1

The effects of prevention: the EDIE-NL study and meta-

analytical outcomes Hoe eerder, hoe beter 30 mei 2013, Den Haag

Mark van der Gaag

STATISTICAL POWER IS THE ISSUE IN

PREVENTION RESEARCH

PLAYING TRICKS WITH EPIDEMIOLOGY

2

3

Extended Psychosis Phenotype: 61 cohorts Prevalence 7.2% Incidence 2.5% Persistence 2-5 years 20-30% Transition rate 7.4%

0.6% 20-30% 10%

Transition rate, success rate, and Number Needed to Treat Van Os, J., & Delespaul, P. (2005). Toward a world consensus on prevention of schizophrenia. Dialogues Clin Neurosci, 7(1), 53-67.

4

IT IS ALL ABOUT SAMPLE

ENRICHMENT

5

EDIE

Help-seeking for a co-morbid axis 1 or 2 disorder Young people 14-35 years of age (18-35 in The

Hague) Psychotic-like experiences (Prodromal

Questionnaire, Rachel Loewy) Decline in functioning SOFAS<55

EDIE TRIAL 2008-2012

7

Acknowledgements Early Detection and Evaluation Intervention

David van den Berg Petra Bervoets Nynke Boonstra Marion Bruns Sara Dragt Saskia van Es Sarah Eussen Gitty de Haan Mischa van der Helm Martijn Huijgen Helga Ising Lianne Kampman Rianne Klaassen Don Linszen Aaltsje Malda

Carin Meijer Julia Meijer Roeline Nieboer Dorien Nieman Bianca Raijmakers Judith Rietdijk Marleen Rietveld Nadia van der Spek Annelies van Strater Tinie van de Tang Zhenya Tatkova Jenny van der Werf Swanny Wierenga Lex Wunderink Annemieke Zwart

Alison Yung Tony Morrison Paul French Lucia Valmaggia Rachel Loewy Pim Cuijpers Niels Smits Kees Korrelboom ZonMW grant

PSYCHOLOGICAL MODEL

9

Goals of CBT in UHR

‘Manage’ external risk factors à education and behavioural control

Prevent catastrophising and delusional interpretations to PLEs à CBT for current concerns à Reduce selective attention and avoidance à Metacognitive awareness of risky thinking styles and ways to handle and cope with cognitive biases and PLEs

10

11

Risk management Ø  No cannabis or other drugs Ø  Keep attending school and work Ø  Share thoughts and convictions with others and discuss these

RESULTS 12

13

Parnassia Psychiatric Institute & PsyQ Haaglanden, The Hague

MHS Leiden, Child and Adolescent Department Academic Medical Centre & PsyQ Amsterdam MHS Friesland, Leeuwarden and province

Flowchart of patients

Results of CBTuhr intervention

TAU = State of the art treatment for co-morbid disorder

CBTuhr add-on to TAU

10 transitions

22 transitions

Patient status at 18-month follow-up

The Chi-square linear-by-linear association is significant: Chi2 (df=1)=4,266, p=.039

The NNT to prevent a transition was 9 The NNT to bring a patient into remission was 7

16

ECONOMIC COSTS 17

Cost Utility Analysis Voorlopige uitkomsten, nog aan verandering onderhevig

-10.000

-7.500

-5.000

-2.500

0

2.500

5.000

7.500

-0,15 -0,10 -0,05 - 0,05 0,10 0,15 0,20

Add

ition

al c

osts

(€)

Additional effects

NE: more effective and more expensive 26.1 %

NW: less effective and more expensive 10,9%

SW: less effective and less expensive 11,5%

SE: more effective and less expensive 51.6 %

Cost Uitility Analysis Voorlopige uitkomsten, nog aan verandering onderhevig

CBT uhr Effect: 0.60 Costs: € 6.734

TAU Effect: 0.57 Costs: € 7.445

19

Incremental Cost Effectiveness Ratio Incr effectiveness: 0.03 Incr costs: - € 710 ICER: - € 21.747 Median: - € 14.650

Results Cost Utility and Cost Effectiveness Analyses

CBT uhr is cost-effective:

•  Increased health QALYs for reduced costs •  Reduced psychotic transitions for reduced

societal costs •  Reduced psychotic transitions for reduced

health costs 20

META ANALYSIS 21

Preventing a first episode of psychosis: meta-analysis of randomized controlled prevention trials of 12 month and longer-term follow-ups

Mark van der Gaag1,2, Filip Smit1,3, Andreas Bechdolf4, Paul French5, Don H

Linszen6, Alison R Yung7, Patrick McGorry7, Pim Cuijpers1,3 1) VU University and EMGO Institute of Health and Care Research, Amsterdam, The

Netherlands 2) Parnassia Psychiatric Institute, The Hague, The Netherlands 3) Trimbos Institute, Utrecht, The Netherlands 4) Department of Psychiatry and Psychotherapy, University of Cologne, Germany 5) University of Manchester, Manchester, United Kingdom 6) Academic Medical Center of the University of Amsterdam, The Netherlands 7) University of Melbourne, Australia.

22

23

Study name Statistics for each study Risk ratio and 95% CI

Risk Lower Upper ratio limit limit Z-Value p-Value

McGorry, 2002 0,542 0,226 1,298 -1,374 0,169McGlashan, 2006 0,425 0,168 1,076 -1,806 0,071McGorry, 2012a 0,760 0,285 2,026 -0,549 0,583Amminger, 2008 0,177 0,042 0,750 -2,350 0,019Nordentoft, 2006 0,264 0,079 0,888 -2,152 0,031Bechdolf, 2012 0,054 0,003 0,913 -2,023 0,043Morrison, 2004 0,207 0,046 0,941 -2,039 0,041Addington, 2011 0,128 0,007 2,350 -1,385 0,166McGorry, 2012b 0,742 0,278 1,982 -0,594 0,552Morrison, 2012 0,700 0,274 1,788 -0,745 0,456Van der Gaag, 2012 0,473 0,226 0,988 -1,993 0,046

0,463 0,334 0,642 -4,616 0,000

0,01 0,1 1 10 100

Intervention Control

Study name Statistics for each study Risk ratio and 95% CI

Risk Lower Upper ratio limit limit Z-Value p-Value

McGorry, 2002/70,753 0,387 1,465 -0,836 0,403

Nordentoft, 20060,566 0,278 1,153 -1,567 0,117

Bechdolf, 2012 0,103 0,014 0,783 -2,197 0,028

Morrison, 2004/70,622 0,250 1,543 -1,025 0,305

Morrison, 2012 0,769 0,349 1,697 -0,650 0,516

0,635 0,438 0,919 -2,405 0,016

0,01 0,1 1 10 100

Favors Experimental Favors Control

Results

Early detection and indicated prevention are about to become an evidence-based intervention

CBT that showed a transition reduction of 48% and a NNT of 13 in five RCTs with 672 subjects

CBT uhr is cost-effective: increased health for

reduced costs

26

MARK VAN DER GAAG PhD

HEAD OF PSYCHOSIS RESEARCH PARNASSIA PSYCHIATRIC INSTITUTE, DEN HAAG PROFESSOR OF CLINICAL PSYCHOLOGY AT VU UNIVERSITY, AMSTERDAM

m.vander.gaag@vu.nl

Thank you for your attention!

27