Cost-effectiveness in Personality Disorder

1
Cost-effectiveness in Personality Disorder

• Dr. J.J.V. Busschbach
• Psychotherapeutic centre De Viersprong
• PO Box 74660 AA Halsteren 31 164 63220031
  164 632220 (fax)jan.busschbach_at_devierspong.nl
• Erasmus MC, Rotterdam
• Department of Medical Psychology Psychotherapy
• www.xs4all.nl/jannetvb/busschbach
• Contains the slides of this presentation

2
The usual convention.

• Doubt about the cost-effectiveness
• Treatment of personality disorder is expensive
• Treatment is long
• Effect is low
• Cost-effectiveness is unfavourable
• How to deal with such stigma?

3
Stigma is not unique

• Typical for new interventions
• Especially new pharmacy
• Prozac is example
• Prozac was said to be
• More expensive
• As effective as old medication
• As established in RCT
• Therefore not a cost-effective alternative

4
Stigma versus science

• Reaction of Ely Lilly
• Manufacturer of Prozac
• Two main arguments
• They questioned the randomised trial results
• The generalisability of results for clinical
  practice
• Introducing Outcome Research
• They questioned the assumption about higher costs
• Medication cost may be higher, but total cost may
  be lower
• Introducing Health Economics

5
Outcome Research

• Clinical research
• Does it work?
• Efficacy
• Perfect patient
• No co morbidity
• Randomized Clinical Trial
• Controlled conditions

• Outcome research
• Does it work in practice
• Effectiveness
• Every day patient
• Normal co morbidity
• Trials in a naturalistic setting
• Real life conditions

6
In RCT no differences in efficacy

• Between Prozac and old medication
• No differences between TCA and SSRI
• Citation British Medical Journal
• Randomised, controlled clinical trial (RCTs )
  generally show equal efficacy among
  antidepressants
• Song F et al. BMJ, 1993306683-7

7
But in outcome research

• In practice much better effectiveness
• Drop out ration TCA SSRI 3 1
• Lobowitz, JAMA 19972781186-90
• After drop out, recurrence depression 2 to 4 time
  higher
• Minimal effective dose
• SSRI 98 (Prozac)
• TCA 61
• N 23000, General Practitioner
• De Waal et al, NTVG 19961402131-4
• Randomised trials mask differences compliance!
• Outcome research reveals remarkable results

8
Health economics

• Simon et al, JAMA 19962751897-902
• Six-month health care expenditures
• Total cost, not just medication costs
• Compared
• Desipramine N 181
• Old TCA
• 2361
• Imipramine N 182
• Old TCA
• 2105
• Fluoxetine N 173
• New SSRI Prozac
• 1967
• No statistical significant differences

9
Regression in quasi-experimentcontrolled for
sex, age, prior-period expenditures etc.
Sclar et al, 1994 N 701
10
What can we learn?

• Randomised trials are not the holy grail
• They do serve in efficacy
• But there are higher order measurements
• Effectiveness
• Outcome research
• Randomised trials AND naturalistic studies
• Quasi experimental design
• Cost-effectiveness
• Health economics
• Randomised trials AND naturalistic studies
• Quasi experimental design

11
Where do we stand?

• Favourable results in (randomised) trials
• Psychotherapy versus usual care
• 6 Reviews en 1 meta analysis
• Perry et al, Am J Psychiatry 1999571312-21
• What about cost effectiveness.?
• is psychotherapy in personality disorder worth
  the costs?

12
Existing evidence suggests considerable savings

• New investigations
• Bateman, Fonagy, Am J Psychiatry 2003160169-71
• Reviews
• Gabbard et al. Am J Psychiatry 1997154147-50

13
Problem in cost effectiveness results

• Cost estimates made in trial environment
• No real cost estimates
• No adjustment made for trial situation
• No formal cost-effectiveness study designs
• Typical elements are missing
• Discounting
• Costs and effects in the future are valued lower
• Generic outcome measures
• Quality adjusted life years (QALYs)
• Disease specific outcome do not allow for
  comparisons between different allocations in
  health care

14
What do we need?.

• Naturalistic trial
• To prove the effects in practice
• To estimate costs in practice
• Formal cost-effectiveness study
• Following international guidelines

15
Sceptre hopes to fulfil these demands

• Quasi experimental trial in a naturalistic
  setting
• Introducing outcome research
• The design follows standards in health economics
• Introducing health economics
• But even more than Sceptre we need.

16
Confidence

• Good treatment will be cost effective
• If a treatment works in practice, it will almost
  certainly be cost-effective
• Like Prozac
• In that conviction we need to put our treatments
  to the test.