The Deterioration in Quality and Usefulness of the Notes Written by Psychiatric Nurses

The hospital notes of psychiatric nurses are starting to be as unrevealing of the patient's true condition as the notes by psychiatrists. Both types of notes are relying more and more on useless symptom checklists. As my friend James Woods observed, using a symptom checklist to make a diagnosis is only slightly more accurate than throwing darts at a dartboard - blindfolded.

Whenever a psychiatrist sees a new patient who had been treated by another clinician in the past, it is usually standard operating procedure for the doctor to obtain any prior medical or hospital records that pertain to that patient’s case. However, an ethical psychiatrist does not rely completely on another doctor’s diagnosis or treatment, but does his or her own evaluation to form an opinion. 

The previous doctor might have had things completely wrong. The former chairman of the academic department in which I was the residency training director, Dr. Neil Edwards, used to tell residents in training that the evaluation of each patient new to a clinician should take into account two different possibilities: first, that any prior clinician was competent, thorough and correct. Second, that the prior clinician was none of these things. 

In order to judge the diagnosis made by a mental health professional from the medical records, there has to be more than just a diagnosis on the chart or a quick “mental status” exam from one or two points in time. There has to be a narrative description of the patient. Particularly in a psychiatric hospital setting, there should also be a description somewhere in the chart of the patient’s ongoing behavior on the ward, particularly when the patient is not aware that he or she is being observed, and any major consistent changes in that behavior over time as treatment progresses.

To understand why this is important, let me relay some personal experience. I worked for a time in the main psychiatric emergency room in Memphis, from which many patients were referred for psychiatric hospitalization. After having evaluated a patient in the ER, I would later see the same patient in the hospital on weekends, when I was covering a service for the doctors who were in charge of the patient in the new setting. I would review the patient’s chart for the events and subsequent evaluations that had occurred since admission. One thing I witnessed time and time again – and not just with trainees but with the attending faculty as well – was quite striking. 

In the ER, some patients looked very depressed and showed all the signs (observed characteristics) and complained of all of the symptoms of clinical major depressive disorder. Their movements and speech were slow (psychomotor retardation), and they complained of chronic and persistent changes in appetite, sleep, pleasure, energy, and concentration. Since the patient seemed to meet all of the necessary DSM criteria, the diagnosis of major depressive disorder was therefore made by my resident, and I concurred.

However, in many cases in which I saw the patient in the hospital the very next day, or heard a description of the patient’s behavior on the ward by the nurses on the day following admission, an entirely different picture emerged. The patient was observed to be actively socializing with other patients, friendly and talkative on approach, speaking with normal rate and volume, sleeping on hourly bed checks (although they often had been given a sedative), and eating 100% of served meals!

One major characteristic that distinguishes true major depressive disorder from other types of depression is that its symptoms do not evaporate overnight, nor do they disappear merely because of a change of venue for the patient. They persist and are present in all environmental contexts. They can get to the point where a patient could win the lottery but not crack a smile. If symptoms change radically with a change in venue, then the primary problem for the patient resides in the venue, not in his or her mental disorder.

Clearly, everyone’s initial diagnostic impression of these patients – including my own – was incorrect. 

Yet surprisingly, the patients’ diagnosis was almost never changed during the entire course of their hospitalization. Their admission and discharge diagnoses were identical. Furthermore, they had been given treatments that patients who actually had major depression would be given.

Were my observations of this phenomenon biased in some way? Perhaps not changing a diagnosis in the light of new and conflicting information was something peculiar to doctors practicing in academic settings or only those practicing in Memphis, the city in which I was located. Possible, but not likely. How do I know?  Well, in over 30 years of practice, I have had the opportunity to review medical records of patients who had been psychiatrically hospitalized all over the United States. 

Nowadays, I often do not even bother to even send for hospital records, for they have become next to worthless for understanding a patient's condition. In the doctor’s initial evaluation, there would be no information about the time course, pervasiveness, or persistence of any symptom the patient was alleged to have had. There might be a "description" of symptoms that would consist of meaningless terms such as  “paranoid” or “suicidal” without any explanation of what the term meant as applied to the particular patient, or the context in which it occurred. Did “paranoid” mean delusional, or merely distrustful? No information.

With electronic medical records, I often have access to the notes and evaluations of several previous clinicians. Often no “target symptoms “ are described when a patient is put on a medication, and there is no information later on about what, if any, symptoms, had gone away in response to treatment. No way to know if the patient should have stayed on the medication. Sometimes the meds would be changed, but no reasons would be given for doing so. Did it not work? Or were there intolerable side effects? Or what?

With hospital records, I used to overcome the problem of physicians’ notes lacking any indication of how the patient was behaving on the ward by looking at the nurses’ notes. They used to actually describe such things. The patient might be noted the day after admission to eat, say, 100% of meals, be up and about socializing on the ward, sleeping on one hour bed checks, and pleasant and appropriate on approach.  If such a patient had been diagnosed as major depression or manic by the doctor the day before, I would know that this diagnosis was incorrect.

Imagine my reaction when I recently learned that the major local psychiatric hospital in my area, which is a haven for manufacturing bipolar disorder diagnoses where none exist, was instructing its nursing staff on what and what not to include in its nursing notes. I thought I’d better take a look at them to see what was being included and what was not.

It was far worse than I thought. There were very few narrative nurses notes at all! The main part of the nurses' notes consist of  a checklist which basically completely omitted the type of information I was looking for.

I asked a patient about it. She told me that the nurses would line up all the patients on their ward at certain times of day, take their blood pressures and other vital signs, and ask them if they were suicidal or not. The patient told me that the nurses spent almost zero time observing them on the wards!

I really had to look at even the checklists very carefully to glean anything about how the patient was behaving. Another patient was diagnosed as bipolar, but even on the checklist, the patient's energy was marked normal, affect was marked appropriate, and mood was checked "euthymic" (normal). Bipolar, my ass.

There was also a separate sheet with observations allegedly made every 15 minutes, which of course was inconsistent with what my other patient told me about how often they were closely observed. The patient's behavior on the sheet was again not a narrative description but a number with each digit defined so generally as to be diagnostically meaningless.  Even then, this patient was noted to be sleeping throughout the entire night.

To be fair, there were some narrative notes scribbled on the back of these checklists, often written by a psychiatric technician and not a nurse. Many were illegible, and they contained fairly minimal information. Even so, an allegedly bipolar patient was described as "isolating to room frequently."  [How frequently? Why? In this case the patient later told me that there was a specific reason for going to her room that had nothing to do with a mood disorder], "Pleasant mood, appetite good," and "no distress noted." [Bipolar? Really??]

What is the hospital trying to hide? You can bet that somehow limiting the validity of the information on the patients' charts helps them to maximize their reimbursements from insurance companies - perhaps hiding the fact that they want their doctors to diagnose patients with something serious rather than something the insurance companies might question. Patient welfare be damned.