Direct Cost of Care for Patients With Schizophrenia Treated With Haloperidol in an Emergency Unit

Economic Grand Rounds
Direct Cost of Care for Patients
With Schizophrenia Treated With
Haloperidol in an Emergency Unit
Nancy Ricard, M.Sc.
Luc Sauriol, M.Sc.
Marie-Claire Bélanger, R.N.
Guy Chouinard, M.D., M.Sc.
A 1991 study estimated that some patients with haloperidol. To this end, and the number of visits from general
250,000 Canadians suffered
we reviewed current resource utiliza-
practitioners, psychiatrists, psychiatry
from schizophrenia and that at least
tion in the department.
residents, and nurses.
105,000 adults were currently being
Data on resources used were ob-
treated for this illness (1). In recent
Methods
tained from the emergency unit and
years, the emphasis in schizophrenia
The study was conducted in the
hospital records. Cost data came from
management has shifted from long-
emergency department of the Royal
the May 1996 editions of the General
term inpatient care to community-
Victoria Hospital in Montreal. The
Practitioners’ Manual and the Med-
based resources. Because both fami-
records of psychiatric patients who
ical Specialists’ Manual of the Health
ly members and community provid-
came to the emergency department
Insurance Plan of Quebec. A visit by
ers have been unprepared to offer
between April 22 and September 17,
a general practitioner, which included
care for patients with schizophrenia,
1996, were extracted from emergency
a complete physical examination, was
the number of these patients pre-
room and hospital records. Patients
$31.65. A visit by the psychiatrist was
senting in hospital emergency units
with a DSM-IV diagnosis of schizo-
$75, and a follow-up psychiatric visit
has increased.
phrenia who were treated with oral
was $18.15.
Haloperidol has been widely used
haloperidol (N=19) or injectable hal-
Costs for oral and injectable an-
to treat patients admitted to emer-
operidol (N=1) were included in the
tipsychotics, hypnotics, antiparkinson-
gency units who are experiencing ex-
analysis. Because the study was retro-
ian drugs, and other medications used
acerbation of the symptoms of schizo-
spective, physicians and other health
to treat psychosis were calculated.
phrenia. To assess the cost impact of
professionals were unaware of our
The daily bed cost in the emer-
using new drugs in our emergency
analysis, and thus neither treatment
gency unit was calculated. One day
department, we were interested in
modalities nor the practices of profes-
was defined as at least an eight-hour
determining baseline costs of treating
sionals were influenced.
stay; bed costs were not added for
The study sought to determine the
stays less than eight hours.
proportion of patients with schizo-
The total cost of care in the emer-
When this work was done, Ms. Ricard
phrenia who were treated with
gency unit was calculated based on
was affiliated with the health economics
haloperidol and the mean cost of an
bed cost, costs for services provided
department of Hoechst Marion Roussel
emergency room stay for these pa-
by physicians and psychiatrists, and
Canada Research Inc., where Mr. Lauri-
ol is currently affiliated. Ms. Ricard is
tients. The patients were followed
costs of medications. Not included
now with Biocapital, a biotechnology ven-
from the time of their arrival in the
were costs for interventions by psy-
ture capital firm in Montreal. Ms.
emergency unit until their discharge
chiatric residents and nurses, because
Bélanger and Dr. Chouinard are with
from the unit or hospital. Only costs
their time with patients was not re-
the clinical psychopharmacology unit of
of care provided in the emergency
ported in the medical records. Indi-
the Allan Memorial Institute at the Royal
unit were calculated.
rect costs also were not included.
Victoria Hospital in Montreal. Dr. Choui-
A psychiatric nurse was trained to
nard is also with the department of psy-
use a questionnaire to collect patient
Results
chiatry at the University of Montreal.
data from the charts. Information col-
Of the 257 psychiatric patients who
Send correspondence to Ms. Ricard at
lected included age, gender, principal
visited the emergency unit over the
Biocapital, 3690 Rue de la Montaigne,
diagnosis, length of stay in the emer-
five-month study period, 56 (22 per-
Montreal, Quebec, Canada H3G288 (e-
mail, nricard@biocapital.com). Steven S.
gency department, medications pre-
cent) had a diagnosis of schizophre-
Sharfstein, M.D., is editor of this column.
scribed and quantity administered,
nia. Of the 56 patients, 20 (36 per-
PSYCHIATRIC SERVICES o October 1999 Vol. 50 No. 10
1287

---

Table 1
in medical records can be assessed
using a questionnaire or prospective
Use of health care resources by 20 patients with schizophrenia treated with
methods. However, in a prospective
haloperidol in a hospital emergency unit1
study it is important to ensure that
Cost per
N units
professionals provide the same types
Cost variable
service unit
used
Total cost
and amounts of care and do not mod-
ify their practices as a result of being
General practitioner
observed.
Complete physical examination
$ 31.65
6
$ 189.90
Because of the small sample size, it
Psychiatrist
Initial visit
75.00
19
1,425.00
was not relevant to adjust for vari-
Follow-up visit
18.15
19
344.85
ables such as disposition of patients
Medications2
318.29
after treatment, noncompliance, level
Bed cost in the emergency unit
251.85
29
7,303.65
of education, presence of a caregiver,
Total
$9,581.69
socioeconomic status, and urban or
1 Costs for care provided by psychiatry residents and by nurses were not calculated. Costs are re-
rural residence. Information on most
ported in Canadian dollars.
of these variables was not reported in
2 Oral and injectable antipsychotics, hypnotics, antiparkinsonian drugs, and other medications used
the medical records. However, they
to treat psychosis
could have affected the study results.
Aside from the basic cost of an
cent) received haloperidol as a treat-
mean daily bed cost in the emergency
emergency stay, services delivered by
ment, 11 (20 percent) received anoth-
unit was $251.65 per day; the total
psychiatrists represented a major cost
er antipsychotic, and 25 (45 percent)
cost for the 20 patients included
factor. Psychiatrists remained the
received no medication.
charges for 29 days. The mean±SD
most costly resource utilized. Nursing
Among the 20 patients treated with
cost for the 20 haloperidol-treated
services, such as patient monitoring,
haloperidol, 23 admissions to the
patients was $416.60 per emergency
recording of progress notes, and ad-
emergency unit were noted. Of the
visit. The total cost of medications for
ministration of medication, may have
20 patients, 18 (90 percent) were
these patients was $318.29.
accounted for more of the resources
men. The mean±SD age of the 20 pa-
utilized, but nursing services provid-
tients was 46±12 years (range=23 to
Discussion and conclusions
ed were not translated into monetary
74 years), with no significant differ-
This retrospective study identified
values.
ence between men and women. The
health care resources used by patients
The results of this study represent a
mean±SD length of stay in the emer-
in our emergency unit who were ex-
starting point. More research is need-
gency unit of the haloperidol-treated
periencing schizophrenic relapses
ed to better understand resource uti-
patients was 25±23 hours (17±12
and were treated with haloperidol.
lization in an emergency setting for
hours for men and 54±29 hours for
By using medical records, it was
patients with schizophrenia who are
women). Half of the patients treated
possible to evaluate the costs of med-
experiencing relapses. Costs for resi-
with haloperidol stayed for less than
ication, the daily cost of an emer-
dent and nursing care should be
17 hours, and three stayed for two
gency stay, and the fees for general
added to the model. The model could
hours. The mean length of stay for
practitioners and psychiatrists. Be-
be refined to account for differences
the haloperidol-treated patients tend-
cause the psychiatry residents and
in practice patterns and to determine
ed to increase with age, although the
nurses receive a fixed annual wage,
direct costs of resources used in other
differences between age groups were
the time they spent with each patient
emergency units. Calculating indirect
not statistically significant. A total of
was not documented. However, be-
costs, such as the social consequences
205 nurse interventions were record-
cause time spent caring for patients
of symptom exacerbation, would add
ed for the 20 patients, although, as
with acute exacerbation of symptoms
another dimension to the model. Fu-
noted, costs for these services were
of schizophrenia is a major cost factor,
ture studies should also compare
not included in the total cost.
it would be useful to obtain this infor-
costs for patients treated with oral or
Only three of the haloperidol-treat-
mation. One way to do so would be to
injectable haloperidol or for those
ed patients (15 percent) were trans-
conduct a time-and-motion study.
treated with other antipsychotics and
ferred to the hospital, for 29, 36, and
Data on the amount of time each
with any new drugs that may be de-
50 days. Whether they were returned
nurse and resident spend with each
veloped. o
home or were institutionalized re-
patient would provide a better under-
mains unknown because this informa-
standing of the pattern of care provid-
Acknowledgment
tion was not documented in the med-
ed to this patient group in an emer-
The authors thank Yvonne Hindle, R.N.,
ical records.
gency setting.
for assistance with data collection.
As Table 1 shows, six of the patients
The study demonstrated that it is
received a complete physical exami-
possible to consult medical records to
Reference
nation by a general practitioner, and
assess most costs, and that a complete
1. Schizophrenia: A Handbook for Families.
19 had both a consultation and a fol-
and expensive prospective study is
Ottawa, Health Canada and the Schizo-
low-up visit with a psychiatrist. The
not necessary. Costs not documented
phrenia Society of Canada, 1991
1288
PSYCHIATRIC SERVICES o October 1999 Vol. 50 No. 10

---