TNH & FPECA Data Request Proposal to the Minimum Data Set (MDS) Version 2.0

topsalmonAI and Robotics

Feb 23, 2014 (3 years and 1 month ago)

580 views

TNH & FPECA

Data Request Proposal to the Minimum Data Set (MDS)
Version 2.0



















Characteristics of Florida Nursing Home Residents





































EXECUTIVE SUMMARY

This proposal is a consolidation of several distinct studi
es that will be conducted using Minimum Data Set (MDS) version 2.0.


Researchers at the School of Aging Studies (SAS) and the Florida Mental Health Institute (FMHI) at the University of South
Florida as well as Florida researchers involved in the
Teaching
Nursing Home
[1]

have combined expertise and resources to
make a series of coordinated Minimum Data Set version 2.0 requests to more effectively use the available
resources as well as
more efficiently work with data from the Centers for Medicare & Medicaid Services (CMS).





These studies has been reviewed and funded by Florida Medicaid and have been reviewed or exempted from review by the
University of South Flori
da IRB.


A variety of methodologies (OLS regression, group
-
mean differences, correlations, logistic
regression, time series analysis) will be employed to analyze the Minimum Data Set and link it with other data sets to study
the
following issues:





The development of wandering and other behavior problems in nursing home residents;



Trajectories of health and well
-
being among minority residents of long
-
term care facilities;



The impact of autonomy
-
enhancing interventions on function, de
pression, and overall costs to Medicaid and
Medicare for nursing home residents;



Assess the relationships between utilization of psychotropic medication and adverse events;



The provision of psychological services in nursing homes and health

care expenditures;



Compare pre
-

and post
-
SB1202 nursing facility resident characteristics;



Compare characteristics and trajectories of NF residents with dementia on quality of care and quality of life;



Improve the assessment of qua
lity of care for residents of LTC facilities through an analysis of falls
-
related injuries;



Evaluate end
-
of
-
life care outcomes among nursing home residents.



To accomplish the above objectives, we are initially requesting one year (calendar year 1
998) of MDS version 2.0, and to assess
the longitudinal and forecasting analyses, we plan requests for subsequent 5 years.



Each of the research studies summarized in this document will have important and profound consequences for state and
national polic
y
-
makers.


With CMS approval, findings from these studies will be widely disseminated through several means:





Internally published newspapers and reports by FPECA, TNH and FMHI





Formal reports



Scholarly publications in refereed jo
urnals



Posting of findings on the internet web pages of FPECA, TNH and FMHI



Both SAS and FMHI have been conducting research and influencing policy for many years and this new direction of inquiry,
using MDS version 2.0, will complement the ongoi
ng research and evaluation within both centers.

INTRODUCTION

Title:
Characteristics of Florida Nursing Home Residents



Objectives:

The aims of this project are to conduct a series of studies using Minimum Data Set (MDS) version 2.0, that address several
pressing policy issues facing leaders in the State of Florida and nationally.


All of these studies have been reviewed and funded
by Florida Medicaid, and all are either reviewed or exempted from review by the University of South Florida IRB.


A variety of

methodologies (OLS regression, Group
-
mean differences, Correlations, Logistic regression, Time Series Analysis) will be
employed to analyze MDS version 2.0 and link it with other data sets to:





The development of wandering and other behavior prob
lems in nursing home residents;



Trajectories of health and well
-
being among minority residents of long
-
term care facilities;



The impact of autonomy
-
enhancing interventions on function, depression, and overall costs to Medicaid and
Medicare
for nursing home residents;



Assess the relationships between utilization of psychotropic medication and adverse events;



The provision of psychological services in nursing homes and health care expenditures;



Compare pre
-

and post
-
SB1
202 nursing facility resident characteristics;



Compare characteristics and trajectories of NF residents with dementia on quality of care and quality of life;



Improve the assessment of quality of care for residents of LTC facilities through
an analysis of falls
-
related injuries;



Evaluate end
-
of
-
life care outcomes among nursing home residents.



To accomplish the above objectives, we are initially requesting one year (calendar year 1998) of MDS version 2.0, and to asse
ss
the longitudin
al and forecasting analyses, we plan to make additional requests for MDS version 2.0 annually to supplement this
initial data year with subsequent data years over a five
-
year period.


We are attempting to combine several studies under one
coordinated serie
s of data requests in order to more efficiently use the resources that we can allocate toward the studies
outlined, as well as more efficiently work with the Centers for Medicare & Medicaid Services (CMS) in the data acquisition
process.



Background:




Both the Florida School of Aging Studies (SAS) and the Florida Mental Health Institute (FMHI) have been conducting research
and influencing policy for many years and this new direction of inquiry, using MDS version 2.0, will complement the ongoi
ng
research and evaluation within both centers.


SAS and FMHI work collaboratively to assist Florida Medicaid, the Florida
Department of Elder Affairs, and the Florida Department of Children and Families to improve program monitoring and
evaluation of serv
ices offered to older Floridians and consumers of substance abuse and mental health services.




SAS and FMHI are constituent units of the University of South Florida. They share a commitment to assist Florida government
with the development of

actuarially sound payment rates for Medicaid services, conduct better program evaluations through the
development of outcome measures, and facilitate program monitoring through refined performance
-
based budgeting criteria.



Faculty from SAS have been act
ive in the development and evolution of the Teaching Nursing Home, an AHCA
-
funded effort
of gerontology researchers, clinicians and providers whose mission is “to improve and expand capacity of Florida's healthcare

system to respond to the medical, psychol
ogical, and social needs of the increasing population of frail older citizens.” The
request for data reflects the commitment of researchers at SAS,


TNH and AHCA to work together to share data results and
analyze quality improvement in Florida long
-
term ca
re settings.





SAS and FMHI receive contracts from the State of Florida for the analysis of Medicaid data and relevant administrative data
from state agencies.


These contracts include Federal matches from CMS, owing to the nature of the research, which
is intended
to improve quality of services and reduce the costs of providing services to Medicaid clients.



SAS contracts with Florida Medicaid also outlines several tasks that will benefit from Medicare data, including:





Probabilistic linkage of

Medicaid claims and eligibility information with Florida death certificates and
information on long
-
term care needs and services provided to older Floridians by Florida Medicaid, the Florida
Department of Elder Affairs, and the Florida Department of Child
ren and Families.



Develop risk
-
based capitation rates for long
-
term care.



Develop a comprehensive array of long
-
term care services.



Track outcome measures and performance
-
based budgeting measures across different long
-
term care prog
rams.



Uncover inappropriate care and cost
-
shifting.





FMHI contracts with Florida Medicaid specifically outlines several studies that will benefit from the inclusion of
Medicare data, including:





Examination of the interactio
n of Medicaid and Medicare for several cohorts (elders, SMI, etc.) with a focus on
the dual eligible population.



Examination of mental health issues in ALFs (particularly for those on OSS).



Access for minorities.



Co
-
morbidity.




Interaction/integration of Medicare, Medicaid, and Adult Protective Services.



In addition to contracts, SAS and FMHI are actively engaged in research supported by grants. Grant sources include the
National Institute of Mental Health, the Robert Woods
Johnson Foundation, and the University of South Florida AoA
(Administration on Aging), Hartford Foundation, and OSHA Grant.



Importance:



Access to MDS version 2.0 will have a direct and profound influence on the ability of SAS and FMHI to assist legisla
tors and
agency policymakers in the conduct of state policy and planned state initiatives.




Each of the research studies highlighted in this document will have important consequences for state policymakers.


Each study
will answer questions t
hat will help them ensure access to care for low
-
income and disabled Floridians that is of good quality
and is cost
-
effective.



RESEARCH QUESTIONS AND METHODS



Hypotheses/Study Issues:

All the proposed studies are initially baseline in nature.


Thus as
we explore and understand the data sets, and immerse
ourselves further into the integration of the multiple data sets that will be used in these studies, additional questions and

specific
hypotheses will evolve and emerge.


Subsequent analyses examining th
ese additional hypotheses resulting from initial analyses
will be proposed and carried out in future years.


The current initial questions that we intend to address in the several studies
proposed here are outlined below in the Study Design section.



Stud
y Design and Analytic Plan:

A description of the primary question(s), data sources and essential study design is provided in the table next page for each

proposed study component.


Please see Appendix A for more detail on each funded project.

Research Ques
tions and Methods



Study Title

Study Design and Methodology

Relevant Variables

Key
Personnel

1. The
development of
wandering and
other behavior
problems in
nursing home
residents

Description:

Uses the Minimum Data Set (MDS)
II.

Study Design:

1. Compare
“wanderers” and "nonwanderers" on
the initial assessment on the following variables:


presence short term memory problem, gender,
degree and types of functional impairment,
dementia (Alzheimer's disease versus other
dementia), presence of long term memory,

problems in communication, cognition (speech,
lethargy, cognitive ability
-
daily), presence of
negative affect/mood, behavior problems such as
physical and verbal abusiveness to others, general
medical problems, medications, and mode of
locomotion (self
-
pr
opelling wheelchair users
versus nonusers/walking residents)

2. Compare individuals exhibiting no wandering
on their initial MDS assessment at any time with
individuals who are wanderers and those who
develop wandering sometime after admission



Subjects:


The study population will be all nursing home
residents age 65+ who entered a nursing home in
the State of Florida during the period for which
the MDS are obtained (1998) and any subsequent
years of MDS that are obtained at a later date.


Data will be li
nked to Medicare and Medicaid
cost data for a three year period including and
following admission to the nursing home.



Analyses:

This will involve basic statistics to determine
strength of relationship and prediction
(correlation, multiple regression), g
roup mean
differences (analysis of variance, co
-
variance,
MANOVA), and possibly, hierarchical linear
modeling.

Relevant data

includes: demographic
characteristics, presence short term memory
problem, degree and types of functional
impairment, dementia (Alz
heimer's disease
versus other dementia), presence of long term
memory, problems in communication, cognition
(speech, lethargy, cognitive ability
-
daily),
presence of negative affect/mood, behavior
problems such as physical and verbal
abusiveness to others,
general medical
problems, medications, and mode of locomotion
(self
-
propelling wheelchair users versus
nonusers/walking residents).



Relevant variables includes
:

AA2
-
AA8a,b

AB1
-
AB3, AB7
-
8, AC1a
-
y

A3
-
A8

B1
-
B6

C1
-
C7

D1
-
D3

E1a
-
p; E2
-
3; E4a
-
e; E5

F1
-
F3

G1a
-
j
(A)(B); G2(Bathing)
-
G6; G8

H1a
-
b (Continence); H3

I1a
-
rr(Diseases);I2a
-
m(Infections); I3a
-
e; ICD
-
9

J1a
-
p; J2a&b; J3a
-
j; J4a&b; J5

M1; M2a
-
b; M6a
-
g

N1a
-
d;N2; N3a
-
e; N4a
-
m; N5

O1; O4a
-
e

P1b(a)
-
(e); P2a
-
f; P3a
-
k; P4a
-
e; P5
-
P9

Q1a
-
c; Q2

R1a
-
c

T1a
-
d; T2a
-
e

V1
-
18.
-

a&b



Schonfeld,

Molinari,

Brown

2.Trajectories
of health and
well
-
being in
Minority
Residents of
Long Term
Care Facilities

Description:

Uses the Minimum Data Set (MDS) II.

Study Design:

This is a broad
-
gauged exploratory study that wil
l assess
initial and subsequent health and well being among long
term care residents from diverse ethnic/minority
backgrounds.



Subjects:

The proposed project will examine the records of all
African American, Asian American, and Hispanic
American resident
s for whom MDS II data were collected
during the period extending from 1999 through 2003.


These records will be contrasted with a random sample of
5000 Anglo American.



Analyses:

Because so little is known concerning minority residents of
long term car
e facilities, save that relatively few reside in
LTC facilities for any length of time, the first set of
analyses will simply examine characteristics of each of the
three minority populations of interest.


Basic distributions
will be examined for evidence
of distinct patterns or
profiles. Evidence for differences among subgroups within
each broad category of minority status will also be
evaluated (e.g., according to nation of birth, length of stay
in the United States, education and economic status, health
status at admission, and source of admission). Of particular
interest will be information on mental health and overall
quality of life.


Second, comparisons will be made across
the minority groups and with the Anglo American sample.
These comparisons will
be conducted not only for the
entire groups but also according to the critical parameters
noted above. Third, trajectories of change will be identified
and assessed in terms of implications for future
interventions.

Relevant data

includes:


demographic
cha
racteristics at entry, diagnoses,
physical and mental health, medications,
ADLS/IADLs, pain management,
cognition, RAP summary form, discharge
and reentry tracking data, cost of care,
and length of stay. Data points to include
one full 12 month period (or
to
termination of long term care, defined as
departure due to death or for a period of
three months or more) subsequent to
admission except for data on (1) cost of
care, for which a two year period is
requested, and (2) overall length of stay,
for which al
l available data will be
requested. A complete listing of variables
is appended.



Relevant variables include:

A10A through A10h

A3A through A9G

AA1A through AB9

AC1C through AC1Y





B1 through B6

C1 through C7

D1 through D3

E1A t
hrough E5

F1A through FAC_INT_ID

G1AA_SELF
-
BED Through G9

H1A Through H4

I1AA Through MDS_SUBM_SEQ_NBR

N1A through R4_DISCHAGE_DT

T1AA through T3STATE

U01AA Through U18RA

VA01A through VA18B

Chiriboga
,

Jang,

Watson,

Molinari,

Brown,

Schonfeld
,

Branch,


Kearns,

Schneider

3. The impact of
autonomy
-
enhancing
interventions on function,
depression, and overall
costs to Medicaid and
Medicare for nursing
home residents.

Description:

Uses the Minimum
Data Set (MDS) II.

Research Questions:

1. What is the rela
tionship between
personal control, as expressed by
sense of initiative to changes in
function, depression, and overall
costs to Medicare and Medicaid?



2. What is the relationship between
rehabilitation support, as expressed
by use of appliances and progr
ams
for continence, physical therapies,
emotional therapies, and special
treatments and procedures, to
changes in function, depression, and
overall costs to Medicare and
Medicaid?



3. Controlling for predisposing,
enabling, need, personal control, and
reh
abilitation support, changes in
function, and depression, what
factors in these models most
contribute to overall costs of
Medicare and Medicaid?



General Design:

A cohort of nursing
home residents who entered the
nursing home in 1998 or 1999 will
be foll
owed for three years (1999
-
2002) from time of entry to
understand the importance of
personal control and rehabilitation in
improving function, reducing
depression, and decreasing overall
costs to Medicare and Medicaid.



Subjects:

Population of nursing ho
me
residents age 65+ who entered a
nursing home in 1998 or 1999 in the
State of Florida.


Data will be linked
to Medicare and Medicaid cost data.


Records will be linked using
identifier information and then de
-
identified for analyses.



Analyses:


A serie
s of hierarchical linear
models using a pooled cross
-
sectional time series design and
correcting for the autoregressive
nature of health and disability will
be used to test a health care
utilization and satisfaction model
(Andersen and Newman, 1973;
Anders
en, 1995).


Each set of
variables (predisposing, enabling,
need, psychological and
rehabilitation mediators) will be
added in steps to predict three
dependent variables: Change in
Function, Change in Depression, and
Costs to Medicare and Medicaid.


The fin
al dependent variable will be
based on linked data not including in
the MDS.

Relevant data

includes; demographic characteristics,
Medical Record Number, current payment sources,
reason for assessment, comatose, memory, cognitive
skills for daily decision m
aking, health conditions
(hearing, vision, etc.), communication making self
understood, depression, sense of initiative, unsettled
relationships, past roles, physical functioning and
structural problems, bathing, balance, range of motion,
modes of locomoti
on, modes of transfer, ADL,
continence, appliances and programs, diseases, oral
problems, oral status and disease prevention, time
awake, average time in activities, preferred activity
settings, general activity preferences, prefers changes
in daily routin
e, number of medications, therapies,
intervention programs for mood, behavior, cognitive
loss, nursing home rehabilitation/ restorative care,
devices and restraints, discharge potential, overall
change in care needs, participation in assessment,
special tr
eatments and procedures, and walking when
most self
-
sufficient.



Relevant variables include:

AA2
-
AA8a,b

AB1
-
AB3; AB7; AB8

AC1a
-
y

A3
-
A8

B1
-
B2(memory); B4

C1
-
C2; C4

D1

E1a
-
p(depression);F1a
-
g; F2a
-
h; F3a
-
d



G1a
-
j(A)(B); G2(Bathing)G3
-
G6; G8

H1a
-
b; H3

I1a
-
rr(diseases)

K1a
-
d

L1a
-
g

N1a
-
d; N2; N3a
-
e; N4a
-
m; N5

O1

P1b(a)
-
(e); P2a
-
f; P3a
-
k

P4a
-
e

Q1a
-
c; Q2

R1a
-
c

T1a
-
d; T2a
-
e

Salmon,

Mitchell

4. Relationships
between utilization
of psychotropic
medication and
adverse events

Description:

Uses the Minimum Data Set
(MDS) II.

Study Design and Specific Aim:

The objective of this analysis is to describe the characteristics and
trajectories of nursing home (NF) residents with and without dementia
at psychotropic medication prescribed over the course of their NF stay.


My

aim is to evaluate utilization of psychotropic medication, and the
relationship between psychotropic medication use, changes in
functional and cognitive status over time, and adverse outcomes such
as falls, fractures, use of physical restraints, and decre
ased time spent
in social activity.


I would like to use data from twelve consecutive
quarters.


Additionally, I would like to compare outcomes for residents
in special care units with those not receiving special care.



Relevant data

includes;
demographic

characteristics,
depression, falls, need for
restraints, ADL, behavior
symptoms, cognitions, and
psychotropic medication
use.



Relevant variables
include:

AA2
-
AA8a,b; AB1, 2,
AB6, 7, 9;

AC1K, AC1
m
-
l, AC1 s
-
x;

A5,
A7 a
-
j
, A8, A9 a
-
g, A10
a
-
j;

B1, B2 a
-
b
, B3 a
-
e, B4, B5
a
-
f, B6;

E1GP1

a
-
p, E 2
-
5;

F1 b
-
d, F a
-
h, Fa
-
d; G1
-
G9;

I
1 m
, q, t, u, y, I1 bb, ee,
ff, gg;

K5 a; J4 a
-
e, Ji;

N2, N5, N6;

O1, O4 a
-
e

Becker

5. The provision
of psychological
services in nursing
homes and health
care expenditures

Descrip
tion:

Uses the Minimum Data Set (MDS) II.

Study Design:

This is an exploratory study designed to answer a basic
question: Does the provision of psychological services in
Florida nursing homes reduce overall health care
utilization and expenditures? The res
earch plan has 2
related aspects. One is to identify those nursing homes that
provide mental health services and compare them with a
'matched' group of nursing homes that do not provide
mental health services. Two, we will identify residents
who have eithe
r mental health diagnoses and/or those who
score high on certain mental health variables (i.e.,
mentally impaired). We will then compare health service
utilization and follow
-
up mental health indices between
those mentally impaired who receive mental healt
h
treatment and those who do not.



Subjects
: The prabalistic sample will be drawn from all the
nursing homes in the State of Florida who admitted
residents during the period for which the MDS data was
obtained (1998) and any subsequent years that the MDS
is
obtained. Data will be linked to Medicare and Medicaid
cost.



Analyses
: Probalistic matching for nursing homes who do
and do not offer mental health services will be conducted
using the Minimum Data Set (MDS) II.

Relevant data

includes; demographic
cha
racteristics (age, gender, LOS),
health service utilization (# of
prescriptions for psychotropic meds (it
would be good if we could track the # of
new prescriptions after visits by mh
professionals); # of psychiatric
hospitalizations; # of visits by mh
pro
fessionals; # of psychiatric sentinel
events), and mental health related
variables (diagnoses; depression score,
mental status score).



Relevant variables
:

A4B; A5 AA1A;AA1B;AA1C;AA1D;
AA2;AA3;AB2; AB7;AB9
(Demographics)

AB9 (Prior MH hx);

B2
-
B6 (Cogniti
on)

E1; E2; E3; (Mood)

E4; E5 (Beh Probs)

J1E; J1I

I1BB; I1CC; I1DD; I1EE; I1FF; I1GG
(Psych symptoms); 1Q; 1R; 1T; 1U; 1Y

P1BEB (Psychosocial Therapy)***

P2A (Behavior); P2B (Eval); P2C
(Group); P2D;2E; 2F (Intervention
Program)

U (Medication)

VA07A (Well
-
being); VA08A (Mood);
VA09A (Behavior)

Molinari

6. Pre and Post
SB1202 Nursing
Facility Resident
Characteristics

Description:

Uses the Minimum Data Set (MDS) II.

Study Design:

Quasi
-
experimental design using interrupted time series analysis of
secondary
data to determine changes in resident clinical condition trends
after introduction of SB1202.



Subjects:

Entire Florida nursing facility resident population and the specific
nursing facility resident subpopulation of Hillsborough County, Florida
from 1999

through 2003



Research Questions:

1. What are the demographic and clinical characteristics of residents
involved in nursing facility litigation?



2. Will successive individual resident assessments reveal deterioration in
cognitive, psychosocial, mood, p
hysical functioning, disease diagnoses,
health, continence, oral/nutritional status, skin conditions or increased
pain levels prior to lawsuit filings?

3. Will specific clinical conditions decrease in frequency and severity
after SB1202’s implementation?



Analyses:

The PIs are very much aware of the stigma of lawsuits against nursing
homes and the level of concern by providers and regulators about
protecting the anonymity of nursing homes and their residents.


Therefore, the following procedures will be
followed to assure data
confidentiality: 1) Lawsuit data will be de
-
identified except for the
Social Security number.


The following fields will be removed: court ID,
plaintiff name, birth year, address, and defendant and attorney names. 2)
MDS data will b
e de
-
identified except for Social Security number. The
following fields will not be included in the data extract for analysis:
nursing home identifier, other resident identifiers, year of birth. County
will be retained. 3) The two de
-
identified databases w
ill be linked
through the Social Security number of the resident by the Director of the
State Data Center on Aging who does not have access to the lawsuit
data. 4) The merged database will replace Social Security numbers with
an encrypted code for analysis

purposes. In addition to these steps, the
PIs will submit a separate human subjects application to the USF
Institutional Review Board to assure that all procedures will protect the
confidentiality of all residents and facilities.



Interrupted time series

analysis will be used, observing frequency and
severity of resident clinical characteristics for two
-
year periods before
and after passage of SB1202.


Analysis of variance will be used to
discover main and interaction effects of the following MDS variable
s on
filed lawsuits in pre and post SB1202 two
-
year periods

Relevant data

includes;
demographic
characteristics, clinical
characteristics, cognitive
skills, psychosocial
status, mood, physical
functioning, disease
diagnoes, health,
continence,
oral/nutriti
onal status,
skin conditions, and pain
levels.



Relevant variables
include:

AA2
-
AA4; AB7
(demographic variables)

A5; A9; A10

B1(comatose);
B2(memory); B4; B6

C1(hearing);
C2(communication)

D1(vision)

E1a
-
p (depression); E3;
E5

F1a
-
g; F2a
-
h; F3a
-
d



G1a
-
j(A)(B);
G2(bathing)
-
G6; G8
-
G9

H1a
-
b(continence); H3
-
H4.



I1a
-
rr(diseases); I2a
-
m(infections)

J1a
-
p; J2(a)(b); J3a
-
j;
J4a
-
e; J5a
-
d

K1a
-
d; K3(a)(b); K4a
-
d;
K5a
-
I; K6(a)(b)



M1a
-
d; M2a
-
b; M3;
M4a
-
h; M5a
-
j; M6a
-
g

N1a
-
d; N2(average t
ime
in activities)

O1(number of
medications)
-
O4.a
-
e

P1a(a)
-
(s); P1b(a)
-
(e);
P2a
-
f; P3a
-
k.; P4a
-
e; P5
-
P6(Er Visit(s)); P7
-
P9

Q1a
-
c; Q2

R1a
-
c (participation in
assessment)

T1a
-
d; T2a
-
e

Hedgecock
,

Johnson,

Mitchell

7. Characteristics and
trajectories of NF
r
esidents with
dementia on quality
of care and quality of
life

Description:

Uses the
Minimum Data Set (MDS) II
from twelve consecutive
quarters.

Study Design:

1. Describe the characteristics
and trajectories of nursing
home (NF) residents with and
without d
ementia at initial
admission to a NF and over the
course of their stay.

2. Evaluate changes in
functional and cognitive status
over time and the relationship
between changes in cognitive
and functional status and
adverse outcomes such as falls,
fractures,

use of physical
restraints, and decreased time
spent in social activity.

3. Compare outcome for
residents in special care units
with those not receiving
special care.



Analyses:

Using the MDS variables that
identify dementia on the initial
assessment o
r identify dementia
during subsequent assessments
we will compare persons with
dementia on the following
variables.

Relevant data includes demographic characteristics, ADL,
comatose, memory, cognition, mood
-
verbal, sleep, sadness,
interest level, mood pers
istence, change in mood, behavioral
symptoms, change in behavior, psychosocial structured activities,
unsettled relationships, past roles, health conditions (hip fracture,
Alzheimer’s, CVA, dementia, Parkinsons, TIA, depression, manic
depression, and Schiz
ophrenia), falls within the last 30 days,
falls, hallucinations, time spent in social activities, hospital stays
in last 90 days, ER stays in last 90 days, number of medications,
types of medication, AD special care unit, and psychical
restraints.

Hyer

Bec
ker

8. Improving the
Assessment of
Quality of Care for
Residents of LTC
Facilities through
an Analysis of
Falls
-
Related
Injuries in LTC
Facilities (in
Florida)

Description:

Uses the
Minimum Data Set (MDS) II
from twelve consecutive
quarters.

Study design:

1.Facility ranking or grouping
for quality of care for its long
-
term residents based on
meultiple measures of quality
from State
-
published
measures (ie Nursing home
guide, Watch list and gold
seal facilities, deficiencies)
available for Florida LTC
facili
ties.

Compare the above rankings
or groupings (univariate and
other types of analyses) with
facility rankings based on the
following:



1) Facility rank based on
prevalence of falls
-
related
injury (specifically, fractures
occurring in LTC residents
while

in the facility; other
adverse events)



2) prevalence of selected
risk factors (specified in the
falls rap)



3) fall
-
related injury
prevalence adjusted based on
the mean of a combination of
falls and falls
-
related injury


risk factor prevalences



Ana
lyses:

Since this is an exploratory
study, it is likely that initially
we would use very basic
techniques such as univariate
analysis and non
-
parametric
rank tests; we will also
consider using age and
disease burden to further
adjust the injury risk at a
f
acility. However, as detailed
in our proposal outline, we
would pursue a positive result
with the univariate analysis
by a more complex series of
analyses.



Relevant data include demographic characteristics and
followings

AA2 (gender); AA3

(age
-
set for th
e first quarter of 8 quarters);
sex (AA4); race/ethnicity (AA5); AA8 a,b(reason for
admission, special assessment code); AB1 (date of entry);


AB2(admitted from); AB6( occupation); AB7(education); AB9
(mental health history); AC1 m
-
l (ADL patterns); G1a, a
A, aB,
(ADL bed); G1b, aA, aB, (ADL transfer); G1c, aA, aB, (ADL
walk in room)G1d, aA, aB, (ADL walk in corridor); G1e, aA,
aB, (ADL locomotion on unit); G1f, aA, aB, (ADL locomotion
off unit); G1g, aA, aB, (ADL

dress); G1h, aA, aB, (ADL eat);
G1i, aA, aB,

(ADL toilet); G1j, aA, aB, (ADL personal
hygiene); G2, (bathing); G2 A,B, (bathing);


G3 a,b, (balance);
G4 a,f, (orthopedic/Nagi items); G5 a
-
3, (modes of
locomotion) ; (crutch, walker, or cane), G6 a
-
f, (transfer aids);
G7, (task segmentation); G8 a
-
e,
(rehab potential) G9, (change
in ADLs);


I1o, osteoporosis,


I1m, fracture, I1 m,

(
hip
fracture); I1 q, (Alzheimer's); I1 t, (CVA); I1 u (other
dementia); I1 y (Parkinson’s);


I1 bb (TIA);


I1 ee (depression);
I1 ff (manic depression); I1 gg (Schizophrenia
);


K5 a, (Falls
within the last 30 days);


J4 a
-
e, (falls); J1f, dizziness;


O1, #
meds; O4 a
-
e, (type of meds);


P1
-
an, (AD special care unit);
P4c, use of trunk restraints



Roos,

Hyer,

Polivka
-
West,
Polivka,

Lai

9. End
-
of
-
life
care outcomes
among nu
rsing
home residents

Description:

Uses the Minimum Data Set (MDS)
II.

Study Design:

Nursing home residents who died
between 1998 and 2002 and had at least two MDS
assessments available to identify and understand
predictors of end
-
of
-
life care outcomes.





Subjects:



Population of nursing home residents age 65+
who deceased between 1998 and 2002 in the State
of Florida



Research Questions:

1.What is the relationship between patient
characteristics (demographic, functional and
clinical characteristics, an
d cognitive status), and
patient preference [presence of do
-
not
-
resuscitate
order (DNRO) and do
-
not
-
hospitalize order
(DNHO)] and three end
-
of
-
life care outcomes
(hospitalization, pain management and place of
death)?

2.What is the role of hospice care in t
hree end
-
of
-
life care outcomes controlling for patient related
characteristics?

3.What is the overall cost of end
-
of
-
life care of
nursing home residents who use and do not use
hospice care after controlling for patient related
characteristics?





Analy
ses:

A series of logistic regression models
will be conducted to test three major research
questions.

Relevant data
includes; demographic
characteristics, clinical, functional, cognitive,
nutritional, patient preference (DNRO, DNHO),
family characteristic
s, pain management (reported
frequency and intensity of pain), number of
hospitalization, experience of hospital death, patient
related variables, hospice use, and overall end
-
of
-
life cost in the last year of life



Relevant variables include;

AA2
-
AA8a,b

A
B1; AB2; AB5; AB7; AB8; AC1a
-
y

A3
-
A10(advanced directives)

B1
-
B2; B4; B6

C1
-
C2; C4; C6

D1(vision)

E1a
-
p(depression)

G1a
-
j(A)(B); G2(bathing)
-
G6; G8
-
G9

H1a
-
b(continence)

I1a
-
rr(diseases)

J1 a
-
e; J2a
-
b; J3a
-
j; J5a
-
d

K1a
-
d; K3a
-
b; K4a
-
d

O1(number of medicatio
ns)

P1; P5
-
P6(ER visits)

Q2(overall change in care needs)

R1a
-
c(participation in assessment)

Kwak,

Mitchell


77



Database Management:

Data will be stored on a secure server in the Sate Data Center on Aging. The secure server is not part of any domain at

the
University of South Florida and all data access is directly from the terminal. No network access to data is permitted. Extrac
ts
will be prepared for each team and made available to the team leader on physical media (
e.g.
, CD
-
ROM, DVD).



EVALUATION AN
D ANALYSIS PLAN

Analysis Plan and Analytic Methods:

Summaries of specific analysis plans and analytic methods for each of the studies are summarized in the table contained in th
e
STUDY DESIGN section.



INITIAL YEAR WORK PLAN

Activity / Tasks

Resources

/P
ersonnel

Time
Frame

Outcome


Receive data


Raw data organization


Initial fidelity checks performed

Mitchell,

Data Analyst

Months 1
-
5

Data organized and ready for integration
and analysis


Data extracts for each sub
-
study are created and organized f
or analysis


Analyses for each sub
-
study are assessed

Entire Project
Team

Months 6
-
10

Initial analyses completed for each sub
-
study




Reports are written and submitted to contracting agencies


Scholarly articles are written and submitted to refereed jo
urnals

Entire Project
Team

Months 10
-
12

Reports and articles submitted and second
year data request provided to CMS


Year two data request is constructed and submitted to CMS




Continue use of data in forecasting studies, other longitudinal studies and
future proposed studies (appropriate re
-
use agreements will be executed)

Entire Project
Team

Months 13
-
60

Additional completed analyses for CMS
approved studies



The work plan is divided three broad phases.


During the first stage (months 1
-
3), the State

Data Center on Aging will receive
the data, compile it on their system and perform the appropriate fidelity checks to ensure data quality.


It will be structured so
that extracting the needed data sets for each of the sub
-
studies is facilitated.


During t
he first stage (months 3
-
10), the actual
data extracts will be constructed for the year one analyses and distributed to the various primary analytic personnel.


The
analyses for the sub
-
studies will also be assessed during this phase.


During the first sta
ge (months 10
-
12), the required reports
will be written for the funding agencies, and work will begin on drafting scholarly articles based upon the first year analys
es.

QUALIFICATIONS OF KEY STAFF



Marion Becker, Ph.D.
is a tenured Professor in the Depart
ment of Mental Health Law and Policy at the Louis de la Parte
Florida Mental Health Institute, a college located at the University of South Florida (USF). She also holds appointments in t
he
USF School of Social Work and the Colleges of Public Health and Nu
rsing. Dr. Becker is a psychiatric nurse with a PhD in
Social Welfare from the University of Wisconsin
-
Madison. Her research focuses on the problems of providing high quality
cost
-
effective behavioral health services and quality of life outcomes for vulner
able populations. Dr. Becker is the developer of
the Wisconsin Quality of Life Index (W
-
QLI), a core development in mental health research. In 1997 she received the National
Alliance for Mental Health (NAMI) Research Award for her quality of life outcomes
research. Dr Becker has served as
Principle Investigator and Co
-
PI on numerous outcome studies. Most recently she served as Co
-
PI, and Lead Evaluator for the
Triad Women’s Project, a multi
-
million dollar, competitive federal grant designed to create and ev
aluate specialized
interventions for women with alcohol, drug abuse and mental disorders who have histories of interpersonal violence. Other
recent research includes studies examining the predictors of successful permanency planning, ways to improve servic
es to
families involved in the child welfare system, and the economic value of antipsychotic medications.


Along with that research
she continues her focus on quality of life outcomes across the life span and linking outcomes research to clinical practice.



Lisa Brown, Ph.D
. is an Assistant Professor in the Department of Aging and Mental Health at the Florida Mental Health
Institute. She is a graduate of the Pacific Graduate School of Psychology, Palo Alto, California. Dr. Brown completed her
internship an
d postdoctoral fellowship at the James A. Haley Veteran's Hospital, Tampa, Florida, where she specialized in
neuropsychology and geropsychology. Dr. Brown's research interests include indirect self
-
destructive behaviors, suicide
assessment and intervention
s for older adults, the effects of terrorism among older adults, mood and anxiety disorders,
dementia, and long
-
term care.



David A. Chiriboga
,
Ph.D
. is a professor in the Department of Aging and Mental Health, Florida Mental Health Institute,
University
of South Florida. His prior positions were at the University of Texas Medical Branch, and the University of
California San Francisco, and he holds a 1972 PhD from the University of Chicago. His research has three overlapping and
ongoing themes. The first,
beginning with longitudinal data collected from 1968 through 1997, involves study of the influence
of stress exposure for mental health over periods of 20 or more years. The second, beginning in 1976, involves the longitudin
al
study of acculturation and di
fferential mental health issues in minority and majority populations, especially Mexican American
elders. The third, beginning in 1995, involves the use of multimedia and distance technologies for health care and training.

Yuri Jang, Ph.D
. is Assistant Pr
ofessor in the Department of Aging and Mental Health (AMH) at the Louis de la Parte Florida
Mental Health Institute (FMHI). Dr. Jang received her doctoral degree in Aging Studies from the University of South Florida i
n
2001. After her postdoctoral training

at the USF Institute on Aging and University of Georgia Gerontology Center, she joined
the Department of Aging and Mental Health in 2003. Dr. Jang was awarded Minority Fellowships from the Gerontological
Society of America (GSA) and the American Psycholog
ical Association (APA). Dr. Jang's areas of interest include positive
adaptation in aging, personality, stress and coping, mental health, minority aging, and cross
-
cultural research. Her recent
projects focus on 1) mental health among older residents in As
sisted Living Facilities and 2) service utilization among ethnic
minority elders.



Deborah Hedgecock

is a Ph.D. candidate in the University of South Florida’s School of Aging Studies. She earned her B.A. in
psychology in 1999 from USF as well. Her interes
ts and research are in the areas of quality of life and policy in long
-
term care,
including nursing home litigation, insurance coverage and costs and their effects on individuals, nursing homes, and public
-
policy making. She has worked as a graduate resear
ch assistant for the Florida Policy Exchange Center on Aging from 2000
through 2003. She served as a staff researcher on Florida’s Task Force on the Availability and Affordability of Long
-
Term Care
in 2000, co
-
principal investigator on a statewide survey t
hat examined the initial impact of Senate Bill 1202 on the lawsuit and
liability insurance experience of Florida nursing homes from January through October 5, 2001, researcher on the Florida Polic
y
Exchange Center on Aging’s 2003 Assisted Living Research S
tudy, and co
-
principle investigator and author of The Nursing
Home Liability Insurance Crisis: A Case Study of Georgia conducted by Medstat, Boston, MA (in review).



Kathryn Hyer, DrPA, MPP
. is the Director of the University of South Florida Training Acad
emy on Aging and a faculty
member in the Departments of Gerontology and Health Policy and Management. Dr. Hyer is the Co
-
Principal Investigator of
the Resource Center for the Geriatric Interdisciplinary Team Training Program at New York University, a $12 m
illion initiative
of the John A. Hartford Foundation of New York City. In 1995, Dr. Hyer served as an Advisor for Assistant Secretary of
Planning and Evaluation for Department of Health and Human Services project for “Innovative Interdisciplinary Education

and
Training Programs for Professionals Caring for Persons with Disabilities.” From 1990
-
1994, Dr. Hyer was the Vice President of
Business Development and Research at the Visiting Nurse Service of New York, a $400 million dollar home health care agency
wh
ere she was responsible for developing the $1 million grant
-
funded program to create the Community Nursing Organization
capitation program funded by DHHS. Dr. Hyer holds a Masters degree in Public Policy from the Kennedy School of
Government at Harvard Uni
versity and doctorate in Public Administration from Arizona State University. She has contributed
more than 25 publications to the literature in the field of geriatrics and public policy and served as presenter/panelist at
numerous conferences throughout t
he country. Dr. Hyer received federal funding for her dissertation,
HMOs and the elderly:
Adjusting Medicare’s AAPCC.



Christopher E. Johnson
, Ph.D. is a leader in health services research related to nursing homes and rehabilitation quality of
care outcom
es.


He recently completed work examining quality indicators and the impact of litigation on nursing homes in
Florida that was crucial in the passing of a long
-
term care reform bill in Florida that addressed concerns about decreasing
quality and the impact

of litigation on the financial viability of the long
-
term care system.


Dr. Johnson and colleagues at the
Institute for Child Health Policy at the University of Florida have studied how organizational structure impacts the provisio
n of
services to childre
n under state insurance programs.


One paper from his stream of research about medical group practices won
two national best paper awards, including one from the prestigious Academy for Health Services Research and Policy (now
known as AcademyHealth) durin
g the 2001 Annual Meeting in Atlanta, Georgia.


Dr. Johnson is currently studying the role that
social networks play in determining how nursing staff conceptualizes adverse events in Department of Veterans Affairs nursing

homes.


He most recently won a pre
stigious Merit Review Entry
-
level Program career development award from the VA’s Health
Services Research & Development division to study nursing home impacts on quality of care for stroke rehabilitation patients.



William Kearns, Ph.D
. received his Ph.D.

in psychology from the University of South Florida in 1989. He has served on the
faculty of the Louis de la Parte Florida Mental Health Institute from 1990 to the present and was the director of the Compute
r
Support Center from 1992 to 2003. He joined the

Department of Aging and Mental Health in October, 2003 as an Associate in
Research. His interests include the long term adverse effects of psychological stress on senior citizens psychological and
physical health, and using automation to facilitate improv
ed access to mental health services and educational materials. He
serves as USF's executive liaison to the Internet2 Project, a consortium of over 200 Carnegie Research I institutions nationw
ide
developing enhanced network services supporting research and
education.



Jung Kwak
, M.S.W. is a doctoral student in the Aging Studies program at the University of South Florida. Before entering this
program, she earned a B.S. in business administration, double majoring in finance and marketing, and an M.S.W. from t
he
University of South Carolina. Her current research interest is in cultural diversity and end of life decision
-
making among
ethnically diverse groups of older people and their families. Her work experience includes working in both clinical and polic
y
set
tings. Currently she is the principal investigator on pilot project examining knowledge and preferences for end
-
of
-
life care
and hospice among Korean American elders and their caregivers
.

Glenn E. Mitchell II,

PhD.

Received his Ph.D. from the University of

Iowa. He was a member of the faculty at Florida State
University and currently serves as a faculty member and administrator at the University of South Florida, where he directs th
e
State Data Center on Aging. Dr. Mitchell has numerous peer
-
reviewed public
ations in health policy, gerontology, biostatistics,
and political science. He has received numerous grants and awards. Dr. Mitchell has been the principle investigator on many
state contracts, including three current contracts with Florida Medicaid and Fl
orida Department of Elder Affairs.



Victor Molinari,

PhD, ABPP (Clin)

is a Professor in the Dept of Aging and Mental Health (AMH). He is the past national
coordinator for the Psychologists in Long Term Care (PLTC) and the president of the American Psychol
ogical Association's
Division 12, Section 2 (Clinical Geropsychology). He is a fellow of APA 's Division 12 (Clinical Psychology) and the
Behavioral and Social Sciences section of The Gerontological Society of America. He has been appointed as AMH's
repres
entative to the Florida Mental Health Institute's APA
-
approved psychology internship program, and is the major preceptor
for the joint USF/Tampa VA geropsychology post
-
doctoral fellowship. His major research interests include mental health
outcomes in long

term care sites, reminiscence therapy, and personality disorder in older adults.



Larry Polivka, Ph.D
. has served as Director of the Florida Policy Exchange Center on Aging, which is part of the School of
Aging Studies at the University of South Florida,

since September of 1992.


Since August 2003, he has served as Associate
Director of the USF School of Aging Studies.


Dr.

Polivka worked at the State of Florida’s Health and Rehabilitative Services
as Assistant Secretary for Aging and Adult Services from
August 1989 through September 1992 and as Policy Coordinator for
Health and Human Services, Office of Planning and Budgeting, Executive Office of the Governor from 1986 through August
1989.


Dr. Polivka’s primary research interests are in long
-
term care, h
ousing, ethics and politics of care,
globalization/population aging, cultures of aging, and the arts/humanities and aging.



Bernard A. Roos, M.D.
, the University of Miami School of Medicine director for geriatrics programs, is professor of medicine
and ne
urology, director of the Geriatric Research, Education, and Clinical Center (GRECC), based at the VA Medical Center,
and chief academic officer at the Miami Jewish Home and Hospital for the Aged, where he directs the Stein Gerontological
Institute. He head
s the Division of Gerontology and Geriatric Medicine of the Department of Medicine. In 1993 he was named
recipient of one of the first endowed chairs in the country for research in aging, the University of Miami's Chester Cassel C
hair
for Research in Geron
tology. Since 1999 he has been the founding director of the State of Florida’s Teaching Nursing Home
Program, based at the Stein Gerontological Institute.


His MD from the University of Chicago School of Medicine. He took
residency training in internal med
icine at New York's Mt. Sinai Medical Center and a fellowship at the University of California,
San Diego. He held academic appointments at UC San Diego and Case Western Reserve University and came to Miami in 1989
from the University of Washington in Seatt
le.



Nancy Ross, Ph.D.
has over 30 years of experience in program evaluation and health services research primarily with the State
of Florida.


For two years, she was employed by the U.S. General Accounting Agency.


She received her doctorate in Political

Science from Florida State University in 1986.


In addition to research, she was instrumental in obtaining numerous grants
including a $1.7 million Robert Wood Johnson Foundation grant to implement health care reform in Florida. She has completed
over 50
internal research reports, and published articles on the following subjects: quality of care related to health plan
performance, the tracer approach to quality of care in assisted living facilities, attributes needed for purchasing coalition
s,
effects of l
ow birth weight on need for special education services, and the relationship of evaluation and research to policy.

In her current position, she has been responsible for contracting for independent evaluations of Agency programs, designing
Florida’s perform
ance based budgeting outcome measures, preparing federal waiver requests such as securing funding for
Healthy Start Coalitions and family planning, and completing the federally required Title XXI evaluation and annual report. A
t
present she has a federal P
ayment Accuracy Grant to test methodologies for assessing the medical necessity and
appropriateness of payments made on behalf of Medicaid recipients.


She serves on several national committees related to
measuring and evaluating health care provided to ch
ildren.



Jennifer R. Salmon, Ph.D.

is an assistant professor in the School of Aging Studies and director of the Center for Housing and
Long
-
Term Care at the Florida Policy Exchange Center on Aging at the University of South Florida.


Dr. Salmon’s research

has
focused on issues of personal control and personal meaning in long
-
term care settings including nursing homes, assisted living
facilities, and in
-
home care.


She has identified predictors of receiving care in these three settings and, with colleagues,

has
identified outcomes of Medicaid programs that support clients in these settings.


She teaches Long
-
Term Care Administration
and Housing for the Elderly at USF.


Dr. Salmon is experienced in hierarchical statistical models, including hierarchical
multi
ple linear regression and hierarchical multinomial logistic regression.



Myra Schneider, Ph.D
. is Visiting Assistant in Research in the Department of Aging and Mental Health at the Louis de la Parte
Florida Mental Health Institute. Dr. Schneider received
her doctorate in Public Health, having previously obtained a master’s
degree in clinical social work. She has been involved in departmental research including the integration of primary care
medicine with behavioral health care for older adults, and a stud
y of older African American mental health care preferences.
She is a member of the Institute on Aging, and of the Prevention Science and Methodology Group. Current research interests
include psychosocial aspects of mental health/health in older adults; men
tal health issues in minority older adults; stress, aging,
and chronic disease; behavioral health; integration of primary care and behavioral health care; and psychosocial epidemiology
.

Lawrence Schonfeld, Ph.D.

is a Professor in the Department of Aging an
d Mental Health at the Louis de la Parte Florida
Mental Health Institute, University of South Florida.


He previously served as Department Chair (1989
-
1992), Assistant Chair
(1987
-
1989), and director of day treatment programs for older adults with substanc
e abuse problems or mental illness.


Dr.
Schonfeld has published over 40 book chapters, journal articles, treatment manuals, and curricula on mental health and
substance abuse issues related to older adults.


He works closely with local, state, and nationa
l organizations on issues related to
development, implementation, and evaluation of substance abuse programs for older adults.


Current research focuses on
development of brief interventions for older alcohol abusers and medication insusers, identification

and management of
behavior problems in long term care facilities, and the impact of elder mistreatment on service utilization.







Appendix A



Sub
-
proposals

1. TITLE
: The development of wandering and other behavior problems in nursing home residents.



INVESTIGATOR:


Lawrence Schonfeld, Ph.D., Department of Aging and Mental Health, Florida Mental Health Institute,
University of South Florida.



BACKGROUND:


Nursing home residents with dementia often experience disruptive behavioral problems.


Wanderin
g is
among the most problematic of these behaviors and creates a potential safety problem for both the wanderer and facility. Thos
e
who frequently wander may become lost and are at higher risk for injuries from falls or fractures (Cesari et al., 2002; Colo
mbo
et al., 2001; Colon
-
Emeric et al., 2003; Kiely et al., 1998).


Other negative behaviors are correlated with wandering such as
non
-
aggressive agitation (Cohen
-
Mansfeld et al., 1991; Colombo et al., 2001), screaming and calling out (Snyder, 1978),
physic
al aggression (Dawson & Reid, 1987), depression (Hope et al.; Kiely et al., 2000; Klein et al., 1999) and disturbed
nighttime sleep (Hope et al.; Klein et al. 1999).


Although some studies suggest that wandering is not related to gender (Burns
et al.; Cohe
n
-
Mansfeld et al., 1991; Hope et al., 2001; Jackson et al., 1997; Snyder, 1978; Yang et al., 1999), two studies report
that wandering was more prevalent among males


(Kiely et al., 2000; Klein et al.).


The present study will help determine
whether variabl
es that have been found to be associated with wandering in the literature are useful for describing and predicting
wandering.



SPECIFIC AIMS/QUESTIONS
:


This is an exploratory study of wandering behavior in which MDS data will be linked with
Medicare and
Medicaid data to determine types characteristics of wandering and healthcare services and costs associated with
wandering (versus not wandering).


Specific aims are :



Aim 1:


To compare individuals coded as "wanderers" to those residents who do not exhib
it such behaviors on:


activities of
daily living, cognition and mental status, physical health, and behavior problems, and on costs to Medicare and Medicaid, bas
ed
on data recorded at admission and on a quarterly basis of available state wide MDS data for

Florida.



Aim 2:


Using the same variables from Aim 1, compare three groups of patients based on wandering:


a) those with history of
wandering on admission, b) those with emerging wandering after admission, and those who never develop wandering.



HYPOT
HESIS:

Hypothesis 1:


Nursing home residents who are judged as wanderers at admission or those who develop wandering at a later
MDS assessment will utilize greater healthcare services and incur greater costs than non
-
wanderers.



Hypothesis 2:


Wandering w
ill be highly correlated with the presence of other behavior problems as measured by the MDS.



Hypothesis 3:


Individuals judged as wanderers at admission to the nursing home will demonstrate greater cognitive decline,
mortality, and healthcare services i
n comparison to those who either do not wander at admission or those who wander at a later
assessment.



Hypothesis 4:


Wandering will be positively associated with poor communication and comprehension skills, and negatively
associated with social activiti
es.



METHODOLOGY:



General Design
:



For Aim 1, using the MDS variable that identifies how often a person wanders within a seven
-
day period, we will compare
“wanderers” and "nonwanderers" on the initial assessment on the following variables:


presence s
hort term memory problem,
gender, degree and types of functional impairment, dementia (Alzheimer's disease versus other dementia), presence of long
term memory, problems in communication, cognition (speech, lethargy, cognitive ability
-
daily), presence of n
egative
affect/mood, behavior problems such as physical and verbal abusiveness to others, general medical problems, medications, and
mode of locomotion (self
-
propelling wheelchair users versus nonusers/walking residents).


For Aim 2, we will compare
indivi
duals exhibiting no wandering on their initial MDS assessment at any time with individuals who are wanderers and those
who develop wandering sometime after admission.





Subjects
:


The study population will be all nursing home residents age 65+ who entere
d a nursing home in the State of Florida
during the period for which the MDS are obtained (1998) and any subsequent years of MDS that are obtained at a later date.


Data will be linked to Medicare and Medicaid cost data for a three year period including an
d following admission to the nursing
home.



Instrumentation/Measurements
: The following variables from the MDS 2.0 will be used and will be compared to Medicare
and Medicaid healthcare utilization and cost variables:



Admission, Quarterly, and Annual Ass
essments

AA8=1,2,5,6,7,8 and B1=0)

AA2. Gender

AA3. Birthdate

AA4. Race/ethnicity

AA5. Social Security and Medicare Number

AA6. Facility Provider Number (or encrypted code)

AA7. Medicaid Number

AA8a. Primary Reason for Assessment

AA8b. Codes for Assessment
s Required for Medicare PPS or State



Background Information at Admission

AB1. Data of Entry

AB3. Lived Alone

AB2. Admitted From

AB7. Education

AB8. Language

AC1a
-
y. Customary Routine



Full and Quarterly Assessment Forms

A3. Assessment Date

A4. Date of
Re
-
Entry

A5. Marital Status

A6. Medical Record Number

A7. Current Payment Sources

A8. Reason for Assessment



B1. Comatose

B2. Memory

B3. Memory/Recall Ability

B4. Cognitive Skills for Daily Decision Making

B5.


Indicators of Delirium

B6.


Change in Cognit
ive Status



C1. Hearing

C2. Communication

C3. Modes of Expression

C4. Making Self Understood

C5. Speech Clarity

C6. Ability to Understand Others

C7. Change in Communication



D1. Vision

D2.


Visual Limitations/Difficulties

D3.


Visual Appliances



E1a
-
p.
Depression

E2.


Mood Persistence

E3.


Change in Mood

E4a
-
e Columns A & B:


Behavioral symptoms

E5.


Change in Behavioral Symptoms



F1.


Sense of Social Involvement

F2.


Unsettled Relationships

F3.


Past Roles



G1a
-
j(A)(B) Physical Functioning and Stru
ctural Problems

G2. Bathing

G3. Balance

G4. Range of Motion

G5. Modes of Locomotion

G6. Modes of Transfer

G8. ADL Functional Rehabilitation Potential



H1a
-
b. Continence

H3. Appliances and Programs



I1a
-
rr. Diseases

I2a
-
m. Infections

I3a
-
e.


O
ther Diagnoses
-

ICD
-
9 Codes



J1a
-
p.


Problem Conditions

J2a&b. Pain symptoms

J3a
-
j.


Pain site

J4a&b.


Accidents

J5.


Stability of Conditions



M1.


Ulcers (skin)

M2a
-
b.


Type of Ulc
er

M6a
-
g.


Foot ulcers



N1a
-
d. Time Awake

N2. Average Time in Activities

N3a
-
e. Preferred Activity Settings

N4a
-
m. General Activity Preferences

N5. Prefers Changes in Daily Routine



O1. Number of Medications

O4a
-
e.


Received the Following Med
ications



P1b(a)
-
(e). Therapies

P2a
-
f. Intervention Programs for Mood, Behavior, Cognitive Loss

P3a
-
k. Nursing Home Rehabilitation/Restorative Care

P4a
-
e. Devices and Restraints

P5.


Hospital Stays

P6.


ER Visits

P7.


Physician Visits

P8.


Physicians Orders

P9.


Abnormal Lab Values



Q1a
-
c. Discharge Potential

Q2. Overall Change in Care Needs



R1a
-
c. Participation in Assessment



T1a
-
d. Special Treatments and Procedures

T2a
-
e. Walking When Most
Self
-
Sufficient



V1
-
18.
-

a&b


Resident Assessment Protocol Summary (triggered and planned)





ANALYSES
:


This will involve basic statistics to determine strength of relationship and prediction (correlation, multiple
regression), group mean
differences (analysis of variance, co
-
variance, MANOVA), and possibly, hierarchical linear modeling.

2. TITLE
:

Trajectories of health and well
-
being in Minority Residents of Long Term Care Facilities



Faculty:
David Chiriboga, Yuri Jang, Mary Watson, Vict
or Molinari, Lisa Brown, Larry Schonfeld, Larry Branch, William
Kearns and Myra Schneider.



Measures.


This is a broad
-
gauged exploratory study that will assess initial and subsequent health and well being among long
term care residents from diverse ethn
ic/minority backgrounds. Measures will include:


demographic characteristics at entry,
diagnoses, physical and mental health, medications, ADLS/IADLs, pain management, cognition, RAP summary form,
discharge and reentry tracking data, cost of care, and leng
th of stay. Data points to include one full 12 month period (or to
termination of long term care, defined as departure due to death or for a period of three months or more) subsequent to
admission except for data on (1) cost of care, for which a two year p
eriod is requested, and (2) overall length of stay, for which
all available data will be requested. A complete listing of variables is appended.



Sample Population of Interest
. The proposed project will examine the records of all African American, Asian A
merican, and
Hispanic American residents for whom MDS II data were collected during the period extending from 1999 through 2003.


These records will be contrasted with a random sample of 5000 Anglo American.



Analyses.

Because so little is known concerni
ng minority residents of long term care facilities, save that relatively few reside in
LTC facilities for any length of time, the first set of analyses will simply examine characteristics of each of the three min
ority
populations of interest.


Basic distri
butions will be examined for evidence of distinct patterns or profiles. Evidence for
differences among subgroups within each broad category of minority status will also be evaluated (e.g., according to nation o
f
birth, length of stay in the United States,
education and economic status, health status at admission, and source of admission).
Of particular interest will be information on mental health and overall quality of life.


Second, comparisons will be made across
the minority groups and with the Anglo Am
erican sample. These comparisons will be conducted not only for the entire groups
but also according to the critical parameters noted above. Third, trajectories of change will be identified and assessed in t
erms
of implications for future interventions.



List of MDS II Measures



A10A through A10h








living will info

A3A through A9G








demographics, payment sources, guardianship

AA1A through AB9_MH_HISTO
RY


Personal identifiers, demographics

AC1C through AC1Y








Activities

B1 through B6_CHANG_COG
-
STAT


cognition

C1 through C7_CHANGE_IN_COMM


Communications

CARE_LOCK

D1

through D3_VISUAL
-
APP




Vision

E1A_NEG_STAT through






Mood, mental health

E5_CHANGE_BEHAVE

EFFECTIVE DATE

F1A_ACTIVE_OTHERS






Psychosocial well
-
being

t
hrough FAC_INT_ID

G1AA_SELF
-
BED








Physical: adls

Through G9_CHG_ADL

H1A_BOWEL_CONTROL

Through H4_CHANGE_URINARY




B&B

I1AA_SEIZURE_DIS

Through MDS_SUBM_SEQ_NBR





Dx

N1A_MORNING through






Other health information

R4_DISCHAGE_DT

T1AA_REC_THPY_DAYS through




Special treatments

T3STATE_CALC_RUG

TARGET_DATE








date of discharge/reentry/assessment

U01AA_MED_AMT








Meds

Through U18RA_MED_ROUTE




RAP summary data

VA01A_DELURIUM_TR through

VA18B

3. TITLE
:

The impact of autono
my
-
enhancing interventions on function, depression, and overall costs to Medicaid and
Medicare for nursing home residents.



PRINCIPAL INVESTIGATOR:

Jennifer R. Salmon, Ph.D., Assistant Professor, School of Aging Studies, Director, Center for
Housing and L
ong
-
Term Care, Florida Policy Exchange Center on Aging, University of South Florida.



CO
-
INVESTIGATOR
: Glenn Mitchell, Ph.D., Director, State Data Center on Aging, Florida Policy Exchange Center on
Aging, University of South Florida.



BACKGROUND:

Persona
l control, or “self rule,” has long been shown to have positive social and psychological benefits.


It
makes significant contributions to resident life satisfaction and satisfaction with care in nursing homes and other long
-
term care
settings (Salmon, 2001
; Kane et al., 1997).


One way to enhance personal control of nursing home residents, is to first asking
about their values and preferences, which has been shown to affect satisfaction with services (Kane, Degenholtz & Kane,
1999).


Another way to improve
personal control is to increase social contacts which also improves well
-
being in institutions
(Baltes, 1994).


Finally, self
-
care and the use of assistive devices, two practical ways to enhance personal control, lead to better
quality of life and combats
learned helplessness in institutional settings (Teitelman & Priddy, 1988; Baltes, 1994).


Although the
MDS is a comprehensive database on nursing home residents’ physical and mental health, it only indirectly addresses the need
for and access to autonomy
-
e
nhancing activities in the nursing homes: such as speech, physical, respiratory, and psychological
therapies and nursing rehabilitation and restorative care interventions (range of motion, training in activities of daily liv
ing).


This study will test the
impact of these interventions on function and depression, controlling for predisposing, enabling, and
need characteristics at intake, and the potential mediating effect of the degree of initiation and involvement in activities
(a
potential measure of perso
nal control from the MDS).



SPECIFIC AIMS/QUESTIONS
: The aims of this study are to understand the relationship between disability, personal
control, rehabilitation and depression over a 3 year period in a nursing home.


Research questions include: 1) what

is the
relationship between personal control, as expressed by sense of initiative to changes in function, depression, and overall co
sts to
Medicare and Medicaid?


2) what is the relationship between rehabilitation support, as expressed by use of appliance
s and
programs for continence, physical therapies, emotional therapies, and special treatments and procedures, to changes in functi
on,
depression, and overall costs to Medicare and Medicaid?


3) Controlling for predisposing, enabling, need, personal contro
l, and
rehabilitation support, changes in function, and depression, what factors in these models most contribute to overall costs of

Medicare and Medicaid?



HYPOTHESIS:

Nursing home residents who have higher levels of cognitive and psychological autonomy
and rehabilitation support will have
higher function, lower depression, and lower overall costs to Medicare and Medicaid, controlling for predisposing, enabling,
and need variables.



METHODOLOGY:



General Design
: A cohort of nursing home residents who en
tered the nursing home in 1998 or 1999 will be followed for three
years (1999
-
2002) from time of entry to understand the importance of personal control and rehabilitation in improving function,
reducing depression, and decreasing overall costs to Medicare
and Medicaid.



Subjects
:


Population of nursing home residents age 65+ who entered a nursing home in 1998 or 1999 in the State of Florida.


Data will be linked to Medicare and Medicaid cost data.


Records will be linked using identifier information and th
en de
-
identified for analyses.



Instrumentation/Measurements
: The following variables from the MDS 2.0 will be used:



Admission, Quarterly, and Annual Assessments

AA8=1,2,5,6,7,8 and B1=0)

AA2. Gender

AA3. Birthdate

AA4. Race/ethnicity

AA5. Social Securi
ty and Medicare Number

AA6. Facility Provider Number (or encrypted code)

AA7. Medicaid Number

AA8a. Primary Reason for Assessment

AA8b. Codes for Assessments Required for Medicare PPS or State

Background Information at Admission

AB1. Data of Entry

AB3. Liv
ed Alone

AB2. Admitted From

AB7. Education

AB8. Language

AC1a
-
y. Customary Routine

Full and Quarterly (RUG III, 1997 Update) Assessment Forms
(For those residents with A8=1,2,5,6,7,8 and B1=0)

A3. Assessment Date

A4. Date of Re
-
Entry

A5. Marital Status

A6.

Medical Record Number

A7. Current Payment Sources

A8. Reason for Assessment

B1. Comatose

B2. Memory

B4. Cognitive Skills for Daily Decision Making

C1. Hearing

C2. Communication

C4. Making Self Understood

D1. Vision

E1a
-
p. Depression

F1a
-
g. Sense of Initia
tive

F2a
-
h. Unsettled Relationships

F3a
-
d. Past Roles



G1a
-
j(A)(B) Physical Functioning and Structural Problems

G2. Bathing

G3. Balance

G4. Range of Motion

G5. Modes of Locomotion

G6. Modes of Transfer

G8. ADL Functional Rehabilitation Potential

H1a
-
b. C
ontinence

H3. Appliances and Programs

I1a
-
rr. Diseases

K1a
-
d. Oral Problems

L1a
-
g. Oral Status and Disease Prevention

N1a
-
d. Time Awake

N2. Average Time in Activities

N3a
-
e. Preferred Activity Settings

N4a
-
m. General Activity Preferences

N5. Prefers Change
s in Daily Routine

O1. Number of Medications

P1b(a)
-
(e). Therapies

P2a
-
f. Intervention Programs for Mood, Behavior, Cognitive Loss

P3a
-
k. Nursing Home Rehabilitation/Restorative Care

P4a
-
e. Devices and Restraints

Q1a
-
c. Discharge Potential

Q2. Overall Chan
ge in Care Needs

R1a
-
c. Participation in Assessment

T1a
-
d. Special Treatments and Procedures

T2a
-
e. Walking When Most Self
-
Sufficient



ANALYSES
: A series of hierarchical linear models using a pooled cross
-
sectional time series design and correcting for th
e
autoregressive nature of health and disability will be used to test a health care utilization and satisfaction model (Anderse
n and
Newman, 1973; Andersen, 1995).


Each set of variables (predisposing, enabling, need, psychological and rehabilitation
media
tors) will be added in steps to predict three dependent variables: Change in Function, Change in Depression, and Costs to
Medicare and Medicaid.


The final dependent variable will be based on linked data not including in the MDS.



The requested variables
listed above are listed below in Table 1 and grouped by model components and assessment form.



IMPORTANCE:
Federal and state governments are interested in containing Medicare and Medicaid costs for acute and long
-
term care.


In addition, CMS and the State

Agency for Healthcare Administration is interested in improving quality of life for
residents of nursing homes.


This study will look at how procedures that are in place in nursing homes can improve quality of
life and save money for Medicare and Medicaid

by improving function and socialization of residents.



DISSEMINATION:

Results of these analyses will be submitted to peer reviewed journals in healthcare financing and long
-
term care for potential publication.

Table 1

Model Components and Assessment Form



Model Variables

Tracking Form: Admission,
Discharge, Re
-
Entry

Admission

Face Sheet

Full

Assessment

(admission and
annual)

Quarterly
Assessment

Predisposing









AA2. Gender








AA3. Birthdate (Compute Current Age)








AA4. Race/ethnicity








AA5. Social Security Number and Medicare
Numbers (or encrypted code).








AA6. Facility Provider Number (or encrypted
code)








AA8a. Primary Reason for Assessment








AA8b. Codes for Assessments Required for
Medicare PPS or State








AB1. Data of Entry (Compute Days in NH)








AB2. Admitted From








AB3. Lived Alone








AB7. Education








AB8. Language








A3. Assessment Date







A4. Date of Re
-
Entry







A8. Reason for Assessment








Inde
pendent Variables: Enabling









A5. Marital Status








A7. Current Payment Sources








F2a
-
h. Unsettled Relationships









F3a
-
d. Past Roles











Independent Variables: Need









B2. Memory







C1. Hearing








C2. Co
mmunication








D1. Vision








G8. ADL Functional Rehabilitation Potential








H1a
-
b. Continence







I1a
-
rr. Diseases







K1a
-
d. Oral Problems







L1a
-
g. Oral Status and Disease Prevention








O1. Number of Medications







Q1a
-
c. Discharge Potential








Q2. Overall Change in Care Needs







T2a
-
e. Walking When Most Self
-
Sufficient








Mediating Variables:









Personal Control









AC1a
-
y. Customary Routine








B4. Cognitive Skills for Daily

Decision
Making







F1a
-
g. Sense of Initiative









C4. Making Self Understood







N1a
-
d. Time Awake







N2. Average Time in Activities







N3a
-
e. Preferred Activity Settings








N4a
-
m. General Activity Preferences








N5
. Prefers Changes in Daily Routine








R1a
-
c. Participation in Assessment








Mediating Variables: Rehab Support









H3a
-
j. Appliances and Programs







P1b(a)
-
(e). Therapies







P2a
-
f. Intervention Programs for Mood,
Behavior, Cogn
itive Loss








P3a
-
k. Nursing Home
Rehabilitation/Restorative Care







P4a
-
e. Devices and Restraints







T1a
-
d. Special Treatments and Procedures








Dependent Variables: (separate equation
for each DV):









Count of variables code
d level 2
-
3:

G1a
-
j(A)(B) Physical Functioning and
Structural Problems







Count of variables coded level 2
-
3:

G2. Bathing

G3. Balance

G4. Range of Motion

G5. Modes of Locomotion

G6. Modes of Transfer







[G2, G3, G4, G6
only]

E1a
-
p. Depression







Total Medicare and Medicaid Costs

From Florida State Data Center on Aging linked files



4. Title
:


Relationships between utilization of psychotropic medication and adverse events.



Investigator
: Marion A. Becker, PhD., Department of Mental Healt
h Law and Policy



Specific aim:
The objective of this analysis is to describe the characteristics and trajectories of nursing home (NF) residents
with and without dementia at psychotropic medication prescribed over the course of their NF stay.


My aim is
to evaluate
utilization of psychotropic medication, and the relationship between psychotropic medication use, changes in functional and
cognitive status over time, and adverse outcomes such as falls, fractures, use of physical restraints, and decreased tim
e spent in
social activity.


I would like to use data from twelve consecutive quarters.


Additionally, I would like to compare outcomes for
residents in special care units with those not receiving special care.



Variables needed:

AA2

gender

AA3

age
-
set f
or the first quarter of 8 quarters sex

AA4

race/ethnicity

AA8 a,b

reason for admission, special assessment code

AB1

date of entry

AB2

admitted from

AB6

occupation

AB7

education

AB9

mental health history

AC1K

Alcohol use

AC1 m
-
l

ADL patterns

AC1 s
-
x

involvement w/ others

A5

marstat

A7 a
-
j

source of payment

A8

reason for assessment

A9 a
-
g

responsibility/legal guardian

A10 a
-
j

advance directives





B1

comatose

B2 a
-
b

short term/long term memory

B3 a
-
e

memory questions

B4

cog skills/daily d
/m

B5 a
-
f

delirium

B6

change in cognition





E1GP1 a
-
i

mood
-

verbal

E1GP2 j
-
k

sleep

E1GP3 l
-
n

sadness

E1GP4 o
-
p

interest level





E2

mood persistence

E3

change in mood

E4 a
-
e

behavioral symptoms

E5

change in behavior





F1 b
-
d

psychosocial

structured activities

F2 a
-
h

unsettled relationships

F3 a
-
d

past roles





G1a, aA, aB

ADL bed

G1b, aA, aB

ADL transfer

G1c, aA, aB

ADL walk in room

G1d, aA, aB

ADL walk in corridor

G1e, aA, aB

ADL locomotion on unit

G1f, aA, aB

ADL locomotion of
f unit

G1g, aA, aB

ADL dress

G1h, aA, aB

ADL eat

G1i, aA, aB

ADL toilet

G1j, aA, aB

ADL personal hygiene

G2

bathing

G2 A,B

bathing

G3 a,b

balance

G4 a,f

orthopedic/Nagi items

G5 a
-
3

modes of locomotion

G6 a
-
f

transfer aids

G7

task segmentation

G8 a
-
e

rehab potential

G9

change in ADLs





I
1 m

hip fracture

I
1 q

Alzheimer's

I
1 t

CVA

I
1 u

other dementia

I
1 y

Parkinson’s

I
1 bb

TIA

I
1 ee

depression

I
1 ff

manic depression

I
1 gg

Schizophrenia





K5 a

Falls within the last 30 days





J4

a
-
e

falls

Ji

hallucinations





N2

time spent in social activities

N5

hospital stays in last 90 days

N6

ER stays in last 90 days

O1

# meds

O4 a
-
e

type of meds





P1
-
an

AD special care unit

P4 0
-
2 a
-
e

physical restraints



5. Title
: The provis
ion of psychological services in nursing homes and health care expenditures



Investigator
: Victor Molinari, PhD; Department of Aging and Mental Health; Florida Mental Health Institute; University of
South Florida.



Background
: The move to a nursing home
represents the introduction of several significant psychological stressors for the
caregiver. It has been long established that family members do not abandon their older family members and do provide most of
the community care that older adults need (Brody
, 1985). Studies have documented the decline in psychological well being that
also occurs for family members during the transition from community to institutional living (Greenfield, 1984; Lieberman,
Prock & Tobin, 1968). Placement in an institutional sett
ing often occurs after years of caregiving by family members who
finally become overwhelmed (Pruchno, Michaels, & Potashnik, 1990). Although family distress is predictive of
institutionalization (Lieberman & Kramer, 1991), NH placement does not necessarily

reduce caregiver burden (Lieberman &
Fisher, 2001; Stephens, Ogrocki, & Kinney, 1991; Whitlach, Schur, Noelker, Ejaz & Loman, 2001). Relief and respite may
certainly follow placement, yet the experience also brings new feelings of guilt and demoralization

(George & Gwyther, 1986;
Grau & Teresi, 1993; Pushkar Gold et al., 1995). Attention to potential adjustment issues common to newly admitted NH
residents may serve an important primary prevention function against the development or exacerbation of mental i
llness in NH
residents.





Two
-
thirds of NH residents have been reported to have a mental disorder (Burns et al., 1993), as many as 40% suffer
from depression (Randall, 1993), and 3.5%


20% suffer from symptoms of anxiety (Parmelee, Katz, & L
awton, 1993).
Depression in the elderly has been linked with many adverse outcomes, including more frequent medical office visits and
multiple medications (Luber, Alexopoulos, & Charlson, 1996), longer hospital stays for NH residents (Samuels & Katz, 1995)
,
poorer self
-
rated health (Mulsant, Ganguli, & Seaberg, 1997), and higher rates of mortality (Lebowitz, 1996). Depressive
symptomatology can also negatively impact functioning (Cross
-
Dunham & Sager, 1994), exacerbate medical illness (Miller et
al., 1996)
and heighten physical disability (Lyness et al., 1993). Further, there is high co
-
morbidity between depression and
anxiety in NH residents, and anxiety is also associated with decreased physical health status, increased functional disabilit
y, and
increased

rates of cognitive impairment (Parmelee, Katz, & Lawton, 1993). At least half of the residents in NHs suffer from
dementia (Kim & Rovner, 1996), so interventions for NH residents must take account of the high rate of cognitive impairment.

Despite the doc
umentation of such high rates of psychiatric problems in nursing home settings (indeed nursing homes
have been called psychiatric institutions), there is very little mental health treatment. In one study, only 4.5% of nursing
home
residents with a psychiat
ric diagnosis received treatment in a one
-
month period. Even though most of this treatment is
psychopharmacological, half of all nursing homes do not have access to a psychiatrist and three
-
quarters if nursing homes don’t
have access to behavioral consulta
nts. We need to determine how much psychological care is received by residents of nursing
homes, and what factors are associated with the receipt of this care.



Specific Aims/ Questions
:
This is an exploratory study designed to answer a basic question: Do
es the provision of
psychological services in Florida nursing homes reduce overall health care utilization and expenditures? The research plan ha
s
2 related aspects. One, to identify those nursing homes that provide mental health services, and compare them

with a 'matched'
group of nursing homes that do not provide mental health services. Two, to identify residents who have either mental health
diagnoses and/or those who score high on certain mental health variables (i.e., mentally impaired). We will then c
ompare health
service utilization and follow
-
up mental health indices between those mentally impaired who receive mental health treatment
and those who do not.



Hypotheses:

1)


There will be a very low percentage of nursing homes that provide mental h
ealth services.

2)


Provision of mental health services will be associated with nursing homes that are more expensive and have less
Medicaid residents.

3)


Provision of mental health services will be associated with behavioral problems associated w
ith the diagnoses of a)
dementia and b) psychosis.

4)


Those residents with mental health problems who receive treatment will score more positively on mental health
indices than those who do not receive treatment.



Methodology
:



General Design
: Prob
alistic matching for nursing homes who do and do not offer mental health services will be conducted
using the Minimum Data Set (MDS) II. Relevant data potentially associated with provision of mental health services include:
demographic characteristics (age
, gender, LOS), health service utilization (# of prescriptions for psychotropic meds (it would be
good if we could track the # of new prescriptions after visits by mh professionals); # of psychiatric hospitalizations; # of
visits
by mh professionals; # of
psychiatric sentinel events), and mental health related variables (diagnoses; depression score, mental
status score).



Subjects
: The prabalistic sample will be drawn from all the nursing homes in the State of Florida who admitted residents during
the peri
od for which the MDS data was obtained (1998) and any subsequent years that the MDS is obtained. Data will be linked
to Medicare and Medicaid cost.



MDS variables
:

A4B; A5 AA1A;AA1B;AA1C;AA1D; AA2;AA3;AB2; AB7;AB9 (Demographics)

AB9 (Prior MH hx);

B2
-
B6
(Cognition)

E1; E2; E3; (Mood)

E4; E5 (Beh Probs)

J1E; J1I

I1BB; I1CC; I1DD; I1EE; I1FF; I1GG (Psych symptoms); 1Q; 1R; 1T; 1U; 1Y

P1BEB (Psychosocial Therapy)***

P2A (Behavior); P2B (Eval); P2C (Group); P2D;2E; 2F (Intervention Program)

U (Medication)

VA0
7A (Well
-
being); VA08A(Mood); VA09A (Behavior)





6. TITLE
:
Pre and Post SB1202 Nursing Facility Resident Characteristics



CO
-
INVESTIGATORS:

Deborah K. Hedgecock, Ph.D.
c
, Florida Policy Exchange Center on Aging, School of Aging Studies; Christopher E.
Jo
hnson, Ph.D., Research Health Scientist, Rehabilitation Outcomes Research Center, Department of Veterans Affairs,
Gainesville, FL; and Glenn E. Mitchell II, Ph.D. Director, State Data Center on Aging, Policy Center, School of Aging Studies
,
University of S
outh Florida.



BACKGROUND:

The function of litigation in the improvement, survival or declination of nursing facilities is of great interest to
providers, consumers, insurers and legislators, particularly in light of the fact that nationwide 1.6 million
persons currently use
long
-
term care services provided by approximately 17,000 such facilities. Empirical studies examining numbers, complaints,
allegations, contents and outcomes of lawsuits filed against nursing facilities are scant, but are emerging as
an important area of
inquiry with an accentuation on nursing facility quality of care (Kapp, 2000). Part of the issue may involve how quality of c
are
is defined. Although 79 percent of cases in one nursing facility litigation study alleged “failure to prov
ide adequate and
appropriate health care,” the expression “quality of care” as such was not used in filed complaints (Hedgecock et al., 2003).

Researchers must then discover what details can be determined about allegations in nursing facility legal actions

that become
descriptors of care issues or that identify consequences of improper care or health deterioration.



Unacceptable pressure ulcer rates observed in nursing facilities at the time of state survey inspections result in
deficiency cit
ations; are considered a manifestation of poor quality care; and have been considered as a possible lawsuit pattern
indicator. Development and complications from pressure ulcers have been noted in resident
-
care related lawsuits (Bennett,
O'Sullivan, DeVito
, & Remsburg, 2000; Hedgecock et al., 2003). However, deficiencies in quality indicators, including
pressure sore rates, have been found to have no significant effect on the number of nursing facility lawsuits filed (Johnson
et al.,
2003).



In
dividual lawsuits filed within court systems are publicly accessible and may contain valuable details that can be used
in primary data analysis to explore litigation trends and correlations of numerous variables, e.g., complaints, allegations,
plaintiff st
atistics, lawsuit duration or settlement amounts, across multiple settings including individual facilities, cities,
counties, regions or states.

Even more uncommon are studies examining effects of tort reforms on common complaints and allegations frequentl
y
seen in nursing facility litigation. Yet without in
-
depth data collection and analysis, simply reporting numbers of filed lawsuits
will not necessarily identify the complete impact of preventative or curtailing measures instituted by legislative actions.

With
the passage of legislation to improve nursing facility quality through incremental increases in per resident per day staffing

hours
and other tort reform measures to affect a decrease nursing home litigation ("Senate 1202: Relating to Long
-
term
-
care
Facilities," 2001)(SB1202) in 2001, Florida has been placed in a unique position to provide an arena for such data collection
.



SPECIFIC AIMS/QUESTIONS
:

The aims of this study are to identify possible trends or patterns in characteristics of residents i
nvolved in nursing
facility litigation seeking possible correlations with trends of lawsuit complaints and allegations in one Florida region, an
d
identifying any differences in these patterns after passage of SB1202.


Research questions include: 1) what ar
e the demographic
and clinical characteristics of residents involved in nursing facility litigation?


2) Will successive individual resident
assessments reveal deterioration in cognitive, psychosocial, mood, physical functioning, disease diagnoses, health,

continence,
oral/nutritional status, skin conditions or increased pain levels prior to lawsuit filings? 3) Will specific clinical conditi
ons
decrease in frequency and severity after SB1202’s implementation?



HYPOTHESIS:

There will be a correlation betwe
en residents’ serious clinical conditions and complaints and allegations in filed lawsuits
with a change in the pattern of clinical conditions manifested in a decrease in more serious health conditions after the
implementation of SB1202.



METHODOLOGY:

Gen
eral Design
:



Quasi
-
experimental design using interrupted time series analysis of secondary data to determine changes in resident
clinical condition trends after introduction of SB1202.



Subjects
:



Entire Florida nursing facility resident p
opulation and the specific nursing facility resident subpopulation of
Hillsborough County, Florida from 1999 through 2003.





Instrumentation/Measurements
:

The following variables from the Minimum Data Set (MDS)


Version 2.0 For Nursing Home Resident As
sessment
and Care Screening

(September 2000), or applicable earlier MDS versions will be used:



Basic Assessment Tracking Form
(For those residents with AA8= 1,2,3,5 and B1=0)

AA2.




Gender

AA3.




Birthd
ate

AA4.




Race/ethnicity

AA5.




Social Security and Medicare Number

AA6.




Facility Provider Number (or encrypted code)

AA7.




Medicaid Number

AA8a.





Primary Reason for Assessment



Background (Face Sheet) Information at Admission

AB1.




Date of Entry

AB2.




Admitted From

AB5a
-
f.


Residential History 5 Years Prior
To Entry

AB7.




Education

AB8.




Language



Full Assessment Form
(For those residents with A8= 1,2,3,5,6,7,8,9 and B1=0)

A3.




Assessment Date

A4.




Dat
e of Re
-
Entry

A5.




Marital Status

A6.




Medical Record Number

A7.




Current Payment Sources

A8.




Reason for Assessment

A9.




Responsibility/Legal Guardian

A10.




Advanced Directives

B1.




Comatose

B2.




Memory

B4.




Cognitive Skills for Daily Decision Making

B6.




Change in Cognitive Status

C1.




Hearing

C2.




Communication

C4.




Making Self Understood

D1.




Vision

E1a
-
p.




Depression

E3.




Change in Mood

E5.




Change in Behavioral Symptoms

F1a
-
g.




Sense of Initiative

F2a
-
h.




Unsettled Relationships

F3a
-
d.





Past Roles



G1a
-
j(A)(B)


Physical Functioning and Structural Problems

G2.




Bathing

G3.




Balance

G4.




Range of Motion

G5.




Modes of Locomotion

G6.




Modes of Transfer

G8.




ADL Functional Rehabilitation Potential

G9.




Change in ADL Function

H1a
-
b.




Continence

H3.





Appliances and Programs

H4.




Change in Urinary Continence

I1a
-
rr.




Diseases

I2a
-
m.




Infections

J1a
-
p.




Problem Conditions

J2(a)(b)



Symptoms

J3a
-
j.




Pain Site

J4a
-
e.




Accidents

J5a
-
d.




Stability of Conditions

K1a
-
d.




Oral Problems

K3(a)(b)


Weight Change

K4a
-
d
.




Nutritional Problems

K5a
-
I.




Nutritional Approaches

K6(a)(b)


Parenteral or Enteral Intake

M1a
-
d.




Ulcers (Due to any cause)

M2a
-
b.




Typ
e of Ulcer

M3.




History of Resolved Ulcers

M4a
-
h.




Other Skin Problems or Lesions Present

M5a
-
j.




Skin Treatments

M6a
-
g.




Foot Problems and Care

N1a
-
d.




Time Awake

N2.




Average Time in Activities

N3a
-
e.




Preferred Activity Settings

N4a
-
m.


General Activity Preferences

N5.




Prefers Chang
es in Daily Routine

O1.




Number of Medications

O2.




New Medications

O3.




Injections

O4.a
-
e.




Days Received the Following Medication

P1a(a)
-
(s).



Special Care

P1b(a)
-
(e).


Therapies

P2a
-
f.




Intervention Programs for Mood, Behavior, Cognitive Loss

P3a
-
k.




Nursing Home Rehabilitation/Restorative Care

P4a
-
e.





Devices and Restraints

P5.




Hospital Stay(s)

P6.




Emergency Room (Er) Visit(s)

P7.




Physician Visits

P8.




Physician Orders

P9.





Abnormal Lab Values

Q1a
-
c.




Discharge Potential

Q2.




Overall Change in Care Needs

R1a
-
c.




Participation in Assessment

T1a
-
d.




Special Treat
ments and Procedures

T2a
-
e.




Walking When Most Self
-
Sufficient



ANALYSES
:



The PIs are very much aware of the stigma of lawsuits against nursing homes and the level of concern by providers
and regulators about prot
ecting the anonymity of nursing homes and their residents.


Therefore, the following procedures will
be followed to assure data confidentiality:





1)


Lawsuit data will be de
-
identified except for the Social Security number.


The following f
ields will be removed:
court ID, plaintiff name, birth year, address, and defendant and attorney names.

2)


MDS data will be de
-
identified except for Social Security number. The following fields will not be included in the
data extract for ana
lysis: nursing home identifier, other resident identifiers, year of birth. County will be retained.

3)


The two de
-
identified databases will be linked through the Social Security number of the resident by the Director of
the State Data Center
on Aging who does not have access to the lawsuit data.

4)


The merged database will replace Social Security numbers with an encrypted code for analysis purposes.



In addition to these steps, the PIs will submit a separate human subjects appli
cation to the USF Institutional Review
Board to assure that all procedures will protect the confidentiality of all residents and facilities.



Interrupted time series analysis will be used, observing frequency and severity of resident clinical characterist
ics for
two
-
year periods before and after passage of SB1202.


Analysis of variance will be used to discover main and interaction
effects of the following MDS variables on filed lawsuits in pre and post SB1202 two
-
year periods:



Basic Assessment Tracking F
orm
(For those residents with AA8= 1,2,3,5 and B1=0)

AA2.




Gender

AA3.




Birthdate

AA4.




Race/ethnicity





Background (Face Sheet) Information at Admission

AB7.




Education





Full Assessment Form
(For those residents with A8= 1,2,3,5,6,7,8,9 and B1=0)

A5.




Marital Status

A9.




Responsibility/Legal Guardian

A10.




Advanced
Directives

B1.




Comatose

B2.




Memory

B4.




Cognitive Skills for Daily Decision Making

B6.




Change in Cognitive Status

C1.




Hearing

C2.




Communication

D1.




Vision

E1a
-
p.




Depression

E3.




Change in Mood

E5.




Change in Behavioral Symptoms

F1a
-
g.





Sense of Initiative

F2a
-
h.




Unsettled Relationships

F3a
-
d.




Past Roles



G1a
-
j(A)(B)


Physical Functioning and Structural Problems

G2.




Bathing

G3.




Balance

G4.




Range of Motion

G5.




Modes of Locomotion

G6.




Modes of Transfer

G8.




ADL Functional Rehabilita
tion Potential

G9.




Change in ADL Function

H1a
-
b.




Continence

H3.




Appliances and Programs

H4.




Change in Urinary Continence

I1a
-
rr.





Diseases

I2a
-
m.




Infections

J1a
-
p.




Problem Conditions

J2(a)(b)


Symptoms

J3a
-
j.




Pain Site

J4a
-
e.




Accidents

J5a
-
d.




Stability of Conditions

K1a
-
d.




Oral Problems

K3(a)(b)


Weight Change

K4a
-
d.




Nutritional Problems

K5a
-
I.




Nutritional Approaches

K6(a)(b)



Parenteral or Enteral Intake

M1a
-
d.




Ulcers (Due to any cause)

M2a
-
b.




Type of Ulcer

M3.




History of Resolved Ulcers

M4a
-
h.




Other Skin Problems or L
esions Present

M5a
-
j.




Skin Treatments

M6a
-
g.




Foot Problems and Care

N1a
-
d.




Time Awake

N2.




Average Time in Activities

O1.





Number of Medications

O2.




New Medications

O3.




Injections

O4.a
-
e.




Days Received the Following Medication

P1a(a)
-
(s).


Special Care

P1b(a)
-
(e).


The
rapies

P2a
-
f.




Intervention Programs for Mood, Behavior, Cognitive Loss

P3a
-
k.




Nursing Home Rehabilitation/Restorative Care

P4a
-
e.




Devices and Restraints

P5.





Hospital Stay(s)

P6.




Emergency Room (Er) Visit(s)

P7.




Physician Visits

P8.




Physician Orders

P9.




Abnormal Lab Values

Q1a
-
c.




Discharge Potential

Q2.




Overall Change in Care Needs

R1a
-
c.




Participation in Assessment

T1a
-
d.




Special Treatments and Procedures

T2a
-
e.





Walking When Most Self
-
Sufficient



References



Bennett, R. G., O'Sullivan, J., DeVito, E. M., & Remsburg, R. (2000). The increasing medical malpractice risk related to
pressure ulcers in the United States.
Journal of the American Geriatrics Societ
y, 48
(1), 73
-
81.

Hedgecock, D. K., Johnson, C. E., Oakley, M. L., Salmon, J. R., Polivka, L., & Hyer, K. (2003). Nature and extent of nursing
facility lawsuits.
Long
-
Term Care Interface, 4
(11).

Johnson, C. E., Hedgecock, D. K., Oakley, M. L., Dobalian, A.,

Salmon, J. R., Hyer, K., et al. (2003). Predictors of lawsuit
activity against nursing homes in Hillsborough county Florida.
The Gerontologist
(In press).

Kapp, M. B. (2000). Quality of care and quality of life in nursing facilities: What's regulation got
to do with it?
McGeorge Law
Review, 31
(Spring), 707
-
731.

Senate 1202: Relating to Long
-
term
-
care Facilities, Fla. S., 2001 Sess.(2001).





7. Title
:


Characteristics and trajectories of NF residents with dementia on quality of care and quality of life.



Investigator:

Kathryn Hyer, PhD, MPP, School of Aging Studies and Marion A. Becker, PhD., Department of Mental Health
Law and Policy



Specific aim:

The objective of this analysis is to describe the characteristics and trajectories of nursing home (NF) re
sidents
with and without dementia over the course of their NF stay.


Our objective is to evaluate quality of care and quality of life
measures for residents with dementia at admission and those who develop dementia over the course of their stay in the NF.


Specifically we will analyze dementia’s impact on changes in functional and cognitive status over time, and adverse outcomes
such as falls, fractures, use of physical restraints, and decreased time spent in social activity.


We would like to use data from

twelve consecutive quarters to begin to identify the quality of life and quality of care impact of dementia in NH residents.


We
will also compare outcomes for residents with dementia who are in special care units with those not receiving special care.


A

final component will be to assess the cost of care and the level of care required over these trajectories.



Variables needed:

AA2

gender

AA3

age
-
set for the first quarter of 8 quarters sex

AA4

race/ethnicity

AA8 a,b

reason for admission, special asses
sment code

AB1

date of entry

AB2

admitted from

AB6

occupation

AB7

education

AB9

mental health history

AC1K

Alcohol use

AC1 m
-
l

ADL patterns

AC1 s
-
x

involvement w/ others

A5

marstat

A7 a
-
j

source of payment

A8

reason for assessment

A9 a
-
g

respon
sibility/legal guardian

A10 a
-
j

advance directives





B1

comatose

B2 a
-
b

short term/long term memory

B3 a
-
e

memory questions

B4

cog skills/daily d/m

B5 a
-
f

delirium

B6

change in cognition





E1GP1 a
-
i

mood
-

verbal

E1GP2 j
-
k

sleep

E1GP3 l
-
n

sa
dness

E1GP4 o
-
p

interest level





E2

mood persistence

E3

change in mood

E4 a
-
e

behavioral symptoms

E5

change in behavior





F1 b
-
d

psychosocial structured activities

F2 a
-
h

unsettled relationships

F3 a
-
d

past roles





G1a, aA, aB

ADL bed

G1
b, aA, aB

ADL transfer

G1c, aA, aB

ADL walk in room

G1d, aA, aB

ADL walk in corridor

G1e, aA, aB

ADL locomotion on unit

G1f, aA, aB

ADL locomotion off unit

G1g, aA, aB

ADL dress

G1h, aA, aB

ADL eat

G1i, aA, aB

ADL toilet

G1j, aA, aB

ADL personal hy
giene

G2

bathing

G2 A,B

bathing

G3 a,b

balance

G4 a,f

orthopedic/Nagi items

G5 a
-
3

modes of locomotion

G6 a
-
f

transfer aids

G7

task segmentation

G8 a
-
e

rehab potential

G9

change in ADLs





I
1 m

hip fracture

I
1 q

Alzheimer's

I
1 t

CVA

I
1 u

oth
er dementia

I
1 y

Parkinson’s

I
1 bb

TIA

I
1 ee

depression

I
1 ff

manic depression

I
1 gg

Schizophrenia





K5 a

Falls within the last 30 days





J4 a
-
e

falls

Ji

hallucinations





N2

time spent in social activities

N5

hospital stays in last 90 da
ys

N6

ER stays in last 90 days

O1

# meds

O4 a
-
e

type of meds





P1
-
an

AD special care unit

P4 0
-
2 a
-
e

physical restraints



METHODOLOGY:



General design:


Using the MDS variables that identify dementia on the initial assessment or identify dement
ia during
subsequent assessments we will compare persons with dementia on the following variables.



8. TITLE
: Quality of Care Assessment and Improvement for Falls
-
Related Injuries in LTC Facilities (in Florida)



Investigator:
Bernard A. Roos, MD, profess
or of medicine and neurology and director,

Geriatrics Institute, University of Miami, GRECC, VA Medical Center, and Stein Gerontological Institute at the Miami Jewish
Home & Hospital





BACKGROUND:



Chronic resident care in the institutional

LTC setting encompasses many more dimensions of concern than seen in
post
-
acute or subacute care. Florida’s


Teaching Nursing Home (TNH), supported by the State of Florida’s Agency for Health
Care Administration (AHCA) is focused on the continuing LTC of
frail elders and shares with FPECA and FHCA a strong
commitment to assist Florida with the development of better evaluations of such LTC programs through the development of
best practices and measures that reflect high quality related to performance of LTC

facilities in the continuing care of frail
residents. One of the major concerns in this regard has been falls and especially falls
-
related injuries.



For this study, we are researching the performance of LTC facilities with regard to falls a
nd falls
-
related injuries, the
greatest single cause of litigation in Florida. Such injuries greatly concern LTC facilities because their reporting and the
resultant publicity lead to erosion of reputation and morale for LTC facilities.

Falls
-
related inju
ries are more prevalent in a frail population. More specifically, such injuries are directly related to risk
factors such as dysmobility, osteoporosis, use of anticoagulants, problems with ambulation as reflected by need for locomotio
n
assist, and the use
of psychotropic medications for treatment of depression, anxiety, and other behavioral problems. Thus, it
seems likely that quality of care for long
-
term residents of LTC facilities may not be reflected accurately by reference only to
the prevalence of fal
ls and prevalence of falls
-
related injury. Consider that within any facility the falls
-
related injuries occur in
certain subpopulations, such as persons who are being rehabilitated after knee and hip replacement. This type of finding
highlights our concern

and the motivation for the proposed study. Our pilot study is designed to open the way for more accurate
interpretation, with regard to institutional performance, of the prevalence of falls
-
related injuries in the frail long
-
term residents
of our nursing
homes.

We consider these initial analyses of MDS and related data critical to creating the evidential basis for judging when
falls
-
related injuries are likely to reflect negligence linked to facility care practices vs. the expectations for injury in the f
ace of
high risk for falls and/or injury from falls.



HYPOTHESIS AND STUDY AIMS:

We postulate that quality measured in terms of the prevalence of falls
-
related injuries does not accurately reflect the
quality of care at a facility. In fact, there may be

little relationship between falls
-
related injuries and the quality of care. A
corollary of this postulate is that falls
-
related injuries (and perhaps all foci of care that are used to inform consumers) should be
reported in terms of the facility
-
specific
population risk and its change over time.

Accordingly, we propose to develop an approach to adjusting falls
-
related injury data that should more accurately reflect
underlying risk factors that clearly vary within different care arenas and within and betwee
n institutions. This approach would
first of all provide a means of using falls
-
related injury prevalence data to assess LTC performance; and it could open the way
to a better approach for reporting quality measures to be used by consumers and regulators.
This approach, if successful for
falls
-
related injuries, could be added and applied more broadly along with other quality measures to evaluate adverse events in
LTC care.


Ideally, understanding falls
-
related injuries will enable the more efficient and eff
ective targeting of QI education and
training because we will have a more sophisticated understanding of the characteristics, risk factors and implications on qua
lity
of life of falls
-
related injuries.

While our work to create measures of quality is preli
minary, we will explore ways of classifying nursing homes using
publicly reported State criteria to indicate high quality facilities.


One measure would be facilities that have a “Gold Seal”
although there are only a handful of these facilities.


Another i
s ranking Fl. facilities within regions on the States’ Nursing
Home Guide or using the Nursing Home Watch List as a measure of poor quality.


We will also work with the state to create
new measures of quality that we hope they will help us create by sharin
g Nursing home adverse incidents and deficiencies for
the same time periods of our data collection.


Ideally we would use these measures to create rankings of nursing homes using
multiple measures to help validate our falls
-
related measure.



METHODOLOGY:

This proposal uses selected data from Minimum Data Set (MDS) version 2.0.


We need to demonstrate the stability of
our assessments within the observation interval chosen. Hence, we will in these initial studies analyze data generated over
consecutive (at l
east 4 and preferably12) quarters. The approach involves researching falls
-
related injury prevalence, falls risk
as assessed by a LTC facility’s prevalence of specific major falls risk factors (as per falls RAP), and finally falls
-
related injuries
adjusted

for the falls incidence and/or extent of falls risk as reflected by the combination of prevalence for specific falls and
injury risk factors (see table below).



In addition, we will look at age, race (Caucasian, African American, Asian, and Hispanic), co
morbidity, and functional
status of residents. The specific major risk factors for falls that we will consider for our initial analyses include those f
rom the
MDS falls RAP, including types and numbers of drugs, mobility and use of locomotion assist, prese
nce of osteoporosis (see
table below).



Our focus is on the true LTC resident as opposed to the post
-
acute patient undergoing an interval of care in a LTC site.
To test our hypothesis that falls
-
related injuries prevalence in LTC residents may reflect p
opulation risk within a facility as
opposed to reflecting quality of care at a facility, we will compare various measures of quality with falls
-
related rankings of
LTC facilities based on (1) facility
-
wide prevalence of falls
-
related injuries in LTC reside
nts as opposed to post
-
acute or
subacute patients, (2) facility
-
wide risk for falls, fractures, and other injuries (such as cerebral bleed), in LTC residents, and (3)
a risk
-
adjusted normalized injury risk for the LTC residents. The risk adjustment will be

based on consideration of prevalence
for major falls and falls
-
related injury risk factors reported in the MDS dataset.



The initial statistical approach will be a univariate analysis of falls, falls with injuries and to develop models of risk
-
adjusted f
alls. As we understand risk
-
adjusted falls with injury we will seek to test these risk
-
adjusted measures against
measures reflecting overall quality of LTC facilities in an effort to test risk
-
adjusted normalized falls
-
related injury risk. From
this initia
l analysis we would proceed to a multivariate more complex analysis to discover which specific risk factors and
demographics are the major determinants of falls
-
related injury risk.

Specifically, MDS data will be used to:



Rank facilities with rega
rd to prevalence and rates of falls
-
related injury



Rank facilities with regard to prevalence for a group of major risk factors (trunk restraints, psychoactive drugs,
use of locomotion assist device, dizziness or orthostatic hypotension, etc.) for
falls
-
related injuries



Rank facilities based on an adjusted rank (rank based on prevalence of falls
-
related injuries divided by the rank
for prevalence of major risks factors for falls
-
related injuries)



Seek best fit (univariate analysis) f
or rank based on quality measures such as gold seal facilities, Nursing home
guides and Nursing home watch lists with the above 3 ranking approaches.



Proceed to more complex analyses of rank to identify specific risk factors and correlates of LTC q
uality of care
(see statistical analyses section below).



This study has important consequences for quality assessment in general and for regulator and public perception of
quality in particular. Quality of care at large LTC facilities and certain highly
specialized facilities could be misjudged based on
the caring for populations of residents at greater frailty and falls risk with accordingly greater prevalence of falls
-
related injury.


If we are correct that falls
-
related injuries reflect the facility
-
wi
de risk for falls and related injuries inherent to their residents
more than the quality of care, we will have opened an important line of defense against the growing falls
-
related litigation
activity in our state.



We also expect to uncover e
vidence supporting our belief, contrary to prevailing consumer views, that smaller facilities
with less compromised residents experience a relatively high falls and injuries
once the presence of risk factors is taken into
consideration
. For example, use of

antipsychotic drugs is associated with a high rate of falls and injuries; corrected for the
prevalence of such drug use, certain facilities might emerge as better performers than previously understood from the
examination of falls and injuries rates alone
. This type of understanding could be of enormous importance in terms of
developing best practice models and targeting quality improvement efforts and interventions.



In addition to State of Florida’s Agency for Health Care Administration (AH
CA) contract, the faculty involved in the
TNH work are actively engaged in research and related education supported by grants. Grant sources include the NIH, the VA
(Includes the GRECC and VA Merit Review funding), State of Florida DOH funding, the Hartfor
d Foundation, DW Reynolds
Foundation, AAMC, and the Stein Gerontological Institute.



Study Design and Analytic Plan:



Our primary research question(s), the various types and sources of data, and study design are indicated in the following
sec
tion and summarized in the table above. This dataset is obtained from a cross
-
sectional investigation. Thus, caution should
be given for statistical inference on causal relationships.







Database Management and Analysis:



Data Management
:



The quality of this research project is dependent on having valid data and adequate data analysis. SAS (SAS Institute,
Inc., Cary, NC) will be used to manage the data and perform the major statistical analyses. Dr. Shenghan Lai, associate
pr
ofessor and director of Biostatistics Core, Cardiology Research Unit, Johns Hopkins Medical School, will assist in the
management and data analysis. Dr. Lai has more than ten years’ experience in managing and analyzing complicated large
datasets. He has wo
rked closely with Dr. Roos on several projects.





A detailed code
-
book will be developed for this dataset. The data management group in our data center will provide data
management and record maintenance for all data. The security and confid
entiality of the data are of utmost concern, and appropriate
data management procedures and security clearances will be established to insure confidentiality.


The confidentiality and security
of the data files in the computer will be maintained by passwor
d protection on all computer accounts.

Protection against loss of data
is essential to data management. Copies of the study database files will reside on dedicated workstations.


Backup of all database
files and programs will occur daily.



Data Analysis:



a. Univariate linear regression analysis will first be performed to identify crude associations between possible LTC
quality measures (such as “Gold Seal” facilities or regional rankings provided by AHCA on its Nursing Home web site,
numbers
of adverse incidents, if AHCA agrees to provide these on a facility
-
specific level) and the various falls
-
related factors,
including the risk
-
adjusted normalized falls
-
related injury prevalence.



b. Multiple regression analysis will be perfor
med to adjust for potential confounding factors, including functional
status and comorbidity severity of the residents.



c. Mixed model will be used to adjust for the cluster effects of residents in LTC (such as age, race, gender, and
socioeco
nomic status).



d. Based on the above
-
mentioned regression analyses, an approach to developing risk
-
adjusted ranking for falls
-
related
injuries that reflects the general quality of care will be explored.





9. TITLE
:
End
-
of
-
life care outcomes

among nursing home residents.



INVESTIGATOR:

Jung Kwak, MSW, Graduate Student and Glenn Mitchell, Director, State Data Center on Aging, School of
Aging Studies



OBJECTIVES:

The main objective of this study is to identify patient
-
related characteristics

for end
-
of
-
life care outcomes
among the nursing home residents in their last year of life.


Research questions include: (1) what are the relationship between
patient characteristics (demographic, functional and clinical characteristics, and cognitive stat
us), patient preference [presence
of do
-
not
-
resuscitate order (DNRO) and do
-
not
-
hospitalize order (DNHO)], and three end
-
of
-
life care outcomes
(hospitalization, pain management and place of death)?; (2) what is the role of hospice care in three end
-
of
-
life

care outcomes
controlling for patient related characteristics?; (3) what is the overall cost of end
-
of
-
life care of nursing home residents who use
and do not use hospice care after controlling for patient related characteristics?





BACKGROUND:
It
is estimated that between 28 percent to 35 percent of elderly nursing home residents died in acute care
hospitals while nursing home residents constituted about 22 percent of elderly decedents in 1993 (Hogan, Lunney, Gabel, &
Lynn, 2001).


Current literatu
re on end
-
of
-
life care suggests a number of factors that influence end
-
of
-
life care outcomes such
as pain management, hospitalizations and hospital death.


Predictors for these outcomes include demographic and clinical
characteristics of patients, and pati
ent preferences for types of end
-
of
-
life treatments (Berger, Pereira, Baker, O’Mara, & Bolle,
2002).


There is some evidence that hospice care leads to better pain control among nursing home residents than routine, non
-
hospice care (Gage et al., 2000) whil
e cost savings by use of hospice is inconclusive (Buntin & Huskamp, 2002).


However,
these findings are based mostly on samples of individuals in the community setting and those with cancer diagnosis.


Thus, it is
not clear how these patient related factor
s independently affect types and quality of end
-
of
-
life care received and place of death
experienced by nursing home residents and overall end
-
of
-
life care costs in their last year of life.





IMPORTANCE:
This study will allow to identify and examine the
role of demographic and clinical characteristics and
preferences for end
-
of
-
life care among nursing home residents in the actual end
-
of
-
life care outcomes including quality of and
total cost of care received by these residents.


Findings from this study wi
ll provide implications for service providers and
policy makers on end
-
of
-
life care for nursing home residents.





METHODOLOGY:

1.


General Design
: Nursing home residents who died between 1998 and 2002 and had at least two MDS assessments
available t
o identify and understand predictors of end
-
of
-
life care outcomes.





2.


Subjects
:


Population of nursing home residents age 65+ who deceased between 1998 and 2002 in the State of
Florida.





3.


Data Sources:
MDS 2.0, Florida death certificat
e, Florida Medicaid Long
-
Term Care data.



4.


Hypotheses:

(1)


Nursing home residents who are non
-
Hispanic White, younger, diagnosed with cancer and no dementia and
cognitively competent, and who have completed DNRO or DNHO are more likely to experi
ence fewer
hospitalizations, better pain management and hospital deaths.



(2)


Nursing home residents who use hospice are more likely to experience fewer hospitalizations, better pain
management and hospital deaths than those who do not.



(3)


Nursing
home residents who use hospice are likely to incur less end
-
of
-
life cost than those who do not.





5.


Instrumentation/Measurements
:

(1)


The following variables from the MDS 2.0 will be used:

(2)


Admission Data, Quarterly, and Annual Assessments

(For those residents with AA8=1,2,5,6,7,8 and B1=0)

(3)


AA2. Gender

(4)


AA3. Birthdate

(5)


AA4. Race/ethnicity

(6)


AA5. Social Security and Medicare Number

(7)


AA6. Facility Provider Number (or encrypted code)

(8)


AA7. Medicaid Number

(9)


AA
8a. Primary Reason for Assessment

(10)


AA8b. Codes for Assessments Required for Medicare PPS or State

(11)


Background Information at Admission

(12)


AB1. Data of Entry

(13)


AB2. Admitted From

(14)


AB5.
Residential History 5 Years Prior to Entry

(15)


AB7. Education

(16)


AB8. Language

(17)


AC1a
-
y. Customary Routine

(18)


Full and Quarterly (RUG III, 1997 Update) Assessment Forms
(For those residents with A8=1,2,5,
6,7,8
and B1=0)

(19)


A3. Assessment Date

(20)


A4. Date of Re
-
Entry

(21)


A5. Marital Status

(22)


A6. Medical Record Number

(23)


A7. Current Payment Sources

(24)


A8. Reason for Assessment

(25)


A9. Responsibility/Legal Guardian

(26)


A10. Advanced Directives

(27)


B1. Comatose

(28)


B2. Memory

(29)


B4. Cognitive Skills for Daily Decision Making

(30)


B6. Change in Cognitive Status

(31
)


C1. Hearing

(32)


C2. Communication

(33)


C4. Making Self Understood

(34)


C6. Ability to Understand Others

(35)


D1. Vision

(36)


E1a
-
p. Depression

(37)


G1a
-
j(A)(B) Physical Functio
ning and Structural Problems

(38)


G2. Bathing

(39)


G3. Balance

(40)


G4. Range of Motion

(41)


G5. Modes of Locomotion

(42)


G6. Modes of Transfer

(43)


G8. ADL Functional Rehabilitation Potenti
al

(44)


G9. Change in ADL Function

(45)


H1a
-
b. Continence

(46)


I1a
-
rr. Diseases

(47)


J1 a
-
e. Problem Conditions

(48)


J2a
-
b. Pain Symptoms

(49)


J3a
-
j. Pain Site

(50)


J5a
-
d. Stabili
ty of Conditions

(51)


K1a
-
d. Oral Problems

(52)


K3a
-
b. Weight Change

(53)


K4a
-
d. Nutritional Problems

(54)


O1. Number of Medications

(55)


P1. Special Treatments, Procedures, and Programs

(56)



P5. Hospital Stay(s)

(57)


P6. Emergency Room (ER) Visit(s)

(58)


Q2. Overall Change in Care Needs

(59)


R1a
-
c. Participation in Assessment



6.


ANALYSES
: A series of logistic regression models will be conducted to tes
t three major hypotheses.

1.


Hypothesis 1.


Independent variables will include variables on demographic, clinical, functional, cognitive,
nutritional, patient preference (DNRO, DNHO), and family characteristics and dependent variables will include p
ain
management (reported frequency and intensity of pain), number of hospitalization, and experience of hospital death.


The place of death information on the deceased will be identified by linking the MDS data set with the Florida death
certificates.


Lo
gistic regression model will be used to identify significant predictors separately for three outcome
measures (pain management, hospitalization, and hospital death) for the first hypothesis.





2.


Hypothesis 2.


Independent variables will include pa
tient related variables and hospice use and same dependent
variables used in Hypothesis 1.


Logistic regression models will be used to estimate the odds of experiencing each of the
three outcome measures (as used in Hypothesis 1 and these three outcomes va
riables will be dichotomized into yes or
no) after controlling for patient related characteristics.





3.


Hypothesis 3. Independent variable will be hospice use and dependent variable will be the overall end
-
of
-
life cost in
the last year of life (
which will be derived from the linked data (from the Data Center).


Logistic regression model will
be used to explore the relationship between hospice use and end
-
of
-
life expenditure after controlling for confounders.





Appendix A



Data Security Polici
es of the State Data Center on Aging

Policy Memorandum

Date:


1/26/00

To:


Larry Polivka, Director, Florida Policy Exchange Center on Aging;



Scott Hinton, Assistant Director, State Data Center on Aging

Cc:


Staff and G
raduate Students, Florida Policy Exchange Center on Aging

From:


Glenn E. Mitchell II, Director, State Data Center on Aging

RE:


State Data Center on Aging
--

Policies Relating to Protection of Confidential Data

Priority:



[Urgent]

The State Data Center on Aging (SDCA) is a repository for both confidential data and non
-
confidential data. Confidential data is
housed in the SDCA under contract with government agencies.

The following policies are designed to preserve confid
entiality. State and federal law impose severe penalties for the release or
misuse of confidential data.

1.

Confidentiality shall be preserved as required by contract or letter of agreement from a supplying agency and as required by
applicable federal or stat
e law.

a.

Upon receipt of confidential data from an agency or other data provider, the following actions will occur with all due
diligence:

i.

The dataset containing the confidential data items will be encrypted with military grade encryption software
and stored

on removable media.

ii.

The original data media will be returned to its supplier by personal courier or by bonded carrier with
instructions to release only to a specific individual upon signature. Return receipt will also be required. SDCA
staff will maintain

records of data returned to authorizing agencies.

iii.

The encrypted media will be place in the storage vault.

i.

The storage vault will consist of a securely locked room with a locked data storage cabinet. The keys
to the room and the data storage cabinet will n
ot be keyed to any master key. Only the SDCA director
and assistant director and the university police will have copies of the keys.

iv.

Any temporary storage space to create the encrypted copy of confidential data files will be erased with
military grade soft
ware to remove any residual traces of the data.

2.

Access to confidential data, unless otherwise directed by an agency and in accordance with applicable state and federal law,
will be limited to the staff of the SDCA.

a.

The purposes of use will be strictly limi
ted to purposes allowed by contract or letter of agreement between FPECA,
SDCA and the agency providing the data. No other use shall be permitted.

3.

Access to confidential data shall be done on PCs located in the SDCA.

a.

Confidential data may be loaded tempora
rily on PCs to accomplish contracted services.

i.

Data will be copied from removable media to a secure workspace on the PC protected by military grade
encryption software.

ii.

Upon completion of work, the temporary workspace will be erased by military grade encry
ption software to
remove any residual traces of the data.

b.

Confidential data will not be stored on network drives nor in directories shared with any other user.

c.

Confidential data will not be left on the PCs overnight. At the end of work each day, the remova
ble media will be
secured in the data vault and the temporary workspace will be erased by military grade encryption software to remove
any residual traces of the data.

4.

Records of data access noting the date, time, project or purpose, and name of the person

removing and returning confidential
data to the data vault will be maintained by SDCA staff.

5.

Removable media containing confidential data will be physically destroyed as required by the retention schedules specified
by contract or letter of agreement with

the authorizing agency. SDCA staff will maintain records of destruction.

6.

All records relating to the receipt, use, and destruction of confidential data will be made available, upon reasonable notice
, to
the contracting agency, University officials, and ot
hers as directed by the authorizing agency.

Please report any violations of these policies immediately to the directors of FPECA and SDCA.



GEM






[1]

House

Bill 1971 charged AHCA with the establishment of a Teaching Nursing Home Pilot Project at Stein Gerontologic Institute at the

Miami Jewish Home and Hospital for the Aged "to improve and expand capacity of Florida's healthcare system to respond to the
medi
cal,
psychological, and social needs of the increasing population of frail older citizens."