Abstracts of the 4th Biennial Schizophrenia International Research Conference / Schizophrenia Research 153, Supplement 1 (2014) S1–S384 S265

Poster #M206

PERFORMANCE ON THE UCSD PERFORMANCE-BASED SKILLS

ASSESSMENT (UPSA) AND REAL-WORLD OUTCOMES IN SEVERE MENTAL

ILLNESS: A SYSTEMATIC REVIEW OF THE LITERATURE

Amy Duhig1, Julie Myers2, Amber Pitts2, Steven Hass1, Robert Klein2,

Philip D. Harvey3

1AbbVie; 2MDM; 3U of Miami

Background: There is a regulatory agency expectation that pharmacother-

apies developed to treat cognition in schizophrenia must demonstrate

improvement on cognitive and functional co-primary endpoints. The Mea-

surement and Treatment Research to Improve Cognition in Schizophrenia

(MATRICS) initiative conducted the Validation of Intermediate Measures

study (VIM) to assess the psychometric characteristics and practicality of

alternative functional measures. Results of this study identified the UCSD

Performance-based Skills Assessment (UPSA) as the superior co-primary

measure due to its psychometric properties and reasonable acceptability

and practicality. The objective of this study was to systematically evaluate

literature regarding the UPSA’s relationship to real-world functional out-

comes that are relevant to treatment and coverage decisions in today’s

cost-constrained medical environment.

Methods: Systematic searches of articles published between January 2001

and April 2013 were conducted in multiple databases. Results were re-

viewed using predetermined inclusion and exclusion criteria. Included

articles utilized any version of the UPSA and either milestone (direct) or

rating scale measures of real-world outcomes. After relevant studies were

selected, data were extracted (e.g., study design, patient characteristics,

country of study, study size, duration, UPSA version, measures of real-world

outcomes, rater, summary of results relating to the relationships between

the UPSA and real-world outcomes). Results of studies were compared and

contrasted.

Results: Forty studies met the inclusion criteria, with most being conducted

in the U.S. (n=31). Seven studies utilized milestone measures, 27 used rating

scale measures, and six evaluated both milestone and rating scale measures

of real-world outcomes. Overall, the studies reviewed suggest the UPSA and

the UPSA-B are associated with real-world outcomes, whether milestone

or rating scale measures were utilized. Studies evaluating residential and

employment milestones reported significant correlations with UPSA and

UPSA-B scores. Studies of the correlations between the UPSA and various

components of the rating scales generally showed modest associations.

The Specific Level of Function (SLOF) rating scale, a measure of real world

function (personal care, interpersonal relationships, social acceptability, ac-

tivities of community living, work skills), was used in 16 of the 40 studies.

SLOF scores rated by clinicians and informants have been found to be

significantly correlated to both UPSA and UPSA-B scores (r=0.18-0.63). Less

consistent results and weaker associations were reported for self-reports

and in countries outside the US.

Discussion: This review identified substantial evidence indicating the

UPSA is associated with the real-world status/capabilities of patients with

schizophrenia. However, further research is needed to better understand

how social support and healthcare systems in countries other than the US

impact the UPSA and real-world outcomes associations. Taken together, the

evidence reviewed supports the use of the UPSA as a proxy for functional

improvements that are relevant not only to patients, families, physicians

and treatment teams, but also to reimbursement authorities.

Poster #M207

USE OF RELIABLE CHANGE INDEX TO EVALUATE CLINICAL SIGNIFICANCE

IN THE POSITIVE AND NEGATIVE SYNDROME SCALE (PANSS): A CATIE

ANALYSIS

Linda Gao1, Anzalee Khan2,1, Luka Lucic1, Brian Rothman1

1ProPhase LLC; 2Nathan S. Kline Institute for Psychiatric Research

Background: The PANSS is the most widely used measures of psychopathol-

ogy in schizophrenia. It is commonly used in both randomized controlled

trials (RCT) and non-controlled evaluations. RCTs assess clinical efficacy of

an intervention relative to a placebo or control condition by making group

comparisons and evaluating for statistically significant differences. How-

ever, statistical significance does not in itself provide concise information

about a given intervention’s clinically meaningful effects. The process of

defining clinical significance remains a challenge. As an attempt to develop

a standard method of estimating clinically significant change, we propose

adoption of a two-part strategy: The first part of the strategy involves

using the Reliable Change Index (RCI). The second part involves use of

examination of clinical significance (CS). RCI is whether patients changed

sufficiently that the change is unlikely to be due to measurement unreli-

ability. CS change takes the patient from a score typical of schizophrenia

to a score typical of the “normal” population. Studying RCI and CS has

moved the outcomes paradigm from studying treatment groups to studying

individual change within those groups. Assessments must move beyond

symptom focus and evaluate individuals with respect to the complex

broader domains of their functional, real-world, lives in which clinically

significant change is operationalized.

Methods: Data on symptomatology, PANSS, from the Clinical Antipsychotic

Trials of Intervention Effectiveness (CATIE) were analyzed. Three methods

of RCI (Jacobson-Truax, Edwards-Nunnally, and Hageman-Arindell meth-

ods) were compared to CS change (pre to post change of at least 2 SDs

from the original mean, 20% improvement, and change in PANSS remission

criteria).

Results: For the three RCI methods, 29.73%, 31.08% and 52.70% showed

reliable improvements in PANSS scores. For CS, 22.97% showed greater

than 20% improvement, 29.73% improved on the PANSS remission criteria,

and only 8.11% showed CS improvement of 2 SDs from the mean. When

comparing RCIs with CS, only 18.92% of CS improvement also resulted in RCI

significant improvement. Regarding clinically meaningful improvement, the

Hageman-Arindell method was most concordant with all three RCI mea-

sures and with the 20% improvement as this method differentially analyzes

clinically meaningful change at the individual level and at the group level

(i.e., obtaining proportions of patients who have reliably changed and

passed the cutoff point).

Discussion: Reliable and clinically significant change should be reported

in articles to complement the more familiar group summary methods. As-

sessment of clinically meaningful change is useful for evaluating treatment

response. Outcome studies often assess statistically significant change,

which may not be clinically meaningful. Comparisons of the proposed

methods of determining clinically significant PANSS outcomes to biomedi-

cal standards of clinical significance will help determine the validation of

this procedure, and improve the precision and effectiveness of the PANSS

in clinical trials.

Poster #M208

AGEING IN SCHIZOPHRENIA: A SYSTEMATIC REVIEW

Matti Isohanni1, David Cowling2

1Department of Psychiatry, University of Oulu; 2University of Oulu

Background: Schizophrenia is generally a lifelong condition. Despite high

mortality, most survive into old age. Few prospective longitudinal studies

have analysed trajectories from early-mid adulthood into old age. System-

atically review longitudinal studies of the progression of schizophrenia into

old age, focusing on cognition, functioning, co-morbidity, mortality and

quality of life. Advance understanding of the course of schizophrenia and

highlight interventions that improve outcomes and even achieve a state of

wellbeing in later life.

Methods: Electronic search of PubMed, PsychINFO and Scopus. Search

terms: (ageing OR “older adult” OR elderly OR geriatric OR “late life”)

AND (schizophrenia OR schizoaffective OR schizophreniform"). Articles and

books searched manually.

Results: The course in later life is variable. Many remain symptomatic

and mortality and somatic comorbidity increase. Higher rates of decline

in cognitive functioning affect ability to function independently. However,

many individuals have a favourable clinical course and may stop receiving

treatment altogether. 18-27% achieve recovery in old age. Growing evidence

base for interventions that alleviate symptoms, improve social and cognitive

functioning and quality of life. Inequalities remain in the quality and range

of treatment interventions available to older people with schizophrenia.

Discussion: Early introduction of regular psychiatric and somatic assess-

ments and prompt and adequate treatment of symptoms and comorbidities

throughout the life course are essential. Cases of remission/recovery are

often excluded in clinical research. This risks presenting a biased, somewhat

S266 Abstracts of the 4th Biennial Schizophrenia International Research Conference / Schizophrenia Research 153, Supplement 1 (2014) S1–S384

pessimistic description of the illness course into later life. Evidence base

for interventions needs improvement and future trials must include older

participants.

Poster #M209

THE LONGITUDINAL COURSE OF SCHIZOPHRENIA: A SYSTEMATIC REVIEW

Adrian Preda1, Robert Bota2, Steven Potkin1

1University of California Irvine; 2Kaiser Permanente

Background: Naturalistic studies of schizophrenia have been completed

prior to the emergence of neuroleptics in the last part of the last century.

Since the advent of neuroleptics longitudinal studies of schizophrenia have

mostly followed medicated patients. In addition to changes in the medica-

tion background of schizophrenia, a number of changes in the diagnostic

criteria over time brings an additional layer of complexity to any attempt to

systematically organize data about the longitudinal course of schizophrenia.

In this context a systematic review of the literature is essential to prop-

erly understand the longitudinal course of schizophrenia, with important

implications for early diagnosis and intervention, prophylaxis, diagnostic

validity, and prognosis. Further, properly charting the longitudinal course

of schizophrenia is essential to improve our definition of concepts such as

partial response, remission and recovery.

Methods: We completed a systematic literature search for longitudinal,

both retrospective and prospective studies of schizophrenia. To decrease

the heterogeneity of studies span using different diagnostic criteria and

important confounders in the study population (e.g. non-medicated vs.

medicated patients, institutionalization status) we organized our review

based on historical periods which were deemed “homogeneous” in rapport

to important variables of interest (e.g. institutionalization and deinstitu-

tionalization, pre and post neuroleptic periods).

Results: The majority of the longitudinal studies of schizophrenia report

that up to 30-50% of patients present with a stable or favorable course.

Gender, age of onset, duration of illness, core symptoms (positive, negative,

cognitive) can putatively affect prognosis and course.

Discussion: The literature indicated that the course of schizophrenia is one

of stability of symptoms over time for the majority of patients. Stability is

maintained at a lower level of functioning than pre-diagnosis - and as such

is not the optimal outcome - at the same time stability also implies that

gradually deterioration is unlikely for the majority of patients. Of note, the

literature tends to emphasize the less favorable outcome of the minority of

the patients (less than 50%) who show a gradual and progressive deterio-

ration during the course of illness. A “bleaker” interpretation of course and

prognosis dat while justified from a public health perspective might not

be as informative to the clinician aiming to have an objective and, if not

unjustified, appropriately hopeful discussion of course and prognosis with

individual patients.

Poster #M210

THE UCSD PERFORMANCE-BASED SKILLS ASSESSMENT-BRIEF JAPANESE

VERSION (UPSA-B_J): DISCRIMINATIVE VALIDITY FOR SCHIZOPHRENIA

Chika Sumiyoshi1,2, Manabu Takaki3, Yuko Okahisa3, Thomas Patterson4,

Philip D. Harvey5, Tomiki Sumiyoshi6

1Faculty of Human Development and Culture, Fukushima University;2Fukushima University; 3Dept. of Neuropsychiatry, Okayama University

Graduate School of Medicine, Dentistry and Pharmaceutical Sciences; 4Dept. of

Psychiatry, School of Medicine, University of California San Diego; 5Dept. of

Psychiatry and Behavioral Sciences, University of Miami Miller School of

Medicine; 6National Center of Neurology and Psychiatry, Clinical Research Unit

Background: The UCSD Performance-based Skills Assessment (UPSA) has

been developed to assess daily living skills related to neurocognition in

people with psychiatric disorders (Patterson et al, 2001). Specifically, its

brief version (UPSA-B; Mausbach et al, 2007) has been widely used for eval-

uating functional capacity in patients with schizophrenia (Harvey, 2009).

To enhance the clinical utility of this battery, its sensitivity, i.e. an optimal

cut-off point discriminating between normal subjects and patients, needs

to be clarified. The current study investigated the issue using the Japanese

version of UPSA-B (UPSA-B_J; Sumiyoshi et al., 2011).

Methods: Sixty-four Japanese patients meeting DSM-IV-TR criteria for

schizophrenia (M/F=34/30; mean age=35.2) and 113 normal controls

(university students=9/21, mean age=20.6; office workers=71/12, mean

age=34.6) entered the study. The UPSA-B_J and MATRICS Cognitive Consen-

sus Battery Japanese version (MCCB_J) was administered to all participants.

The group differences (students vs. workers vs. patients) were analyzed by

one-way ANOVA for the UPSA-B_J Total score, and by ANCOVA (controlling

for education) for the MCCB T-score. The profiles UPSA-B_J were created

to show task-specific performance in each group. Receiver Operating Char-

acteristic (ROC) curve analysis was conducted for the UPSA Total score.

The optimal cut-off was determined in a manner to maximize the sum of

sensitivity (% of hit) and specificity (% of correct rejection) (Youden, 1950;

Mausbach, 2011).

Results: Overall results: Normal students and workers performed better

than patients both in the MCCB_J (Students > Workers > Patients; F=56.65,

df=2 170, p<0.5) and UPSA-B_J (Students=Workers > Patients: F=38.19

df=2 174, p<0.01). The profiles for UPSA-B_J revealed that scores of the

memory-oriented tasks in the Communication part for normal workers and

patients tended to be worse compared to those for normal students. ROC

analysis: The area under the ROC curve (0.80, CI: 0.73-0.87, p <0.001)

and d’ (1.46) indicated good discriminative power of the UPSA-B_J. The

optimal cut-off was estimated as 81.5 (MAX=100), at which sensitivity and

specificity were 81.3% and 65.5%, respectively.

Discussion: The UPSA-B_J Total score around 80 was found to distinguish

adequately between patients and control subjects. This cut-off seems to

be consistent with a previous study reporting “functional milestones (i.e.

residential of employment status)” in patients were to be around this

point (Mausbach, 2011). Result from MCCB_J and UPSA-B_J indicated that

daily-living skills (especially Communication), as measured by the UPSA-B,

depend on neuropsychological abilities.

References:[1] Sumiyoshi T, Sumiyoshi C, Hemmi C: UCSD Performance-Based Skills Assessment-

Brief (UPSA-B): Japanese version; Administration and Scoring Manual, 2011.

Poster #M211

THE THREE FACES OF SCHIZOPHRENIA: SUB-TYPING SCHIZOPHRENIA

BASED ON RESPONSE AND IMPLICATIONS FOR TREATMENT

Ofer Agid1, Gagan Fervaha1,2, George Foussias1, Cynthia Siu3,

Krysta McDonald2, Gary Remington1

1University of Toronto; 2Centre for Addiction and Mental Health; 3Data Powers

Background: Treatment and classification of schizophrenia continues to

present an enormous challenge, in part due to the disease’s heterogeneous

nature. Of note, response to a particular treatment, or lack thereof, is used

for diagnostic purposes in other areas of medicine (e.g., Insulin/Non-Insulin

Dependent Diabetes Mellitus, certain cancers). The present work set out

to establish if, in fact, antipsychotic treatment response can be used to

subtype distinguishable groups of schizophrenia that may, in turn, be used

to better elucidate differences in underlying pathophysiologic mechanisms.

Methods: Relevant literature was reviewed with specific reference to an-

tipsychotic treatment response. Where available, we focused on the early

stages of the illness and evidence that sequentially followed antipsychotic

trials, including use of clozapine in individuals identified as treatment resis-

tant. This approach aligns with evidence that early and effective treatment

favourably alters clinical and functional outcome measures.

Results: While the first-episode schizophrenia population has very high

symptom response rates, approximately 25-30% do not respond favourably.

For patients experiencing partial or non-response to first-line treatment,

current guidelines indicate a switch to another non-clozapine antipsychotic

before resorting to clozapine as a third (or later) line of treatment. Clini-

cians also often employ dose increase and/or polypharmacy in these cases,

although the benefit of these strategies has been called into question. These

early non-responders, or treatment-resistant patients (TRS), generally re-

spond exclusively to clozapine. While D2 blockade is identified as central

to the antipsychotic effect observed with standard antipsychotics, clozap-

ine’s mechanism(s) of action are currently unclear. In those demonstrating

suboptimal response to clozapine (i.e. URS) we currently have no effective

treatments, making it even more difficult to speculate regarding underlying

pathophysiological differences.

Discussion: Antipsychotic response patterns identify three subtypes of