---

Summary: a decision analysis narrows the roles of the medical expert and jury within a malpractice trial, which can be favorable for the patient and provider.

---

In a recent post, I described how decision analysis defines the patient and provider’s roles within a medical decision (specifying utilities and probabilities, respectively). Decision analysis may also help define the role of the medical expert and jury in a malpractice trial, which I will now discuss.

At a population level, the legal system is essential to building and maintaining a just healthcare system. At an individual level, the legal system is important for protecting patients from unethical or incompetent medical care, which we can define more generally as “malpractice.”

Overall, I take a similar perspective to Brian Forst in his 1974 paper, where he argues that comparing expected utilities gives us a clear way to establish whether malpractice has occurred. However, I will assume that the parties who made the original medical decision already used a decision analysis, and I will go into more detail as to how a decision analysis redefines roles within a trial.

Decision analysis can help detect malpractice.  Recall that, as described in this post, under a decision analysis, a provider inputs the probabilities of outcomes,  Hence, one can analyze these probabilities to detect malpractice. For example, consider a case in which a provider claims that a procedure has a small probability of a complication, when the true complication rate is high.  A malevolent provider might do this for personal gain, and an incompetent provider might do this by accident.  Either way, input by the provider, will be incorrect, and this will be easy to spot.  

Thus, to analyze a medical decision that was made using a decision analysis, is cross-checked against relevant probabilities in the literature. This is the medical “testimony.” Therefore, the role of the medical expert, which just involves referencing the literature for this purpose, is narrow.

Also note that, because decision analysis defines to be the medical expert’s territory, a jury is spared from having to assess it, or from otherwise interpreting, as might usually be the case, a medical testimony.  From the jury’s perspective, there will be certainty with respect to whether an error in the provider’s input, was made. 

In such a case, the jury is freed to play a narrower role: inferring intent. Intent is difficult to infer algorithmically. Note that reveals whether a provider made an error, but one would need to observe multiple decisions over time to algorithmically distinguish malevolence from incompetence (for example, in the case of malevolence, the errors in the probabilities would be consistently skewed to promote self interest, whereas in the case of incompetence, the errors would be random). Since a malpractice trial usually only considers one decision, it is difficult to algorithmically infer intent. It is important, therefore, that the jury help with this.  This is a natural task for a jury—it involves social intuition and consensus.

In contrast, without a decision analysis, a jury would be expected to perform a substantially more unnatural, complex task: to make an overall gestalt assessment about whether a provider is guilty based on a medical outcome and the interpretation of a medical testimony.

In addition to inferring intent, the jury is also freed to focus its efforts on other aspects of the decision, such as whether the patient’s utility, or even autonomy, has been jeopardized (this would be easy, because a decision analysis is transparent about the patient’s utility, , and it would be clear if it were overwritten). 

Consider also the provider’s perspective. Note that a decision analysis shows them which boxes to check.  If these boxes are checked, the provider has done their job. Under the current, outcome-focused system, a provider might be held responsible for something that was mostly due to random chance. Accountability is good, but accountability for things outside of one’s control can cause fear that hinders cognition. Under a process-focused system, which decision analysis promotes, the provider would only be responsible for   This could ultimately lessen fear of frivolous lawsuits, leading to better cognition and improved decision making.

Overall, a decision analysis is algorithmic, and breakdowns in such a process are easy to evaluate. Decision analysis clearly defines the roles in a trial, in the same way it clearly defines roles in medical decision making more generally.