MELD Score Alternative

Q: What is the MELD score and how might we construct an alternative to the MELD score?

Source: Tsuriel
The original MELD score or Model For End-Stage Liver Disease score has a three-dimensional distribution

It uses the following variables:
  1. i) Serum creatinine,
  2. ii) Bilirubin, and
  3. iii) INR (International Normalized Ratio)

The MELD-Na score has a four-dimensional distribution
It uses a fourth variable if the MELD score > 11:
iv) Serum sodium

EXTRA: A MELD Score Calculator
A c-statistic gives the probability that a randomly selected patient who experienced the adverse healthcare outcome (e.g. disease, death) had a higher risk score than a patient who had not experienced this adverse healthcare outcome
∴ The c-statistic measures discrimination
We need to balance optimization (e.g. fewest number of dead patients on the waitlist) with equity-based considerations (e.g. gender, age, height, time spent on waitlist)
Beneficence and justice are two of the four principles of biomedical ethics that have been identified by Beauchamp & Childress (1979)  

Questions to ponder:
Q1: Which risk factors have been omitted by the MELD score?
Q2: How do we go about assigning weights to these additional risk factors?
Q3: How do we balance optimization with equity-based considerations in our alternative measure?
Q4: Can we achieve a higher c-statistic for our alternative measure than the MELD score c-statistic?