was not solely a surrogate for extreme pre-transplant illness acuity (Table 5).
3.6.2. Restrictive Frailty-Definition Sensitivity Analysis
To improve specificity and address potential misclassification from single-code ascertainment, the primary analysis was repeated using a more restrictive administratively defined frailty phenotype requiring at least two qualifying frailty-related ICD-10-CM codes within 12 months before transplantation. After repeated PSM, associations remained directionally consistent with the primary analysis across mortality, AKI, prolonged mechanical ventilation, vasopressor requirement, RRT, and 90-day readmission (Table 6).
3.6.3. MELD 3.0 Sensitivity Analysis
To evaluate whether findings were robust to the use of a contemporary liver disease severity metric, PSM was repeated after substituting MELD 3.0 for MELD-Na in the propensity score model. The direction and magnitude of associations remained broadly consistent with the primary analysis, including for mortality, AKI, prolonged mechanical ventilation, vasopressor requirement, RRT, and 90-day readmission (Table 7).
3.7. Etiology-Stratified Subgroup Analysis
To explore whether associations between administratively defined frailty and early outcomes differed by underlying liver disease etiology, 30-day mortality, 7-day AKI, and 7-day prolonged mechanical ventilation were examined within etiology-defined subgroups from the matched cohort. These analyses were exploratory and were not powered for definitive subgroup-specific conclusions. Associations were directionally consistent across etiologic subgroups, although several subgroup-specific estimates did not reach statistical significance. Interaction p-values were non-significant, indicating no statistically detectable heterogeneity across etiologies. Therefore, these findings should be interpreted as supporting the overall consistency of the association rather than demonstrating clinically meaningful etiology-specific differences (Table 8).
4. Discussion
In this large propensity-matched multicenter cohort of 1460 liver transplant recipients, administratively defined pre-transplant frailty was associated with worse early post-transplant outcomes across multiple domains, including mortality, AKI, prolonged mechanical ventilation, vasopressor requirement, RRT, prolonged hospitalization, and 90-day readmission. Relative risks were numerically larger in the immediate postoperative period and smaller at later time points.
The mortality findings are directionally consistent with the landmark FrAILT study, which reported an adjusted HR of 2.13 (95% CI: 1.39–3.26) for post-transplant mortality in frail recipients defined by LFI ≥ 4.5 [5]. The present study’s 90-day mortality HR of 1.71 (95% CI: 1.28–2.28) is modestly lower, which may reflect differences in frailty ascertainment (EHR-based coding vs. validated performance-based instrument) and the resulting misclassification bias, which would be expected to attenuate effect estimates toward the null. This study builds on the FrAILT findings by showing associations between frailty and more granular early outcomes—AKI, RRT, prolonged mechanical ventilation, and hemodynamic instability—that were not reported in the FrAILT cohort [5]. These outcomes may represent potential targets for future perioperative optimization studies.
A key distinction between the present study and prior prospective frailty studies is that frailty was identified using ICD-10-CM codes rather than direct performance-based testing. Therefore, the exposure in this study should be interpreted as an administratively defined frailty phenotype or claims-based frailty phenotype, not as directly measured physiologic frailty. The included codes likely capture a heterogeneous clinical construct overlapping with frailty, sarcopenia, cachexia, malnutrition, generalized weakness, debility, disability, and mobility limitation. This approach may have lower sensitivity than prospective LFI-based assessment and may preferentially identify patients with clinically apparent functional impairment. Conversely, some included codes may reflect advanced liver disease or disability rather than frailty itself. These considerations may introduce exposure misclassification and should be considered when interpreting the findings. Prior studies have shown that claims-based frailty measures can stratify risk in administrative datasets, but they are not interchangeable with direct performance-based frailty instruments.
The association between frailty/sarcopenia and AKI is consistent with findings from a nationwide analysis of over 170,000 liver transplant recipients, which demonstrated that sarcopenia was independently associated with AKI (aOR 1.4; 95% CI: 1.32–1.49), shock (aOR 2.18), and in-hospital mortality (aOR 2.16) [13]. Similarly, frailty has been identified as an independent predictor of AKI in other surgical populations, including cardiac surgery and emergency laparotomy [14,15]. The recently published LTCI incorporated frailty alongside coronary artery disease, HCC, renal dysfunction, and diabetes into a composite index predicting 3-year post-transplant mortality [6]. The present study complements the LTCI framework by showing that administratively defined frailty was associated with early outcomes even after propensity score matching. Other recipient risk stratification models have similarly identified overlapping comorbidity domains (age, diabetes, renal dysfunction, ventilator status) as predictors of post-transplant mortality, but none have incorporated frailty as a distinct variable or examined its association with granular early postoperative outcomes [16,17].
An important observation was that relative risks were numerically larger at earlier postoperative time points and smaller at later time points across several measured outcomes. This may suggest that administratively defined pre-transplant frailty is most apparent as a risk marker during the immediate physiologic stress of transplantation, when cardiopulmonary reserve, respiratory muscle strength, renal vulnerability, and nutritional reserve may be challenged [18,19]. However, this temporal pattern should be interpreted cautiously. Alternative explanations include survivor bias, as the sickest, frail recipients who experience early mortality are no longer in the risk set at later time points; competing risks, as death may preclude the occurrence or capture of non-fatal outcomes; and differential discharge or follow-up patterns, particularly in a federated EHR database where post-discharge care outside participating healthcare organizations may not be fully captured [10,11]. Because formal time-varying effect modeling was not performed, these findings should be considered descriptive and hypothesis-generating rather than definitive evidence of biological attenuation of frailty-associated risk. These observations may help identify early postoperative outcomes for future studies of perioperative optimization, including early mobilization, nephroprotective strategies, and ventilator-weaning protocols, but whether such interventions modify risk in frail recipients requires prospective evaluation [20,21,22].
The etiology-stratified analysis was exploratory. Although point estimates varied numerically across liver disease etiologies, subgroup comparisons were underpowered for independent significance, and interaction p-values were non-significant. Therefore, there was no statistically detectable heterogeneity in the association between administratively defined frailty and early outcomes across etiologies. These findings should be interpreted as supporting the overall consistency of the association rather than suggesting meaningful etiology-specific differences.
The persistence of associations in the restricted cohort excluding ICU-transplant recipients and those requiring pre-transplant mechanical ventilation or vasopressors supports the robustness of the observed findings. However, this analysis should be interpreted cautiously because residual confounding by unmeasured illness acuity and transplant-specific factors cannot be excluded. The attenuation of effect sizes in the restricted cohort may reflect exclusion of the highest-acuity recipients rather than proving that administratively defined frailty represents a distinct biologic vulnerability state.
Taken together, these findings may have several clinical implications. First, they support further evaluation of structured frailty assessment during pre-transplant evaluation, consistent with the 2025 AASLD/AST Practice Guideline recommendation for LFI measurement [4]. Second, the association of administratively defined frailty with AKI, prolonged mechanical ventilation, and hemodynamic instability may help identify early postoperative outcomes that warrant closer monitoring and future study within perioperative optimization pathways, including nephroprotective strategies, respiratory support planning, early mobilization, and ICU care protocols [20,23]. Third, although the observed temporal pattern should be interpreted cautiously, these findings suggest that the immediate postoperative period may be a particularly important window for studying targeted interventions in vulnerable recipients. This aligns with emerging prehabilitation literature showing that frailty and functional capacity may be modifiable before transplantation, including improvements in LFI, 6 min walk test performance, VO2, and Short Physical Performance Battery measures in selected transplant candidates [8,9,24]. Fourth, the directionally consistent findings across etiologic subgroups suggest that frailty assessment may be relevant across liver disease etiologies, although subgroup analyses were exploratory and not powered for definitive etiology-specific conclusions. Finally, prior evidence that liver transplantation provides survival benefit across frailty strata reinforces that frailty-related risk information should be used to guide optimization and perioperative planning rather than to exclude candidates from transplantation [4,7].
These implications are strengthened by several features of the present study. The TriNetX platform provided a large, geographically diverse, multicenter cohort that is broadly representative of US transplant centers [10,11]. The sample size (730 matched pairs) exceeded the total FrAILT cohort (1166 in total) and enabled assessment of multiple secondary outcomes with adequate statistical power [5]. The comprehensive propensity matching on 18 covariates provided robust confounding adjustment. The multi-time point outcome assessment (7, 30, 90 days) enabled characterization of temporal attenuation patterns. The restricted acuity, restrictive frailty-definition, and MELD 3.0 sensitivity analyses support the robustness of the findings, while the etiology-stratified analysis provides exploratory evidence of directional consistency across liver disease etiologies.
Despite these strengths, several limitations warrant consideration. First, frailty was defined using ICD-10-CM administrative codes rather than validated performance-based instruments such as the LFI [2,3]. Therefore, the exposure should be interpreted as administratively defined frailty rather than directly measured physiologic frailty. Administrative frailty coding likely underestimates true frailty prevalence (15.3% in this study vs. 20–33% in prospective studies using LFI) and may introduce misclassification bias [5,6]. Although such misclassification would generally be expected to bias estimates toward the null, the included codes may also capture overlapping domains of sarcopenia, cachexia, malnutrition, weakness, debility, disability, mobility limitation, and advanced liver disease severity. Sensitivity analysis requiring at least two qualifying frailty-related ICD-10-CM codes yielded directionally consistent findings. Second, TriNetX does not capture several transplant-specific variables that may influence early postoperative outcomes, including donor age, donor type (DCD vs. DBD), graft quality, cold ischemia time, warm ischemia time, intraoperative blood loss, transfusion requirement, surgical complexity, center-level volume, and center-specific perioperative practices [16]. These factors may influence AKI, ICU length of stay, mechanical ventilation duration, vasopressor requirement, and early mortality through ischemia–reperfusion injury, hemodynamic instability, transfusion exposure, operative complexity, and variation in institutional care pathways. Therefore, residual confounding by unmeasured donor, intraoperative, graft-related, and center-level factors cannot be excluded. Third, several outcomes were identified using structured EHR and administrative data rather than manual chart adjudication. AKI was defined using ICD-10-CM codes rather than KDIGO creatinine- or urine output-based criteria [12]. This likely underestimates mild AKI and preferentially captures clinically recognized moderate-to-severe AKI. Vasopressor requirement may be underestimated if medication administration records are incomplete or overestimated if vasopressors were administered briefly for transient perioperative hypotension. Mechanical ventilation duration may be imprecisely captured because ventilator flow-sheet data and exact extubation timing are not uniformly available. Because these outcome definitions were applied uniformly to both cohorts, misclassification is likely to be largely nondifferential, although the exact direction and magnitude of bias cannot be determined. Fourth, the federated EHR design may result in incomplete follow-up if patients receive post-transplant care at non-participating institutions [10,11]. This limitation is particularly relevant for post-discharge outcomes such as readmission and later time point events. Fifth, MELD-Na rather than MELD 3.0 was retained as the primary composite liver disease severity variable to maintain comparability with prior transplant outcomes and frailty literature [5,6,16,17]. However, MELD 3.0 has been adopted for contemporary allocation and may better reflect current liver disease severity assessment [25]. In this study, sensitivity analysis substituting MELD 3.0 for MELD-Na in the propensity score model yielded directionally consistent findings. Sixth, the observed pattern of numerically smaller relative risks at later time points was descriptive. Formal time-varying effect modeling was not performed, and this pattern may reflect survivor bias, competing risks, differential discharge patterns, or incomplete capture of post-discharge outcomes rather than true biological attenuation of frailty-associated risk. Finally, the etiology-stratified analyses were exploratory and underpowered for definitive subgroup comparisons. Non-significant interaction p-values indicate no statistically detectable heterogeneity across etiologies. Given the observational design and administrative data source, all findings should be interpreted as associative rather than causal.
5. Conclusions
In this large multicenter propensity-matched cohort, administratively defined pre-transplant frailty was associated with worse early outcomes after liver transplantation, including higher mortality, AKI, prolonged mechanical ventilation, RRT, hemodynamic instability, longer hospitalization, and 90-day readmission. Relative risks were numerically larger in the immediate postoperative period and smaller at later time points, although this temporal pattern requires confirmation using formal time-varying analyses. Associations were directionally consistent across liver disease etiologies, with no statistically significant evidence of heterogeneity. Sensitivity analyses using a more restrictive frailty definition and substituting MELD 3.0 for MELD-Na yielded consistent findings. Because frailty and postoperative outcomes were identified using structured EHR and administrative data, these findings should be interpreted as associative and hypothesis-generating. Prospective studies using validated frailty instruments and granular donor, intraoperative, graft-related, and center-level variables are needed to confirm these findings and evaluate whether targeted prehabilitation or perioperative optimization can modify early post-transplant risk.
Author Contributions
Conceptualization, N.A. and H.A.; methodology, N.A. and H.A.; software, N.A.; formal analysis, N.A.; investigation, N.A., M.A. and S.I.; data curation, N.A.; writing—original draft preparation, N.A.; writing—review and editing, N.A., M.A., S.I. and H.A.; visualization, N.A.; supervision, H.A.; project administration, N.A. All authors have read and agreed to the published version of the manuscript.
Funding
This research received no external funding.
Institutional Review Board Statement
Ethical review and approval were waived for this study because it used de-identified data from the TriNetX US Collaborative Research Network and did not involve direct patient contact or access to identifiable personal health information.
Informed Consent Statement
Patient consent was waived because the study used de-identified data from the TriNetX US Collaborative Research Network.
Data Availability Statement
The data that support the findings of this study are available from TriNetX, but restrictions apply to the availability of these data, which were used under license for the current study and are therefore not publicly available. Data may be available from the authors upon reasonable request and with permission of TriNetX.
Conflicts of Interest
The authors declare no conflicts of interest.
Abbreviations
The following abbreviations are used in this manuscript:
| AASLD | American Association for the Study of Liver Diseases |
| AKI | Acute kidney injury |
| AST | American Society of Transplantation |
| CI | Confidence interval |
| CKD | Chronic kidney disease |
| CPT | Current Procedural Terminology |
| DBD | Donation after brain death |
| DCD | Donation after circulatory death |
| EHR | Electronic health record |
| HCC | Hepatocellular carcinoma |
| HCO | Healthcare organization |
| HIPAA | Health Insurance Portability and Accountability Act |
| HR | Hazard ratio |
| ICD-10-CM | International Classification of Diseases, Tenth Revision, Clinical Modification |
| ICD-10-PCS | International Classification of Diseases, Tenth Revision, Procedure Coding System |
| ICU | Intensive care unit |
| INR | International normalized ratio |
| KDIGO | Kidney Disease: Improving Global Outcomes |
| LFI | Liver Frailty Index |
| LOS | Length of stay |
| LTCI | Liver Transplant Comorbidity Index |
| MASLD | Metabolic dysfunction-associated steatotic liver disease |
| MELD 3.0 | Model for End-Stage Liver Disease 3.0 |
| MELD-Na | Model for End-Stage Liver Disease–Sodium |
| OPTN | Organ Procurement and Transplantation Network |
| PSM | Propensity score matching |
| RD | Risk difference |
| RR | Relative risk |
| RRT | Renal replacement therapy |
| SMD | Standardized mean difference |
| STROBE | Strengthening the Reporting of Observational Studies in Epidemiology |
References
- Lai, J.C.; Tandon, P.; Bernal, W.; Tapper, E.B.; Ekong, U.; Dasarathy, S.; Carey, E.J. Malnutrition, frailty, and sarcopenia in patients with cirrhosis: 2021 practice guidance by the American Association for the Study of Liver Diseases. Hepatology 2021, 74, 1611–1644. [Google Scholar] [CrossRef]
- Lai, J.C.; Rahimi, R.S.; Verna, E.C.; Kappus, M.R.; Dunn, M.A.; McAdams-DeMarco, M.; Haugen, C.E.; Volk, M.L.; Duarte-Rojo, A.; Ganger, D.R.; et al. Frailty associated with waitlist mortality independent of ascites and hepatic encephalopathy in a multicenter study. Gastroenterology 2019, 156, 1675–1682. [Google Scholar] [CrossRef]
- Kardashian, A.; Ge, J.; McCulloch, C.E.; Kappus, M.R.; Dunn, M.A.; Duarte-Rojo, A.; Volk, M.L.; Rahimi, R.S.; Verna, E.C.; Ganger, D.; et al. Identifying an optimal Liver Frailty Index cutoff to predict waitlist mortality in liver transplant candidates. Hepatology 2021, 73, 1132–1139. [Google Scholar] [CrossRef]
- Dove, L.; Chadha, R.M.; Lai, J.C.; DiMartini, A.; Liapakis, A.; Parikh, N.D.; Firpi-Morell, R.; Conteh, L.; Fallon, M.; Trotter, J.; et al. AASLD/AST practice guideline on adult liver transplantation: Candidate evaluation. Hepatology 2026, 83, 1609–1645. [Google Scholar] [CrossRef]
- Lai, J.C.; Shui, A.M.; Duarte-Rojo, A.; Ganger, D.R.; Rahimi, R.S.; Huang, C.-Y.; Yao, F.; Kappus, M.; Boyarsky, B.; McAdams-DeMarco, M.; et al. Frailty, mortality, and health care utilization after liver transplantation: From the multicenter Functional Assessment in Liver Transplantation (FrAILT) study. Hepatology 2022, 75, 1471–1479. [Google Scholar] [CrossRef]
- Lai, J.C.; Shui, A.M.; Molinari, M.; Rahimi, R.S.; Ladner, D.P.; Ganger, D.R.; Kappus, M.; King, E.A.; Tevar, A.D.; Volk, M.L.; et al. The Liver Transplant Comorbidity Index (LTCI): A composite index of ambulatory pre-LT factors to identify patients at increased risk of post-LT mortality. Hepatology 2025, 83, 294–303. [Google Scholar] [CrossRef]
- Wang, M.; Chiou, S.H.; Ganger, D.; Ruck, J.; Huang, C.-Y.; Kappus, M.R.; King, E.A.; Ladner, D.P.; Rahimi, R.S.; Duarte-Rojo, A.; et al. Liver transplantation provides survival benefit at all levels of frailty: From the multicenter Functional Assessment in Liver Transplantation study. Hepatology 2025, 81, 1269–1275. [Google Scholar] [CrossRef]
- Lin, F.-P.; Visina, J.M.; Bloomer, P.M.; Dunn, M.A.; Josbeno, D.A.; Zhang, X.; Clemente-Sanchez, A.; Tevar, A.D.; Hughes, C.B.; Jakicic, J.M.; et al. Prehabilitation-driven changes in frailty metrics predict mortality in patients with advanced liver disease. Am. J. Gastroenterol. 2021, 116, 2105–2117. [Google Scholar] [CrossRef]
- Abow-Mohamed, I.; Martel, M.; Janaudis-Ferreira, T.; Gillis, C.; Geraci, O.; Albader, F.; Chen, Y.-I.; Boruff, J.; Carli, F.; Tandon, P.; et al. Feasibility, safety, and effectiveness of prehabilitation programs in patients awaiting liver transplantation: A systematic review and meta-analysis. Am. J. Transplant. 2026, 26, 1110–1120. [Google Scholar] [CrossRef]
- Ludwig, R.J.; Anson, M.; Zirpel, H.; Thaci, D.; Olbrich, H.; Bieber, K.; Kridin, K.; Dempfle, A.; Curman, P.; Zhao, S.S.; et al. A comprehensive review of methodologies and application to use the real-world data and analytics platform TriNetX. Front. Pharmacol. 2025, 16, 1516126. [Google Scholar] [CrossRef]
- Stein, E.; Hüser, M.; Amirian, E.S.; Palchuk, M.B.; Brown, J.S. TriNetX Dataworks-USA: Overview of a multi-purpose, de-identified, federated electronic health record real-world data and analytics network and comparison to the US census. Pharmacoepidemiol. Drug Saf. 2025, 34, e70198. [Google Scholar] [CrossRef]
- Kidney Disease: Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group. KDIGO clinical practice guideline for acute kidney injury. Kidney Int. Suppl. 2012, 2, 1–138. [Google Scholar]
- Lin, J.; Dhruve, R.; Singh, C.; Singh Bhupal, S.; Sohal, A.; Roytman, M. Impact of sarcopenia on outcomes in liver transplant recipients: A nationwide analysis. Eur. J. Gastroenterol. Hepatol. 2025, 38, 361–367. [Google Scholar] [CrossRef]
- Bai, Y.-X.; Wang, Z.-H.; Lv, Y.; Liu, J.; Xu, Z.-Z.; Feng, Y.-Q.; Liu, G.-Y.; Yin, P.; Wang, Y.-T.; Dong, N.-G.; et al. Association between frailty and acute kidney injury after cardiac surgery: Unraveling the moderation effect of body fat through an international, retrospective, multicohort study. Int. J. Surg. 2025, 111, 761–770. [Google Scholar] [CrossRef]
- Proaño-Zamudio, J.A.; Arda, Y.; Perkins, T.M.; Nzenwa, I.C.; Argandykov, D.; Rafaqat, W.; Paranjape, C.N.; Velmahos, G.C.; Kaafarani, H.M.A.; Hwabejire, J.O. Frailty impacts outcomes for all older adult age groups in emergency laparotomy. Am. J. Surg. 2025, 249, 116571. [Google Scholar] [CrossRef]
- Asrani, S.K.; Saracino, G.; O’Leary, J.G.; Gonzalez, S.; Kim, P.T.; McKenna, G.J.; Klintmalm, G.; Trotter, J. Recipient characteristics and morbidity and mortality after liver transplantation. J. Hepatol. 2018, 69, 43–50. [Google Scholar] [CrossRef]
- Molinari, M.; Ayloo, S.; Tsung, A.; Jorgensen, D.; Tevar, A.; Rahman, S.H.; Jonassaint, N. Prediction of perioperative mortality of cadaveric liver transplant recipients during their evaluations. Transplantation 2019, 103, e297–e307. [Google Scholar] [CrossRef]
- Carey, E.J.; Lai, J.C.; Sonnenday, C.; Tapper, E.B.; Tandon, P.; Duarte-Rojo, A.; Dunn, M.A.; Tsien, C.; Kallwitz, E.R.; Ng, V.; et al. A North American expert opinion statement on sarcopenia in liver transplantation. Hepatology 2019, 70, 1816–1829. [Google Scholar] [CrossRef]
- Ooi, P.H.; Hager, A.; Mazurak, V.C.; Dajani, K.; Bhargava, R.; Gilmour, S.M.; Mager, D.R. Sarcopenia in chronic liver disease: Impact on outcomes. Liver Transpl. 2019, 25, 1422–1438. [Google Scholar] [CrossRef]
- Pollok, J.M.; Tinguely, P.; Berenguer, M.; Niemann, C.U.; Raptis, D.A.; Spiro, M. Enhanced recovery for liver transplantation: Recommendations from the 2022 International Liver Transplantation Society consensus conference. Lancet Gastroenterol. Hepatol. 2023, 8, 81–94. [Google Scholar] [CrossRef] [PubMed]
- Feizpour, C.A.; Patel, M.S.; Syed, M.A.; Carrasco, A.; Shah, J.; Hanish, S.; Sosa, L.; Fogus, S.; Bennett, S.; Shi, C.; et al. Enhanced recovery in liver transplantation: A value-based approach to complex surgical care. Surgery 2021, 170, 1830–1837. [Google Scholar] [CrossRef]
- He, Y.; Zhang, L. Efficacy of perioperative application of enhanced recovery after surgery on elderly recipients underwent liver transplantation. World J. Surg. 2025, 49, 1618–1624. [Google Scholar] [CrossRef]
- Campbell, N.; Beattie, C.; Gillies, M.A. Advances in the postoperative care of the liver transplant recipient. Curr. Opin. Crit. Care 2026, 32, 179–186. [Google Scholar] [CrossRef] [PubMed]
- Thuluvath, A.J.; Polineni, P.; Morrissey, S.; Belfanti, K.; Nizamuddin, M.; Siddiqui, O.; Daud, A.; Simpson, D.C.; Levitsky, J.; Flores, A.M.; et al. Home-based LIver FrailTy Intervention (LIFT) in transplant candidates: A feasibility study. Transplantation 2025, 109, e202–e212. [Google Scholar] [CrossRef] [PubMed]
- Kim, W.R.; Mannalithara, A.; Heimbach, J.K.; Kamath, P.S.; Asrani, S.K.; Biggins, S.W.; Wood, N.L.; Gentry, S.E.; Kwong, A.J. MELD 3.0: The Model for End-Stage Liver Disease Updated for the Modern Era. Gastroenterology 2021, 161, 1887–1895.e4. [Google Scholar] [CrossRef] [PubMed]
Figure 1.
Study flow diagram showing cohort selection, exclusion criteria, propensity score matching, and outcome analyses.
Figure 1.
Study flow diagram showing cohort selection, exclusion criteria, propensity score matching, and outcome analyses.
Table 1.
Baseline demographic and clinical characteristics before and after propensity score matching.
Table 1.
Baseline demographic and clinical characteristics before and after propensity score matching.
| Characteristic | Before PSM Frailty (N = 742) | Before PSM No Frailty (N = 4118) | p Value | SMD | After PSM Frailty (N = 730) | After PSM No Frailty (N = 730) | p Value | SMD |
|---|---|---|---|---|---|---|---|---|
| Age, years | 59.8 ± 10.7 | 56.4 ± 11.3 | <0.001 | 0.307 | 59.7 ± 10.7 | 59.5 ± 10.9 | 0.742 | 0.018 |
| Female | 286 (38.5%) | 1344 (32.6%) | 0.002 | 0.124 | 281 (38.5%) | 276 (37.8%) | 0.791 | 0.014 |
| White | 501 (67.5%) | 2751 (66.8%) | 0.718 | 0.015 | 494 (67.7%) | 489 (67.0%) | 0.779 | 0.015 |
| Black | 108 (14.6%) | 587 (14.3%) | 0.834 | 0.009 | 106 (14.5%) | 102 (14.0%) | 0.786 | 0.016 |
| Hispanic | 92 (12.4%) | 536 (13.0%) | 0.656 | 0.018 | 90 (12.3%) | 95 (13.0%) | 0.694 | 0.022 |
| Type 2 diabetes | 311 (41.9%) | 1325 (32.2%) | <0.001 | 0.202 | 304 (41.6%) | 298 (40.8%) | 0.760 | 0.016 |
| Hypertension | 422 (56.9%) | 2077 (50.4%) | 0.001 | 0.131 | 416 (57.0%) | 410 (56.2%) | 0.770 | 0.016 |
| CKD | 247 (33.3%) | 1039 (25.2%) | <0.001 | 0.179 | 242 (33.2%) | 236 (32.3%) | 0.714 | 0.018 |
| Hepatic encephalopathy | 388 (52.3%) | 1729 (42.0%) | <0.001 | 0.209 | 381 (52.2%) | 374 (51.2%) | 0.698 | 0.020 |
| Ascites | 471 (63.5%) | 2241 (54.4%) | <0.001 | 0.187 | 463 (63.4%) | 455 (62.3%) | 0.677 | 0.023 |
| Hepatocellular carcinoma | 141 (19.0%) | 896 (21.8%) | 0.091 | 0.069 | 140 (19.2%) | 144 (19.7%) | 0.821 | 0.013 |
| Pre-transplant dialysis | 86 (11.6%) | 307 (7.5%) | <0.001 | 0.141 | 82 (11.2%) | 79 (10.8%) | 0.810 | 0.013 |
| MELD-Na | 24.8 ± 7.3 | 22.6 ± 6.8 | <0.001 | 0.312 | 24.7 ± 7.2 | 24.5 ± 7.1 | 0.645 | 0.028 |
| MELD 3.0 | 25.6 ± 7.5 | 23.1 ± 7.0 | <0.001 | 0.345 | 25.4 ± 7.4 | 25.2 ± 7.3 | 0.604 | 0.027 |
| Creatinine, mg/dL | 1.42 ± 0.72 | 1.26 ± 0.65 | <0.001 | 0.232 | 1.41 ± 0.71 | 1.39 ± 0.69 | 0.650 | 0.029 |
| Bilirubin, mg/dL | 7.6 ± 6.3 | 6.5 ± 5.7 | <0.001 | 0.184 | 7.5 ± 6.2 | 7.4 ± 6.0 | 0.820 | 0.016 |
| INR | 1.91 ± 0.58 | 1.78 ± 0.52 | <0.001 | 0.235 | 1.90 ± 0.58 | 1.89 ± 0.56 | 0.761 | 0.018 |
| Sodium, mmol/L | 131.2 ± 5.7 | 132.5 ± 5.3 | <0.001 | 0.237 | 131.3 ± 5.6 | 131.5 ± 5.4 | 0.528 | 0.036 |
| Albumin, g/dL | 2.68 ± 0.49 | 2.83 ± 0.46 | <0.001 | 0.316 | 2.69 ± 0.49 | 2.70 ± 0.47 | 0.711 | 0.021 |
Abbreviations: CKD = chronic kidney disease; INR = international normalized ratio; MELD-Na = Model for End-Stage Liver Disease–Sodium; MELD 3.0 = Model for End-Stage Liver Disease 3.0; PSM = propensity score matching; SMD = standardized mean difference. Continuous variables are presented as mean ± standard deviation and categorical variables as n (%).
Table 2.
Relative risks and risk differences for clinical outcomes after propensity score matching.
| Outcome | Time Point | Frailty (n/N) | No Frailty (n/N) | RD (95% CI) | RR (95% CI) | p Value |
|---|---|---|---|---|---|---|
| All-cause mortality | 7 Days | 18/730 | 7/730 | 0.015 (0.002, 0.029) | 2.571 (1.102–5.998) | 0.021 |
| 30 Days | 47/730 | 24/730 | 0.032 (0.010, 0.055) | 1.958 (1.224–3.133) | 0.004 | |
| 90 Days | 79/730 | 48/730 | 0.042 (0.014, 0.070) | 1.646 (1.182–2.293) | 0.003 | |
| Acute kidney injury | 7 Days | 219/730 | 154/730 | 0.089 (0.043, 0.135) | 1.422 (1.196–1.692) | <0.001 |
| 30 Days | 290/730 | 224/730 | 0.090 (0.041, 0.140) | 1.295 (1.132–1.481) | <0.001 | |
| 90 Days | 331/730 | 275/730 | 0.076 (0.026, 0.126) | 1.204 (1.083–1.340) | 0.001 | |
| Prolonged mechanical ventilation | 7 Days | 196/730 | 121/730 | 0.103 (0.060, 0.146) | 1.620 (1.334–1.967) | <0.001 |
| 30 Days | 218/730 | 147/730 | 0.097 (0.051, 0.143) | 1.483 (1.240–1.774) | <0.001 | |
| Vasopressor requirement/hemodynamic instability | 7 Days | 241/730 | 162/730 | 0.108 (0.062, 0.154) | 1.488 (1.251–1.771) | <0.001 |
| 30 Days | 260/730 | 190/730 | 0.096 (0.048, 0.145) | 1.368 (1.173–1.595) | <0.001 | |
| Renal replacement therapy | 7 Days | 51/730 | 29/730 | 0.030 (0.008, 0.053) | 1.759 (1.127–2.747) | 0.012 |
| 30 Days | 74/730 | 49/730 | 0.034 (0.008, 0.060) | 1.510 (1.056–2.159) | 0.022 | |
| 90-day readmission | 90 Days | 167/730 | 128/730 | 0.053 (0.014, 0.092) | 1.305 (1.073–1.587) | 0.007 |
Abbreviations: CI = confidence interval; RD = risk difference; RR = relative risk.
Table 3.
Length of stay and selected continuous outcomes after propensity score matching.
| Outcome | Frailty (N = 730) | No Frailty (N = 730) | Mean Difference (95% CI) | p Value |
|---|---|---|---|---|
| Hospital length of stay, days | 18.6 ± 9.7 | 14.9 ± 8.3 | 3.7 (2.8 to 4.6) | <0.001 |
| ICU length of stay, days | 7.3 ± 4.8 | 5.9 ± 4.1 | 1.4 (0.9 to 1.9) | <0.001 |
| Peak creatinine during admission, mg/dL | 2.10 ± 1.02 | 1.80 ± 0.91 | 0.30 (0.20 to 0.40) | <0.001 |
| Change in creatinine from baseline to peak, mg/dL | +0.68 ± 0.74 | +0.41 ± 0.63 | 0.27 (0.20 to 0.34) | <0.001 |
| Ventilator days among ventilated patients | 5.8 ± 4.3 | 4.1 ± 3.6 | 1.7 (1.0 to 2.4) | <0.001 |
Abbreviations: CI = confidence interval; ICU = intensive care unit. Continuous variables are presented as mean ± standard deviation.
Table 4.
Cox proportional hazards models for outcomes associated with administratively defined pre-transplant frailty in the propensity-matched cohort.
Table 4.
Cox proportional hazards models for outcomes associated with administratively defined pre-transplant frailty in the propensity-matched cohort.
| Outcome | HR | Coefficient | SE | z | p Value | 95% CI |
|---|---|---|---|---|---|---|
| All-cause mortality | 1.71 | 0.536 | 0.148 | 3.62 | <0.001 | 1.28–2.28 |
| Acute kidney injury | 1.38 | 0.322 | 0.067 | 4.81 | <0.001 | 1.21–1.57 |
| Prolonged mechanical ventilation | 1.59 | 0.464 | 0.081 | 5.73 | <0.001 | 1.36–1.86 |
| Vasopressor requirement/hemodynamic instability | 1.44 | 0.365 | 0.074 | 4.93 | <0.001 | 1.24–1.67 |
| Renal replacement therapy | 1.47 | 0.385 | 0.139 | 2.77 | 0.006 | 1.12–1.94 |
| 90-day readmission | 1.31 | 0.270 | 0.088 | 3.07 | 0.002 | 1.10–1.56 |
Abbreviations: CI = confidence interval; HR = hazard ratio; SE = standard error.
Table 5.
Restricted subgroup analysis excluding recipients transplanted from the ICU or requiring pre-transplant mechanical ventilation or vasopressors (post-PSM, n = 540 per group).
Table 5.
Restricted subgroup analysis excluding recipients transplanted from the ICU or requiring pre-transplant mechanical ventilation or vasopressors (post-PSM, n = 540 per group).
| Outcome | Time Point | Events Frailty | Events No Frailty | RR (95% CI) | HR (95% CI) | p Value |
|---|---|---|---|---|---|---|
| All-cause mortality | 30 Days | 29 | 17 | 1.74 (1.01–2.99) | 1.52 (1.08–2.15) | 0.017 |
| AKI | 7 Days | 141 | 108 | 1.31 (1.08–1.59) | 1.27 (1.08–1.49) | 0.003 |
| Prolonged mechanical ventilation | 7 Days | 126 | 92 | 1.37 (1.09–1.73) | 1.34 (1.10–1.63) | 0.004 |
| Vasopressor requirement | 7 Days | 151 | 117 | 1.29 (1.07–1.57) | 1.26 (1.05–1.51) | 0.010 |
Abbreviations: AKI = acute kidney injury; CI = confidence interval; HR = hazard ratio; ICU = intensive care unit; PSM = propensity score matching; RR = relative risk.
Table 6.
Sensitivity analysis using a restrictive administratively defined frailty phenotype requiring ≥2 qualifying frailty-related ICD-10-CM codes within 12 months before liver transplantation.
Table 6.
Sensitivity analysis using a restrictive administratively defined frailty phenotype requiring ≥2 qualifying frailty-related ICD-10-CM codes within 12 months before liver transplantation.
| Outcome | Time Point | Frailty n/N | No Frailty n/N | RD, % | RR | 95% CI | p Value |
|---|---|---|---|---|---|---|---|
| All-cause mortality | 7 days | 14/512 | 6/512 | 1.6 | 2.33 | 0.90–6.03 | 0.071 |
| All-cause mortality | 30 days | 36/512 | 20/512 | 3.1 | 1.80 | 1.06–3.05 | 0.027 |
| All-cause mortality | 90 days | 61/512 | 39/512 | 4.3 | 1.56 | 1.06–2.30 | 0.021 |
| AKI | 7 days | 158/512 | 116/512 | 8.2 | 1.36 | 1.12–1.66 | 0.002 |
| AKI | 30 days | 204/512 | 163/512 | 8.0 | 1.25 | 1.07–1.46 | 0.005 |
| AKI | 90 days | 236/512 | 202/512 | 6.6 | 1.17 | 1.01–1.35 | 0.034 |
| Prolonged mechanical ventilation | 7 days | 143/512 | 96/512 | 9.2 | 1.49 | 1.19–1.87 | <0.001 |
| Vasopressor requirement/hemodynamic instability | 7 days | 169/512 | 122/512 | 9.2 | 1.39 | 1.14–1.68 | 0.001 |
| RRT | 7 days | 38/512 | 24/512 | 2.7 | 1.58 | 0.96–2.59 | 0.068 |
| 90-day readmission | 90 days | 118/512 | 92/512 | 5.1 | 1.28 | 1.01–1.62 | 0.039 |
Abbreviations: AKI = acute kidney injury; CI = confidence interval; ICD-10-CM = International Classification of Diseases, Tenth Revision, Clinical Modification; RD = risk difference; RR = relative risk; RRT = renal replacement therapy.
Table 7.
Sensitivity analysis using MELD 3.0 instead of MELD-Na in the propensity score matching model.
Table 7.
Sensitivity analysis using MELD 3.0 instead of MELD-Na in the propensity score matching model.
| Outcome | Time Point | Frailty n/N | No Frailty n/N | RD, % | RR | 95% CI | p Value |
|---|---|---|---|---|---|---|---|
| All-cause mortality | 7 days | 17/718 | 7/718 | 1.4 | 2.43 | 1.02–5.78 | 0.038 |
| All-cause mortality | 30 days | 45/718 | 24/718 | 2.9 | 1.88 | 1.17–3.02 | 0.008 |
| All-cause mortality | 90 days | 76/718 | 49/718 | 3.8 | 1.55 | 1.10–2.18 | 0.010 |
| AKI | 7 days | 214/718 | 153/718 | 8.5 | 1.40 | 1.17–1.67 | <0.001 |
| AKI | 30 days | 283/718 | 222/718 | 8.5 | 1.27 | 1.11–1.46 | <0.001 |
| AKI | 90 days | 324/718 | 273/718 | 7.1 | 1.19 | 1.06–1.33 | 0.003 |
| Prolonged mechanical ventilation | 7 days | 191/718 | 122/718 | 9.6 | 1.57 | 1.29–1.91 | <0.001 |
| Prolonged mechanical ventilation | 30 days | 212/718 | 148/718 | 8.9 | 1.43 | 1.20–1.71 | <0.001 |
| Vasopressor requirement/hemodynamic instability | 7 days | 236/718 | 164/718 | 10.0 | 1.44 | 1.21–1.71 | <0.001 |
| Vasopressor requirement/hemodynamic instability | 30 days | 254/718 | 192/718 | 8.6 | 1.32 | 1.13–1.54 | <0.001 |
| RRT | 7 days | 49/718 | 30/718 | 2.6 | 1.63 | 1.04–2.56 | 0.031 |
| RRT | 30 days | 71/718 | 51/718 | 2.8 | 1.39 | 0.98–1.98 | 0.063 |
| 90-day readmission | 90 days | 164/718 | 129/718 | 4.9 | 1.27 | 1.04–1.55 | 0.017 |
Abbreviations: AKI = acute kidney injury; CI = confidence interval; MELD = Model for End-Stage Liver Disease; RD = risk difference; RR = relative risk; RRT = renal replacement therapy.
Table 8.
Etiology-Stratified Subgroup Analysis: 30-Day Mortality, 7-Day AKI, and 7-Day Prolonged Mechanical Ventilation by Liver Disease Etiology (Matched Cohort).
Table 8.
Etiology-Stratified Subgroup Analysis: 30-Day Mortality, 7-Day AKI, and 7-Day Prolonged Mechanical Ventilation by Liver Disease Etiology (Matched Cohort).
| Etiology | n (F/NF) | 30-Day Mortality Events (F vs. NF) | 30-Day Mortality RR (95% CI) | 30-Day Mortality p | 7-Day AKI Events (F vs. NF) | 7-Day AKI RR (95% CI) | 7-Day AKI p | 7-Day MV Events (F vs. NF) | 7-Day MV RR (95% CI) | 7-Day MV p |
|---|---|---|---|---|---|---|---|---|---|---|
| Alcohol-associated | 263/259 | 18 vs. 9 | 1.97 (0.90–4.31) | 0.084 | 86 vs. 54 | 1.57 (1.17–2.11) | 0.003 | 74 vs. 43 | 1.70 (1.22–2.36) | 0.002 |
| MASLD | 117/114 | 10 vs. 4 | 2.44 (0.79–7.53) | 0.108 | 38 vs. 24 | 1.54 (0.98–2.43) | 0.058 | 30 vs. 18 | 1.62 (0.96–2.75) | 0.068 |
| Viral hepatitis | 175/179 | 12 vs. 7 | 1.75 (0.71–4.33) | 0.218 | 58 vs. 44 | 1.35 (0.97–1.88) | 0.076 | 51 vs. 34 | 1.53 (1.05–2.24) | 0.026 |
| Autoimmune/cholestatic | 81/84 | 4 vs. 2 | 2.07 (0.39–11.0) | 0.389 | 21 vs. 17 | 1.28 (0.73–2.25) | 0.387 | 22 vs. 14 | 1.63 (0.89–2.98) | 0.109 |
| Other/cryptogenic | 94/94 | 3 vs. 2 | 1.50 (0.26–8.76) | 0.651 | 16 vs. 15 | 1.07 (0.56–2.03) | 0.844 | 19 vs. 12 | 1.58 (0.82–3.07) | 0.170 |
| Total (matched cohort) | 730/730 | 47 vs. 24 | 1.96 (1.22–3.13) | 0.004 | 219 vs. 154 | 1.42 (1.20–1.69) | 0.001 | 196 vs. 121 | 1.62 (1.33–1.97) | 0.001 |
| Interaction p value | 0.872 | 0.641 | 0.914 |
Abbreviations: AKI = acute kidney injury; CI = confidence interval; F = frailty; MASLD = metabolic dysfunction-associated steatotic liver disease; MV = mechanical ventilation; NF = no frailty; RR = relative risk. Interaction p-values test for heterogeneity of the frailty effect across etiologic subgroups using the Breslow-Day test. Non-significant interaction p-values indicate that the frailty-outcome association was consistent across etiologies. Individual subgroups were underpowered for independent significance in several strata.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2026 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license.
Share and Cite
MDPI and ACS Style
Albusta, N.; Abdulla, M.; Isa, S.; Alrahma, H. Impact of Pre-Transplant Frailty on Early Outcomes Following Liver Transplantation: A Propensity-Matched Multicenter Cohort Study. J. Clin. Med. 2026, 15, 4003. https://doi.org/10.3390/jcm15114003
AMA Style
Albusta N, Abdulla M, Isa S, Alrahma H. Impact of Pre-Transplant Frailty on Early Outcomes Following Liver Transplantation: A Propensity-Matched Multicenter Cohort Study. Journal of Clinical Medicine. 2026; 15(11):4003. https://doi.org/10.3390/jcm15114003
Chicago/Turabian StyleAlbusta, Noor, Mohamed Abdulla, Sara Isa, and Hussain Alrahma. 2026. "Impact of Pre-Transplant Frailty on Early Outcomes Following Liver Transplantation: A Propensity-Matched Multicenter Cohort Study" Journal of Clinical Medicine 15, no. 11: 4003. https://doi.org/10.3390/jcm15114003
APA StyleAlbusta, N., Abdulla, M., Isa, S., & Alrahma, H. (2026). Impact of Pre-Transplant Frailty on Early Outcomes Following Liver Transplantation: A Propensity-Matched Multicenter Cohort Study. Journal of Clinical Medicine, 15(11), 4003. https://doi.org/10.3390/jcm15114003
Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.
