close

Вход

Забыли?

вход по аккаунту

?

Antiepileptic drug use in nursing home admissions.

код для вставкиСкачать
Antiepileptic Drug Use in Nursing
Home Admissions
Judith Garrard, PhD,1,2 Susan Harms, RPh, MPH, PhD,1 Nancy Hardie, MS, MPH,2 Lynn E. Eberly, PhD,3
Nicole Nitz, MS,1 Patricia Bland, MS,1 Cynthia R. Gross, PhD,2 and Ilo E. Leppik, MD2,4,5
Although 1 of 10 nursing home residents is taking an antiepileptic drug (AED), no study to our knowledge has determined whether most residents are already receiving AED treatment when they are admitted or are given these drugs
afterward. That differentiation was the focus of this study. The study group consisted of 10,318 residents, 65 years and
older, admitted to 510 nursing homes located throughout the United States during the first quarter of 1999. AED
prevalence at admission was 7.7%; three fifths had an epilepsy/seizure indication. In a multivariate analysis, factors
associated with AED use at admission included epilepsy/seizure, bipolar depression, age group, and cognitive performance. In the follow-up cohort (N ⴝ 9,516), postadmission initiation of AEDs was 2.7%; one fifth had an epilepsy/
seizure indication. In the multivariate analysis, factors associated with postadmission AED initiation included epilepsy/
seizure indication, bipolar depression, age group, peripheral vascular disease, and cognitive performance. This rate of
AED postadmission initiation within the first 3 months of admission was much higher than expected, suggesting that
new symptoms may develop after admission. Results also show that the rate of AED use in nursing homes is not static.
Ann Neurol 2003;54:75– 85
The incidence of epilepsy forms a U-shaped curve over
a normal life span, beginning with a high level in infancy and a higher level at age 60 years.1–3 Antiepileptic
drugs (AEDs) are the primary treatment for epilepsy,4,5
although they are prescribed for other conditions, for example, neuropathic pain,6,7 headaches,7,8 bipolar depression,9 mood disorders,10 and behavioral problems.6,11 In
the community, the prevalences of epilepsy and AED
treatment among the elderly are both approximately
1%12,13; among nursing home (NH) residents, however,
epilepsy/seizure has a prevalence of 4.7 to 6.3%14 and
AED use is 10.5 to 11%.15 Whether elderly persons enter NHs with that level of AED use or are given AEDs
after admission has not been reported. AED use at admission suggests that the treatment was initiated in
health care delivery sites other than the NH, and that
the treatment regimen may differ by the existence of the
condition before admission or the specific medications
prescribed. If AED treatment is initiated after admission
to the NH, then the resident’s symptoms, diagnosis, or
need for treatment may have changed. From a public
health standpoint, the distinction is important in understanding how the initiation of AEDs change over time as
the elderly enter and reside in a nursing home. For exFrom the 1Division of Health Services Research and Policy, School
of Public Health; 2Department of Experimental and Clinical Pharmacology, College of Pharmacy; 3Division of Biostatistics, School of
Public Health, University of Minnesota; 4MINCEP Epilepsy Care;
and 5Department of Neurology, Medical School, University of Minnesota, Minneapolis, MN.
Received Feb 12, 2002, and in revised form Oct 15, 2002, and Mar
7, 2003. Accepted for publication Mar 7, 2003.
ample, the initiation of AEDs subsequent to admission
may suggest a greater need for seizure control, increased
need for treatment of other conditions such as neuropathy, the diagnosis of a new condition for which an
AED has been shown to be clinically effective, or recognition of a previously untreated or undetected condition.
AEDs are used by an estimated 100,000 to 150,000
of the 1.5 million NH residents in the United States.
Risks of AED use by elderly in any setting include
falls,16 cognitive impairment,17 and adverse drug
events.18 Population studies about AED use in NHs
either are nonexistent or lack sufficient power to detect
meaningful relationships with factors such as advancing
age,19,20 educational level, cognitive ability, acute illness (stroke),21 and geographic region. Without such
population level research, public health policy and clinical care decisions are difficult. The purpose of this
study was to address this need, as defined by the following research questions.
• Antiepileptic drug use at admission. What was
the prevalence of AED use among elderly nursing
home admissions, stratified by epilepsy/seizure?
Which AEDs were used most frequently by el-
Current address for Nicole Nitz: Lilly Research Laboratories, Indianapolis, IN.
Address correspondence to Dr Garrard, Division of Health Services
Research and Policy, School of Public Health, MMC #729, University of Minnesota, 420 Delaware Street SE, Minneapolis, MN
55455. E-mail: jgarrard@umn.edu
© 2003 American Neurological Association
Published by Wiley-Liss, Inc., through Wiley Subscription Services
75
derly people at admission to nursing homes?
• Antiepileptic drug initiation during follow-up.
Of those admitted who were not taking AEDs
upon entry, what was the incidence of AED use
within 3 months of admission date, stratified by
epilepsy/seizure? What was the timing of AED
medication orders during the follow-up period?
Which AEDs initiated after admission were used
most frequently?
• Factors associated with antiepileptic drug use at
Admission and during Follow-up. What factors
(demographic, clinical, comorbidity, functional
status, reimbursement source) were associated
with use of an AED at either admission or during
follow-up?
Subjects and Methods
Methodological Design
OVERVIEW. This was an epidemiological study using secondary source data from physicians’ medication orders and
standardized patient assessments generated during the first 6
months of 1999. Electronic data were obtained from Beverly
Enterprises, Inc, a publicly traded corporation that owns or
manages the largest chain of long-term care facilities in the
United States.22,23 Beverly residents were distributed geographically by census region in states located in the northeastern (10%), southern (39%), midwestern (24%), and
western (26%) parts of the United States.
Two
groups were studied: (1) an Admissions Group, consisting of
all people (N ⫽ 10,318) 65 years and older admitted between January 1 and March 31, 1999 to all 510 Beverly
facilities in 31 U.S. states; and (2) a Follow-up Cohort (N ⫽
9,516) of all in the Admissions Group who were not using
an AED at entry. The Follow-up Cohort was followed for 3
months after their individual admission dates or until NH
discharge, whichever occurred first.
ADMISSIONS GROUP AND FOLLOW-UP COHORT.
Our intent was to capture the
experience of all newly admitted NH residents within a specified time period, whether admitted for short-term, postacute, or long-term care.24,25 In 1999, the average length of
stay for Medicare-reimbursed postacute care was 29.3 days.25
The average length of stay for all NH admissions in the
United States is highly asymmetric, with approximately 45%
staying less than 3 months.26 In this study, 56.6% of all
admissions stayed less than 3 months.
Medicaid, a jointly funded federal-state health insurance
program for the indigent, reimburses qualified NH expenses
including medications. In this study, as in other health services research, Medicaid status was used as a surrogate measure for low income.
MEDICARE AND MEDICAID.
Data Sources
OVERVIEW. Data from the Minimum Data Set (MDS)
and physicians’ orders (POs), including all orders for medications, were obtained for all subjects. The MDS is widely
76
Annals of Neurology
Vol 54
No 1
July 2003
used in gerontological research,27,28 although it is rarely used
in studies of epilepsy or AED treatment. Medication data
from POs are less commonly used because the information is
not available in most data sets. In this study, the MDS variables were from a single point in time (upon admission), and
PO data were available daily throughout the study period.
MINIMUM DATA SET (MDS) AND PHYSICIANS’ ORDERS (PO).
In 1990, federal regulations required that all residents in
Medicare/Medicaid certified NHs have an MDS evaluation
by a qualified nurse upon admission and annually or more
often if medical conditions changed. MDS validity and reliability are well established.29 –31 MDS variables include
resident demographics, clinical indicators (eg, depression,
dizziness, hip fractures), and functional capabilities, including activities of daily living. Diagnoses are recorded in the
form of both International Classification of Diseases–9
(ICD-9) codes and a checklist of common conditions. By
federal regulation, a PO is required for administration of
any medication, including nonprescription drugs, treatments such as physical therapy, diet, laboratory orders, and
all procedures.
Major Variables
ANTIEPILEPTIC DRUGS. The 16 study AEDs included carbamazepine (CBZ), ethotoin (EHN), ethosuximide (ESM),
felbamate (FBM), gabapentin (GBP), lamotrigine (LTG),
mephenytoin (MHT), mephobarbital (MPB), methsuximide
(MSM), phenobarbital (PB), phenytoin (PHT), primidone
(PRM), phensuximide (PSX), tiagabine (TGB), topiramate
(TPM), and valproic acid/valproate sodium (VPA).32 AEDs
not on the market during the study period were levetiracetam (LEV), zonisamide (ZSM), and oxcarbazepine (OXC).
AED use at admission was defined as a medication order for one or more AEDs recorded
at or within 2 days of NH entry by the Admissions Group.
For the Follow-up Cohort, AED use was defined as the first
medication order for one or more AEDs from 2 days after
admission to the end of the patient’s follow-up period. If the
first AED treatment during the follow-up period consisted of
two or more AEDs used simultaneously and orders were
dated within 2 days of one another, then multiple AEDs
were recorded as the incident medication.
ANTIEPILEPTIC DRUG USE.
EPILEPSY/SEIZURE INDICATION. Indication of epilepsy/seizure was based on documentation in one or more of five
sources: an ICD-9 code of 345.0, 780.3, or 779.0 in the
MDS or one of these ICD-9 codes or use of the words,
“convulsion,” “seizure,” or “epilepsy” in the PO. When both
an indication and an AED order were present, we assumed
that the medication was being used for seizure control.
Statistical Analysis
Logistic regression analysis was used to examine factors associated with the two response variables, (1) any AED use at
admission, and (2) incident AED use during the follow-up
period. Confounding by condition was controlled by entering epilepsy/seizure as the first factor in the regression model
and including two-way interactions between epilepsy/seizure
and all other factors. In all analyses, a minimum significance
level of 0.01 was used. Variables, defined within six groupings, were the following.
Patient characteristic variables included gender, age group at admission (65–
74, 75– 84, 85⫹ years) based on birth date, race/ethnicity
(American Indian/Alaskan native, Asian/Pacific Islander,
Black not of Hispanic origin, Hispanic, White not of Hispanic origin), educational level, alcohol and tobacco use
within the past year, month of admission, epilepsy/seizure
indication, living situation before admission, geographic location by U.S. Census region, month of admission, and
Medicare and Medicaid status.
PATIENT CHARACTERISTIC VARIABLES.
MUSCULOSKELETAL CONDITIONS. Musculoskeletal conditions included arthritis, hip fracture, amputation, osteoporosis, pathological bone fracture, and a summary score of activities of daily living.
CARDIOVASCULAR CONDITIONS. Cardiovascular conditions included arteriosclerotic heart disease, cardiac dysrhythmias, congestive heart failure, deep vein thrombosis, hypotension, hypertension, peripheral vascular disease, other
cardiovascular diseases.
NEUROLOGICAL CONDITIONS OTHER THAN EPILEPSY/SEIZURE. Neurological conditions (other than epilepsy or sei-
zure) included Alzheimer’s disease, aphasia, cerebral palsy,
cerebrovascular accident (stroke).
Psychiatric and mood
conditions included anxiety disorder, depression, manic depression (bipolar), schizophrenia, dementia, and the score on
the MDS Cognition Scale (MDS-COGS),33 a validated measure of cognitive performance in NH residents34 (not applicable to a comatose patient).
PSYCHIATRIC/MOOD CONDITIONS.
Other conditions included pain status, syncope, use of devices and restraints, accidents.
Because of the possibility of multicolinearity, initial models were fit separately using variables within each of the six
groupings, with “any AED” as the response variable. All twoway interactions were tested. If an interaction was statistically
significant, contributing factors were included in the model
whether significant or not. Significant factors at level 0.01
from each initial model then were entered in the final regression model to explore their combined association with AED
use. Results of the final models are reported. Reported p values were not adjusted for multiple comparisons.
Results
Overview
Of the 10,318 NH residents in the Admissions Group,
802 had AED use at entry, with a prevalence of
7.77%. Of these, 57.7% had an epilepsy/seizure indication (Fig 1). In the Follow-up Cohort (N ⫽ 9,516),
260 (2.73%) were given an AED, of whom 21.2% had
an epilepsy/seizure indication. Regardless of AED use,
5.83% (N ⫽ 602) of the Admissions Group and
1.47% (N ⫽ 140) of the Follow-up Cohort had an
epilepsy/seizure indication at admission.
Baseline characteristics by AED status are shown in
Table 1 for the Admissions Group and Table 2 for the
Follow-up Cohort. Approximately half (52%) of the
Admissions Group were admitted for short-term postacute care. The Follow-up Cohort had similar distributions by gender, race/ethnicity, age group, and sources
of payment (Medicare, Medicaid, or both).
Antiepileptic Drugs Used at Admission
Commonly used AEDs at admission were PHT (N ⫽
414; 52% of the 802 admissions with AED use), VPA
Fig 1. Percentage of elderly with documentation of epilepsy/seizure (Epi/Sz) indication by antiepileptic drug (AED) use at admission
and AED use initiated during follow-up period.
Garrard et al: AEDs Nursing Home Admissions
77
Table 1. Baseline Characteristics of Admissions Group at Entry to Nursing Homes by AED Use at Admission
Characteristic
Sample size
Gender
Female
Male
Age group (yr)
65–74
75–84
ⱖ85
Race/ethnicity
Native American
Asian/Pacific Islander
Black, not Hispanic
Hispanic
White, not Hispanic
Education
No schooling
No high school degree
High school graduate
Trade school, some college
College degree
Graduate school
Data missing
Prior residence
Private home
Acute care hospital
Other NH
Other
Data missing
Geographic region
Northeastern
Southern
Midwestern
Western
Data missing
Payment source
Medicare (per diem only)
Yes
No
Data missing
Medicaid recipient
Yes
No
Data missing
Documentation of epilepsy/seizure
Yes
No
Bipolar depression
Yes
No
Cognitive performance (MDS-COGS)
(score: degree of impairment)
0–1: Intact to mild
2–4: Mild to moderate
5–8: Moderate to severe
9–10: Severe to very severe
On AED at
Admission, n (%)
Not on AED at
Admission, n (%)
Total (N)
802 (7.8%)
9,516 (92.2%)
10,318
475 (7%)
327 (9%)
6,290 (93%)
3,226 (91%)
6,765
3,553
248 (13%)
390 (9%)
164 (4%)
1,730 (87%)
4,096 (91%)
3,690 (96%)
1,978
4,486
3,854
5 (7%)
10 (5%)
89 (10%)
13 (6%)
685 (8%)
65 (93%)
193 (95%)
827 (90%)
201 (94%)
8,230 (92%)
70
203
916
214
8,915
11 (9%)
308 (7%)
310 (8%)
105 (7%)
44 (8%)
22 (9%)
2
107 (91%)
3,910 (93%)
3,356 (92%)
1,427 (93%)
479 (92%)
224 (91%)
13
118
4,218
3,666
1,532
523
246
15
71 (6%)
599 (7%)
68 (12%)
63 (12%)
1
1,070 (94%)
7,474 (93%)
497 (88%)
473 (88%)
2
1,141
8,073
565
536
3
86 (8%)
352 (9%)
182 (7%)
181 (7%)
1
981 (92%)
3,685 (91%)
2,337 (93%)
2,511 (93%)
2
1,067
4,037
2,519
2,692
3
399 (7%)
400 (8%)
3
4,997 (93%)
4,509 (92%)
10
5,396
4,909
13
194 (9%)
582 (7%)
26
1,856 (91%)
7,382 (93%)
278
2,050
7,964
304
462 (77%)
340 (4%)
140 (23%)
9,376 (96%)
602
9,716
40 (35%)
761 (8%)
73 (65%)
9,434 (92%)
113
10,195
241 (6%)
209 (7%)
258 (9%)
91 (14%)
3,665 (94%)
2,669 (93%)
2,611 (91%)
562 (86%)
3,906
2,878
2,869
653
AED ⫽ antiepileptic drug; NH ⫽ nursing home; MDS-COGS ⫽ Minimum Data Set Cognition Scale.
(N ⫽ 154; 19%), GBP (N ⫽ 132; 17%), and CBZ
(N ⫽ 84; 11%). Collectively, these drugs constituted
89% of all of the AED orders at admission (Table 3).
78
Annals of Neurology
Vol 54
No 1
July 2003
When stratified by epilepsy/seizure indication, only
three drugs, PHT, CBZ, and PB, had a majority of
users with this indication. The number of AEDs orders
Table 2. Characteristics of Follow-up Cohort at Entry to Nursing Homes by Initiation of AED during Follow-up
Characteristic
Sample size
Gender
Female
Male
Age group (yr)
65–74
75–84
ⱖ85
Race/ethnicity
Native American
Asian/Pacific Islander
Black, not Hispanic
Hispanic
White, not Hispanic
Education
No schooling
No high school degree
High school graduate
Trade school, some college
College degree
Graduate school
Data missing
Prior residence
Private home
Acute care hospital
Other NH
Other
Geographic region
Northeastern
Southern
Midwestern
Western
Data missing
Payment source
Medicare (per diem only)
Yes
No
Data missing
Medicaid
Yes
No
Data missing
Documentation of: epilepsy/seizure
Yes
No
Bipolar depression
Yes
No
Cognitive performance (MDS-COGS)
(score: degree of impairment)
0–1: Intact to mild
2–4: Mild to moderate
5–8: Moderate to severe
9–10: Severe to very severe
Incident AED during
Follow-up, n (%)
No Incident AED
during Follow-up, n (%)
Total (N)
260 (3%)
9,256 (97%)
9,516
155 (3%)
105 (3%)
6,135 (97%)
3,121 (97%)
6,290
3,226
73 (4%)
120 (3%)
67 (2%)
1,657 (96%)
3,976 (97%)
3,623 (98%)
1,730
4,096
3,690
3 (5%)
1 (⬍1%)
30 (4%)
4 (2%)
222 (3%)
62 (95%)
192 (99%)
797 (96%)
197 (98%)
8,008 (97%)
65
193
827
201
8,230
3 (3%)
87 (2%)
117 (4%)
36 (3%)
11 (2%)
6 (3%)
0
104 (97%)
3,823 (98%)
3,239 (96%)
1,391 (97%)
468 (98%)
218 (97%)
13
107
3,910
3,356
1,427
479
224
13
33 (3%)
184 (3%)
24 (5%)
19 (4%)
1,037 (97%)
7,290 (97%)
473 (95%)
454 (96%)
1,070
7,474
497
473
31 (3%)
113 (3%)
58 (3%)
58 (2%)
0
950 (97%)
3,572 (97%)
2,279 (97%)
2,453 (98%)
2
981
3,685
2,337
2,511
2
127 (3%)
133 (3%)
0
4,870 (97%)
4,376 (97%)
10
4,997
4,509
10
64 (3%)
187 (3%)
9
1,792 (97%)
7,195 (97%)
269
1,856
7,382
278
55 (39%)
205 (2%)
85 (61%)
9,171 (98%)
140
9,376
8 (11%)
260 (3%)
65 (89%)
9,247 (97%)
73
9,507
68 (2%)
69 (3%)
102 (4%)
21 (4%)
3,597 (98%)
2,600 (97%)
2,509 (96%)
541 (96%)
3,665
2,669
2,611
562
AED ⫽ antiepileptic drug; NH ⫽ nursing home; MDS-COGS ⫽ Minimum Data Set Cognition Scale.
Garrard et al: AEDs Nursing Home Admissions
79
Table 3. Percentage and Number of Medication Orders for an AED Taken at Admission or Initiated during Follow-up Period by
Documentation of Epilepsy/Seizure Indication
AEDs Initiated during Follow-up (312
AED orders for 260 subjects),
Epilepsy/Seizure Indication
AED Orders Present at Admission (882
AED orders for 802 subjects),
Epilepsy/Seizure Indication
AED
Phenytoin
Valproale sodium
Gabapentin
Carbamazepine
Phenobarbital
Primidone
Other
Any AED
Present (%)
Absent (%)
n
Present
Absent
n
79
40
21
63
82
35
75
60
21
60
79
37
18
65
25
40
414
154
132
84
60
34
4
882*
40
8
13
11
70
17
0
22
60
92
87
89
30
83
100
78
100
87
80
28
10
6
1
312a
The number of medication orders for AEDs at admission (N ⫽ 882) exceeds the total number of admissions with AED use (N ⫽ 802)
because of simultaneous use of two or more AEDs. The number of AEDs initiated during follow-up (N ⫽ 312) exceeds the number of subjects
(N ⫽ 260).
a
There were no users of the following AEDs at either admission or follow-up: ethotoin, ethosuximide, felbamate, methsuximide, phensuximide,
and tiagabine. The following “other” AEDs were ordered for only one or two individuals: mephenytoin, topiramate, and lamotrigine.
AED ⫽ antiepileptic drug.
(N ⫽ 882) exceeded total number of admissions with
AED use (N ⫽ 802) because of poly-AED orders.
Antiepileptic Drugs Initiated during Follow-up
During the follow-up period, commonly initiated
AEDs were the same as those at admission, but not at
the same rates: PHT (N ⫽ 100; 39% of 260 residents
with AED initiation), VPA (N ⫽ 87; 34%), GBP
(N ⫽ 80; 31%), and CBZ (N ⫽ 28; 11%). Poly-AED
use accounted for differences between number of AED
orders (N ⫽ 312) and number of residents with an
AED initiation (N ⫽ 260). Although 7% of the AED
orders at admission were for PB, this drug accounted
for only 3% of all AED orders initiated during followup. Among the six AEDs most commonly initiated
during follow-up, only PB had an epilepsy/seizure indication by most (see Table 3). These results raised additional questions about the initiation of AEDs during
the follow-up period, each of which was the subject of
a subanalysis.
Were
most of the AEDs during follow-up initiated immediately after admission, with the possibility that epilepsy/
seizure was either under diagnosed at admission or
there was an increase in need for seizure management
within a few days of admission? AED incidence rates
were calculated for each week, adjusted for NH attrition. Results showed that AED initiations were distributed throughout the follow-up period, regardless of indication (Fig 2). Among residents with an indication
(N ⫽ 55), two thirds (67%) were given the drug
within 4 weeks of admission compared with half
(57%) of those without an indication (N ⫽ 205).
TIMING OF ANTIEPILEPTIC DRUG INITIATIONS.
80
Annals of Neurology
Vol 54
No 1
July 2003
NEW DOCUMENTATION OF EPILEPSY/SEIZURE. Was an
indication documented for the first time during the
follow-up period, and an AED order initiated for that
reason? Of the 52 residents with a new epilepsy indication, 33 (63%) had orders for one or more study
AEDs initiated. Dates of AED orders were distributed
throughout the follow-up period. Nineteen of the 52
cases not treated with a study AED during the
follow-up period had instructions in the POs such as
“monitoring for seizures,” or “seizure precautions.”
EPILEPSY/SEIZURE INDICATION BUT NO ANTIEPILEPTIC
DRUG ORDERS. Were there residents with an indica-
tion but no AED treatment? Approximately one fifth
(23%; N ⫽ 140) of admissions with an indication
(N ⫽ 602) did not have POs for an AED at entry.
These 140 people became part of the Follow-up Cohort and approximately half (N ⫽ 74) eventually were
treated. By the end of the study period, 85 residents
with an indication at either admission (N ⫽ 66) or
during follow-up (N ⫽ 19) did not have an AED order. All POs for this subgroup were examined to determine if any medications other than the 16 study
AEDs were being used for seizure management. Results
showed that 12 residents (14%) had an order for either
lorazepam or clonazepam, of which four had a seizure
indication on the orders. Thus, 73 of the remaining 85
orders (86%) had no PO for drugs prescribed for management of seizures.
Factors Associated with Use or Initiation of
Antiepileptic Drugs
ANTIEPILEPTIC DRUG USE AT ADMISSION. In the regression analysis of the Admissions Group, four factors
Fig 2. Antiepileptic drug (AED) initiation rate (%) by week since admission by epilepsy/seizure (Epi/Sz) indication.
were associated with AED use: epilepsy/seizure, bipolar
depression, age group, and cognitive performance. Epilepsy/seizure interacted with age group and cognitive
performance, and in both interactions, epilepsy/seizure
was the dominant factor (Table 4).
The epilepsy/seizure by age group interaction indicates that when epilepsy/seizure was not present, AED
use declined significantly as age increased. Compared
with the reference group of residents who were 65 to
74 years old with no epilepsy/seizure, those 75 to 84
years old with no epilepsy were one third less likely to
be taking an AED at admission (odds ratio [OR], 0.67;
p ⬍ 0.01), and the oldest group (85 years and older)
was two thirds less likely (OR, 0.33; p ⬍ 0.0000).
Thus, there was an inverse relationship between AED
use and age group if epilepsy/seizure was not present.
For those with an epilepsy/seizure indication, however, the odds of AED use in all three age groups were
extremely high compared with the reference group: 65
to 74 years: OR, 81.57 ( p ⬍ 0.0001); 75 to 84 years:
OR, 81.04 ( p ⬍ 00001), and 85 years and older: OR,
90.97 ( p ⬍ 0.0001). When epilepsy/seizure was
present, AED use in these three age groups did not
differ statistically.
Epilepsy/seizure indication and cognitive performance (MDS-COGS) also had a significant interaction. For every one point higher on the MDS-COGS
(in the direction of worsening cognitive performance),
admissions without an indication were 9% more likely
to be using an AED (OR, 1.09; p ⬍ 0.0001). Among
those with an epilepsy/seizure indication, however, a
one-point higher MDS-COGS was associated with a
2% decrease in AED use (OR, 0.98; p ⬍ 0.0001). As
above, the odds of AED use associated with the epilepsy/
seizure indication were extremely high, whether the
score on the MDS-COGS was 0 (OR, 81.57; p ⬍
0.0000), indicating no cognitive impairment, or 10
(OR, 29.64; p ⬍ 0.0001), indicating severe cognitive
impairment.
Bipolar depression was independently associated
with AED use (OR, 10.87; p ⬍ 0.0001). Admissions
with this indication were 11 times more likely to be
using an AED upon entry compared with those without bipolar depression. This regression model was
based on use of “any AED,” and we further examined
which specific AEDs were used by admissions with either epilepsy/seizure only (N ⫽ 585), bipolar depression only (N ⫽ 105), or both (N ⫽ 8) as an indication. The four most commonly used AEDs (including
poly-AED use) for admissions with epilepsy/seizure
were PHT (54% of the 585 residents; N ⫽ 315), VPA
(10%; N ⫽ 57), CBZ (9%; N ⫽ 52), and GBP (5%;
N ⫽ 27). Among the 105 residents with bipolar depression, commonly used AEDs were VPA (21%; N ⫽
22), GBP (8%; N ⫽ 8), and CBZ (4%; N ⫽ 4). PHT
was not used by any residents with a bipolar indication. Eight residents with both indications used VPA
(N ⫽ 4), PHT (N ⫽ 4), and CBZ (N ⫽ 1). The
remainder of the admissions group with AED use had
neither indication (38%; N ⫽ 308). Although other
Garrard et al: AEDs Nursing Home Admissions
81
Table 4. Factors Associated with AED Use at Admission to Nursing Home: Results of Final Logistic Regression Model
OR
95% CI
Statistical
Significance (p)
1.00 (reference group)
0.67
0.33
81.57
81.04
90.97
0.51–0.87
0.24–0.45
50.90–130.71
55.41–118.52
51.25–161.46
⬍0.01
⬍0.0001
⬍0.0001
⬍0.0001
⬍0.0001
1.09
0.98
10.87
1.05–1.13
0.98–0.98
7.03–16.82
⬍0.0001
⬍0.0001
⬍0.0001
Factor
Epi/Sz, age group (yr)
Epi/Sz (no), 65–74
Epi/Sz (no), 75–84
Epi/Sz (no), 85⫹
Epi/Sz (yes), 65–74
Epi/Sz (yes), 75–84
Epi/Sz (yes), 85⫹
Epi/Sz, MDS-COGSa
Epi/Sz (no), MDS-COGS (one-point increase)
Epi/Sz (yes), MDS-COGS (one-point increase)
Bipolar depression (present)
a
The MDS-COGS is a continuous measure ranging from 0 (no cognitive impairment) to 10 (maximum cognitive impairment); that is, each
one-point increase in the score is equivalent to one point worse in cognitive performance.
AED ⫽ antiepileptic drug; OR ⫽ odds ratio; CI ⫽ confidence interval; Epi/Sz ⫽ epilepsy/seizure; MDS-COGS ⫽ Minimum Data Set
Cognitive Performance Scale.
diagnoses, for example, neuropathic pain, headaches,
behavioral problems, were included in the initial analyses, they did not reach statistical significance in the
final analysis.
INITIATION OF ANTIEPILEPTIC DRUGS DURING FOLLOWUP. Five factors were associated with initiation of
shown in Table 5) when cognitive performance was
normal (MDS-COGS ⫽ 0), the odds of AED use associated with PVD was 3.26 ( p ⬍ 0.0001), but, at its
worst (MDS-COGS ⫽ 10), the odds of AED use associated with PVD was 0.29 ( p ⬍ 0.0001). The PVD
indication was the dominant factor in this pairwise interaction.
AEDs during the follow-up period: epilepsy/seizure, bipolar depression, age group, cognitive performance
(MDS-COGS), and peripheral vascular disease (PVD).
The multivariate analysis shows that there was a significant interaction between PVD and MDS-COGS (Table 5).
Epilepsy/seizure had the strongest independent association ( p ⬍ 0.00001) with AED use during follow-up
of any of the significant factors in this analysis. If residents had this indication, the odds were over 25 to 1
that they would be given an AED during the follow-up
period ( p ⬍ 0.0001). Bipolar depression was also independently and significantly associated with an incident AED during follow-up. Residents with this indication were over four times more likely to be taking an
AED (OR, 4.85; p ⬍ 0.0001) than those without bipolar depression.
Age group in the Follow-up Cohort was inversely
related to having an AED initiated after admission.
Compared with the reference group of young-old
(65–74 years), those in the old group (75– 84) were
one third less likely to have an AED initiated (OR,
0.68; p ⬍ 0.05), and the oldest old (85 years and
older) were half as likely (OR, 0.47; p ⬍ 0.0001).
Among residents without PVD, a one-point higher
MDS-COGS (indicating lower cognitive performance)
was associated with a 13% higher likelihood of using
an AED (OR, 1.13; p ⬍ 0.00001). Among those with
PVD, there was not significant association with MDSCOGS (see Table 5). Upon further examination (not
Discussion
This study is significant in being the first to our
knowledge (1) to examine issues of whether NH patients enter the facilities with an AED or are given
AEDs after entry; (2) to describe which AEDs are being prescribed for NH elderly in current clinical practice; and (3) to identify factors associated with AED
use by a nationwide sample of NH residents based on
multivariate analysis.
The timing of the initiation of AEDs after admission
is described for up to 3 months for each individual;
such a follow-up has not been described previously.
The results of this kind of study are useful not only to
other researchers concerning incidence and prevalence
rates, but also to health planners and policy makers in
long-term care planning and health providers who need
data about clinical profiles of NH admissions.
This data-rich study of a very large NH population
corroborates previous findings, with added details
about which AEDs are used. Although an AED prevalence of 10.5 to 11%6,35 has been reported previously,
this study further shows that most NH patients enter
with AED treatment (prevalence of 7.77%), whereas
an additional 2.73% are given AEDs after admission.
Thus, AED treatment is not static. Results also show
that a significant number of people (one fifth of admissions) with an epilepsy/seizure indication are not
being treated. Upon follow-up, more than three fifths
of the residents with the indication had not had an
82
Annals of Neurology
Vol 54
No 1
July 2003
Table 5. Factors Associated with AED Use during the Postadmission Follow-up Period: Results of Final Logistic Regression Model
OR
95% CI
Statistical
Significance (p)
25.48
4.85
17.44–37.23
2.28–10.33
⬍0.0001
⬍0.0001
1.00 (reference group)
0.68
0.47
0.50–0.94
0.33–0.67
1.13
0.89
1.08–1.18
0.77–1.03
⬍0.05
⬍0.0001
⬍0.0018
⬍0.0001
0.11
Factor
Epilepsy/seizure (present)
Bipolar depression (present)
Age group (yr)
65–74
75–84
85⫹
PVD MDS-COGS
PVD (no), MDS-COGS (one-point increase)
PVD (yes), MDS-COGS (one-point increase)
The MDS-COGS is a continuous measure ranging from 0 (no cognitive impairment) to 10 (maximum cognitive impairment); that is, each
one-point increase in the score is equivalent to one point worse in cognitive performance.
AED ⫽ antiepileptic drug; OR ⫽ odds ratio; CI ⫽ confidence interval; PVD ⫽ peripheral vascular disease; MDS-COGS ⫽ minimum Data
Set Cognition Scale.
AED order by the end of the study period. These findings raise issues about the basis of an epilepsy/seizure
diagnosis and the appropriateness of treating or not
treating.
Factors Associated with Antiepileptic Drug Use
Within age group, the association between AED use
and epilepsy/seizure was extremely high, as could be
expected. At admission, AED use was inversely related
to age if epilepsy was not present but was nearly constant if epilepsy/seizure was present. During follow-up,
age and initiation of AED use were also inversely related regardless of epilepsy/seizure status. The literature
shows that incidence and prevalence rates of epilepsy
increase with age (across the entire age span) in the
community-dwelling population from approximately
0.68% for people 65 to 74 years old36 to 1.50% for
those 75 years and older.3 Do the results of this study
of this NH admissions group suggest under treatment
of epilepsy/seizure among the oldest old in NHs? This
question deserves further study; however, the results
from the Admissions Group analysis suggest that health
providers do respond to the need for AED treatment
for the oldest old NH admissions when epilepsy/seizure
indications are present.
This interpretation may not apply to the Follow-up
Cohort, however. Despite population trends showing
an increase in epilepsy in the older age groups, in this
study the incidence of AED use during follow-up declined as age group increased, independent of the presence of epilepsy/seizure indications. Furthermore, none
of the two-way interactions between age group and all
other factors in the final model were statistically significant. It is possible that AEDs were initiated as treatment for conditions other than epilepsy/seizure for residents in the Follow-up Cohort, and more of those
conditions were in the younger age groups. Note the
differences between use of VPA by admissions (19%)
versus follow-up residents (34%) and, similarly, the use
of GBP for admissions (11%) versus follow-up residents (31%). VPA often is used for treatment of psychiatric conditions, and GBP for treatment of pain.
Thus, the proportional differences for these AEDs before and after admission may be caused by increased
recognition or diagnosis of these underlying conditions
after admission.
Among admissions, and to a lesser extent among
those in the Follow-up Cohort, the strong positive association between AED use and bipolar depression suggests that AEDs are being used as a treatment option,
possibly as adjunctive therapy for the manic phase.37–39
Declining cognitive performance was associated with
AED use in both groups; however, this factor did not
operate independently. Among admissions, AED use
increased by 9% for each point on the MDS-COGS
scale in the worsening direction if epilepsy/seizure was
not present. One interpretation might be that the
AEDs are being used to treat behavioral problems,
which, in turn, are associated with declining cognitive
performance, perhaps caused by dementia. Among
those with epilepsy, AED use, however, was slightly
lower for residents with higher (worse) MDS-COGS
scores.
Among follow-up residents, cognitive performance
was also linked to PVD. The initiation of an AED was
more than three times greater if PVD was present at
admission for those with no cognitive impairment but
was more than three times lower for those with severe
cognitive impairment. When PVD was absent, there
was a modest but significant increase in initiation of an
AED of 13% if the MDS-COGS was worse by one
point. In contrast, when PVD was present, there was a
modest decrease in AED initiation by 11% for every
one point worse in MDS-COGS. The clinical role of
this statistical interaction between PVD and cognitive
impairment in the use or nonuse of AEDs after admis-
Garrard et al: AEDs Nursing Home Admissions
83
sion is not clear, and further research is needed about
this relationship and what it means clinically.
Factors Not Associated with Antiepileptic Drug Use
Although statistically significant results from multivariate analyses are important in understanding the pharmacoepidemiology of AED use/initiation, equally important findings are those that were not statistically
significant. Based on previous research, we hypothesized that factors such as gender,3,40 race,41 educational
level,42 geographic region,41 financial dependency (ie,
Medicaid recipients),43 alcohol use/abuse,43 and postacute care might be associated with AED use. We also
hypothesized that other conditions would be highly related to AED use, such as neuropathic pain,6,7 dementia or Alzheimer’s disease,44 – 46 and injury.47– 49 In this
study, all of these conditions were included in the analyses, but none were statistically significant in the final
model.
Strengths and Weaknesses
The strengths of this study included the large numbers
of NHs and subjects sufficient to achieve statistical
power in examining rare events. Although the NHs in
this sample were distributed throughout the United
States, this was not a probability-based sample of all
free-standing NHs (N ⫽ 12,859) in the United States
in 1999. Therefore, generalization is not possible. Another potential weakness was the definition of AED use
or initiation. AED data were based on POs, and,
whereas the order itself is believed to be valid, it is
possible that a small percentage of medications may
have been ordered, but not administered, resulting in a
slight overestimate of drug use or initiation. A third
possible weakness was our assumption that if both an
epilepsy/seizure indication and an AED order were
present, then the patient was being treated for seizure
control. Alternatively, the condition might not have
been active and/or the AED was used for treatment of
some other condition. Given these possibilities, and the
inability to query health providers about their reasons
for the prescription, our measure may not accurately
reflect all AEDs used for seizure management. Finally,
a limitation of this study is our reliance on the diagnosis of epilepsy/seizure in the record. The diagnostic
criteria used to make this diagnosis by different health
care providers are not known and may not be uniform.
Thus, this study should not be regarded as a study of
epilepsy/seizures but rather an epidemiological study of
drugs defined here as AEDs.
Research Needed
More research is needed about AED use in NHs. Such
research would include studies of relationships between
AED use and patient outcomes, for example, falls, hip
fractures, changes in cognitive performance, and activ-
84
Annals of Neurology
Vol 54
No 1
July 2003
ities of daily living. Little is known about the use of
mono- versus poly-AED use and their effects. No studies have examined dosing of AEDs, taking into account
the pharmacokinetics of elderly with compromised kidney and hepatic functioning. All of these studies are
possible with secondary source data; however, other
studies using primary source data, including randomized control trials, also are needed to improve the care
of the growing numbers of elderly in long-term care.
This research was supported by an NIH grant to the Epilepsy Clinical Research Program (NIH, NINDS P50-NS16308, NINDS
2P50-NS 16308-22A1, I.E.L.).
We are grateful to the leadership and staff of Beverly Enterprises,
Inc., for their cooperation in making the database available for this
study. They are not responsible for the content or the opinions expressed in this article. We also greatly appreciate the support from
and expertise of the late Dr F. Annegers, who was a member of this
project’s advisory group.
Portions of this article were presented at national and international
meetings of the American Society of Epilepsy, Los Angeles, CA,
12/00; American Pharmaceutical Outcomes Research Conference,
Seattle, WA, 5/00; and the International Society of Pharmacoepidemiology, Barcelona, Spain, 7/00.
References
1. Hauser WA. Seizure disorders: the changes with age. Epilepsia
1992;33(suppl 4):S6 –S14.
2. Cockerell OC, Eckle I, Goodridge DM, et al. Epilepsy in a
population of 6000 re-examined: secular trends in first attendance rates, prevalence, and prognosis. J Neurol Neurosurg Psychiatry 1995;58:570 –576.
3. Hauser WA, Annegers JF, Rocca WA. Descriptive epidemiology of epilepsy: contributions of population-based studies from
Rochester, Minnesota. Mayo Clin Proc 1996;71:576 –586.
4. Wallace H, Shorvon S, Tallis R. Age-specific incidence and
prevalence rates of treated epilepsy in an unselected population
of 2,052,922 and age-specific fertility rates of women with epilepsy. Lancet 1998;352:1970 –1973.
5. Reynolds EH. Early treatment and prognosis of epilepsy. Epilepsia 1987;28:97–106.
6. Cloyd JC, Lackner TE, Leppik IE. Antiepileptics in the elderly.
Pharmacoepidemiology and pharmacokinetics. Arch Family
Med 1994;3:589 –598.
7. Ross EL. The evolving role of antiepileptic drugs in treating
neuropathic pain. Neurology 2000;55:S41–S46.
8. Di Trapani G, Mei D, Marra C, et al. Gabapentin in the prophylaxis of migraine: a double-blind randomized placebocontrolled study. Clin Ter 2000;151:145–148.
9. Calabrese J, Bowden C, Sachs G, et al. A double-blind placebocontrolled study of lamotrigine monotherapy in outpatients
with bipolar I depression. Lamictal 602 Study Group. J Clin
Psychiatry 1999;60:79 – 88.
10. Harden CL, Lazar LM, Pick LH, et al. A beneficial effect on
mood in partial epilepsy patients treated with gabapentin. Epilepsia 1999;40:1129 –1134.
11. Gleason RP, Schneider LS. Carbamazepine treatment of agitation in Alzheimer’s outpatients refractory to neuroleptics. J Clin
Psychiatry 1990;51:115–118.
12. de la Court A, Breteler MM, Meinardi H, et al. Prevalence of
epilepsy in the elderly: the Rotterdam Study. Epilepsia 1996;
37:141–147.
13. Rochat P, Hallas J, Gaist D, Friis ML. Antiepileptic drug
utilization: a Danish prescription database analysis. Acta Neurol
Scand 2001;104:6 –11.
14. Moore T, White A. Drug utilization and MDS-based outcomes
for elderly nursing home residents—briefing paper. Cambridge,
MA: Abt Associates, 1998.
15. Garrard J, Hardie NA, Harms SL, et al. Factors associated with
existing and newly initiated antiepileptic drug use in nursing
home admissions. Paper presented at: Sixteenth Annual International Conference on Pharmacoepidemiology; 2000; Barcelona, Spain.
16. Tromp AM, Smith JH, Deeg DJ, et al. Predictors for falls and
fractures in the Longitudinal Aging Study Amsterdam. J Bone
Min Res 1998;13:1932–1939.
17. Aldenkamp AP, Baker GA. The Neurotoxicity Scale. II. Results
of a patient-based scale assessing neurotoxicity in patients with
epilepsy. Epilepsy Res 1997;27:165–173.
18. Field TS, Gurwitz JH, Avorn J, et al. Risk factors for adverse
drug events among nursing home residents. Arch Intern Med
2001;161:1629 –1634.
19. Annegers JF, Hauser WA, Lee JR, Rocca WA. Incidence of
acute symptomatic seizures in Rochester, Minnesota,
1935–1984. Epilepsia 1995;36:327–333.
20. Lackner TE, Cloyd JC, Thomas LW, Leppik IE. Antiepileptic
drug use in nursing home residents: effect of age, gender, and
comedication on patterns of use. Epilepsia 1998;39:1083–1087.
21. Garrard J, Cloyd J, Gross C, et al. Factors associated with antiepileptic drug use among elderly nursing home residents. J
Gerontol A Biol Sci Med Sci 2000;55:M384 –M392.
22. Vickery K. Top nursing facility chains. Provider 2000;26:
37– 41.
23. Abelson R. Bringing discipline (and scorecards) to nursing
homes. New York Times. July 7, 2002:1, 10.
24. Garrard J, Buchanan JL, Ratner ER, et al. Differences between
nursing home admissions and residents. J Gerontol 1993;48:
S301–S309.
25. Liu K, McBride T, Coughlin T. Risk of entering nursing
homes for long versus short stays. Med Care 1994;32:315–327.
26. Spence DA, Wiener JM. Nursing home length of stay patterns:
results from the 1985 National Nursing Home Survey. Gerontologist 1990;30:16 –20.
27. Fries BE, Schroll M, Hawes C, et al. Approaching crossnational comparisons of nursing home residents. Age Ageing
1997;26(suppl 2):13–18.
28. Rantz MJ, Popejoy L, Mehr DR, et al. Verifying nursing home
care quality using Minimum Data Set quality indicators and
other quality measures. J Nurs Care Qual 1997;12:54 – 62.
29. Snowden M, McCormick W, Russo J, et al. Validity and responsiveness of the Minimum Data Set. J Am Geriatr Soc
1999;47:1000 –1004.
30. Casten R, Lawton MP, Parmelee PA, Kleban MH. Psychometric characteristics of the minimum data set. I. Confirmatory factor analysis. J Am Geriatr Soc 1998;46:726 –735.
31. Lawton MP, Casten R, Parmelee PA, et al. Psychometric characteristics of the Minimum Data Set. II. Validity. J Am Geriatr
Soc 1998;46:736 –744.
32. Leppik IE. Issues in the treatment of epilepsy. Epilepsia 2001;
42:1– 6.
33. Hartmaier SL, Sloane PD, Guess HA, Koch GG. The MDS
Cognition Scale: a valid instrument for identifying and staging
nursing home residents with dementia using the Minimum
Data Set. J Am Geriatr Soc 1994;42:1173–1179.
34. Gruber-Baldini AL, Zimmerman SI, Mortimore E, Magaziner
J. The validity of the Minimum Data Set in measuring the
cognitive impairment of persons admitted to nursing homes.
J Am Geriatr Soc 2000;48:1601–1606.
35. Schachter SC, Cramer GW, Thompson GD, et al. An evaluation of antiepileptic drug therapy in nursing facilities. J Am
Geriatr Soc 1998;46:1137–1141.
36. Hauser WA, Annegers JF, Kurland LT. Prevalence of epilepsy
in Rochester, Minnesota: 1940 –1980. Epilepsia 1991;32:
429 – 445.
37. McElroy SL, Keck PE, Pope HG, Hudson JI. Valproate in the
treatment of bipolar disorder: literature review and clinical
guidelines. J Clin Psychopharmacol 1992;12(1 suppl):42S–52S.
38. Chengappa KN, Rathore D, Levine J, et al. Topiramate as
add-on treatment for patients with bipolar mania. Bipolar Disord 1999;1:42–53.
39. McElroy SL, Suppes T, Keck PE, et al. Open-label adjunctive
topiramate in the treatment of bipolar disorders. Biol Psychiatry
2000;47:1025–1033.
40. Reife R, Pledger G, Wu SC. Topiramate as add-on therapy:
pooled analysis of randomized controlled trials in adults. Epilepsia 2000;41(suppl 1):S66 –S71.
41. Harms SL, Hardie N, Garrard J, et al. Association of race and
geographic region with antiepileptic drug use among elderly
nursing home residents. Paper presented at: Annual Meeting of
the American Epilepsy Society; 2000; Los Angeles, CA.
42. Neugebauer R, Paik M, Hauser WA, et al. Stressful life events
and seizure frequency in patients with epilepsy. Epilepsia 1994;
35:336 –343.
43. Krumholz A, Grufferman S, Orr ST, Stern BJ. Seizures and
seizure care in an emergency department. Epilepsia 1989;30:
175–181.
44. Faught E. Epidemiology and drug treatment of epilepsy in elderly people. Drugs Aging 1999;15:255–269.
45. Hauser WA, Morris ML, Heston LL, Anderson VE. Seizures
and myoclonus in patients with Alzheimer’s disease. Neurology
1986;36:1226 –1230.
46. Thomas RJ. Seizures and epilepsy in the elderly. Arch Intern
Med 1997;157:605– 617.
47. Neufeld MY, Vishne T, Chistik V, Korczyn AD. Life-long history of injuries related to seizures. Epilepsy Res 1999;34:
123–127.
48. Stephen LJ, Brodie MJ. Epilepsy in elderly people. Lancet
2000;355:1441–1446.
49. Buck D, Baker GA, Jacoby A, et al. Patients’ experiences of
injury as a result of epilepsy. Epilepsia 1997;38:439 – 444.
Garrard et al: AEDs Nursing Home Admissions
85
Документ
Категория
Без категории
Просмотров
2
Размер файла
190 Кб
Теги
drug, nursing, home, admission, use, antiepileptic
1/--страниц
Пожаловаться на содержимое документа