Abstract
Liver allocation is currently determined by the Model for End-Stage Liver Disease (MELDNa) score, an objective metric based on four laboratory measurements. Despite its intended objectivity, sex disparities in liver transplantation have increased during the MELD era. This study utilizes electronic health record (EHR) data from Vanderbilt University Medical Center (VUMC) and the All of Us Research Project to describe sex differences in MELD labs and propose a sex-adjusted scoring system. By mapping female scores to male scores of equal clinical disease severity, the researchers developed a sex-adjusted MELDNa map. Simulations using Liver Simulated Allocation Modeling (LSAM) demonstrate that this adjustment modestly increases female transplantation rates and decreases overall waitlist mortality, highlighting the necessity of sex-based modifications to ensure equitable organ allocation.
Introduction
In the United States, liver transplant prioritization follows the "sickest first" principle. Since 2002, the Model for End-Stage Liver Disease (MELD) score—and its current iteration, MELDNa—has served as the primary tool for ranking patients. The score is calculated using four laboratory values: creatinine, bilirubin, sodium, and the International Normalized Ratio (INR).
While the formula is applied equally to all candidates, it does not result in equitable outcomes. Women on the liver transplant waitlist are more likely than men to die or be removed due to becoming too ill for surgery. A well-documented factor is creatinine, a marker of kidney function influenced by muscle mass. Because women typically have lower muscle mass, their creatinine levels often underestimate their degree of renal impairment, leading to lower MELDNa scores compared to men with similar levels of organ failure. This research seeks to provide empirical evidence for a sex-adjusted MELDNa score to rectify these systemic disadvantages.
Methodology
The researchers employed a multi-stage approach to analyze sex-based disparities and develop a corrective mapping tool:
- Phenotyping and Data Extraction: Longitudinal lab data and ICD billing codes were extracted from two large-scale databases: Vanderbilt University Medical Center (VUMC), containing over 3.2 million records, and the NIH’s All of Us Research Project, a diverse multi-site cohort.
- Defining Disease Severity: To establish a "ground truth" for sickness independent of lab scores, the study used "decompensation traits." These are clinical manifestations of liver failure, such as jaundice, internal bleeding (variceal hemorrhage), and fluid accumulation (ascites).
- Quantile Matching: The team created a sex-adjusted MELDNa map by matching male and female patients based on age, race, and the number of decompensation traits. By comparing the distribution (quantiles) of MELDNa scores between these matched groups, they determined the score adjustment required to align female priority with male disease severity.
- Simulation: The proposed scores were tested using Liver Simulated Allocation Modeling (LSAM), the standard discrete event simulation software used by national policy-makers to predict the impact of allocation changes.
Key Insights
Pervasive Lab Disparities
While previous literature focused largely on creatinine, this study revealed that all four laboratory components of the MELDNa score exhibit significant sex differences. These differences consistently disadvantage female patients, resulting in lower scores despite equivalent clinical status.
The "Sicker at Transplant" Phenomenon
The data confirmed that women who successfully undergo transplantation typically present with significantly more clinical decompensation traits than their male counterparts. This suggests that the current system requires women to reach a higher threshold of physical illness before they are granted the same priority as men.
Mathematical Magnification
The MELDNa formula utilizes natural logarithms of laboratory values. This mathematical structure tends to magnify small biological differences in baseline lab levels, which further widens the gap between male and female scores at critical stages of disease progression.
Results
The analysis of 623,931 individuals from VUMC and 56,715 from All of Us yielded the following findings:
- Score Adjustment: The proposed mapping increases female MELDNa scores by an average of 1.6 points (with adjustments typically ranging from 0 to 2 points).
- Transplant Rates: In LSAM simulations, the sex-adjusted score increased the female transplantation rate from 23.0% to 24.1%.
- Mortality Impact: Implementing the sex-adjusted score reduced female waitlist mortality from 15% to 14%. Notably, the simulations also suggested a slight decrease in overall deaths across the entire waitlist, indicating that more accurate prioritization benefits the system as a whole.
- Validation: The sex-based differences in lab markers were successfully replicated across both the VUMC and All of Us datasets, confirming that the findings are not specific to a single geographic location or hospital system.
Discussion and Limitations
This study provides a practical, data-driven framework for adjusting MELDNa scores to account for biological sex. However, several challenges remain. EHR data, while robust, is not perfectly linked to real-time national transplant waitlists, which can complicate the tracking of patient listing and de-listing.
Furthermore, while a sex-adjusted score addresses the "priority gap," it does not solve the "size mismatch" problem. Women are generally smaller than men, and there is a persistent shortage of smaller donor livers. Achieving true equity may require combining sex-adjusted scoring with other policy shifts, such as increasing the use of split-liver transplants or improving access to pediatric donors for small-stature adults.
Conclusion
The current MELDNa system, while objective in its application, is biased in its outcomes. By implementing a sex-adjusted MELDNa map, the transplant community can better reflect the true severity of liver disease in female patients, potentially saving lives and making the allocation process more equitable for all.
Summary updated