1. About OMICS Group OMICS Group International is an amalgamation of Open Access publications and worldwide international science conferences and events. Established in the year 2007 with the sole aim of making the information on Sciences and technology ‘Open Access’, OMICS Group publishes 400 online open access scholarly journals in all aspects of Science, Engineering, Management and Technology journals. OMICS Group has been instrumental in taking the knowledge on Science & technology to the doorsteps of ordinary men and women. Research Scholars, Students, Libraries, Educational Institutions, Research centers and the industry are main stakeholders that benefitted greatly from this knowledge dissemination. OMICS Group also organizes 300 International conferences annually across the globe, where knowledge transfer takes place through debates, round table discussions, poster presentations, workshops, symposia and exhibitions.Open Access publicationsscholarly journalsInternational conferences

2. About OMICS Group Conferences OMICS Group International is a pioneer and leading science event organizer, which publishes around 400 open access journals and conducts over 300 Medical, Clinical, Engineering, Life Sciences, Pharma scientific conferences all over the globe annually with the support of more than 1000 scientific associations and 30,000 editorial board members and 3.5 million followers to its credit. OMICS Group has organized 500 conferences, workshops and national symposiums across the major cities including San Francisco, Las Vegas, San Antonio, Omaha, Orlando, Raleigh, Santa Clara, Chicago, Philadelphia, Baltimore, United Kingdom, Valencia, Dubai, Beijing, Hyderabad, Bengaluru and Mumbai.

3. Stepped Wedge Design: An RCT for Developmental Growth in Phelan-McDermid Children Edwin R. van den Heuvel Professor of Statistics e.r.v.d.heuvel@tue.nl Department of Mathematics and Computer Science Eindhoven University of Technology October 2014

4. Content Introduction Type of Trial Designs Statistical Model Simulation –Parameter Settings –Results Conclusions (4)

5. Children with Phelan-McDermid syndrome develop slower (~70%) than normal children The syndrome is caused by a deletion of gene 22q13.3 It is a rare disease (~50 known children in Netherlands) A pilot study showed that insuline can enhance mental development An RCT is needed to investigate a possible effect of intranasal insuline versus placebo Introduction (5)

6. First type of designs that were thought off were –Cross-over design: issues with carry-over? –Parallel group design: possible –Matched-pairs design: possible Parents were not cooperative to participate if their child would not receive the intervention In a stepped wedge design all participants will switch during the trial (one-directional cross-over) We felt it would only be ethical when this design would have similar power than other trials Expected number of participants was ~20 Introduction (6)

7. Stepped wedge designs: All patients are first treated with the control Patients or group of patients are changing at different time points It uses intra- and inter-individual variation of patients to test for treatment effect It could take longer than parallel group designs Example: Three groups with four periods Introduction (7) 7 6 SWD-S3 7

8. Prerequisites: Five time moments: 0, 6, 12, 18, and 24 months –More time moments would be to much of a burden –Shorter times affect test-retest variability Development is measured by Bayley III –Scores are transformed to developmental age Possible types of designs –Parallel group design (PGD) –Matched-Pairs design (MPD) –Stepped Wedge Designs (SWD) –Delayed Start Design (DSD) Type of Trial Designs (8)

9. Parallel Group Design: At time zero patients are randomly allocated to either placebo or intranasal insuline Ramdomization is balanced (ratio 1:1) –Dotted line is placebo –Solid line is intranasal insuline –Ratio of total treatment time on both treatments is 1:1 Type of Trial Designs (9) 06121824

10. Matched-Pairs Design: Patients are treated the first six months at placebo Developmental growth or change is matched Per pair placebo or intranasal insuline is randomly allocated Ratio of total treatment in placebo vs. Insuline: 5:3 Type of Trial Designs (10)

11. Stepped wedge designs: Treatment switches coincide with measurements Ratio of total treatment in placebo vs. Insuline: –SWD-S2a:1:1 –SWD-S2b:3:5 –SWD-S3:1:1 Type of Trial Designs (11)

12. Delayed start designs: Trial starts as a parallel group design At a specific time point, (part of) the control group switches to treatment Common trials for Alzheimer and Parkinson’s disease The goal is to –Investigate treatment effect –Whether disease progression can be slowed down The second goal is less important for our RCT Type of Trial Designs (12)

13. Delayed start designs: Ratio of total treatment in placebo vs. Insuline: –DSD-S1a:1:3 –Others:1:1 Type of Trial Designs (13)

14. Type of Trial Design (14) Ratio 1:1 Ratio 3:5 Ratio 1:1 Ratio 5:3 Ratio 1:1 Ratio 1:3 Ratio 1:1

15. Linear Mixed Model Random coefficients model Y ij :developmental age of child I at time point j Z 0,i :developmental age at baseline for child i Z 1,i :developmental growth for child i t ij :moments of measurement for child i x ij :moments of treatment switch for child i  ij :residuals  :treatment effect Statistical Model (15)

16. Model for Stepped Wedge Design Statistical Model (16) 06121824 Child 1: Placebo Child 2: Insuline Child 2: Placebo Child 1: Insuline  : treatment effect Developmental age

17. Analysis of Model Statistical model can be analyzed with standard software for mixed models Requires two regression time variables: –One for placebo time = t ij –One for the new treatment: time2 = t ij -x ij when t ij > x ij time2 = 0 when t ij ≤ x ij SAS codes PROC MIXED DATA=SIMULATIONDATA METHOD=REML; CLASS INDIVIDUAL; MODEL Y = TIME TIME2/SOLUTION DDFM=SAT; RANDOM INT TIME/SUBJECT=INDIVIDUAL TYPE=UNR; RUN; Statistical Model (17)

18. Parameter Settings: Settings were selected on the basis of real data and on literature on test-retest variation of the Bayley III test  0 = 21  1 = 0.3  0 =  10  0 =  45  1 =  0.1  = 0.25; 0.50; 0.75  = 0; 0.15; 0.30; 0.45 Number of simulations: 50000 –Makes it possible to distinguish a difference of 1% in power Simulation (18)

19. Type 1 error: Type 1 error rates are somewhat inflated –The null-distribution of the Wald test statistic is not perfectly chi- squared distributed SWD-S2b and DSD-S1a have only limited inflation of type 1 error rates Simulation (19)

20. Power: SWD-S3 and DSD-S3 seem to perform quite good MPD-A1 is best for  =0.75 PGD, SWD-S2b, and DSD-S1a are less powerful designs for testing treatment effect Simulation (20)

21. Stepped wedge design is competitive with the more classical designs and delayed start designs Stepped wedge designs are considered ethical since not participant is withhold the new treatment The number of steps does not seem to affect the power strongly Imbalance in total treatment time seem to reduce power Power values are not very large with only 20 participants Conclusions (21)