Original Research Reports

Detection of Depression in the Stroke Patient

DANIEL S.P. SCHUBERT, M.D., PH.D., CYNTHIA TAYLOR, D.O.

SUK LEE, M.D., ASKIN MENTARI, M.D.

WILBERFORCE TAMAKLO, M.D.

The literature sURRests the hypothesis that nonpsychiatrists will underrecoRnize depres-
sion in el'Oluations ofstroke patients. On a medical rehabilitation ward. 15 stroke pa-
tients were naluatedfor depression by psychiatric interview and self-report. Charts
were examinedfor detection of depression by the rehabilitation team. The hypothesis
was supported: in contrast to psychiatric interview (68% depressed) and self-report
(Beck Depression Inventory. 50% depressed). none of the patients were described as
depressed in chart notes by the rehahilitationteam (excludinR the psychiatrists). Psy-
chiatrists should develop ongoing interactions with primary care physicians 10 im-
prove detection of poststroke depression and other depressions on medical wards.

P oststroke depression has been well docu-

mented by Robinson and colleagues (see
reviews by Robinson and Starkstein)1.2 and
other investigators. J - 14 Robinson and Price l5
found about one-third of their stroke patients
were depressed. whereas Sinyor et al. 12 found
nearly half (47%) of their stroke patients had
clinically significant depression. Other inves-
tigators. however. have found lower rates of
depression. 16.17 Prior investigators agree that
poststroke depression is most frequent in the
acute poststroke phase while the patient is in the
hospital (i.e.• the first weeks after stroke).2.6
Over the next 6 months after stroke. the rate of
depression decreases slightly in some studies6
but up to 20% in others. 2 The rates of poststroke
Received May 21. 1991; revised August 19. 1991;
accepted September 4. 1991. From the School of Medicine.
Case Western Reserve Universily and MetroHealth Medical
and Rehabilitation Cenler. Cleveland. OH. Address reprint
requests to Dr. Schubert. Department of Psychiatry.
MetroHealth Medical Center. 3395 Scranton Road. Cleve-
land. OH 44109.
Copyright © 1992 The Academy of Psychosomatic
Medicine.
depression at different times vary with the cri-
teria used for defining depression. nature of the
sample. etc. 2•6 Because significant depression
has been found in stroke patients. depression
should be detected by the nonpsychiatric physi-
cian.
Several studies have found that nonpsy-
chiatric physicians underdiagnose depression.
Depression was undetected by nonpsychiatric
physicians in 70%-80% of patients who were
diagnosed as depressed by Research Diagnostic
Criteria. 18-20 Depression in tum may be associ-
ated with decreased physical functioning. 21
Other studies of detection of depression are
considered in greater detail in another article
(Schubert DSP. unpublished observations). This
literature (Schubert DSP. unpublished observa-
tions) suggests that depression will be un-
derrecognized in evaluations of stroke patients
by nonpsychiatrists.
The aim of our study was to assess detec-
tion of depression in stroke patients by a non-
psychiatric rehabilitation team and to compare
this with patients' self-reports and diagnoses by
psychiatrists.

290 PSYCHOSOMATICS

METHODS
Subjects were 15 patients ages 47-72 who were
admitted to MetroHealth Medical and Rehabil-
itation Center for rehabilitation. All were trans-
ferred to a physical medicine and rehabilitation
ward after acute treatment for stroke at other
Cleveland area acute hospitals. All had CT ev-
idence of a stroke. Subjects and setting are
described further in a previous article. 21
Measures
In addition to a psychiatric interview using
DSM-III-R 22 for diagnosis, all subjects com-
pleted the Beck Depression Inventory (BDI)23.24
(see House et at. 25 and Dam et at. 26 for use of
BDI in stroke patients) and the Mini-Mental
State Exam. 27
DSM-III-R diagnoses were made using an
unstructured interview. 22 Clinical interview re-
liability of DSM-III-R diagnoses has been de-
scribed as "extremely good" by Hyler et at. 28
The psychiatrist (W.T.) who gave the unstruc-
tured interview was blind to BDI results.
Detection of Depression
All admission and progress notes were ex-
amined by the physical medicine and rehabili-
tation physician authors (C.T., S.L., or A.M.) for
the words "depressed" or "depression" to indi-
cate depressed mood or diagnosis of depres-
sion. 29 All members of the rehabilitation team
documented evaluations and progress in the
progress notes. In addition, team meeting sum-
mary notes were written in the chart by a re-
corder. Neither the discharge summary nor old
records from other admissions were included in
this chart evaluation. The only notes not
searched for in the chart evaluation were those
of the psychiatric consultants. None of the
nonpsychiatric team was interviewed about
presence of depression in the patients. The psy-
chiatrist was not present on rounds.
VOLUME 33· NUMBER 3· SUMMER 1992
Schubert et al.
RESULTS
Rating of Depression by
Rehabilitation Team
Chart evaluation revealed no notation of
depression, possible depression, use of the term
"depressed," or similar documentation by the
rehabilitation team for any of the subjects dur-
ing the index admission. As indicated above,
notes by all members of the rehabilitation team
were searched for notation of depression with
the exception of the psychiatrists' notes.
The authors (A.M., C.T., S.L.) who at-
tended combined staff rounds recalled occa-
sional discussion of possible depression, but
these were never noted in the chart. The reha-
bilitation team seemed to have evaluated these
discussed depressions as normal, transient parts
of the poststroke syndrome.
Psychiatric Evaluation of Depression and
Comparison With Rehabilitation Team
As indicated above, the psychiatrists used
the standard psychiatric diagnostic criteria of
DSM-III-R 22 for diagnosis. Thirteen of the sub-
jects received psychiatric diagnoses with some
form of depression. Of these, eight received a
diagnosis of adjustment disorder with depressed
mood and five received a diagnosis of major
depression.
The total rate of psychiatrist-diagnosed de-
pression was 68% (13/19) with major depres-
sion accounting for 26% and adjustment
disorder with depressed mood accounting for
42%. (Percents are all rounded to the nearest
percentage.) This total rate (68%) of psychia-
trist-diagnosed depression was significantly
higher than the rate of 0% diagnosed by the
rehabilitation team excluding the psychiatrist
(X 2 = 16.5, df= I, P<O.OOI). All chi squares
were two-tailed and corrected for continuity.
The rate of diagnosis of major depression (26%
[5/19]) by the psychiatrist was not significantly
greater than the rehabilitation team's rate (0%
[0119]).
291
Depression in the Stroke Patient
Comparison of BDI and Rehabilitation
Team's Evaluation of Depression
The BDI identified 50% (7/14) of the
patients as at least mildly depressed (BDI score
~ 12), whereas the rehabilitation team identi-
fied no one as depressed by chart note (see Table
I). This BDI rate of 50% is significantly higher
than the rehabilitation team rate of 0% (X 2 =6.8,
df = I, P < 0.02). Because the rehabilitation
team may have looked for depressions other
than mild, the contrast was also made for mod-
erate to severe depression. The BDI identified
36% (5/14)ofthe subjects as at least moderately
depressed (BDI score ~ 21). This BDI rate of
36% is also significantly greater than the 0%
rate identified by the rehabilitation team (X 2 =
3.9, df = I, P < 0.05).
DISCUSSION
Comparative Detection of Depression by
Various Assessment Methods
The BDI detected significantly more de-
pression than the rehabilitation team. Similarly,
diagnosis of depression by the psychiatrist
(using DSM-III-R) detected more depression
than the rehabilitation team. These results sup-
port our hypothesis.
TABLE I. Evaluation of depression by the Beck
Depression Inventory in contrast to
documentation by the rehabilitation
team in a study of detection of de-
pression in stroke patients
Evaluation of Depression
Rehabilitation
Degree of Depression BOI Team
Moderate or severe depression
(BDI = 21 - 40) 5 0
Mild depression
(BDI = 12 - 20) 2 0
No depression
(BDI =0-11) 7 14
TOlal 14 14
NOll': BDI = Beck Depression Inventory
292
This study supports prior work that indi-
cates that nonpsychiatric physicians diagnose
20%-60% of their patients who have depres-
sion. Schubert (unpublished obervations) re-
viewed rates of recognition of depression by
nonpsychiatric physicians and discusses differ-
ences by type of setting, criteria for depression
used, etc. When noted, the depressions are fre-
quently evaluated as mild and transitory and
thus not in need of treatment. Although this
evaluation may be correct for a small number of
patients,2.6 poststroke depressions frequently
are severe enough to warrant treatment (they are
considered major depression by DSM-III-R),
and many respond to antidepressant medica-
tion.IO·30.31
Freeling 32 reported data that may suggest
why false negatives occur. Freeling contrasted
patients with recognized vs. unrecognized de-
pression. Those with unrecognized depression
showed less evidence of overt depressed mood
(e.g., reported sadness, depression, crying),
were less aware of being ill or depressed, re-
sponded in the short-term with better or worse
mood, got better or worse in response to envi-
ronmental change, and regarded depressed
mood as distinct from normal experience of
depression or sadness. Patients with unrecog-
nized depression also more frequently had a
current illness of more than a year and were
more likely to have a concurrent nonpsychiatric
medical illness.
Differences From Other Studies
BDI scores over 13 accounted for 42% of
the present sample, including all levels of de-
pression: mild, moderate, and severe. BDI
scores of more than 20 accounted for 35% of the
present sample, corresponding to moderate and
severe depression with 7% in the severe cate-
gory. These rates are higher than in other BDI
studies of general medical inpatient pop-
ulations. Rates for BDI scores over 13 were
reported as 22% by Schwab et al.,33 as 24% on
admission and later as 29% by Moffic and
Paykel,34 and as 32% by Cavanaugh. 35 Cavan-
augh also reported that 14% of her sample had
PSYCHOSOMATICS

8DI scores of more than 20 (moderate to severe
depression) and Schwab et al. 33 found 7% with
scores greater than 20 and 1.4% with scores
greater than 30 (severe depression).
Thus poststroke patients in our sample have
higher rates of self-reported depression than
general medical patients reported by other in-
vestigators. This may be due to a variety of
characteristics of the stroke group: I) older age,
2) lower social class status with associated lack
of stress-buffering social supports, 3) greater
physical disability (see Schubert et aI. 21 ), 4)
relatively longer-term dysfunction (vs. acute
medical illness), and 5) central nervous system
dysfunction (vs. other bodily dysfunction).
References
I. Robinson RG. Starkstein SE: Current research in affec-
tive disorders following stroke. Journal ofNeuropsychi-
atry and Clinical Neurosciences 2: 1-14. 1990
2. Starkstein SE. Robinson RG: Affective disorders and
cerebral vascular disease. 8r} Psychiatry 154: 170-182.
1989
3. Finklestein S. Benowitz L1. Baldessarini R. et al: Mood.
vegetative disturbance and DST after stroke. Ann Neurol
12:463-468.1982
4. Folstein MF. Maiberger R. McHugh PR: Mood disorders
as a specific complication of stroke.} Neural Neurosurg
Psychiatry 41: 1018-1020. 1977
5. Feibel JH. SpringerCJ: Depression and failure to resume
social activity after stroke (abstract). Arch Phys Med
RehabiI63:276. 1982
6. House A: Depression after stroke. 8M} 294:76-78.1987
7. Lim ML. Erbahim SBJ: Depression after stroke: a hospi-
tal treatment survey. Postgrad Med} 59:489--491. 1983
8. Messner M. Messner E: Mood disorders following
stroke. Compr Psychiatry 29:22-27.1988
9. Morris PLP. Raphael B: Depressive disorders associated
with physical illness: the impact of stroke. Gen Hosp
Psychiatry 9:324-330. 1987
10. Reding MJ. Ono LA. Winter SE. et al: Antidepressant
therapy after stroke: a double-blind trial. Arch Neurol
43:763-765. 1986
II. Robins AH: Are stroke patients more depressed than
other disabled subjects? } Chronic Dis 29:479-482. 1976
12. Sinyor D. Amato P. Ka10upek 00. et al: Poststroke
depression: relation to functional impairment. coping
and rehabilitation outcome. Stroke 17:1102-1107. 1986
13. Sinyor D. Jacques P. Kaloupek 00. et al: Poststroke
depression and lesion location: an attempted replication.
8rain 109:537-546. 1986
VOLUME 33· NUMBER 3· SUMMER 1992
Schubert et al.
SUMMARY
Past literature indicates that depression is
underrecognized by nonpsychiatrists. In the
sample of 15 stroke patients, a psychiatrist di-
agnosed 68% as depressed and the 8DI in-
dicated 50% were depressed. Chart notes by
nonpsychiatrists, in contrast, recorded 0% de-
pression. Depression in nonpsychiatric medical
patients may be underrecognized because of
their concurrent nonpsychiatric medical iIlness.
The authors thank Anna Zehnder for typing
several drafts ofthis article.
14. Wade DT. Legh J. Hewer RA: Depressed mood after
stroke. 8r} Psychiatry 151 :200-205. 1987
15. Robinson RG. Price TR: Post-stroke depressive disor-
ders: a follow-up study of 103 outpatients. Stroke
13:635-641. 1982
16. Rosse RB. Ciolino CP: Effect of conical lesion location
on psychiatric consultation referral for depressed stroke
patients. tnt } Psychiatry Med 15:311-320.1985
17. House A. Dennis M. Mogridge L. et al: Mood disorders
in the year after stroke. 8r} Psychiatry 158:83-92. 1991
18. Jones LR. Badger LW. Ficken RD. et al: Inside the
hidden mental health network: examining mental health
care delivery of primary care physicians. Gen Hasp
Psychiatry 9:287-293.1987
19. Koenig HG. Meador KG. Cohen HJ. et al: Detection and
treatment of major depression in older medically ill hos-
pitalized patients. tnt} Psychiatry Med 18: 17-31. 1988
20. Schul berg HC. Saul M. McClelland M. et al: Assessing
depression in primary medical and psychiatric practice.
Arch Gen Psychiatry 42: 1164-1170. 1985
21. Schuben DSP. Taylor C. Lee S. et al: Physical conse-
quences of depression in the stroke patient. Gen Hosp
Psychiatry 14:69-76. 1992
22. American Psychiatric Association: Diagnostic and
Statistical Manual of Mental Disorders. 3rd Edition.
Rel·ised. Washington. DC. American Psychiatric Asso-
ciation. 1987
23. Beck AT: An inventory for measuring depression. Arch
Gen Psychiatry 4:561-571. 1961
24. Beck AT: Depression: Causes and Treatment. Philadel-
phia. University of Pennsylvania Press. 1967
25. House A. Dennis M. Mogridge L. et al: Mood disorders
in the year after first stroke. 8r} Psychiatry 158:83-92.
1991
293
Depression in the Stroke Patient
26. Dam H. Pedersen HE. Ahlgren P: Depression among
patienl~ with stroke. Acta Psychiatr Scand 80:118-124.
1989
27. Folstein M. Folstein SE. McHugh P: Mini-Mental State:
a practical method for grading the cognitive state of
patients for the clinician. J Psychiatr Res 12:189-198.
1975
28. Hyler SE. Williams JBW. Spitzer RL: Reliability in the
DSM-III field trials. Arch Gen Psychiatry 39: 1275-
1278. 1982
29. Nielsen A. Williams T: Depression in ambulatory med-
ical patients. Arch Gen Psychiatry 37:999-1004.1980
30. Lipsey JR. Robinson RG. Pearlson GD: Nonriptyline
treatment of post-stroke depression. Lancet 1:297-300.
1984
294
31. Finklestein SF. Weintraub RJ. Karmauz N: Antidepres-
sant therapy for post-stroke depression: retrospective
study. Arch Phys Med Rehabi/68:772-776. 1987
32. Freeling P. Rao BM. Paykel ES. et al: Unrecognized
depression in general practice. BM} 290: 1880-1883.
1985
33. Schwab JJ. Bialow M. Brown JM. et al: Diagnosing
depression in medical inpatients. Ann Intern Med
67:695-707. 1967
34. Moffic HS. Paykel ES: Depression in medical inpatients.
BrJ Psychiatry 126:346-353. 1975
35. Cavanaugh S: The prevalence of emotional and cogni-
tive dysfunction in a general medical population: using
the MMSE. GHQ and BDI. Gen Hasp Psychiatry 5: 15-
24. 1983
PSYCHOSOMATICS