What happened?

This is a hot take: no one was guilty of mal-practice.

Did she have Bipolar?

Why do we think she was Bipolar? Because she was restless, 'agitated', and not sleeping? Bipolar is defined as hypomanic or manic episodes. Labeling restless patients as bipolar is exactly what leads to over-diagnosis and inappropriate medication. I actually do not see any evidence that she was genuinely bipolar. The forensic psychiatrist may have said so, but why? Because she killed her kids = psychosis = bipolar? That feels like a convenient diagnosis made after the fact. Everything I've read about her state of mind after the killings was that she was calm, cooperative, goal directed, focused on recovery, etc. There is nothing in the history that seems like mania to me. I don't think anyone before or after the killings saw mania. She was thought maybe to be in a "mixed state" by one of the NPs, a trash can diagnosis if I ever saw one. There is then the notion that well she did not respond to SSRIs. So? She has Bipolar then? Many people do not respond to SSRIs. There are meta analysis that SSRIs are only marginally better than placebo. Antidepressant treatment unfortunately is largely garbage to begin with. I think it is most likely she had post partum depression and anxiety, unlikely she did or currently has bipolar. The psychosis part is up in the air. But I am seeing zero evidence from clinicians or her family before or after that she was psychotic.

MA prescription drug monitoring program

As far as I know this system only tracks controlled substances, hence it would have missed all the antidepressants she was prescribed. Unfortunately we live on a disjointed medical system. It should be corrected but this is a matter of political will, not individual clinician responsibility. I have seen numerous patients who cannot even recall their doctors name or where they practice hence making it impossible to send a discharge summary. Or they simply refuse. Yes, in my state a patient can refuse to disclose their medical history including psych notes to another provider including other psych providers. The only thing I absolutely am guaranteed access to is their medication history based on the prescription drug monitoring program (PDMP). However as said above I believe MA PDMP only includes controlled substances because it serves as an overdose prevention measure, not for continuity of care.

Autonomy

95% of patients at my hospital decline IOP or PHP on discharge although we have a program with plenty of availability. It would be impossible to compel people to go if you legally could. Not to mention patient autonomy still matters. It's easy to say in hindsight her rights should have stripped and she should have forced into a long term hospitalization until she was definitively better, but in real life it does not work like this because people have rights and we have moved away from mass institutionalization of the 1960s.

PPV of SI and HI

SI and HI are incredibly difficult to predict. All of the instruments developed and used by EDs and other high acuity environments have been definitively shown to have no predictive value for determining who is an imminent suicide or homicide risk. It is the reality of the situation. One thing to be sure of is there are a million Lindsay Clancy's out there. Case histories that seem almost identical and yet a catastrophic outcome as we saw in this case is immensely rare. Suicidal and homicidal ideations have an incredibly low positive predictive value (PPV). In the studies I have read both are estimated to be < 1%. This means < 1% people with suicidal or homicidal ideation will act out these thoughts. SI especially is incredibly common in the general population. If you hospitalized every patient with SI not only would that violate the autonomy of millions of individuals, 99% will never act on those thoughts, but you would need 100x more hospital beds than we currently have. This is why this is not criteria for involuntary commitment in and of itself. We look at do they have a plan, intention, means, etc. Even then it largely feels like a guessing game.

The bottom line. This work is hard. You are balancing patient safety v autonomy. You have multiple diagnostic considerations without clear cut answers. You are not practicing in a theoretical perfect environment where everybody is admitted to the hospital for exactly as long as they need followed immediately by partial hospitalization. You are practicing in an environment where patients decline your recommendations, minimize their symptoms, seek treatment from other providers often without your knowledge, family disagrees with your diagnosis, stop taking medications, use drugs and alcohol or otherwise self defeating behavior. And even if none of these occur and we will assume none of the above apply to Lindsay, the nature of the illness may remain so obfuscating the unbelievable can still happen.