modelling the costs of illness and the costs of relapse in
TRANSCRIPT
Modelling the Costs of Illness and
the Costs of Relapse in the
Management of Schizophrenia
in the UKLaunois R., Présenté G. - Réseau d’Evaluation en Economie de la Santé (REES France), Paris (France) - Hansen K.,
Toumi M. - Lundbeck, Paris (France)
ISPOR Fourth European Congress
CANNES (France) - 11 / 13 November 2001
REES France : Réseau d ’Evaluation en Economie de la Santé
http://smbh7.smbh.univ-paris13.fr
ISPOR - 11/13 Nov. 2001 PL-2238/01 2
AIMS OF THE STUDY
▪ To compare the toxicities of atypical treatments for
schizophrenic patients
▪ To study their consequences on compliance and relapse
▪ To specify management pathways in UK for patients under
treatment
▪ To analyze the mean cost incurred by relapsing and not
relapsing patients treated with atypical antipsychotics in
the British society
▪ To analyze the total 10 years and life time costs incurred by
the National Health Service
ISPOR - 11/13 Nov. 2001 PL-2238/01 3
METHODS
▪ Simulation model
▪ Clinical states
▪ Management structures
▪ Clinical events
▪ 10 years and life time cost
Mentally ill Patients’ Accomodation in UK
L-S Hospital
Collective
Residence
staffed
FH
DH
NH
DCC
CMHC
AMBULATORY VISITS
Private
Residence
Non-staffed
ISPOR - 11/13 Nov. 2001 PL-2238/01 5
Classification Algorithm
Living in a staffed accommodation?
Schizophrenic Patient
Long-stay
inpatient
hospitalisation
Intensive
Collective Care
ICC
Full-time hospitalisation>180 days?
Collective care Personal care
Staff resident ratio
> 0,5?
Indicateur de prise
en charge > 120?
yes no
In group home
settings?
Mild
Collective Care
MCC
Intensive
Personal
Care IPC
Mild
Personal
Care MPC
no
no
no
yes
yes
yes
Heavy Hospital
Management
Home Hostel
Intensive management
Standard management
Patient’s Care Management Groups
Patient’s residence
IPC
Inte
nsi
ty o
f ca
re
ICC
MPC MCC
L-S Hospital
ISPOR - 11/13 Nov. 2001 PL-2238/01 7
Clinical States
▪ Relapse (5 states)
▪ Non-Relapse (5 states)
▪ Chronic (5 states)
▪ Lost to follow up (1 state)
▪ Death (1 state)
ISPOR - 11/13 Nov. 2001 PL-2238/01 8
Simulation Model
▪ 17 Markov states (R+, R-, chro, DO, DC)
▪ 8 160 variables
▪ 18 213 nodes
▪ Software Data for Windows version 3.15
ISPOR - 11/13 Nov. 2001 PL-2238/01 9
Markov Model in Schizophrenia
Sertindole
Olanzapine
Risperidone
Others
ISPOR - 11/13 Nov. 2001 PL-2238/01 10
Markov Model in Schizophrenia
Hospital
ICC
IPC
MPC
MPC
DO
Death
ISPOR - 11/13 Nov. 2001 PL-2238/01 11
Markov Model in Schizophrenia
EPS
Drowsiness
Weight gain
Sexual dysfunction
M tox
Comp
Comp +
Comp -
Relapse
R+
R-
GO
VV
DO
DC
ISPOR - 11/13 Nov. 2001 PL-2238/01 12
Markov Model in Schizophrenia
Hospital
ICC
IPC
MCC
In
Amb
Int
Mild
Resid
Home MPC
Resid
Home
Death
DO
Hospital
ISPOR - 11/13 Nov. 2001 PL-2238/01 13
Markov Model in Schizophrenia
Hospital
ICC
IPC
MCC
DO
EPS
Drowsiness
Weight gain
Sexual dysfunction
M tox
Comp
Comp +
Comp -
Relapse
R+
In
Amb
Int
Mild
Resid
Home
Hospital
Death
Hospital
ICC
IPC
MPC
MPC
DO
Death
R-
GO
VV
MPC
Resid
Home
DO
DC
Sertindole
Risperidone
Others
Olanzapine
ISPOR - 11/13 Nov. 2001 PL-2238/01 14
Frequency of Side Effects
Short term studies
Sertindole Olanzapine Risperidone
% % %
EPS 19 21 28
Drowsiness 8.2 26 7.9
Weight gain 20.6 30 15.7
Sexual dysfunction 2.8 1 0
Total 50.6 78 51.6
ISPOR - 11/13 Nov. 2001 PL-2238/01 15
Compliance Rate Assumptions
All products
Range
EPS 0.20 0.10-0.60
Somnolence 0.20 0.10-0.60
Weight Gain 0.40 0.10-0.60
Sexual disturbance 0.40 0.10-0.60
Minor toxicity 0.80 0.40-0.90
ISPOR - 11/13 Nov. 2001 PL-2238/01 16
Non-compliance and Effectiveness
Authors Date n Annual rate
Probabili ty
per s ix
months
Del Giudice 1975 57 76% 0.680
Weiden 1995 373 76% 0.510
Glaser 1996 nd 75% 0.509
Baldessarin i 1990 3939 74% 0.490
Kis sl ing 1993 1224 72% 0.470
Johnss on 1983 60 65% 0.400
Gilbert 1993 1224 61% 0.375
Davis 1995 278 46% 0.265
Relapse rate
ISPOR - 11/13 Nov. 2001 PL-2238/01 17
Compliance and Effectiveness
Authors Date n Annual rateProbabili ty per
six m onths
Del Giudice 1975 25 44% 0,250
Weiden 1995 285 35% 0,194
23% 0,117
Baldessarin i 1990 3939 30% 0,160
Davis 1995 249 29% 0,157
Glaser 1996 nd 20% 0,105
Gilbert 1993 3114 20% 0,105
Kis sl ing 1993 270 16% 0,080
Johnss on 1983 60 16% 0,080
Relapse rate
ISPOR - 11/13 Nov. 2001 PL-2238/01 18
Prices of Anti-Psychotic Treatment
Daily cost Six monthly cost
Haloperidol
15 mg / day 0.36 £ 01 65.61 £ 01
Sertindole
16 mg / day 3.20 £ 01 582.40 £ 01
Olanzapine
15 mg / day 5.28 £ 01 951.21 £ 01
Risperidone
5 mg / day 3.24 £ 01 590.86 £ 01
ISPOR - 11/13 Nov. 2001 PL-2238/01 19
Mean Consumption by Patient Per Six Months £01
L-S Hospital Intensive
Residential
Intensive
home Mild Residential
Mild
home
COHORTE STUDY RELAPSE (n=77)
Complete hospitalisation days 3 182.00 0.00 14 805.00 6 382.00
Day hospitalisation days 75.86 0.00 0.00 134.48
Ambulatory contacts 1 273.19 0.00 532.50 1 717.55
TOTAL 4 476.00 0.00 15 337.00 8 235.00
COHORTE STUDY NON-
RELAPSE (n=68)
Complete hospitalisation (days) 0.00 0.00 0.00 0.00
Day hospitalisation (days) 0.00 0.00 1 888.00 100.00
Ambulatory contacts 1 239.70 0.00 2 114.00 1 918.00
TOTAL 1 240.00 0.00 4 012.00 2 020.00
CHRONIC 25 662.00
ISPOR - 11/13 Nov. 2001 PL-2238/01 20
Discount Rate Assumptions
▪ Discount rate
– 10 year : 3%
– Lifetime : 6%
▪ Cost of management with other atypical
antipsychotics
– 10 year: Average (Sertindole, Risperidone &
Olanzapine + 1.5%
– Lifetime: Average (Sertindole, Risperidone &
Olanzapine + 5%
ISPOR - 11/13 Nov. 2001 PL-2238/01 21
Point Prevalence for Psychoses
▪ Adults aged 16-64 is 4/1000
Source: British survey, Office of Population Censuses and Surveys
Population
United kingdom 52 000 000
Age group 16-64 45 000 000
Prevalence 180 000
ISPOR - 11/13 Nov. 2001 PL-2238/01 22
RESULTS
▪ 10 years cost per patient, per clinical state
▪ Overall 10 years cost per patient
▪ Life years cost per patient, per clinical state
▪ Overall life years cost per patient
ISPOR - 11/13 Nov. 2001 PL-2238/01 23
Projected 10 Years Cost
per Patient by Clinical State £01
Cohorte Olanzapine Risperidone Sertindole Others
Relapse 53 097 48 670 47 959 50 657
Non-relapse 15 444 16 241 16 513 16 307
TOTAL 68 541 64 911 64 472 66 964
ISPOR - 11/13 Nov. 2001 PL-2238/01 24
Overall 10 Years Cost
per Patient £01
Drugs Market share % Weighted cost
Olanzapine 33.38 22 878
Risperidone 28.31 18 378
Sertindole 12.32 7 940
Others 25.99 17 407
Total 100 66 603
24 327
2 053
24 19714 508
1 518
0
Predicted 10 Years Cost per Patient by Clinical Status and Type of Management £01
R + R -50 57716 026
ISPOR - 11/13 Nov. 2001 PL-2238/01 26
10 Years Cost for non relapsing
Patient by Type of Consumption £01
2 725
12 462
838
Amb services
78%
Day Hosp
5%
Ful Hosp
0%Drugs
17%
2 725
12 462
838
16 026 £
ISPOR - 11/13 Nov. 2001 PL-2238/01 27
10 Years Cost for relapsing
Patient by Type of Consumption £01
2 725
12 462
838
Amb services
11%Day Hosp
1%
Ful Hosp
87%
Drugs
17%
44 440
4075 370
50 577 £
360
ISPOR - 11/13 Nov. 2001 PL-2238/01 28
The Overall 10 Years Cost of the
Schizophrenic Population £01
£ 9,3 milliards
ISPOR - 11/13 Nov. 2001 PL-2238/01 29
Projected Life Years Cost
per Patient by Clinical State £01
Cohorte Olanzapine Risperidone Sertindole Others
Relapse 78 966 72 041 71 563 76 395
Non-relapse 22 273 24 014 23 854 22 639
TOTAL 101 239 96 055 95 417 99 034
ISPOR - 11/13 Nov. 2001 PL-2238/01 30
Overall Life Years Cost
per Patient £01
Drugs Market share % Weighted cost
Olanzapine 33.38 33 792
Risperidone 28.31 27 195
Sertindole 12.32 11 751
Others 25.99 25 743
Total 100 98 481
ISPOR - 11/13 Nov. 2001 PL-2238/01 31
The Overall Life Time Cost of the
Schizophrenic Population £01
£ 13,6 milliards
ISPOR - 11/13 Nov. 2001 PL-2238/01 32
Incremental Cost Effectiveness
Ratio for UK
Strategy Cost Marginal
Cost Effectiveness
Marginal
Effectiveness ΔC/ΔE
Sertindole 64 472 8.0520 8 006
Risperidone 64 911 439 7.9466 - 0.1055 8 168 Dominated
Haloperidol 68 499 4 027 6.7176 - 1.3344 10 197 Dominated
Olanzapine 68 541 4 069 7.3069 - 7.3069 9 380 Dominated
Cost £ 01 – Effectiveness : semesters without relapse – Time window : 10 years
ISPOR - 11/13 Nov. 2001 PL-2238/01 33
Sensitivity Analysis
▪ Varying toxicity rates
▪ Varying compliance rates
▪ Varying relapse rates
▪ Varying hospitalisation rates
▪ Varying intensive home management rates
▪ Varying incremental costs between relapsers
and non-relapsers
▪ Varying discounting rate
ISPOR - 11/13 Nov. 2001 PL-2238/01 34
CONCLUSION
▪ Although sertindole demonstrates a better cost-effectiveness ratio than other typical or atypicalantipsychotics,
▪ The outcomes are modest and question the decisionmaking process on the sole basis of cost effectivenessanalysis.
▪ It seems quite difficult on the current available data todefine tight recommendations on which antipsychoticshould be recommended as first line treatment.