Transcript Challenges in end-stage heart failure: Compliance

Challenges in end-stage
heart failure: Compliance
Fabienne Dobbels, PhD
Heart failure: A chronic disease
• Requires ongoing management over a period of years
• Cannot be cured
• May lead to disability, or the short- or long-term reduction
of a person’s activity
• Goal of treatment:
= to improve patients’ ability to live a productive and
pain free life
 to get rid of the disease!!!
Treatment of heart failure:
A complex therapeutic regimen
Management
of heart failure
Prescription
of multiple drugs
- fluid restrictions
- diet (< salt)
- weighing +exercise
-…
Optimal treatment
of co-morbidities
Management of patient
with heart failure:
Psychological
Dimension
Behavioral
Dimension
Physical
Dimension
Goal = to optimize outcomes!
Non-adherence: the Achilles heel of
heart failure treatment
 Definition
 Prevalence
 Consequences
 Risk factors
 Interventions
Peter Paul Rubens 1630
Eve and the Apple in the
Garden of Eden....
...the first case of nonadherence?
Compliance = adherence = concordance
=
“The extent to which a person’s behavior
corresponds with the agreed recommendations from a
health care provider”
(Sabate. WHO report 2003)
=
“Is a behavioral process, strongly influenced by the
environment in which the patient lives, including the
healthcare practices and system.
Adherence assumes that a patient has the knowledge,
motivation, skills and resources required to follow the
recommendations of a healthcare professional.
(AHA expert panel. Miller et al. 1997)
MEDICATION ADHERENCE
TAKING
DOSING
TIMING
DRUG HOLIDAYS
Age 65 - 74 years: 8.5 medications daily
Age 75 – 84 years: 7.9 medications daily
Age > 84 years:
7.0 medications daily
(Soumerai et al. Arch Int Med 2006; 166: 1829)
Prevalence of non-adherence (NA)
in elderly with heart failure
• Medication taking:
1 - 90%
• Fluid restrictions:
27 - 77%
• Sodium restrictions:
27 – 87%
• Daily weighing:
21 – 88%
Large variation depending on operational definition
and measurement method used
(van der Wal et al. Int J Cardiol 2008; 125: 203)
Prevalence of adherence
Disease
HIV
Arthritis
Cancer
Cardiovascular disease*
End-stage renal disease
Pulmonary disease
Diabetes
Mean (%)
adherence
88.3
81.2
79.1
76.6
70.0
68.8
67.5
Random effects
95% CI
(78.9; 95.2)
(71.9; 89.0)
(75.9; 84.2)
(73.4; 79.8)
(56.8; 81.6)
(61.1; 76.2)
(58.5; 75.8)
* Numbers for hypertension similar to other cardiovascular diseases
(DiMatteo MR. Med Care 2004; 42(3): 200-209)
Estimated NA of elderly patients with hear
disease in Belgium
Estimations (2004):
- 10 318 000 inhabitants
- 1 754 060 (17%) > 65 years
- 261 355 (14.9%) serious heart
disease or heart attack in past
12 months
- 23.4% non-adherent
61 680 PATIENTS WITH HEART DISEASE NONADHERENT!!!
(Belgian Health Interview Survey 2004, www.iph.fgov.be)
(DiMatteo MR. Med Care 2004; 2: 200)
NA is a prevalent problem:
so what???
“Drugs don’t work in patients
who don’t take them”
(C Everett Koop M.D.)
Clinical
Economic
consequences consequences
NA associated with poor clinical
outcomes in heart failure
• Absence of the intended effect of the drugs
• Higher number of hospitalizations
• More visits to the emergency department
• Adverse effects (rebound effect)
(Hope et al. Am J Health-Syst Pharm 2004; 61: 2043)
(Vinson et al. Am J Geriatr Soc 1990; 38: 1290)
OVERALL
Risk difference
NA and outcome of medical treatment:
A meta-analysis (63 studies)
1
0,8
0,6
0,4
0,2
0
-0,2
-0,4
-0,6
-0,8
-1
Risk difference(%)
OR (95% CI)
Overall
0.26 [0.20 – 0.32]
2.88 [2.23 – 3.73]
Hypercholesterolemia
0.25 [0.13 – 0.35]
2.81 [1.67 – 4.71]
Hypertension
0.30 [0.12 – 0.46]
3.44 [1.70 – 7.37]
Heart disease
0.10 [-0.02 – 0.22]
1.49 [0.91 – 2.42]
(DiMatteo et al. Med Care 2002; 40: 794)
Good adherence reduces mortality risk
in chronic illness populations
TOTAL
0.2 0.5
OR= 0.56 [0.50 – 0.63]
2
5
1
(Simpson et al. 2006; 333: e-pub June 21)
Impact of NA assessed by MEMS on
event-free survival (N= 137)
% of prescribed
doses taken
Median= 95.4%
% of days the correct
number of prescribed
doses were taken
Median= 90.3%
% of doses taken
on schedule (within 25%
of prescribed time interval)
Median= 76.0%
(Wu et al. J Cardiac Fail 2008; 14: 203)
Economic
consequences
Economic consequences of NA
Direct costs
•  cost of non-taken medication
•  cost for treatment of morbidity
•  cost of avoidable hospitalizations
Indirect costs
• Missed work days
• Cost for transportation, household, home care
•  quality of life
•  cost of evolving more potent medications
Non-drug medical costs within 1 year
Adherent
( 80% taking)
Non-adherent
(< 80% taking)
Diabetes
$ 6377
$ 9363 - $ 15 186
Hypertension
$ 6570
$ 7658 - $ 10 286
Hypercholesterolemia
$ 4780
$ 5509 - $ $ 9849
One study in heart failure: No difference in costs
(Muzbek et al. Int J Clin Pract 2008; 62: 338)
Noncompliance: a major and important problem
Can / will health care provider do something...
Or will we expel patients from Paradise?
What can be done to improve adherence ?
Identifying
patients
at risk for NA
Implementation
of interventions
Determinants
of NA
Randomized
controlled
trials
Measurement of nonadherence
Clinical
nonadherence
A. Direct methods
- observation
- assay
- objective tests
Sub-clinical
nonadherence
B. Indirect methods
– pill count
– prescription refill
– clinical judgement
– electronic monitoring
– self-report
No gold standard: combine measures to increase accuracy
(Osterberg et al. N Engl J Med 2005;353)
5 interrelated categories of
determinants
Socio-economic
factors
Condition related
factors
Health
professional and
setting-related
Patient related
factors
Treatment related
factors
(Sabate E. WHO report 2003)
Determinants in patients with HF
Socio-economic
Condition related
Poor socioeconomic status
Low education/illiterate
Depression
Cognitive dysfunction
Cost of medication
Higher co-morbidity
Poor social support
Poor social support
Living alone
Treatment related
Patient related
Complex regimen
Side effects
Poor knowledge
Lack of motivation
Lifelong duration of treatment
Health beliefs/ attitudes
Frequent changes
Interference with socialization
(van der Wal et al. Int J Cardiol 2008; 125: 203)
Impact of the health care provider
and setting related factors
Macro level
Policy
Meso level
Health care organizations
and community
Micro level
patient-provider
interaction
- reimbursement and insurance policy
- no funding for chronic disease
management programs or prevention
- Short consultations
- Lack of follow-up / cooperation with
community services
- Uni-disciplinary treatment
- Poor knowledge about adherence
- lack of trust
- poor communication style
Interventions…
Effectiveness of adherence-enhancing
interventions: RCT’s
intervention
control
random
Education
ES= .20
Behavior
ES= .22
Affective
ES= .20
Education + behavior
ES= .35
0
20
40
60
80
100
Absolute difference (%)
(Roter et al. Med Care 1998; 36: 1138-1161)
Typical reaction if treatment is not
working: the radar syndrome
The patient appears…
Find the problem and fix it, by:
- increasing the dose
- switching to another drug
- adding another drug
But nonadherence frequently ignored!
Disease management programs in
heart failure populations
• Integrated programs with focus
on
- detailed assessment of
the patient
DISEASE
MANAGEMENT
- patient education about
treatment regimen
- optimizing medications
Education
Monitoring
- regular monitoring by
health professionals
(Health and health care 2010 – The Forecast – the Challenge
Institute of the future 2003)
Efficacy of disease management
programs: a meta-analysis
Re-hospitalization *
Mode
Personal
Phone
-10.5
-3.6
[-14.7; -6.2]
[-6.8; -0.3] **
team
Single group
2-3 disciplines
multidisciplinary
-7.5
-2.5
-18.1
[-10.7; -4.4]
[-8.7; -3.8] **
[-23.4; -12.9] **
Transition
Yes
No
-8.6
-6.1
[-12.7; -4.4]
[-9.8; -2.5]
Follow-up
3 months
3-9 months
> 12 months
-10.9
-6.2
-9.0
[-17; -4.9]
[-12; -0.4]
[-13.9; -4]
* Risk Difference; negative value in favor of program
** significant difference with reference value; pooled relative risk 0.84 [0.77; 0.92]
(Göhler et al. J Cardiac Fail 2006; 12: 554)
Cost-effectiveness of disease
management programs
• Mean age at onset 67 years (35% female)
– Quality adjusted life expectancy:
2.64 years for standard care
2.83 years for disease management program
– Additional lifetime cost for 84 days difference:
1700 Euro
(i.e. 9800 Euro per QALY gained)
Beneficial impact on clinical outcomes but expensive…
(Göhler et al. Eur J Heart Failure 2008; e-pub)
Problem of disease
management problems
20%
3o
2o
1o
Patient preferences
Readiness for treatment
Compliance
Symptom management
…
Providing
professional
patient care
The majority of care is taking place
outside the hospital setting
20%
80%
3o
2o
1o
Patient preferences
Readiness for treatment
Compliance
Symptom management
…
Providing
professional
patient care
Fostering patient
self-management
“The most effective approaches
have been shown to be
multidimensional and multilevel –
targeting more than one factor with
more than one intervention”
(Haynes et al. Cochrane Reviews 2008)
Tackling Nonadherence:
A Multidimensional and Multilevel Approach
Social/
Health
economic
system/
factors
HCT-factors
Conditionrelated
factors
Patient-related
factors
Macro
Meso
Micro
Therapyrelated
factors
+
Patient
Multilevel interventions
Macro level
Policy
Meso level
Health care organizations
and community
Micro level
patient-provider
interaction
- Health care system changes
allowing self-management support
and chronic care
- Multi-disciplinary treatment
- follow-up organized with focus on
chronic illness
- Engagement of community resources
- development of adherence
counseling toolkit
- Training in fostering self-management
- training in motivational interviewing
From disease management to
self-management programs
=
A set of things patients can do for themselves
to follow the prescribed therapy, to avoid health
deterioration and preserve function
SELFMANAGEMENT
PROBLEM
SOLVING
DECISION
MAKING
RESOURCE
UTILIZATION
BUILDING
PATIENT/
PROVIDER
RELATION
TAKING
ACTION
Remember the definition???
=
“Is a behavioral process, strongly influenced by the
environment in which the patient lives, including the
healthcare practices and system.
Adherence assumes that a patient has the knowledge,
motivation, skills and resources required to follow the
recommendations of a healthcare professional.
(AHA expert panel. Miller et al. 1997)
Efficacy and cost of HF selfmanagement programs
Odds ratio [95% CI]
All-cause readmission in 1 year (5 studies)
0.59
[0.44; 0.80]
Readmissions due to HF (3 studies)
0.44
[0.27; 0.71]
Mortality (3 studies)
0.95
[0.57; 1.51](NS)
Adherence (2 studies)
Both significant 
Cost saving (3 studies) after subtracting the intervention cost:
$ 1300 - $ 7515 saved per patient annually
ONLY POSSIBLE IF YOU HAVE A TRAINED TEAM!!!
(Jovicic et al. BMC Cardiovascular Disorders 2006; 6: 43)
Conclusion
• HF is a chronic disease requiring a complex management
• Nonadherence is a prevalent problem, resulting in poor
clinical and economical outcomes
• Risk factors are multi-factorial
• Interventions should be multidimensional, targeting more
than 1 risk factor with more than 1 intervention
• A multilevel approach is mandatory, integrating
interventions at the patient, health care professional, team
and policy level
Take home message
“Changing systems of care and applying
multidimensional + multilevel adherenceenhancing interventions to improve selfmanagement may have a far greater impact
on the health of heart failure patients than any
improvement in specific medical treatments”
(Haynes et al. Cochrane review 2008)
Increasing adherence with
heart failure treatment...
ITHET
IS NEVER
TOO
IS NOOIT
TEEARLY!
LAAT!
IT IS NEVER TOO LATE!
KEEP ON BELIEVING THAT
PEOPLE CAN CHANGE!
Some numbers…
• 3.7% of the Belgian population reported with a serious
heart problem or heart attack in the last 12 months (2004)
• 21.9% treated by GP alone
• 38.9% treated by specialist alone
• 25.7% treated by both GP and specialist
•
•
•
•
•
•
•
89.3% of these patients use medicines for this problem
Use of cholesterol reducing agents: 6.4%
Use of cardiac glycosides: 0.7%
Use of anti-arrhytmics: 0.8%
Use of ace-inhibitors: 3.8%
Use of diuretics: 4%
Use of beta-blocking agents: 8.7%
• > 65 years: 3.4 drugs on average
• 14.9% of > 65 years serious cardiac
disease or heart attack in past 12 months
– 22.3% cholesterol reducing agent
– 3.9% cardiac glycosides
– 4.2% anti-arrhytmics
– 14.1% ace-inhibitor
– 18.7% diuretics
– 26.2% beta-blocking agents
Belgian Health Interview Survey 2004, Scientific Institute for Public Health
www.iph.fgov.be
Percentage of nonadherence for
different therapeutic aspects
diet
40,7
appointment
34,1
health behavior
30,3
exercise
28
screening
27,3
medication
20,6
0
10
20
30
40
50
(DiMatteo MR. Med Care 2004; 42(3): 200-209)