Q: How do we model the state transitions of a patient on the liver transplant waitlist? | ||
---|---|---|
FIG. 1: Stages of Liver Damage |
Reasons for liver transplant include: i) Fibrosis or cirrhosis arising from an unhealthy liver. Decompensated cirrhosis (DC) refers to an acute deterioration in liver function in a patient with cirrhosis ii) Acute and acute-on-chronic liver failure (respectively: ALF and ACLF) iii) Hepatocellular carcinoma (HCC) or liver cancer |
|
There are two possibilities for liver transplant: OPTION 1: Deceased donor liver transplant (DDLT) DDLT takes place within a day of the organ harvest to minimize further deterioration of the donated organ OPTION 2: Living donor liver transplant (LDLT) LDLT typically involves a liver donor who volunteers and a liver recipient (usually a close relative) Due to the elective nature of LDLT, the procedure can take place at a time which optimizes the recovery process for both donor and recipient |
FIG. 2: Liver Transplant Image Source: Mayo Clinic |
|
FIG. 3: Allowable State Transitions for Liver Transplant Patients |
We must first identify the following: i) The unique initial states of all patients on the liver transplant waitlist (e.g. DC, ALF, ACLF, HCC, etc), including disease-combinations ii) The possible states into which these patients can transition (e.g. out of transplant criteria or OOC, recovered or WELL) iii) Death (DEATH) as the sole absorbing state of our model Thereafter, we must accept the following Markovian Assumption with respect to progression from one state si to another state sj: There is no memory of prior states and each transition depends only on the current state Last but not least, we must identify all allowable state transitions, including self-loop transitions and transitions into DEATH as the absorbing state |
|
Markov transition probabilities are computed relative to an anonymized dataset of liver transplant waitlist patients from: i) NUH (Singapore) ii) SGH (Singapore iii) Queen Mary Hospital (Hong Kong) |
FIG. 4: Logos of Hospital Partners |