Is decision support supporting the problem? The short answer is “yes.”
The more complex answer is "it doesn't have to."
But first, the problem: transplant teams need data to be able to accept or reject an organ offer. They take in huge amounts of data for each and every organ offer they receive. The data include over 800 variables from the donor captured per OPTN policy, hundreds more from the recipient EHR, and thousands more within the rich contextual meta-data captured within text messaging and phone calls.
The industry at large has tried to support decision making with all of this data. Unfortunately, in its current form, it’s only supporting the problem because transplant teams have no uniform way to share data among their team.
This often results in miscommunication, repeating information to multiple parties, and data getting lost in group text threads.
Take liver transplantation, for instance. The transplant surgeon must consider multiple risk factors for the donor organ, including cause of death, age, and cold ischemia time. They also need to consider the recipient, because not every organ would be a good match for every recipient.
Then, there are other complicating factors that influence their decision, such as practice variation, cognitive bias, policy and regulatory changes, and other competing transplant centers. They also consider how a transplant will reflect on them and their center. When a patient dies waiting, it's God's fault [organ shortage]. When a patient dies after getting a transplant, it's the surgeon or transplant center's fault. This results in a bias toward conservative organ acceptance practices
The final piece that is often left out of the decision making process is patient preference, and not because transplant teams want to leave them out. A busy transplant center often doesn’t have the time or platform to collect patient concerns and preferences.
For decision support to truly support organ offer decisions, transplant teams need a better way to share data internally and with OPOs as well as to ensure patients stay involved in the process.
How to support organ offer decisions
To improve decision support, transplant teams need a tool like TXP Chat that allows them to house all organ offer data in one centralized location.
Data should be accessible and actionable
Currently the data transplant surgeons need to accept or reject an organ offer is difficult to access. For transplant teams, this complicates an already intricate process. Transplant teams should have anytime, anywhere access to data to get to a decision faster and to communicate more effectively.
Insights should be outcome-driven
Technology is advancing rapidly in ways that can benefit transplant teams and their patients.
In particular, AI has been used on multiple platforms to match user data with their interests or needs. An obvious example of this is social media. It connects users with suggested connections and ad content based on their interests and Internet searches.
The technology already exists to analyze patient and organ data and put it into a useable format. Transplant teams can use this analysis to match the right patient to the right organ to lower the risk of graft failure and improve patient outcomes.
Patient should be involved in decision making
Transplant patients who are involved with their organ offer decision have improved survival rates. When an organ offer decision needs to be made fast, however, the transplant team often doesn’t have time to find out how much risk a patient is comfortable with or to succinctly communicate risk/benefit tradeoffs. Even if a patient communicated their preferences ahead of time, that information may not be readily available when an organ offer comes in.
AI can be used here as well. Patient preferences can be stored along with other patient information, and when an organ offer comes in, preferences can be included in the analysis to see if an organ is the right fit for a patient.
Is your decision support supporting your organ offer decisions? Download this study on predicting better kidney transplant outcomes with machine learning.