Impact of possible tardive dyskinesia on physical wellness and social functioning: results from the real-world RE-KINECT study

Abstract

Background:
Tardive dyskinesia (TD) is a persistent and potentially disabling movement disorder associated with antipsychotic use. Data from RE-KINECT, a real-world study of antipsychotic-treated outpatients, were analyzed to assess the effects of possible TD on patient health and social functioning.

Methods:
Analyses were conducted in Cohort 1 (patients with no abnormal involuntary movements) and Cohort 2 (patients with possible TD per clinician judgment). Assessments included: EuroQoL's EQ-5D-5L utility (health); Sheehan Disability Scale (SDS) total score (social functioning); patient- and clinician-rated severity of possible TD ("none", "some", "a lot"); and patient-rated impact of possible TD ("none", "some", "a lot"). Regression models were used to analyze the following: associations between higher (worse) severity/impact scores and lower (worse) EQ-5D-5L utility (indicated by negative regression coefficients); and associations between higher (worse) severity/impact scores and higher (worse) SDS total score (indicated by positive regression coefficients).

Results:
In Cohort 2 patients who were aware of their abnormal movements, patient-rated TD impact was highly and significantly associated with EQ-5D-5L utility (regression coefficient: - 0.023, P < 0.001) and SDS total score (1.027, P < 0.001). Patient-rated severity was also significantly associated with EQ-5D-5L utility (- 0.028, P < 0.05). Clinician-rated severity was moderately associated with both EQ-5D-5L and SDS, but these associations were not statistically significant.

Conclusions:
Patients were consistent in evaluating the impacts of possible TD on their lives, whether based on subjective ratings ("none", "some", "a lot") or standardized instruments (EQ-5D-5L, SDS). Clinician-rated severity of TD may not always correlate with patient perceptions of the significance of TD.

Full text

Tanneretal.
Journal of Patient-Reported Outcomes (2023) 7:21
https://doi.org/10.1186/s41687-023-00551-5
RESEARCH
© The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.
Open Access
Journal of Patient-
Reported Outcomes
Impact ofpossible tardive dyskinesia
onphysical wellness andsocial functioning:
results fromthereal-world RE-KINECT study
Caroline M. Tanner1,2*, Stanley N. Caroff3,4, Andrew J. Cutler5, William R. Lenderking6, Huda Shalhoub6,
Véronique Pagé7, Ericha G. Franey8, Michael Serbin8 and Chuck Yonan8
Abstract
Background Tardive dyskinesia (TD) is a persistent and potentially disabling movement disorder associated with
antipsychotic use. Data from RE-KINECT, a real-world study of antipsychotic-treated outpatients, were analyzed to
assess the effects of possible TD on patient health and social functioning.
Methods Analyses were conducted in Cohort 1 (patients with no abnormal involuntary movements) and Cohort 2
(patients with possible TD per clinician judgment). Assessments included: EuroQoL’s EQ-5D-5L utility (health); Sheehan
Disability Scale (SDS) total score (social functioning); patient- and clinician-rated severity of possible TD (“none”, “some”,
“a lot”); and patient-rated impact of possible TD (“none”, “some”, “a lot”). Regression models were used to analyze the
following: associations between higher (worse) severity/impact scores and lower (worse) EQ-5D-5L utility (indicated
by negative regression coefficients); and associations between higher (worse) severity/impact scores and higher
(worse) SDS total score (indicated by positive regression coefficients).
Results In Cohort 2 patients who were aware of their abnormal movements, patient-rated TD impact was highly and
significantly associated with EQ-5D-5L utility (regression coefficient: − 0.023, P < 0.001) and SDS total score (1.027,
P < 0.001). Patient-rated severity was also significantly associated with EQ-5D-5L utility (− 0.028, P < 0.05). Clinician-
rated severity was moderately associated with both EQ-5D-5L and SDS, but these associations were not statistically
significant.
Conclusions Patients were consistent in evaluating the impacts of possible TD on their lives, whether based on sub-
jective ratings (“none”, “some”, “a lot”) or standardized instruments (EQ-5D-5L, SDS). Clinician-rated severity of TD may
not always correlate with patient perceptions of the significance of TD.
Keywords Tardive dyskinesia, Quality of life, Function, Real-world evidence, Antipsychotic agents
Huda Shalhoub and Chuck Yonan: At time of the study
*Correspondence:
Caroline M. Tanner
Caroline.Tanner@ucsf.edu
Full list of author information is available at the end of the article

Page 2 of 11
Tanneretal. Journal of Patient-Reported Outcomes (2023) 7:21
Background
Tardive dyskinesia (TD) is a persistent and potentially
disabling movement disorder that is associated with
exposure to antipsychotics and other dopamine receptor
blocking agents [1, 2]. While second-generation (“atypi-
cal”) antipsychotics have been associated with a lower
incidence of TD, prevalence studies have shown the con-
tinued importance of TD asaclinical consideration with
these therapies due to their expanding use, especially in
non-psychotic conditions [3, 4]. erefore, it is important
that all patients taking any antipsychotic medication be
screened regularly for potential TD symptoms, includ-
ing abnormal involuntary movements in the face/mouth,
neck/trunk, upper extremities, and lower extremities [5].
Early case reports of TD suggested that it may be
reversible if diagnosed early in some patients who could
be safely withdrawn from antipsychotics and followed
over time; therefore, when appropriate, tapering off
antipsychotic medications could be tried when TD first
emerges [6]. However, antipsychotic discontinuation is
not a practical option for patients with chronic psychotic
disorders, and the overall evidence for discontinuation
leading to resolution of TD is insufficient, especially in
established cases [7]. In these cases, use of an approved
TD medication (i.e., vesicular monoamine transporter
2 inhibitor) is appropriate. Moreover, current treatment
guidelines indicate that factors such as patient prefer-
ence, impairment, or psychosocial functioning should be
considered when planning treatment [8]. Consistent with
these recommendations, results from a modified Delphi
consensus study indicated that improvements in patient-
reported distress, functional impairments, and health-
related quality of life (HRQoL) should be considered
when determining treatment success in patients with TD
[9].
e impact of TD can be wide-ranging. e cumula-
tive effects of TD on patients with a serious mental illness
may include worsened psychopathology, higher rates of
comorbidities, increased risk of mortality, and poorer
treatment outcomes [10–12]. TD can negatively affect
motor functions such as speech, gait, and respiration, as
well as cognitive functions such as verbal memory and
processing [11, 13, 14]. TD can also lead to feelings of
stigmatization, social withdrawal, loss of employment,
and higher healthcare resource utilization [12, 15, 16].
Although many clinicians, patients, and caregivers are
aware of these negative impacts, quantitative assessments
of HRQoL in TD are very limited. RE-KINECT was a
real-world study that included psychiatric outpatients
who had been treated with antipsychotics. Previously
published results from this study indicated that 27.6% of
these patients had abnormal involuntary movements that
were consistent with possible TD, and that the impact of
these movements on both patients and caregivers was
considerable [17, 18]. Relevant to the current analyses,
RE-KINECT included two established patient-reported
HRQoL measures: EuroQoL’s 5-dimension 5-level ques-
tionnaire (EQ-5D-5L) and the Sheehan Disability Scale
(SDS). Analyses based on these outcomes were used
to further explore the impact of TD on patients’ over-
all health, physical wellness, and social functioning. e
findings from these analyses are intended to provide
the type of information needed for appropriate health
technology assessments, as needed for the evaluation
of medications used to treat TD. In a broader sense, the
quantitative analyses in this report are useful for under-
standing the impact of TD beyond subjective impres-
sions. Moreover, they allow for comparisons among
different types of patients (e.g., those with no abnormal
involuntary movements versus those with possible TD).
Methods
Study design
RE-KINECT was a prospective, real-world, observa-
tional, multicenter study conducted at 37 outpatient
psychiatry clinics (e.g., research institutions, community
health centers, private practices) in the United States
from April 2017 to January 2018 [17]. In brief, this study
included adults with ≥ 3months of lifetime exposure to
antipsychotic medication and ≥ 1 clinician-confirmed
psychiatric disorder according to Diagnostic and Statis-
tical Manual of Mental Disorders, Fifth Edition (DSM-5)
criteria [19]. e 3month lifetime antipsychotic exposure
requirement was based on the Schooler-Kane criteria
for TD research and DSM-IV-TR [20, 21], and it is con-
sistent with current American Psychiatric Association
guidelines (DSM-5-TR) [22]. All patients provided writ-
ten informed consent prior to participation; Institutional
Review Board approval was obtained ateach site.
Patient cohorts
Patients were assigned to 1 of 2 cohorts based on clini-
cian assessment (Fig.1). Training for TD screening (via
videos) was provided to all clinical site personnel to pro-
mote inter-rater reliability. Cohort 1 was defined in the
study protocol as patients who had no abnormal involun-
tary movements or whose movements were not consist-
ent with possible TD based on clinician assessment. To
avoid any potential overlap with patients with possible
TD (as described below), this analysis focused on a modi-
fied Cohort 1 (no abnormal involuntary movements),
which excluded patients with non-TD involuntary move-
ments such as tremor.
Cohort 2 was defined in the study protocol as patients
who had abnormal involuntary movements that were
confirmed by their clinician as possible TD. All patients

Page 3 of 11
Tanneretal. Journal of Patient-Reported Outcomes (2023) 7:21
in Cohort 2 had a clinician-rated severity of possible TD
as “some” or “a lot” in at least 1 of the following body
regions: head/face, neck/trunk, upper extremities, and/or
lower extremities. For this analysis, Cohort 2A (“aware”)
was defined as patients with clinician-confirmed pos-
sible TD who also self-reported having abnormal invol-
untary movements within the past 4 weeks and had a
self-rated severity of “some” or “a lot” in at least 1 of the
4 body regions. Data from Cohort 2NA (“not aware”),
defined as patients with possible TD who self-reported
having no abnormal involuntary movements in the past
4weeks, were analyzed to provide supplementary infor-
mation about the potential effects of patient awareness
on HRQoL.
Assessments
Prior to clinician assessment and cohort assignment,
alleligible and consenting patients were asked to com-
plete the EQ-5D-5L [23], which assesses problems in
5-dimensions (mobility, self-care, usual activities, pain/
discomfort, anxiety/depression) that can be used to cal-
culate a utility score, which ranges from 0 (health state
equivalent to death) to 1 (perfect health). e EQ-5D-5L
also includes a visual analog scale (VAS) that measures
current health state, which ranges from 0 to 100 (best
possible health). Patients were also asked to complete
the SDS [24], which assesses disruption in 3 social func-
tioning domains (work/school, social life, family life/
home responsibilities) and includes a total score (sum of
domain scores) which ranges from 0 (no disruption) to 30
(extreme disruption).
For severity of possible TD, clinicians and patients were
asked to “rate the severity of visible, uncontrollable move-
ments” for each of 4 body regions (head/face, neck/trunk,
upper extremities, and/or lower extremities) using sim-
ple descriptors of “none”, “some”, or “a lot”. For impact of
possible TD, patients who were aware of their abnormal
involuntary movements (Cohort 2A) were asked to rate
how much “over the past 4weeks” did these movements
“impact your ability” to perform each of 7 different activi-
ties/functions (usual activities, talking, eating, breath-
ing, being productive, self-care, socializing), also using
the descriptors of “none”, “some”, or “a lot”. For regression
analyses, these descriptors were assigned values of 0, 1,
Fig. 1 Overview of possible TD symptom screen and cohort assignment. EQ-5D-5L, EuroQoL 5-dimension 5-level questionnaire; SDS, Sheehan
Disability Scale; TD, tardive dyskinesia

Page 4 of 11
Tanneretal. Journal of Patient-Reported Outcomes (2023) 7:21
and 2, respectively, with the summary score for severity
ranging from 0 (“none” in all 4 regions) to 8 (“a lot” in
all 4 regions) and the summary score for impact rang-
ing from 0 (“none” in all 7 activities) to 14 (“a lot” in all 7
activities).
Analyses
All analyses were intended to be exploratory, and no
adjustments were made for multiple comparisons.
Patients’ demographics, clinical characteristics, EQ-
5D-5L scores, and SDS scores were analyzed for Cohorts
2 and 2A versus modified Cohort 1, as well as for Cohort
2A versus Cohort 2NA. Chi-squared tests were used for
categorical variables and t-tests were used for continu-
ous variables. Analyses were adjusted for potentially con-
founding factors as follows: age (18–95years); sex (male
or female); patient-reported overall health status (0 = “no
health problems” to 10 = “health as bad as you can imag-
ine”); clinician-rated severity of psychiatric condition
(0 = “normal, not ill” to 6 = “among the most severely ill”);
clinician-rated functional status of patient (0 = working/
studying/managing household independently to 2 = not
working/studying/managing household in any capacity);
and psychiatric diagnosis (presence or absence of schizo-
phrenia or schizoaffective disorder, presence or absence
of a mood or other disorder). ese factors were based
on clinically relevant characteristics that would likely
affect HRQoL in patients with TD; no statistical process
was used for selection.
Simple linear regressions were used to evaluate EQ-
5D-5L utility and SDS total scores by clinician/patient-
rated severity of possible TD and by patient-rated impact
of possible TD. ese analyses compared Cohorts 2 and
2A to modified Cohort 1, based on mean score differ-
ences between the cohorts. Simple linear regressions
were also used to determine the associations between
possible TD (clinician/patient-rated severity, patient-
rated impact) and HRQoL measures—physical well-
ness (EQ-5D-5L utility) or social function (SDS total) in
Cohort 2, Cohort 2A, and Cohort 2NA. ese analyses
were adjusted for age, sex, patient-reported overall health
status, clinician-rated severity of psychiatric condition,
clinician-rated functional status of patient, and psychiat-
ric diagnosis.
Results
Patient cohorts andbaseline characteristics
A total of 1148 patients from 37 clinical sites in 19 states
were screened, including patients from research institu-
tions, community health centers, and private psychiat-
ric practices. Modified Cohort 1 included 450 patients
who had no abnormal involuntary movements (possible
TD or otherwise) (Fig. 1). e full Cohort 2 included
204 patients who had abnormal involuntary movements
that were clinician-confirmed as possible TD, with a
clinician-rated severity of “some” or “a lot” in ≥ 1 body
region. Cohort 2A (aware) included 110 patients with
self-reported abnormal involuntary movements in the
past 4weeks, with a self-rated severity of “some” or “a
lot” in ≥ 1 body region. Cohort2NA (not aware) included
94 patients who had clinician-confirmed possible TD
but self-reported no abnormal involuntary movements
within the past 4weeks.
Compared to patients with no abnormal involuntary
movements (modified Cohort 1), patients with possible
TD (Cohort 2) were significantly older, with a higher pro-
portion of male patients, higher incidence of schizophre-
nia or schizoaffective disorder, and lower incidence of
mood or other psychiatric disorders (Table1). Per clini-
cian impression, patients with possible TD also had more
severe psychiatric conditions and were less likely to be
engaged in what many would consider “usual activities”
such as working/studying or managing a household inde-
pendently and more likely to not engage in these activi-
ties at all. Per patient self-report, those with possible TD
had worse overall health status than those with no abnor-
mal involuntary movements.
No significant differences between Cohort 2NA and
Cohort 2A were found for age, sex, psychiatric condition,
clinician-rated severity of psychiatric condition, or cli-
nician impression of overall functional status. However,
compared to Cohort 2NA, patients in Cohort 2A had sig-
nificantly worse self-reported overall health (Additional
file1: Appendix TableS1).
EQ‑5D‑5L andSDS scores
When adjusted for age, sex, health status, severity of
psychiatric condition, functional status, and psychiatric
diagnosis, mean baseline EQ-5D-5L scores were lower
in patients with possible TD (Cohort 2) than in patients
with no abnormal involuntary movements (modified
Cohort 1), indicating relatively worse overall health (util-
ity score and visual analog scale [VAS]) (Table2). Mean
SDS dimension and total scores were higher in Cohort 2
compared to modified Cohort 1, indicating relatively
worse social functional status in patients with possi-
ble TD than in patients with no abnormal involuntary
movements; however, none of the differences in mean
scores were statistically significant. Differences between
Cohorts 2A and 2NA for mean SDS total and family/
home life scores indicated statistically significantly worse
social functioning in patients who were aware of their
possible TD than in those who were not aware (Addi-
tional file1: Appendix TableS2).

Page 5 of 11
Tanneretal. Journal of Patient-Reported Outcomes (2023) 7:21
Compared to patients with no abnormal involuntary
movements (modified Cohort 1), a higher percentage of
patients with clinician-confirmed possible TD (Cohorts 2
and 2A) reported having moderate problems (score = 3),
severe problems (score = 4), or extreme problems
(score = 5) in all EQ-5D-5L dimensions (Fig.2). Interest-
ingly, the dimensions of mobility and self-care were sig-
nificantly worse in Cohorts 2 and 2A than in modified
Cohort 1. Compared to patients who were aware of their
possible TD (Cohort 2A), significantly fewer unaware
patients (Cohort 2NA) had problems with mobility and
self-care (Additional file1: Appendix Fig. S1). However,
moderate or severe pain/discomfort were more prevalent
in unaware patients.
Analyses of EQ-5D-5L utility and SDS total scores
by clinician/patient-rated severity and patient-rated
impact further confirmed that patients with possible
TD had worse physical wellness and social function-
ing than those with no abnormal involuntary move-
ments (Table 3). When adjusted for age, sex, health
status, severity of psychiatric condition, functional sta-
tus, and psychiatric diagnosis, linear regressions based
on mean score differences between cohorts indicated
that the largest decrements in EQ-5D-5L utility were in
Cohort 2A patients with “a lot” of self-reported impact
(adjusted mean score difference relative to modified
Cohort 1: −0.121, P < 0.001) or “a lot” of self-reported
severity (difference: − 0.089, P < 0.001). e worst SDS
total scores were also found inaware patients who self-
reported “a lot” of self-reported impact (adjusted mean
score difference: 5.401, P < 0.001) or “a lot” of self-
reported severity (difference: 3.288, P < 0.01).
Table 1 Demographics and clinical characteristics
SD standard deviation; TD tardive dyskinesia
a Modied Cohort 1 includes patients with no visible or self-reported abnormal involuntary movements. Cohort 2 includes all patients with possible TD per clinician
assessment
b For questions or items that allowed more than 1 response (i.e., categories not mutually exclusive), P-values are provided for each response. Chi-squared tests were
used for categorical variables; t-tests were used for continuous variables
c Based on questionnaire responses (i.e., not diagnostic medical codes). Mood or other psychiatric disorders include bipolar disorder, major depressive disorder,
anxiety disorder or symptoms, post-traumatic stress disorder, personality disorder, attention decit hyperactivity disorder, substance use disorder, and other psychotic
disorder
d Per patient self-report. Higher scores indicate worse overall health (range, 0= “no health problems” to 10= “health as bad as you can imagine”)
e Per clinician impression, based on the following options: (1) been independently working or studying full- or near full-time, in usual occupation; or managing
own household; or participating in unpaid or voluntary activities, whether retired or not; (2) been working or studying with assistance in usual occupation and/or
managing own household or participating in unpaid or voluntary activities; or has experienced a signicant reduction in house work; or was in a sheltered situation or
on sick leave; and (3) not been working or studying in any capacity and not managing own household
Modied Cohort 1 (N = 450)aCohort 2 (N = 204)aP‑valueb
Age, mean (SD) 47.5 (14.6) 54.6 (13.6) < 0.001
Sex, n (%)
Male 182 (40.5) 100 (49.0) 0.043
Female 267 (59.5) 104 (51.0)
Lifetime antipsychotic exposure, mean (SD), years 9.5 (9.0) 15.9 (13.9) < 0.0001
Psychiatric condition, n (%)c
Schizophrenia or schizoaffective disorder 146 (32.4) 107 (52.5) < 0.001
Mood or other psychiatric disorder 354 (78.7) 134 (65.7) < 0.001
Severity of psychiatric condition, n (%)
Normal, not ill 47 (10.4) 7 (3.4) < 0.001
Minimally ill 103 (22.9) 27 (13.2)
Mildly ill 111 (24.7) 68 (33.3)
Moderately ill 128 (28.4) 67 (32.8)
Markedly ill 43 (9.6) 26 (12.7)
Severely ill 16 (3.6) 9 (4.4)
Among the most severely ill 2 (0.4) 0
Overall health status, mean (SD)d4.2 (2.8) 4.7 (2.8) 0.031
Overall functional status, n (%)e
Working/studying independently 242 (53.8) 66 (32.4) < 0.001
Working/studying with assistance 100 (22.2) 44 (21.6)
Not working/studying 108 (24.0) 94 (46.1)

Page 6 of 11
Tanneretal. Journal of Patient-Reported Outcomes (2023) 7:21
Association betweenseverity/impact ofpossible TD
andEQ‑5D‑5L utility score
Linear regression analyses conducted in Cohorts 2, 2A,
and 2NA indicated no statistical significance for the
association between clinician-rated severity and EQ-
5D-5L utility in any body region, regardless of whether
patients were aware or not aware of their possible TD
(Additional file1: Appendix TableS3).
In contrast, statistical significance was found for the
association between patient-rated severity of possible
TD and EQ-5D-5L utility in patients who were aware of
their possible TD (Cohort 2A) (Table4). After adjust-
ments for age, sex, health status, severity of psychiatric
condition, functional status, and psychiatric diagno-
sis, the regression coefficient for overall patient-rated
severity (summary score based on all body regions) was
− 0.028 (P < 0.05). Similar coefficients were found for
severity inthe head/face (−0.026) and upper extremi-
ties (− 0.032), indicating comparable magnitudes of
effect on EQ-5D-5L utility; however, these results were
not statistically significant. Patient-rated severity in the
lower extremities had the largest association with EQ-
5D-5L utility, with an adjusted regression coefficient of
− 0.060 (P < 0.05).
Patient-rated impact (not severity) of possible TD had
the greatest association with EQ-5D-5L utility (Table4).
After adjustments, regression coefficients (unstandard-
ized) indicated that for Cohort 2A, EQ-5D-5L was sig-
nificantly associated with the patient-rated impact of TD
on 5 activities (self-care, usual activities, being produc-
tive, socializing, eating) and on overall patient-reported
impact (summary score) (P < 0.05 for all coefficients).
Since the EQ-5D-5L utility is scored from 0 to 1, and
unstandardized coefficients reflect the underlying metric
of the scale, the coefficients presented in Table4 can be
interpreted as percentage changes in overall quality of life
due to the specific impacts.
Association betweenseverity/impact ofpossible TD
andSDS total score
Similar to EQ-5D-5L utility, no statistically significant
associations were found between clinician-rated severity
and SDS total score in any body region (Additional file1:
Appendix Table S3). After adjustments, the regression
coefficient for overall patient-rated severity (summary
score based on all body regions) was 0.725 (P > 0.05) in
Cohort 2A, with severity in head/neck having the greatest
effect on SDS total score (Table4). Although coefficients
based on patient-rated severity were not statistically sig-
nificant, the positive coefficients for SDS indicate a posi-
tive association with patient-rated severity, suggesting
that higher severity was associated with worse disability.
As with EQ-5D-5L utility, patient-rated impact of
possible TD had the largest association with SDS total
score (Table4). In Cohort 2A, SDS total score was sig-
nificantly associated with the patient-rated impact of TD
on 4 activities (self-care, usual activities, being produc-
tive, socializing) and on overall impact (summary score)
(P < 0.05 for all coefficients).
Discussion
In these analyses of RE-KINECT data, which were con-
ducted to address the need for quantitative evaluations
of HRQoL in patients with TD, patient-reported out-
comes (PROs) regarding health and physical wellness
(EQ-5D-5L) and social functioning (SDS) were generally
worse in patients with clinician-confirmed possible TD
(Cohort 2) than inthose with no abnormal involuntary
movements (modified Cohort 1). Cohort 2 was charac-
terized by older age, more male patients, a higher preva-
lence of schizophrenia or schizoaffective disorder, greater
severity of psychiatric conditions, worse overall health,
and less ability to work/study than modified Cohort 1.
After adjusting for these potentially confounding factors,
analyses of patient-reported EQ-5D-5L and SDS mean
scores indicated that wellness and functioning were most
negatively affected in patients who were aware of their
Table 2 Mean EQ-5D-5L and SDS scores
EQ-5D-5L EuroQoL 5-dimension 5-level questionnaire; SD standard deviation;
SDS Sheehan Disability Scale; VAS visual analog scale
a Modied Cohort 1 includes patients with no visible or self-reported abnormal
involuntary movements. Cohort 2 includes all patients with possible TD per
clinician assessment
b Adjusted for age, sex, overall health status, severity of psychiatric condition per
clinician impression, functional status of patient per clinician impression, and
psychiatric diagnosis
c Higher EQ-5D-5L scores indicate better health-related quality of life: VAS
(range, 0 =“worst health you can imagine” to 100 =“best health you can
imagine”); utility (range, 0 =“health state equivalent to death” to 1 =“perfect
health”)
d Higher SDS scores indicate greater disruption due to health condition: domain
scores (range, 0= “not at all” [no disruption to work/school, social life, or family/
home life] to 10 =“extremely ” [extreme disruption]). Total scores (range, 0–30)
were calculated for patients who had a score on ≥ 2 domains. When only 1
domain was missing, theaverage of the patient’s observed score was imputed
Modied Cohort
1aCohort 2aP- valueb
n Mean (SD) n Mean (SD)
EQ-5D-5L scoresc
Health state VAS 446 70.4 (21.4) 204 66.8 (25.1) 0.2068
Utility score 442 0.78 (0.18) 197 0.71 (0.21) 0.0163
SDS scoresd
Work/school 310 3.5 (3.4) 111 4.2 (3.4) 0.2559
Social life 446 3.5 (3.2) 203 4.0 (3.4) 0.8237
Family/home life 445 3.4 (3.2) 203 3.8 (3.3) 0.8449
Total score 445 10.5 (8.8) 203 11.7 (9.3) 0.7245

Page 7 of 11
Tanneretal. Journal of Patient-Reported Outcomes (2023) 7:21
possible TD symptoms and rated those symptoms as
having “a lot” of severity in ≥ 1 body region or “a lot” of
impact on ≥ 1 daily activity. ese findings were consist-
ent with results from regression analyses, which showed
that patient-reported impact had the largest (andmostly
significant) effects on EQ-5D-5L utility and SDS total
scores, followed by patient-reported severity. In con-
trast, no statistically significant associations were found
between clinician-rated severity and the PRO measures
(EQ-5D-5L utility and SDS total scores).
Among patients who were aware of their possible TD
(Cohort 2A), self-reported severity (“some” or “a lot”)
was associated with standard PROs (EQ-5D-5L utility,
SDS total). ese results highlight some important points
about awareness and physical/social impact that should
be considered when diagnosing and treating patients
with TD. Severity terms like “mild” and “severe” (or
“some” and “a lot”) are inherently subjective, and patients
who are aware of their TD might find even “milder”
symptoms to be disruptive, debilitating, or embarrassing.
Foroptimal diagnosis and treatment, it may be impor-
tant to assess patients’ awareness of their own symptoms
and the impact of these symptoms on overall wellness
and ability to function—whether self-reported using a
Fig. 2 Distribution of EQ-5D-5L scores. *P < 0.05; **P < 0.01; ***P < 0.001 versus modified Cohort 1. EQ-5D-5L, EuroQoL 5-Dimension 5-Level
questionnaire; TD, tardive dyskinesia

Page 8 of 11
Tanneretal. Journal of Patient-Reported Outcomes (2023) 7:21
formal instrument (EQ-5D-5L, SDS) or simple descrip-
tors (“none”, “some”, or “a lot” of impact). Clinician rat-
ings of TD severity were less likely to be associated with
EQ-5D-5L and SDS scores, underscoring the importance
of assessing the patient’s perspective.
Assessing the detrimental effects of TD on patients’
physical wellness and social functioning was one of the
challenges in this analysis. In a cross-sectional study
such as RE-KINECT, assessing causality (i.e., the addi-
tive effects of TD) can be hypothesized but not con-
cluded. e analysis might have also been limited by
potential overlapping characteristics between modified
Cohort 1 (patients with no abnormal involuntary move-
ments) and Cohort 2 (patients with possible TD) due to
the fact that patients in both cohorts had been treated
with an antipsychotic for ≥ 3months, primarily for seri-
ous mental illnesses such as schizophrenia/schizoaf-
fective disorder, bipolar disorder, and major depressive
disorder. Given the functional decrements associated
with these conditions [25–27], it was anticipated that
both physical and social wellbeing would be diminished
in all RE-KINECT patients regardless of cohort. Never-
theless and importantly, EQ-5D-5L and SDS scores were
consistently worse inCohort 2 than in modified Cohort 1
as expected (Tables2 and 3). However, these differences
were not all statistically significant, possibly due to some
residual confounding even after adjusting for various
patient characteristics (e.g., age, severity of psychiatric
condition).
Another objective of this analysis was to explore
whether patients who were aware of their possible TD
had worse self-reported physical wellness and social
functioning than those who were not aware. is
appeared to have been the case. Comparisons between
Cohort 2A (aware) and Cohort 2NA (not aware) dem-
onstrated similar severity of psychiatric conditions
between the cohorts. However, the patients in Cohort
2A (aware) were more likely to self-report poorer overall
health (Additional file1: Appendix TableS1). In addition,
Cohort 2A patients had significantly worse SDS scores
(total and family/home life), suggesting that patients
who are aware of their TD felt a greater negative impact
in areas that require interaction with other people. e
results also indicate that in patients who are not aware
of their TD, it is important to assess the impact of TD
on patients’ health using caregiver feedback, including
medical issues that the patient might not associate with
TD (e.g., pain). Especially in patients with advanced psy-
chotic disorders, family and caregiver accounts of TD
movements and impact are valuable in assessing the
overall “severity” of the condition.
e emphasis on patients’ experiences in this analysis
does not discount the need for regular clinician assess-
ments or caregiver input. Although clinician-rated sever-
ity was not significantly associated with EQ-5D-5L or
SDS scores, clinician assessment of severity remains a
crucial part of the treatment plan for determining the
effectiveness of interventions. In patients who are not
Table 3 EQ-5D-5L and SDS scores by clinician and patient ratings of severity or impact
EQ-5D-5L EuroQoL 5-dimension 5-level questionnaire; SDS Sheehan Disability Scale; SE standard error; TD tardive dyskinesia
*P < 0.05; **P < 0.01; ***P < 0.001 (signicantly worse for Cohort 2 subgroup versus modied Cohort 1). Adjusted for age, sex, overall health status, severity of
psychiatric condition per clinician impression, functional status of patient per clinician impression, and psychiatric diagnosis
a Cohort 2 included all patients with possible TD per clinician assessment. By denition, all Cohort 2 patients had a clinician-rated severity of “some” or “a lot” in ≥ 1
body region (N = 204)
b Cohort 2A included patients who were aware of their possible TD. By denition, all Cohort 2A patients rated their severity as “some” or “a lot” in ≥ 1 body region
(N = 110). Not all aware patients reported having “some” or “a lot” of impact on daily activities; 27 aware patients either reported “none” for all 7 activities or did not
provide a response
c Modied Cohort 1 included patients who had no abnormal involuntary movements (N = 450)
d Based on linear regression analyses of mean score dierences between Cohort 2 subgroups and modied Cohort 1. Negative values for EQ-5D-5L utility score
indicate worse health-related quality of life for Cohort 2 populations. Positive values for SDS total scores indicate worse functioning for Cohort 2 populations
Severity or impact of possible TD Cohort 2 or 2Aab versus modied Cohort 1c
EQ‑5D‑5L utility score (SE)dSDS total score (SE)d
Cohort 2a
“Some” or “a lot” of clinician-rated severity (n = 204) − 0.037 (0.015)* 0.267 (0.756)
“A lot” of clinician-rated severity (n = 68) − 0.044 (0.023) − 0.732 (1.139)
Cohort 2Ab
“Some” or “a lot” of patient-rated severity (n = 110) − 0.036 (0.019) 1.428 (0.931)
“A lot” of patient-rated severity (n = 52) − 0.089 (0.024)*** 3.288 (1.243)**
“Some” or “a lot” of patient-rated impact (n = 83) − 0.042 (0.021)* 1.675 (1.053)
“A lot” of patient-rated impact (n = 33) − 0.121 (0.031)*** 5.401 (1.509)***

Page 9 of 11
Tanneretal. Journal of Patient-Reported Outcomes (2023) 7:21
aware of their symptoms, caregivers can provide infor-
mation about the location and severity of TD symptoms
that are not present during the office visit or are located
in areas that are not easily visible, such as the toes. In
addition, while patients who reported being unaware
ofabnormal movements may rate the impacts of TD on
functioning as less severe, it remains possible that the
abnormal movements could still significantly affect social
and occupational relations with others who are distracted
by the patient’s movements, regardless of their lack of
awareness or denial. Clinician assessments, patient per-
spectives, and caregiver input should all be considered
when evaluating and treating patients with TD.
Limitations for this study have been previously dis-
cussed [17]. In brief, it should be noted that RE-KINECT
was intended to be a screening study of possible TD with
no requirement for “formal” TD diagnoses. Additionally,
the cohorts were not matched for any sociodemographic
factors, and statistically significant differences between
patients with possible TD (Cohort2) and patients with
no abnormal involuntary movements (modified Cohort
1) were found for age, sex, psychiatric conditions, sever-
ity of psychiatric conditions, overall health status, and
overall functional status (Table1). However, the adjust-
ments for age, sex, and psychiatric conditions in this
analysis should have mitigated some of this variability.
Itshould also be noted that although the EQ-5D-5L and
SDS are well-established and standardized instruments
(PROs), neither has been specifically validated for TD.
However, the strong associations found between EQ-
5D-5L/SDS and patient-reported impact of possible TD
indicate that both scales may be appropriate for assessing
HRQoL in patients with TD. Finally, the ordinal values
assigned to various outcomes for the regression analyses
(e.g., 0 = "none", 1 = "some", 2 = "a lot") assumes that
the qualitative differences between responses are simi-
lar, which may or may not be true. e rationale for this
approach was to provide a relatively simple method that
can be easily reproduced. e findings based on this
approach were sufficient to confirm the applicability of
a simple impact assessment (“none”, “some”, or “a lot”) to
real-world practice.
Conclusions
e results of this analysis indicate that physical well-
ness and social functioning were diminished in patients
with possible TD, particularly in those who were aware
of their abnormal involuntary movements and rated
those movements as having “a lot” of impact on daily
activities. Patient-rated severity and impact of possible
TD, but not clinician-rated severity, were significantly
associated with EQ-5D-5L utility or SDS total scores.
ese outcomes suggest that in addition to assessing
the presence and severity of patients’ abnormal move-
ments during usual care visits, clinicians or their staff
may need to ask patients about how TD adversely
affects their HRQoL (particularly as it relates to their
daily activities) and consider these impacts when mak-
ing and evaluating treatment plans. For some patients,
such questions may be as important (or possibly even
more important) than symptom severity. For patients
not aware of their TD, discussion with caregivers may
be needed to determine the impact of TD on patients’
Table 4 Association between patient-rated severity/impact of
possible TD and EQ-5D-5L/SDS (Cohort 2A)
EQ-5D-5L EuroQoL 5-dimension 5-level questionnaire; SD standard deviation;
SDS Sheehan Disability Scale; SE standard error; TD tardive dyskinesia
*P < 0.05; **P < 0.01; ***P < 0.001 for the linear regression coecient, indicating
that the association was statistically signicant. Adjusted for age, sex, overall
health status, severity of psychiatric condition per clinician impression,
functional status of patient per clinician impression, andpsychiatric diagnosis
a Cohort2Aincluded patients who were aware of their possible TD. By denition,
all Cohort 2A patients rated their severity as “some” or “a lot” in ≥ 1 body region
(N = 110). Not all aware patients reported having “some” or “a lot” of impact on
daily activities; 27 aware patients either reported “none” for all 7 activities or did
not provide a response
b Based on patient ratings of “none” (score = 0), “some” (score = 1), or “a lot”
(score = 2), divided by thecohort size (N = 110). For missing values, a score of
0 was assigned. Summary based on summed scores: range, 0 to 8 (severity of
possible TD); range, 0to14 (impact of possible TD)
c Negative regression coecient indicates an association between higher
(worse) severity/impact scores and lower (worse) EQ-5D-5L utility index scores.
Positive regression coecient indicates an association between higher (worse)
severity/impact scores and higher (worse) SDS total scores. For these analyses,
EQ-5D-5L utility and SDS total scores were the dependent variables. Within each
scale, coecients can be compared to each other for relative strength, but they
should not be interpreted as “low” or “high”
Severity
and impact
of possible
TD(Cohort 2A)a
Mean score
(SD)bRegression coecient (SE)c
EQ‑5D‑5L
utility SDS total
Patient-rated severity
Head/face 0.95 (0.75) − 0.026 (0.026) 2.153 (1.136)
Neck/trunk 0.31 (0.60) − 0.013 (0.033) − 1.095 (1.497)
Upper extremi-
ties 0.84 (0.75) − 0.032 (0.026) 1.031 (1.175)
Lower extremi-
ties 0.59 (0.72) − 0.060 (0.027)* 0.539 (1.215)
Summary 2.7 (1.6) − 0.028 (0.012)* 0.725 (0.560)
Patient-rated impact
Usual activities 0.57 (0.70) − 0.094
(0.028)*** 4.544 (1.228)***
Talking 0.53 (0.70) − 0.032 (0.027) 1.992 (1.231)
Eating 0.41 (0.65) − 0.068 (0.030)* 1.987 (1.394)
Breathing 0.11 (0.34) − 0.062 (0.054) 2.927 (2.511)
Being productive 0.59 (0.69) − 0.075 (0.028)** 4.031 (1.232)**
Self-care 0.35 (0.60) − 0.116
(0.032)*** 4.759 (1.429)**
Socializing 0.68 (0.74) − 0.075 (0.026)** 2.818 (1.168)*
Summary 3.2 (3.1) − 0.023
(0.006)*** 1.027 (0.276)***

Page 10 of 11
Tanneretal. Journal of Patient-Reported Outcomes (2023) 7:21
health and daily activities. Such discussions might be
facilitated by specific and descriptive questions that
are specifically focused on social withdrawal (avoiding
social interaction or events, isolation, embarrassment)
or physical disabilities associated with TD symptoms
(difficulty swallowing, breathing, walking, or writing).
Supplementary Information
The online version contains supplementary material available at https:// doi.
org/ 10. 1186/ s41687- 023- 00551-5.
Additional le1. Appendix tables and figures.
Acknowledgements
Medical writing assistance was provided by Mildred Bahn (Prescott Medical
Communications Group; Chicago IL) with support from Neurocrine Bio-
sciences, Inc.
Author contributions
All authors met International Committee of Medical Journal Editors (ICMJE)
recommendations for authorship, including critical review of all drafts and
approval of the final manuscript for submission.
Funding
This study and the data analyses were funded by Neurocrine Biosciences, Inc.
Availability of data and materials
The datasets used and/or analyzed during the current study are available from
the study sponsor (Neurocrine Biosciences, Inc.) upon reasonable request for
research purposes.
Declarations
Ethics approval and consent to participate
All patients provided written and informed consent for participation in the
study. For all participating study sites, protocol and ethics approvals were
provided by Quorum Review IRB (Seattle, WA).
Consent for publication
Not applicable.
Competing interests
CMT receives grants from the Michael J. Fox Foundation, the Parkinson’s
Foundation, the Department of Defense, BioElectron, Roche/Genentech, Bio-
gen, and the National Institutes of Health, compensation for serving on Data
Monitoring Committees from Biotie Therapeutics, Voyager Therapeutics and
Intec Pharma and personal fees for consulting from Neurocrine Biosciences,
Adamas Therapeutics, Biogen, 23andMe, Alexza, Grey Matter, Acadia and CNS
Ratings. SNC has served as a consultant to Neurocrine Biosciences, Inc., Teva
Pharmaceuticals Industries, Ltd., Osmotica Pharmaceuticals and Disper-
sol Technologies, and received separate research grants from Neurocrine
Biosciences, Inc., Pfizer, Inc., and Eagle pharmaceuticals. AJC has served as a
consultant to, received research grants from, and is on the speaker bureau for
ACADIA, Avanir, Neurocrine, Otsuka and Teva, is a consultant to MedAvante-
Prophase, and is on the Board of the Neuroscience Research Institute. WRL is
a current full-time employee of Evidera. HS and VP are former employees of
Evidera. Evidera employees served as consultants on this work. EGF, MS, and
CY are full-time employees of Neurocrine Biosciences, Inc.
Author details
1 Weill Institute for Neurosciences, Department of Neurology, University of Cali-
fornia - San Francisco, 1651 4th Street, San Francisco, CA 94158, USA. 2 Parkin-
son’s Disease Research, Education and Clinical Center, San Francisco Veterans
Affairs Health Care System, 4150 Clement St., San Francisco, CA 94121, USA.
3 Corporal Michael J. Crescenz Veterans Affairs Medical Center, 3900 Woodland
Ave., Philadelphia, PA 19104, USA. 4 Perelman School of Medicine, University
of Pennsylvania, 3400 Civic Center Blvd., Philadelphia, PA 19104, USA. 5 SUNY
Upstate Medical University, 8429 Lorraine Rd., Lakewood Ranch, FL 34202,
USA. 6 Evidera, 500 Totten Pond Rd., Waltham, MA 02451, USA. 7 Evidera, 7575
Trans-Canada Hwy., St-Laurent, QC H4T 1V6, Canada. 8 Neurocrine Biosciences,
Inc., 12780 El Camino Real, San Diego, CA 92130, USA.
Received: 6 April 2022 Accepted: 16 January 2023
References
1. Jain R, Correll CU (2018) Tardive dyskinesia: recognition, patient assess-
ment, and differential diagnosis. J Clin Psychiatry 79(2):nu17034ah17031c
2. Hauser RA, Meyer JM, Factor SA, Comella CL, Tanner CM, Xavier RM,
Caroff SN, Lundt L (2020) Differentiating tardive dyskinesia: a video-based
review of antipsyhotic-induced movement disorders in clinical practice.
CNS Spectr 27:208–217
3. Cloud LJ, Zutshi D, Factor SA (2014) Tardive dyskinesia: therapeutic
options for an increasingly common disorder. Neurotherapeutics
11(1):166–176
4. Carbon M, Hsieh CH, Kane JM, Correll CU (2017) Tardive dyskinesia
prevalence in the period of second-generation antipsychotic use: a
meta-analysis. J Clin Psychiatry 78(3):e264–e278
5. Citrome L, Saklad SR (2020) Revisiting tardive dyskinesia: focusing on the
basics of identification and treatment. J Clin Psychiatry 81:TV18059AH3C
6. Caroff SN (2020) Recent advances in the pharmacology of tardive dyski-
nesia. Clin Psychopharmacol Neurosci 18:493–506
7. Bhidayasiri R, Jitkritsadakul O, Friedman JH, Fahn S (2018) Updating the
recommendations for treatment of tardive syndromes: a systematic review
of new evidence and practical treatment algorithm. J Neurol Sci 389:67–75
8. American Psychiatric Association (2020) Practice guideline for the treat-
ment of patients with schizophrenia. Am J Psychiatry. https:// doi. org/ 10.
1176/ appi. ajp. 2020. 177901
9. Caroff SN, Citrome L, Meyer J, Sajatovic M, Goldberg JF, Jain R, Lundt L,
Lindenmayer JP, Mcevoy JP, Mcintyre RS, Tohen M, Ketter TA (2020) A
modified Delphi consensus study of the screening, diagnosis, and treat-
ment of tardive dyskinesia. J Clin Psychiatry 81:19cs12983
10. Ascher-Svanum H, Zhu B, Faries D, Peng X, Kinon BJ, Tohen M (2008)
Tardive dyskinesia and the 3-year course of schizophrenia: results from a
large, prospective, naturalistic study. J Clin Psychiatry 69:1580–1588
11. Caroff SN, Davis VG, Miller DD, Davis SM, Rosenheck RA, Mcevoy JP,
Campbell EC, Saltz BL, Riggio S, Chakos MH, Swartz MS, Keefe RS, Stroup
TS, Lieberman JA, Investigators C (2011) Treatment outcomes of patients
with tardive dyskinesia and chronic schizophrenia. J Clin Psychiatry
72:295–303
12. Caroff SN, Leong SH, Roberts C, Berkowitz RM, Campbell EC (2020) Cumu-
lative burden of illness in veterans with tardive dyskinesia and serious
mental disorders. J Clin Psychopharmacol 40:38–45
13. Yassa R, Jones BD (1985) Complications of tardive dyskinesia: a review.
Psychosomatics 26:305–313
14. Strassnig M, Rosenfeld A, Harvey PD (2018) Tardive dyskinesia: motor
system impairments, cognition and everyday functioning. CNS Spectr
23:370–377
15. Carroll B, Irwin DE (2019) Health care resource utilization and costs for
patients with tardive dyskinesia. J Manag Care Spec Pharm 25:810–816
16. Mcevoy J, Gandhi SK, Rizio AA, Maher S, Kosinski M, Bjorner JB, Carroll
B (2019) Effect of tardive dyskinesia on quality of life in patients with
bipolar disorder, major depressive disorder, and schizophrenia. Qual Life
Res 28:3303–3312
17. Caroff SN, Yeomans K, Lenderking WR, Cutler AJ, Tanner CM, Shalhoub H,
Pagé V, Chen J, Franey E, Yonan C (2020) RE-KINEC T: a prospective study
of the presence and healthcare burden of tardive dyskinesia in clinical
practice settings. J Clin Psychopharmacol 40:259–268
18. Cutler AJ, Caroff SN, Tanner CM, Shalhoub H, Lenderking WR, Pagé V,
Franey E, Yonan C (2021) Caregiver burden in RE-KINECT: data from ret-
rospective real-world tardive dyskinesia screening study. J Am Psychiatr
Nurses Assoc. https:// doi. org/ 10. 1177/ 10783 90321 10235 65

Page 11 of 11
Tanneretal. Journal of Patient-Reported Outcomes (2023) 7:21
19. American Psychiatric Association (2013) Diagnostic and statistical manual
of mental disorders: DSM-5. American Psychiatric Association, Arlington
20. Schooler NR, Kane JM (1982) Research diagnoses for tardive dyskinesia.
Arch Gen Psychiatry 39:486–487
21. American Psychiatric Association (2000) Diagnostic and statistical
manual of mental disorders: DSM-IV-TR. American Psychiatric Association,
Arlington
22. American Psychiatric Association (2022) Diagnostic and statistical
manual of mental disorders: DSM-5-TR. American Psychiatric Association,
Arlington
23. Herdman M, Gudex C, Lloyd A, Janssen M, Kind P, Parkin D, Bonsel G,
Badia X (2011) Development and preliminary testing of the new five-level
version of EQ-5D (EQ-5D-5L). Qual Life Res 20:1727–1736
24. Sheehan DV, Harnett-Sheehan K, Raj BA (1996) The measurement of dis-
ability. Int Clin Psychopharmacol 11(Suppl 3):89–95
25. Lenert LA, Sturley AP, Rapaport MH, Chavez S, Mohr PE, Rupnow M (2004)
Public preferences for health states with schizophrenia and a mapping
function to estimate utilities from positive and negative symptom scale
scores. Schizophr Res 71:155–165
26. Martin-Subero M, Berk L, Dodd S, Kamalesh V, Maes M, Kulkarni J, De
Castella A, Fitzgerald PB, Berk M (2014) Quality of life in bipolar and
schizoaffective disorder—a naturalistic approach. Compr Psychiatry
55:1540–1545
27. Porcelli S, Kasper S, Zohar J, Souery D, Montgomery S, Ferentinos P,
Rujescu D, Mendlewicz J, Merlo Pich E, Pollentier S, Penninx B, Serretti A
(2020) Social dysfunction in mood disorders and schizophrenia: clinical
modulators in four independent samples. Prog Neuropsychopharmacol
Biol Psychiatry 99:109835
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations.