Surgeon here.
I'm about more surprised by the discussion here than from the article itself.
> 2. There are many consults, but the ICU attending is king (or queen).
There's a concept called doctor's autonomy. The attending physician has the primary "guard" of the patient care, so unless dynamics of power, consultations are more like suggestions than law. So, the final care is generally dependent on the attending physician, for good or worse, be lack of confidence in the other physician be his perceived better understanding of the disease.
> 3. Sometimes nurses are the footsoldiers of the ICU regent, and sometimes they’re governors.
I saw examples of nursing saving and harming patients while disobeying orders. They have a co-participation in care and generally have studied to a degree that enable them to make some decisions.
> 4. Everyone agrees that sleep is important, but nobody has any idea beyond that.
We have decades worth of knowledge, but de facto we don't have a systematized and validated way of sleep care. We have studies on daytime nap and on sedatives effects on quality of sleep, but no full truths. Some day we'll have a better care.
> 6. The ICU staff is literally constantly changing.
The institutional memory are the patient medical records. If the Haloperidol adverse reaction was not noted in there, it was a fault of the care providers. Sometimes nurses chooses to ignore, and the repercussions should be analyzed case by case. The cited whiteboard worked as an "expanded" medical record, as registering that trigger could be seen as too tangential to a disease focused medical record.
> 7. The ICU is great at managing acute issues, and struggles a lot more with longterm issues.
Long term issues are not the concern of ICU. If it's not critical, the care can and maybe should be postponed until better. Of course, we have to be prudent, for example bowel function could be potentially urgent if not intervened early. Frequently I could and should not treat patients depression on an ICU, but it's reasonable to treat intrusive symptoms of early post-traumatic stress disorder, for example.
Free T4 is the method used to assess thyroid hormone supplementation, not TSH.
Delirium, delusions, illusions and hallucinations have a non-pharmacological and pharmacological treatment, and antipsychotics are not the only ones used.
> 8. The ICU is a good place to not die, but a bad place to recover.
The ICU is meant to give patients a better opportunity to not be critical anymore. When they're not critical, we start to deescalate our measures, such as monitoring and IV lines, for example.
People are different, and so are doctors. As the good, so the bad sprouts everywhere.