The power of study design 2014 10-27
•Download as PPTX, PDF•
2 likes•1,061 views
Study design for laboratory developed tests (LDT) and in vitro diagnostics (IVD) can make or break product and business success. This top-down description of how to pick the right study design was delivered at the 5th Clinical Affairs & Regulatory Approvals For Diagnostics, October 27-29, 2014, in Alexandria, Virginia.
Recommended
More Related Content
What's hot
What's hot (20)
Similar to The power of study design 2014 10-27
Similar to The power of study design 2014 10-27 (20)
Recently uploaded
Recently uploaded (20)
The power of study design 2014 10-27
- 1. The Power of Study Design 5th Clinical Affairs & Regulatory Approvals For Diagnostics October 27-29 2014 – Alexandria, VA Lyssa Friedman 1October 27, 2014 Lyssa Friedman
- 2. October 27, 2014 Lyssa Friedman 2 Golden Gate National Recreation Area
- 3. Key Points Study design: Things to consider FDA regulation of LDTs: Keep doing what you’re doing well Publication success Case Study Successes and lessons learned October 27, 2014 Lyssa Friedman 3
- 4. Study goals = company goals Successful product launch On time In scope Revenues Reimbursement October 27, 2014 Lyssa Friedman 4
- 5. Start with the goal and work backward October 27, 2014 Lyssa Friedman 5 Business goals Patient population Regulatory requirements Test requirements Study design Publication strategy
- 6. • Business strategy • Unmet clinical need • Target customer(s) • Market and market size • Messaging strategy • Geography – States (CLIA) – US/global • Reimbursement plan • Revenue goals October 27, 2014 Lyssa Friedman 6 Business goals
- 7. • Disease/health condition • Unmet clinical need • Gold standard • Current landscape – Diagnostic tools – Best practices – Clinical guidelines October 27, 2014 Lyssa Friedman 7 Patient population
- 8. • Risk classification • LDT regulated by CLIA • IVD – 510(k) – PMA – de novo – HDE • Companion diagnostic • Kit manufacturing October 27, 2014 Lyssa Friedman 8 Regulatory requirements
- 9. • Performance characteristics – Sensitivity – Specificity – Negative predictive value – Positive predictive value • Test results – Continuous score – Discrete result • Clinical significance October 27, 2014 Lyssa Friedman 9 Test requirements
- 10. Health Economics Analytic Validation Clinical Validation Diagnostic Clinical Performanc e Clinical Outcomes Clinical Utility October 27, 2014 Lyssa Friedman 10 Study design Publication strategy
- 11. Things to consider October 27, 2014Lyssa Friedman 11
- 12. Health Economics • Payer modeling • Verify unmet need • Verify market October 27, 2014 Lyssa Friedman 12 Study design Why commercialize? What business benefit?
- 13. Health Economics: Consider the payer Market Payer landscape Target price Revenue targets Product positioning COGS October 27, 2014 Lyssa Friedman 13
- 14. Analytic Validation • Verify assay robustness • Commercial readiness • CLIA registration October 27, 2014 Lyssa Friedman 14 Study design Will the assay work? Is it submission-ready?
- 15. Analytic Validation Consider the customer Variability in handling at point of collection Interruptions in current practice flow Requisition form design Patient report design October 27, 2014 Lyssa Friedman 15
- 16. Clinical Validation Diagnostic Clinical Performance • Establish test performance • Gold standard • Intended use • Support • FDA submission • CLIA registration (some states) • Publication plan • Physician adoption • Guideline acceptance • Payer adoption October 27, 2014 Lyssa Friedman 16 Study design Will the data be published? Will physicians order the test? Will it withstand regulatory scrutiny? Will payers reimburse?
- 17. Clinical Validation/Diagnostic Performance Consider the business and regulatory needs Engage agency prior to initiating study Clarify intended use Identify a reference (gold) standard Consistent with practice Recognized by community/literature Enroll subjects that represent target patients Select sites that support intended use Pre-specify Sample collection/handling procedures Analysis methods Use of archived samples (if needed) October 27, 2014 Lyssa Friedman 17
- 18. Clinical Outcomes Clinical Utility • Patient outcome based on test use and results • Randomized control trial • Changes in physician behavior October 27, 2014 Lyssa Friedman 18 Study design What difference does it make to patient outcomes? What difference does it make to healthcare costs?
- 19. Clinical Outcomes/Clinical Utility Consider the payer Consider the patient Effect on patient outcomes Ethics of randomized controlled trial when marketed test available Effect on healthcare costs October 27, 2014 Lyssa Friedman 19
- 20. October 27, 2014 Lyssa Friedman 20 Study design Publication strategy Physician adoption Guideline acceptance Positive coverage decisions Customer engagement Revenues
- 21. 21 • ACCE Evaluation Process for Genetic Testing • Analytic validity • Clinical validity • Clinical utility • Ethical/legal/social • Publications in these areas build the reimbursement case Centers for Disease Control (CDC) Office of Public Health Genomics Evaluation of Genomic Applications in Practice and Prevention (EGAPP™) Initiative Publication strategy October 27, 2014 Lyssa Friedman
- 22. FDA Regulation of LDTs October 27, 2014Lyssa Friedman 22
- 23. FDA regulation of LDTs October 27, 2014 Lyssa Friedman 23 IVD LDT • CLIA (if LDT) • FDA • CLIA • Physician adoption • Reimbursement • FDA • Physician adoption • Reimbursement • Quality system • SOPs • Quality for CLIA • SOPs • Physician adoption • Economic modeling • Clinical utility studies • FDA • Physician adoption • Economic modeling • Clinical utility studies Analytic validation Clinical validation Documentation Intended use
- 24. Keep doing what you’re doing well Well-designed studies Clean samples and clinical annotation Clearly defined gold standard SOPs and pre-specified protocols Culture of design review October 27, 2014 Lyssa Friedman 24
- 25. Site Selection October 27, 2014Lyssa Friedman 25
- 26. Site Selection What are the goals? Study completion Accrual numbers Adherence to Good Clinical Practices Pristine samples Verifiable data Strategic relationship building October 27, 2014 Lyssa Friedman 26 Key opinion leader Collaborator Influencer Geographic significance
- 27. Case Study • 142 genes, microarray • Gold standard: histopathology (central panel) • Optimized for sensitivity/NPV – safe to avoid surgery • Company launched 2008 (IVDMIA) – design control implemented • Product launched as LDT 2011 • Key publications – 2010: first clinical validation – 2011: economic model/cost effectiveness (replacement for surgery) – 2012: second clinical validation, clinical utility (change in recommendation for surgery) – 2013: second clinical utility, limits of gold standard October 27, 2014 Lyssa Friedman 27 Thyroid nodule Fine needle aspirate biopsy Cytology Indeterminate 15-30% Diagnostic surgery Benign >50% Afirma GEC 94% NPV
- 28. Case Study: Success/Challenges Things that worked • Funded on business plan – agnostic regarding clinical platform • All-hands focus on rapid first product to market • Aggressive publication plan – never missed an abstract deadline! • Small niche market – close relationship with major KOLs and influencers, who were clinical sites, early customers, authors/speakers • “Boutique” approach to managing clinical sites, even if challenging • Two clinical validation studies using one amended clinical sample collection protocol • IVDMIA era – design control Lessons learned • “Tarnished” gold standard required publication, customer education • NPV as key message – refuted based on practice variations in malignancy prevalence significantly October 27, 2014 Lyssa Friedman 28
- 29. What does success look like? Design a business plan that drives R&D and product development Build early relationships with KOLs and influencers Strategize cross-functionally from the beginning Stick to a long-term strategic publication plan Commit to a quality system (LDTs) Create a culture of design review with external reviewers (LDTs) Adhere to a milestone-based plan aligned with corporate goals Maintain relationships with strategic sites – even if challenging Never miss an abstract submission deadline Partner with marketing/sales group to transition sites to early customers October 27, 2014 Lyssa Friedman 29
- 30. What are lessons learned? Be willing to change thinking quickly and flexibly Challenge market research assumptions when clinical operations offers counter-intel Don’t underestimate the competition and the naysayers Stress test the product messaging with regard to test performance October 27, 2014 Lyssa Friedman 30
Editor's Notes
- As VP clinical operations at Veracyte, I helped bring the company’s beachhead product to market As an independent consultant, I help early-stage diagnostic companies design a roadmap for the product development lifecycle. This includes developing reimbursement and publication strategies, and designing studies to help achieve goals quickly and efficiently. The companies I work with have developed or are developing LDTs, and we are navigating the shifting landscape regarding FDA oversight of LDTs My point of view is supporting startup LDTs who have traditionally been regulated under CLIA only, and for the most part my comments are focused on a US market
- HDE: Humanitarian Device Exemption
- example: urine test
- Needed: SOPs pre-specified protocols and plans Recommend culture of design or phase review – avoid drinking the Kool-aid
- Keep doing what you’re doing well
- Clinical and medical affairs milestones that tracked both cross-functionally and upward to company goals Clinical operations guiding principles: Clinical operations is the brand These sites will be our customers in about 5 minutes Handheld strategic sites to maximize their success
- Anticipate what others will say and publish about the space Veracyte based marketing on high NPV, which had very high number. This was in an environment where clinicians assumed the gold standard, histopathology, was 100%, though our data showed (and was later published) that pathology concordance was 96-97% This was criticized at meetings and in publications as being too dependent on disease prevalence in specific clinical environment. perhaps a sensitivity message could have had more sustained power, despite the fact that it was slightly lower (92%)