Transcript Patterns of Benzodiazepine Use in the Elderly and Risk of Injury

New users of benzodiazepines:
implications for elder patient safety
G. Bartlett, PhD
Family Medicine
McGill University
Outline








Benzodiazepine use in the elderly
Objectives of study
Methods
Population & data sources
Results – new users vs non-users
Results – predictors of new use
Conclusions, Limitations
Future Directions
Benzodiazepines - review

Sedative/hypnotics:



hypnotic, anxiolytic, anticonvulsant, muscular
relaxant, amnesic
high efficacy, rapid onset of action, low toxicity
unique among psychotropics for multiple
indications and relative safety compared with
other sedative/hypnotics
Concerns about Benzodiazepines

psychomotor impairment

paradoxical excitement


tolerance, dependence and
withdrawal effects with long term
use
injuries from falls
Concerns about Benzodiazepines
for the Elderly
Due to changes that occur with normal
aging, elderly demonstrate increased
sensitivity to:




psychomotor impairment
memory impairment
rebound or withdrawal effects
interactions with other
medications/conditions
Use of Benzodiazepines:
Why the Elderly?




insomnia can be a “pathological”
feature associated with age
anxiety due to other illnesses
more likely to suffer acute grief
reactions
fewer complications with
benzodiazepines than with tricyclic
anti-depressants and anti-psychotics
Why are we still discussing
benzodiazepine safety?



Risk from injuries from falls for benzodiazepines
still in dispute
Physicians may be prescribing benzodiazepines
perceived to be safer to higher risk patients
Pre-existing risk factors may be cause
confounding in published studies
What risk factors for falls are present before a
benzodiazepine is prescribed?
Methods

All patients >65 years with no benzodiazepine
script in baseline year

Risk factors for falls assessed in baseline year:







age & sex
clinical characteristics
disabilities & impairments
prior hospitalizations
prior health care use
use of other prescription medication.
5 years of follow-up until first benzodiazepine
script dispensed – product name identified
Benzodiazepines available in QC














Triazolam (Halcion)
Midazolam (Versed) – IV only
Alprazolam (Xanax)
Bromazepam (Lectopam)
Lorazepam (Ativan)
Oxazepam (Serax)
Nitrazepam (Mogadon)
Temazepam (Restoril)
Clobazam (Frisium)
Clonazepam (Rivotril)
Diazepam (Valium)
Flurazepam (Dalmane)
Chlordiazepoxide (Librium)
Clorazepate* (Traxene)
Data Sources:
The Quebec Health Care Databases




Beneficiary Database: all Quebec residents, age,
sex, date of death, address
Pharmaceutical Database: all claims for
prescriptions dispensed to elderly and welfare
recipients in Quebec
Medical Services Claims: all medical services
provided on a fee-for-service basis (90%) to
Quebec residents
Hospitalization Database: all discharges from
Quebec hospitals - dates for hospitalization
Study Sample
Original Study Sample
727,294
all elderly and welfare recipients
1989
Age < 66 or Welfare Recipient
205,547
28.3%
Temp. ID or Non-resident
13,695
1.9%
Died in 1989
18,258
2.5%
Isolated Region
697
0.1%
Final Sample
462,543
63%
76% of Quebec Elderly
Institution/Perm. Hosp.
26,554
3.6%
Study Overview
462,543
>65 years old
209,732 (45.3%)
prevalent users
253,081
non-prevalent users
174,444 (37.7%)
non-users
Baseline
Jan. 1989
Jan. 1990
78,637 (17.0%)
incident users
Follow-up
Dec. 1994
Results - general






average age 73.4 years with 52% women
78,367 (31%) new benzodiazepine users
New users had an almost two-fold increase in use of
anti-depressants and sedatives, cardiac medications,
anti-hypertensive agents, vasodilators and diuretics
9.5% of new users versus 5.6% of non-users filled at
least one prescription for another psychotropic
medication
44% of new users vs 38% of non-users filled at least
one prescription for medications that affect motor
stability
New users were more likely to have depression and
arthritis, and used more health care services than
non-users
H.R.
95% CI
Sex - Men vs Women
0.87
0.86-0.89
No. Prescribing Drs.
1.09
1.09-1.10
No. Hospital Stays
0.95
0.94-0.97
H.R.
95% CI
0.96
0.97
0.95
0.93-0.99
0.95-0.99
0.91-0.99
1.35
1.10
1.08
0.96
1.30-1.41
1.05-1.15
1.06-1.10
0.91-1.02
Alcohol Abuse
0.99
1.01
1.35
0.90-1.10
0.98-1.05
1.18-1.54
Drug Abuse
1.18
1.03-1.37
Any Injury (1989)
Visual Impairment
Stroke
Depression
Neurological Disorders
Arthritis
Seizure
Osteoporosis
Misc. Impairments
Charlson Co-morbidity Index
H.R.
95% CI
Anti-Depressants
Anti-Psychotics
Sedatives
1.67
1.23
1.27
1.61-1.74
1.16-1.31
1.23-1.32
Lithium, L-tryptophan
Cardiac Drugs
Anti–Hypertensive Agents
Vasodilators
1.26
1.05
1.06
1.17
1.09-1.46
1.03-1.07
1.04-1.08
1.15-1.20
Opiod Agonists
1.14
1.06-1.23
Opiod Mixed Partial
Agonists/Antagonists
Non-Thiazide Diuretics
1.11
0.91-1.34
1.08
1.06-1.11
Results – product specific




decreased risk of starting oxazepam and
flurazepam for older ages
women were more to be new users of the
majority of the benzodiazepines except
temazepam and flurazepam
each additional prescribing physician seen
increased by risk of new use by 5-15%
having an fall injury decreased risk for
lorazepam (HR=0.93, p=0.01) and
diazepam (HR=0.86, p=0.04) and an
increased probability for chlordiazepoxide
(HR=1.34, p=0.04)
Results – disabilities and impairments





depression was strongly associated with new use
except triazolam and temazepam -particularly
strong for alprazolam (HR=1.98, p<0.0001) and
clonazepam (HR=2.46, p<0.0001)
weaker but consistently positive increased risk for
arthritis
neurological disorders (including dementia and
Parkinson’s disease) and clonazepam (HR=2.24,
p<0.0001);
alcohol abuse and both oxazepam (HR=1.55
p=0.001) and chlordiazepoxide (HR=12.1,
p<0.0001)
drug abuse with bromazepam (HR=2.34,
p=0.0008).
Results – disabilities and impairments



strongest and most consistent associations
were seen for use of anti-depressants as
well as other psychotropic medications
(anti-psychotics and non-benzodiazepine
sedatives, lithium or l-tryptophan)
filling a prescription for an anti-depressant
significantly increased risk varying from a
23% increase for diazepam (HR=1.23) to
more than tripling the hazard for
clonazepam (HR=3.13)
use of anti-psychotics, other sedatives, and
lithium or l-tryptophan increased risk by
more than double for new clonazepam and
flurazepam use and over five times for
clonazepam (HR=5.19, p<0.0001).
Conclusions



Factors associated with new
benzodiazepine use vary considerably
among the individual products
Physicians appear to be “channeling”
new users based on own criteria – not
necessarily evidence based
Any research on risk needs to account
for these factors by individual products
Limitations




under-diagnosis and under-reporting of
the treatment of certain diseases
anxiety and insomnia were often not
coded in the database making it difficult
to assess the association between these
diagnoses and benzodiazepine use
proxy measure of use (dispensed
prescription)
no prescription information available
during hospitalization
Future Directions





Why benzodiazepines are chosen by
physicians – are other risk factors
accounted for?
Role of risk in guidelines
recommendations…
Methods to reduce risk of falls – smart
alerts?
Investigations of risk from falls – are
other risk factors accounted for?
Is dose adjusted for in high risk patients?
Questions & Comments
[email protected]