Hospital Psychiatry 14, 340-344, 1992
Underrecognition of Tardive Dyskinesia and
E. Hansen, M.D., William L. Brown, M.D.,
Recognition of tardive dyskinesia (TD) and other
neuroleptic, drug-induced, extrapyramidal side effects presents a major
challenge in modern clinical psychopharmacology. Failure to recognize
these disorders can lead to poor patient care and may contribute to
societal pressure for external control of psychiatric practice. This
study reports the occurrence of tardive dyskinesia and drug-induced
parkinsonism (DIP) in 101 inpatients, and documents under recognition of
both disorders by resident physicians. Researchers noted TD in 28% of
cases and residents only described TD (or symptoms of TD) in 12%. The
researcher determined DIP prevalence rate of 26% contrasted with an 11%
rate found by residents. Patients with psychotic disorders were more
likely than other patients to have researcher-identified TD, whereas DIP
(researcher cases) occurred more often in patients with affective
diagnoses. Residents tended to miss milder cases of TD, and to miss DIP
in younger patients and in patients with affective disorders. Improved
teaching and clinical exams are recommended to improve recognition.
tardive dyskinesia (TD) and other neuroleptic, drug-induced,
extrapyramidal side effects (EPS) remains problematic despite the major
impact these disorders have on psychiatric patients. This is especially
evident in the wide range of prevalence figures reported for TD-from as
low as 5% to more than 70% [1-4]. This variation in TD prevalence can be
attributed partially to differences in patient characteristics (i.e.,
risk factors), and case definition (e.g., less rigorous criteria lead to
higher rates) [4,5]. Claims that physicians fail to identify and treat
akathisia, drug-induced Parkinsonism (DIP), and dystonia suggest that
these EPS are also underrecognized [6-9].
importance for adequate patient care of recognizing and documenting
acute EPS and TD can not be overemphasized. If acute EPS go untreated,
poor medication compliance results and future treatment may be
compromised  The occurrence of very mild DIP may indicate that
neuroleptic dose is adequate and should not be further increased .
Also, ineffective management of EPS, especially TD, may lead to
increased demands for external regulation of psychiatric practice .
increasing prevalence and public concerns about TD and EPS, many
clinicians still report that they do not observe TD occurring at rates
comparable to those seen in research settings. This study prospectively
determines TD and DIP recognition rates for clinicians and researchers,
examines factors affecting syndrome occurrence and clinician
recognition, and makes suggestions to Improve clinical evaluation.
Included were psychiatric inpatients (N - 101)
admitted consecutively over a 6-month period to the Portland Department
of Veterans Affairs Medical Center who had taken neuroleptic medication
for a minimum of 3 months at some point prior to admission (determined
by direct patient inquiry and chart review). The first author (TEH) was
a staff psychiatrist on these patients' ward. All patients gave informed
rated patients with the Abnormal Involuntary Movement Scale (AIMS) 
and the Sct. Hans Extrapyramidal Disorders Scale  within 72 hours of
admission. The Sct. Hans scale includes separate evaluations for
dyskinesia (e.g., TD) and parkinsonism on a 0-6 scale. The raters (TEH
and WLB) achieved acceptable interrater reliability (Spearman's rho =
0.89 for total AIMS and rho = 0.83 for total Sct. Hans parkinsonism). A
researcher TD diagnosis required symptoms of moderate severity (i.e.,
AIMS = 3) in one body area or of mild severity (i.e., AIMS - 2) in two
body areas (these are the severity thresholds for Research Diagnosis of
TD) . A researcher DIP diagnosis required that the global Sat. Hans
parkinsonism score be mild or greater. Researchers rated patients blind
to the resident physicians' chart notes and diagnoses.
study conclusion we reviewed charts for age, DSM-III diagnosis, and
resident physician recognition of movement disorders. Twelve first- and
second-year psychiatric residents admitted patients during the study
period. We reviewed each patient's admission evaluation (assessments
coinciding with researcher ratings) and scored either specific
descriptions (e.g., “tremor,” “mouth movements” and so forth) or
provisional diagnoses of TD and DIP as positive cases.
calculated TD and DIP prevalence for researcher and resident diagnoses.
Risk factors analyzed for each disorder included the presence of the
other disorder, age, and DSM-III diagnoses classified along two
dimensions: 1) psychotic (schizophrenia, mania, delusional depression)
versus nonpsychotic, and 2) affective (16 bipolar, and 12 with major or
delusional depression) vs. nonaffecive. Multiple logistic regression
analyses determined which characteristics were associated with resident
physicians failing to identify researcher TD and DIP cases. The
variables entered included age, psychotic and/or affective diagnosis,
severity of TD, severity of DIP, and location of TD (oral vs.
extremity). All p-values reported in the study are two-tailed. As
TD measures were highly correlated (total AIMS and Sct. Hans Dyskinesia
scores, Spearman's rho = 0.98, p < 0.001; total AIMS and global AIMS
scores, rho = 0.93. p < 0.001), we used total AIMS to indicate TO
severity. Total Sct. Hans Parkinsonism score was used to measure DIP
Prevalence and Risk Factors
1 lists characteristics of patients with researcher and
resident-identified TD. Residents recognized TD significantly less often
than did researchers (12% vs. 28%, p < 0.01, McNemar test).
Researchers concurred with all cases noted by the residents. Older age
was significantly associated with ID prevalence for both researchers and
residents (Mann-Whitney U-test; p = 0.01 for researchers, p <
0.05 for residents). Patients with psychotic
diagnoses were more likely than nonpsychotic patients to have
researcher-identified TD) (Chi-square
7.8, df = 1, p <0.005). Effective diagnoses did not affect
researcher-identified TD prevalence (Chi-square, NS). Neither diagnostic
group influenced resident-identified TD prevalence (Chi-square, NS). The
correlation between researcher scores for TD and DIP was not significant
(rho = 0.13, p> 0.10), indicating that DIP did not obscure TD and
Prevalence and Risk Factors
2 lists characteristics of patients with researcher and
resident-identified DIP. Residents recognized DIP significantly less
often than did researchers (11% vs. 26%, p < 0.001, McNemar test).
Researcher concurred with all cases identified as
DIP by the residents. Older age was significantly associated with higher
DIP prevalence determined by researchers and residents (Mann Whitney
U-test, p<0.05 for both groups). Patients with affective
diagnoses were more likely than the others to have researcher-identified
DIP (Chi-Square = 4.76, df
= 1, p < 0.05). Psychotic diagnosis did not influence
either researcher or resident-identified DIP, and affective disorders
did not affect resident-identified DIP.
Regression and TD Underrecognition
regression analysis of resident-connect TD diagnosis (N = 12) vs.
resident-missed diagnosis (N = 16) revealed a trend for residents to
miss less severe TD (p = 0.07). Age, presence of DIP, and
psychiatric diagnosis did not contribute to the classification
difference. To further examine the effect of TD score on resident
recognition, we repeated the logistic regression analysis with separate
oral-facial and extremity scores (calculated from the individual items
on the AIMS exam) entered instead of total score. Lower oral-facial
scores were significantly associated with resident-missed diagnoses (p
but extremity scores were not associated with accuracy of resident
Regression and DIP Underrecognition.
regression analysis of resident-correct DIP diagnosis (N = 11) vs.
resident-missed diagnosis (N = 15) revealed that severity of DIP did not
influence accuracy of resident physician recognition. Resident missed
diagnosis was associated with younger age (p < 0.01) and
affective diagnosis (p < 0.05) when the affective/nonaffective
variable was used in the regression equation and with younger age (p
< 0.01), and nonpsychotic diagnosis (p < 0.05) when the
psychotic/nonpsychotic variable was used. The high negative correlation
(r = -0.64, p < 0.001) between affective and psychotic
diagnosis prevented analysis of the relative impact of each
primarily demonstrates underrecognition of TD and DIP by resident
physicians. The possibility that researchers overdiagnosed TD and DIP
seems unlikely as researcher criteria were more rigorous than those
applied to the resident cases. This study also found prevalence rates
for TD (28%) and DIP (26%) in the range commonly reported for
chronically medicated patients [1-3,16]. The correlation between
increased age and prevalence for both TD and DIP further agrees with
previous reports [16,17].
this study, underrecognition undoubtedly included both failure to
observe and failure to document the syndromes. Factors influencing
documentation were not specifically studied. We speculate, based on
conversations with physicians at all levels of training, that concern
about the significance of TD and a lack of awareness about the clinical
importance of DIP could contribute to underdocumentation.
of TD significantly affected resident underrecognition but only when
oral-facial and extremity scores were analyzed separately. This finding,
which must be viewed as tentative given the small sample size used in
the regression analysis, suggests that residents recognize TD when it
occurs with relatively greater severity in an obvious and expected
location (e.g., the face) but fail to notice TD regardless of severity
when it occurs elsewhere. (Similarly, as severity did not affect DIP
recognition, residents appear to neglect DIP even in more severe cases).
At least for TD, greater sensitivity in examination may improve
recognition rates by enabling residents to see the milder cases.
The presence of
an affective diagnosis significantly influenced DIP recognition, with
researchers more likely to diagnose DIP in such patients than in others,
and residents less likely to do so. The symptoms of depression can be
hard to distinguish from DIP, and misattribution of symptoms could have
occurred with both the researchers and the residents. However, only 12
of the 28 affective disorder patients had depression, so this is a
partial explanation at best. Given that younger age also predicted
failure to recognize DIP, perhaps residents had more trouble recognizing
DIP when they did not expect it. The failure to note an association
between TD and affective disorder in this report differs with literature
which suggests the latter is a TD risk factor .
Two other reports
[6,18] have identified resident physician underrecognition of TD and
DIP. In an evaluation of physical examinations,
two of three TD cases and three of five DIP cases were missed by
psychiatric resident physicians . The frequency of TD (6%) and DIP
(5%) was low, suggesting that the number of patients at risk was small
(the number of patients on neuroleptics was not given). In the other
study , 48 patients were on neuroleptics for at least 1 week, and
researchers examined entire hospital charts. Symptoms consistent with TD
were described in only four of ten (40%) researcher-diagnosed TD cases
(the actual diagnosis was made by a resident in just one TD case), and
DIP was diagnosed 17 of 29 times (59%) . Though methodologic
differences prevent direct comparison to our data, the authors found
severity influenced recognition of oral TD, but not DIP (extremity TD
was interestingly more likely to be recognized if it was less severe).
Comparison between researchers and residents in the current and above reports reflects the prominent role resident physicians have in academic settings. Whether findings with residents can be applied to psychiatrists in general is unclear. Residents in most programs do many more physical exams than general psychiatrists and the intensity of this activity is likely to offset the greater experience of the psychiatrist in practice. On balance, we suspect underrecognition of TD and DIP also occurs with general psychiatrists.
The importance of
accurate recognition of TD and DIP extends beyond work done by
psychiatrists with inpatients. Psychiatric patients will have their
standard neuroleptic medications continued when treated by physicians
for other conditions, for instance, while hospitalized on a medical or
surgical ward. Nonpsychiatrists may prescribe standard neuroleptic
medications because of convenience, economics, or unavailability of
psychiatrists (e.g., as medical backup to nonpsychiatrist mental health
professionals, in nursing homes, and in rural areas). Also, medications
with neuroleptic properties may be used for nonpsychiatric purposes
(e.g., metoclopramide for gastrointestinal tract conditions,
neuroleptics for pain management). Practitioners in all of these
situations need to be aware of TD and DIP so they can obtain informed
consent from their patients. They also need to be able to recognize
these disorders, so that they can safely manage their patients.
and use of more specific clinical exams might improve recognition rates.
One study claimed that resident physicians who took a course in EPS
demonstrated better management of EPS than did residents 2 years
previously who had not taken the course . Improved EPS recognition
after receiving the course in the same year must be studied to control
for changes in sensitivity over time not related to the course.
In conclusion, we
report underrecognition of TD and DIP by psychiatric resident physicians
in an inpatients setting. Possible contributing problems include
insensitive or incomplete exam techniques and reluctance to report
findings. We suggest the following approaches may lead to improved
recognition of EPS:
A specific examination for TD and DIP should be completed in all
patients with past, current, or anticipated neuroleptic treatment,
regardless of diagnosis.
Attention should be focused on the oral cavity for TD (e.g., more
than the standard pharyngeal exam).
Hand movements must be observed at rest with activation maneuvers
(i.e., finger tapping and walking) to facilitate detection of TD in the
extremities. Observed movements should be characterized as tremor (more
likely DIP) or choreoathetoid (more likely TD).
Symptoms of DIP must be specifically assessed (global impressions
will miss symptoms). Symptoms include rigidity, bradykinesia (slowness
of movement), tremor, facial expression,
gait, and posture. Symptoms should not be automatically dismissed
because of resemblance to
Both presence and absence of signs and symptoms should be
work was supported in part by the Veterans Administration Research
Service (Drs. Hansen, Weigel Casey) and National Institute of Mental
Health grant number MH36657 (Dr. Casey). The authors thank Melinda Lowe
and Sandra K. Long for help in manuscript preparation.
Casey DE: Tardive dyskinesia. In Meltzer MY (ed),
Psychopharmacology: The Third Generation of Progress. New York, Raven
Press. 1987. pp 1411-1419
Kane JM, Smith JM: Tardive dyskinesia: prevalence and risk
factors 1959 to 1979. Arch Gen Psychiatry 39:473-481, 1982
Jeste DV, Wyatt RI: Changing epidemiology of tardive dyskinesia: an
overview. Am J Psychiatry 138:297-309,1981
Hansen TE, Casey DE, Vollmer WM: Is there an epidemic of tardive
dyskinesia? In Casey DE. Gardos G(eds), Tardive Dyskinesia: From Dogma
to Reason. New York, American Psychiatric Press, 1986, pp 2-14
Asnis GM, Leopold MA, Duvoisin RC, Schwartz AM: A survey of tardive
dyskinesia in psychiatric outpatients. Am J Psychiatry
Weiden E, Mann JJ. Haas G: Clinical nonrecognition of drug-induced
movement disorders: a cautionary study. Am J Psychiatry 144:1148-1153,
Van Putten T: Why do schizophrenic patients refuse to take their
drugs? Arch Gen Psychiatry 31:67-72, 1974
Rifkin A, Quitkin P, Klein DF: Akinesia: a poorly recognized
drug-induced extrapyramidal behavior disorder. Arch Gen Psychiatry
Keepers GA, Casey DE: Clinical management of acute
neuroleptic-induced extrapyramidal syndromes. Curr Psychiatr Ther
Keepers GA, Hansen TE, Casey DE: Prospective prediction of
vulnerability to neuroleptic-induced extrapyramidal syndromes (abstract)
Proc Soc Biol Psychiatry 66:222, 1986
McEvoy JP: The neuroleptic threshold as a marker of minimum
effective neuroleptic dose. Compr Psychiatry 27:327-335, 1986
Baldessarini RJ, Cohen BM: Regulation of psychiatric practice. Am
J Psychiatry 143:750-751, 1986
Guy W: ECDEU assessment manual for psycho-pharmacology (revised).
Washington, DC, US Department of Health, Education. and Welfare, USGPO
Gerlach J: Tardive dyskinesia. Dan Med Bull 26:209-244.1979
Schooler NR, Kane JM: Research diagnoses for tardive dyskinesia.
Arch Gen Psychiatry 39:486-487, 1982
Ayd FJ Jr: Early-onset neuroleptic-induced extrapyramidal
reactions; a second survey, 1961-1981. In Coyle JT, Enni SJ (eds),
Neuroleptics: Neurochemical, Behavioral and Clinical Perspectives. New
York, Raven Press, 1983, pp 75-92
Casey DE: Tardive dyskinesia and affective disorders. In Gardos
G, Casey DE (eds), Tardive Dyskinesia and Affective Disorders.
Washington, DC, American Psychiatric Press, 1984, pp 1-20
Rigby JC, Oswald AG: An evaluation of the performing and
recording of physical examinations by psychiatric trainees. Br J
Psychiatry 150:533-535, 1986
Dixon L, Weiden PJ, Frances AJ, Rapkin B: Management of
neuroleptic-induced movement disorder: effects of physician training. Am
J Psychiatry 146:104-106, 1989
the Portland Veterans Affairs Medial Center, Oregon Health Sciences
University. Portland Oregon.
© 1992 Elsevier Science Publishing Co., Inc.
The New England Journal of Medicine -- December
3, 1998 -- Vol. 339, No. 23
To the Editor:
A 30-year-old patient, a nurse's aide, was transferred from another hospital with a history of recurrent episodes of septic arthritis of the right ankle. Over a period of 3 years, she had been hospitalized for 168 days and had undergone 15 surgical interventions. A total of 72 wound samples (including biopsy specimens and tissue fragments obtained during debridement of the ankle) and 110 blood specimens were sent for culture. Apart from "conventional" bacteria such as staphylococci, streptococci, and coliforms, cultures of both wound and blood samples grew unusual nonfermenting, gram-negative rods.
(Table 1). These organisms included Alcaligenes denitrificans subspecies xylosoxidans, A. faecalis, flavobacterium species, Comamonas testosteroni, Sphingomonas paucimobilis, and Stenotrophomonas maltophilia as identified by a commercial biochemical test panel (API 20 NE, BioMerieux, Marcy-l'Etoile, France).
Blood specimens for culture were obtained during febrile episodes by the hospital phlebotomy team or the medical staff. The bacteremic episodes were recurrent, and 10 of 13 positive blood cultures contained multiple organisms. All the polymicrobial blood cultures were obtained when the patient had an intravenous catheter in place, but none of the catheters that were removed grew bacteria. Unusual nonfermenting bacteria account for only 0.2 percent of all blood-culture isolates in our hospital.
In the context of the patient's manipulation of thermometers and the recurrent malfunctioning of her intravenous catheters, the microbiologic findings suggested self-induced infection. It was strongly suspected that she rinsed her wounds with water from a toilet, which was considered to be the probable source of contamination. When confronted with this suspicion, the patient denied it. Over a three-year follow-up period, however, no further infections occurred.
Nonfermenting bacteria live in soil, water, and sea water. They are ubiquitous in the hospital, where their growth is supported by a humid atmosphere. (3) Isolation of these microorganisms from the blood generally points to an exogenous source of infection, such as a hemodialysis system. (4) These organisms are opportunistic and affect severely immunocompromised patients such as those with neutropenia. (5)
In this case, the isolation of unusual nonfermenting bacteria, in the absence of other likely explanations, suggested deliberate contamination. Self-induced infection should be suspected if a patient without an underlying malignant or immunosuppressive condition has recurrent polymicrobial infections with unusual gram-negative rods.
Nico E.L. Meessen, M.D., Ph.D.
1. Reich P, Gottfried LA. Factitious disorders in a teaching hospital. Ann Intern Med 1983;99:240-7.
2. Wallach J. Laboratory diagnosis of factitious disorders. Arch Intern Med 1994;154:1690-6.
3. du Moulin G. Minimizing the potential for nosocomial pneumonia: architectural, engineering, and environmental considerations for the intensive care unit. Eur J Clin Microbiol Infect Dis 1989;8:69-74.
4. Reverdy ME, Freney J, Fleurette J, et al. Nosocomial colonization and infection by Achromobacter xylosoxidans. J Clin Microbiol 1984;19:140-3.
5. Jacobs JA, Stobberingh EE, Schouten HC. Fatal infection due to Alcaligenes xylosoxidans subsp. xylosoxidans in a neutropenic host. Clin Microbiol Newslett 1992;14:182-4.
Mothers Against MSBP Allegations
Heart-To-Heart Discussion Group