Physicians in emergency departments often default to an uncomplicated
diagnosis of psychosis, even when the presence of substance abuse should make
them more cautious.
Emergency department (ED) physicians frequently misdiagnose
substance-induced psychotic disorder, an initial error often compounded by
inappropriate follow-up treatment, according to a study in five New York
A more detailed review of their cases found that 25 percent of patients in
the study who were first diagnosed with a primary psychotic disorder actually
had substance-induced psychotic disorder or no psychosis at all, wrote Bella
Schanzer, M.D., M.P.H., now mental health director of the AIDS center at
Montefiore Medical Center in New York, and four colleagues in the October
Psychiatric Services. Schanzer was a research fellow in the
Department of Psychiatry at Columbia University when she did this study.
Referral to inpatient hospitalization, use of antipsychotic medications,
and referral to mental health or substance-abuse treatment varied according to
the diagnosis made in the ED, despite later evidence that in some cases the
initial diagnosis was incorrect.
The subjects received care in emergency areas that were solely focused on
psychiatric illnesses, and all diagnoses were made by physicians, although not
necessarily by psychiatrists, Schanzer told Psychiatric News.
Study results, the researchers said, “[highlight] the challenge of
accurately diagnosing a first psychotic episode when it occurs in the context
of substance use and underscore the potential for negative consequences if a
diagnostic error is made.”
What might account for this pattern?
“Probably a combination of three factors: a complicated clinical
presentation, lack of time in the ED, and gaps in physician training,”
said Schanzer in an interview.
Patients coming to an emergency department in a psychotic state are hardly
articulate reporters of their own medical history, including even whether they
have taken drugs recently.
Also, the medical system doesn't give patients adequate time to come off
drugs, said Schanzer. DSM-IV calls for a month of observation after
the patient ceases substance use, and hospital admissions averaged only 16
Finally, psychiatric training devotes too little time to learning about
addiction, she said. “Psychiatrists too often jump to a diagnosis of
primary psychosis, rather than substance-induced because that's how they were
In addition, patients were often not admitted to the same hospital where
the ED was located. That limited the chance for feedback from the hospital's
more deliberate evaluation to the psychiatrists in the ED.
“We can't even learn from the patients we admitted,” said
Schanzer and her colleagues studied patients from five Manhattan
psychiatric emergency departments who were diagnosed with early-phase
psychosis and had used drugs or alcohol in the previous 30 days.
They administered the Psychiatric Research Interview for Substance and
Mental Disorders, the Positive and Negative Syndrome Scale, and urine
toxicology screens at baseline, six months, and 12 months. Master's-level
clinicians or a physician interviewed patients at all three assessments.
Additional information came from ED records, inpatient hospital records,
caregivers, and reports from outpatient follow-up referrals.
A second set of expert diagnosticians looked over that information and made
a “best-estimate longitudinal diagnosis.” These diagnoses were
divided into substance-induced psychotic disorder, primary psychotic disorder,
or no psychotic disorder.
Of the 302 patients in the total sample, 223 (74 percent) were diagnosed in
the ED with primary psychotic disorder, 53 (18 percent) with substance-induced
psychotic disorder, and 26 (9 percent) with indeterminate symptoms.
The best-estimate diagnoses, however, found that only 195 patients (65
percent) had a primary psychotic condition, 101 (33 percent) had a
substance-induced psychosis, and six (2 percent) had no psychotic
Agreement between the two sets of diagnoses was only fair, wrote the
researchers. Fifty-six patients classified as having primary psychotic
disorder in the ED (false positives) actually had substance-induced disorder
(52) or no psychotic disorder (4), according to the best-estimate standards.
On the other hand, of the 53 patients initially diagnosed with
substance-induced psychotic disorder, 11 (21 percent) had primary psychotic
disorder by the best-estimate procedure.
Schanzer and colleagues expressed surprise at these results. They thought
that ED psychiatrists would more likely diagnose a substance-induced psychotic
disorder, since all the patients were known to have used alcohol, marijuana,
or cocaine during the previous month, and most had positive urine drug
Schanzer suggested that better diagnostic tools and improved training might
redress some of this imbalance.
“I don't know if we need a new diagnostic tool,” said Jon
Berlin, M.D., medical director of crisis services at Milwaukee County
Behavioral Health Division and an assistant clinical professor of psychiatry
at the Medical College of Wisconsin. “The DSM-IV criteria are
well-written and help in differentiating between substance-induced psychotic
disorder and primary psychosis. We might spend more time with residents
training them how to better distinguish the two by getting them to probe for
symptoms that did or did not occur prior to substance use or are more
pronounced than the drug would cause.”
Someone with either type of psychosis needs 24-hour care and antipsychotic
medications, but the focus depends on the etiology, added Berlin in an
interview. A urine drug screen should be mandatory and, if positive, can
prompt the physician to add substance-abuse treatment. “Ideally you pick
up the difference right away, but you can still shift gears.”
Diagnoses in the ED in Schanzer's study governed more than simply how
patients were treated there. When the ED diagnosed a primary psychotic
disorder, patients were more likely to be admitted for an inpatient stay or
started on antipsychotic medications, and not to be referred to outpatient
treatment for substance abuse. Those diagnosed with substance-induced
psychosis were less likely to be hospitalized, to be prescribed antipsychotic
medication, or to receive mental health treatment.
About 93 percent of patients thought to have primary psychotic disorder in
the ED were hospitalized, and 80 percent were prescribed antipsychotic drugs
at discharge, but only 66 percent of those with substance-induced disorder
were sent for hospital admission, and 32 percent received antipsychotics at
Many such patients would benefit from hospitalization, to permit more
accurate diagnoses, said Berlin. However, if they are no longer psychotic and
can function, they usually are released without a definitive diagnosis for
insurance reasons or because inpatient beds are limited. Services available to
these patients are poorly integrated in most places.
“People with substance-induced psychosis need antipsychotic
medications, but drug abuse [treatment] programs often don't have prescribers
for antipsychotics,” he said. “Plus, some psychiatrists don't want
to treat substance-abuse patients, and some substance-abuse centers don't want
The cultural gap between physicians and drug abusers should also be
addressed in training, he said. “Physicians have a hard time being
sympathetic with users,” he said. “They're turned off by their
need for instant gratification. It's the opposite of how doctors
“Diagnosing Psychotic Disorders in the Emergency Department in
the Context of Substance Use” is posted at<http://psychservices.psychiatryonline.org/cgi/content/full/57/10/1468>.▪