Transcript Slide 1

Independent Living Skills Inventory (ILSI) and the Nurses’ Observational Scale for Inpatient Evaluation,
30-item version (NOSIE) predict length of hospitalization
Yuliana E. Gallegos, Charlie A. Davidson, Elizabeth A. Cook, William D. Spaulding, Ph.D.
University of Nebraska-Lincoln
Hypotheses & Methods
Background & Materials
Length of inpatient hospitalization is an important indicator of rehabilitation
progress and efficiency for people with treatment-resistant serious mental illness.
Furthermore, research has demonstrated that ward behavior is predictive of
outcome measures such as length of hospitalization (LOH). An individual’s
psychosocial skill acquisition and milieu behavior are important criteria by
which discharge decisions are made, and these are quantified in the Independent
Living Skills Inventory (ILSI) and the Nurses’ Observation Scale for Inpatient
Evaluation 30-item version (NOSIE), respectively.
The NOSIE (Honigfield, Gillis, & Klett, 1966) is an observational measure
that identifies behavioral strengths and weaknesses of ward patients. In addition
to computing a Total Assets score, its components form six subscales: Social
Competence, Social Interest, Personal Neatness, Irritability, Psychoticism, and
Motor Retardation. The ILSI (Sanchez, 1986) is a functional assessment that
consists of ten domains relevant to independent community living: Personal
Skills, Hygiene, Clothing, Basic Functioning, Interpersonal Skills, Home
Management, Money Management, Cooking Skills, Resource Utilization, and
General Occupational Skills.
The relationship between ILSI and symptom severity was assessed using the
Brief Psychiatric Rating Scale (BPRS; Lukoff et al., 1986). BPRS scores were
transformed to factor dimensional measures of Thought Disturbance (positive
symptoms, approximately) and Anergia (negative symptoms) based on the 16item, four factor structure of the BPRS (Long & Brekke, 1999).
Results
Conclusions
Significant correlations were found between LOH and ILSI total (n=98,
r=.448, p<.001) and NOSIE total (n=147, r=.542, p<.001). Follow-up analyses
showed that this relationship is most relevant for predicting shorter LOH. For
participants with below average LOH, the correlation did not change
substantially (ILSI n=60, NOSIE n=94). However, for those with higher-thanaverage LOH, this correlation was not significant (ILSI n=38, r=.263, p=.11;
NOSIE n=53, r=.214, p=.125). See Figure 1.
ILSI and BPRS data were examined for 36 people from the total sample.
Correlations with ILSI were not found for positive symptoms (r=.098, p=.571)
or negative symptoms (r=.-.087, p=.617), although Type II error cannot be
excluded due to the small sample size. Thus, the ability of symptoms to predict
living skills was not supported, and with no significant correlations, divergent
validity could not be assessed.
This study demonstrates the predictive validity of the NOSIE and ILSI and
bolsters their clinical value. Strong significant correlations were demonstrated
between these measures and length of hospitalization. Although these scales serve
different clinical purposes and assess different behavioral constructs, this study
demonstrates their convergent validity, clinical utility, and the importance of
behavioral and skills assessment to efficient psychiatric rehabilitation.
Future studies should examine relationships between these measures and other
levels of assessment, such as neurocognition and social cognition. NOSIE and
ILSI subscales should be examined in multivariate analyses with outcomes such
as LOH to determine the unique contributions, interactions, and longitudinal
aspects of different skills and behaviors. Finally, the lack of a relationship
between LOH and NOSIE or ILSI should be followed up in similar and other
designs with a larger sample size.
Figure 1: Correlations at month 3 (Z-scores): LOH, NOSIE & ILSI
Table 1 Demographics Information
n (%)
Gender
Female
Male
Age of Admission
Years of Education
Age of Onset (n=162)
Axis I Diagnosis
Paranoid Schizophrenia
Schizophrenia Undifferentiated
Schizoaffective
Bipolar Disorder
Psychotic Disorder NOS
Axis II Diagnosis
Borderline
Schizoid
Paranoid
Antisocial
Personality Disorder NOS
None
Race
Caucasian
African-America
Hispanic
Asian-American
Other
Marital Status
Married
Single
Divorced
Other
It was hypothesized that LOH would be significantly negatively related to ILSI
and NOSIE summary scores, thus demonstrating both the validity and clinical
utility of these measures and the importance of behavioral assessment in
psychiatric rehabilitation.
This study examined archival longitudinal data collected at the Lincoln
Regional Center’s Community Transition Program, a 40-bed inpatient unit.
Individuals in this program spent about 35 to 40 hours per week participating in
structured treatment and rehabilitation activities, such as skills training (e.g., social
skills training, occupational therapy, therapeutic recreation) and contingency
management (e.g., behavior management program). The CTP treated chronic,
treatment-refractory individuals with severe mental disorders.
Data consisted of 3, 6 and 12-month administrations of the NOSIE and ILSI.
Month 3 is presented here for sample size and interest in predictive ability,
although results were not substantially different at the other time points. Data were
cleaned to ensure normality. Pearson’s correlations were used to assess each
scale’s relationship with length of hospitalization, and Steiger’s Z-tests were used
to compare their predictive value. It was hypothesized that both ILSI and NOSIE
would be significantly negatively correlated with LOH.
The relationship between symptoms and living skills was examined in a
divergent validity design. As demonstrated in similar settings (Brill et al., 2009), it
was hypothesized that negative symptoms would be strongly correlated with ILSI,
whereas positive symptoms would be less or non-correlated.
M
SD
39.67
12.51
20.42
11.71
2.17
7.62
78 (53.0)
88 (47.0)
60 (36.1)
48 (28.9)
36 (21.7)
7 (4.2)
12 (7.2)
12 (7.2)
1 (0.6)
11 (6.6)
6 (3.6)
43 (25.9)
84 (50.6)
147 (88.6)
12 (7.2)
2 (1.2)
2 (1.2)
3 (1.8)
8 (4.8)
107 (64.5)
43 (25.9)
8 (4.8)