The Right Data for the Right Questions:

Transcription

1 The Right Data for the Right Questions: Evidentiary Needs as a Guide to Data Source Selection David Thompson, PhD Senior Vice President Louise Parmenter, PhD Global Head of Operations Real-World & Late Phase Research April 2014 Copyright 2013 Quintiles

2 Presentation Objectives Pragmatic Approaches to Real-World Data Source Selection & Use Real-world evidentiary needs across pharmaceutical product lifecycle Strengths & limitations of alternative real-world data sources Evidentiary needs as a guide to optimal choice of data sources With proliferation of Big Data & machine learning tools, the importance of forethought can get lost. But Big Data without intelligent analytics will fail to deliver meaningful insights, regardless of computer processing power. 2

3 Real-World Evidentiary Needs Across the Pharmaceutical Product Lifecycle 3

4 Evidentiary Needs Real-World Data Support Variety of Activities Across the Product Lifecycle Preclinical Phase I-II Phase III Peri-Launch Phase IV Market Sizing Market Landscape Competitor Reconnaisance Registries Safety Surveillance Unmet Need Economic Burden Patient Burden Global BOIs Continuous Monitoring Patient Profilng Disease History Treatment Patterns Target Product Profile Tailored Therapeutics Early Modeling Model Refinements CE/BI Modeling CEAs & BIAs Health Technology Assessments Endpoint Development Endpoint Assessment Labeled Claims New Indications Instrument Validation Piggyback Evaluations Global Value Dossiers Comparative Effectiveness Pricing & Reimbursement Risk Sharing Arrangements 4

5 Strengths & Limitations of Alternative Real-World Data Sources 5

6 Real-World Data Fundamentals Primary Sources as Defined by ISPOR Supplements to traditional RCTs: > Commonly known as trial-based or piggyback evaluations Large simple trials: > Commonly known as pragmatic or naturalistic trials Registries: > Include prospective cohort studies Administrative data: > Also known as claims data Health surveys: > Useful for basic epidemiologic data or macro-level views on utilization Electronic health records & medical chart review: > Electronic health records also called electronic medical records (EMRs) > Better to separate EHR/EMR data from medical chart review 6

7 Classifying Real-World Data Sources Two-by-Two Typology Facilitates Critical Review Retrospective Designs Prospective Designs Primary Data Collection Medical Chart Review RCT Piggybacks Pragmatic Trials Registries Health Surveys Secondary Data Collection Administrative Claims EMR Automated EMR Data Feeds 7

8 Classifying Real-World Data Sources Strengths & Limitations of Retrospective Sources Primary Data Collection Retrospective Designs Medical Chart Review Key Strengths: > Data already exist in charts / computer systems Economy of data collection > Potential for enormous sample sizes almost instantaneously > Data reflect real-world patterns of care, not affected by study protocol > Data mining approaches can uncover key relationships not on clinical radar Secondary Data Collection Administrative Claims EMR Key Limitations: > Data already exist in charts / computer systems What you see is what you get > Potential for enormous sample sizes, yes, but not for products in development > Numerous sources of confounding & bias, not all of which can be controlled 8

9 Classifying Real-World Data Sources Strengths & Limitations of Prospective Sources Primary Data Collection Prospective Designs RCT Piggybacks Pragmatic Trials Registries Health Surveys Key Strengths: > High degree of control over what data are collected and how > For stand-alone studies, fine-tuning of sample size is possible > For stand-alone studies, upfront control of confounding & bias is possible, even for real-world care patterns Secondary Data Collection Automated EMR Data Feeds Key Limitations: > Prospective data more costly than retrospective sometimes by orders of magnitude > For RCT piggybacks, no fine-tuning of sample size is possible and statistical power is usually lacking for RW measures > For RCT piggybacks, protocol-driven care undermines generalizability to real world 9

10 Classifying Real-World Data Sources How Real-World Sources Array Across the Cost Spectrum $250k $500k $750k $1million $10s of millions Claims Distributed Data Networks EMR Chart Review Retro-Pro Hybrids Registries Pragmatic Trials Studies involving different kinds of RWD sources naturally array across the cost spectrum according to time & effort in data collection Retrospective sources congregate on the lower end, purely prospective sources on the higher end Claims Emerging approaches involving retro-to-pros hybrid designs are in between 10

11 Evidentiary Needs as a Guide to Selection of Real-World Data Sources 11

12 The Right Data for the Right Questions What are the right questions? Heterogeneity Complexity Change What is the right data? 12

13 Systematic Approach How to use evidentiary needs to guide optimal choice of real-world data sources Systematically determine evidentiary needs Define research question Prepare protocol synopsis & data elements Determine primary, secondary or hybrid research approach Systematically evaluate realworld data sources Finalize protocol & study plan 7 8 Conduct research Disseminate evidence 13

14 Systematic Approach How to use evidentiary needs to guide optimal choice of real-world data sources Systematically determine evidentiary needs Define research question Prepare protocol synopsis & data elements Determine primary, secondary or hybrid research approach Systematically evaluate realworld data sources Finalize protocol & study plan 7 8 Conduct research Disseminate evidence 14

15 15 Systematically Determine Evidentiary Needs Current Available Evidence Standard of Care Ongoing Country Studies Regulator Needs HTA Needs Prescriber & Patient Insights Commercial Value Proposition Literature review. Internal clinical trials data Physician interviews Research registers Regulatory intelligence Regulatory advice HTA intelligence Payer advice KOL interviews Focus groups Patient surveys Internal evaluation Affiliate surveys Needs Approach

16 Systematic Approach How to use evidentiary needs to guide optimal choice of real-world data sources Systematically determine evidentiary needs Define research question Prepare protocol synopsis & data elements Determine primary, secondary or hybrid research approach Systematically evaluate realworld data sources Finalize protocol & study plan 7 8 Conduct research Disseminate evidence 16

17 Systematically Evaluate Real-world Data Sources Clinical practice Existing RWD sources Recommendations Gain understanding of which data elements are collected in clinical practice Primary research: Interviews with practicing physicians/ KOLs Compile inventory of existing Increasing RWD sources Demand for Safety Data, Product Knowledge Secondary research: identify existing RWD sources Primary research: conduct a survey among identified RWD sources Understand the best resources to collect specific data elements of interest 17

18 Real-world Database Assessment 1. Identify databases 2. Ascertain interest 3. Survey database attributes 4.Score & weight 5. Analyse Local contacts Internal knowledge Literature searches Initial outreach for interest in commercially sponsored research Survey databases: data elements, ownership of the analyses, contracting, price, ability to contact sites etc Values will be allocated to each attribute. Attributes will be weighted Analyse attributes against research interest 18

19 1. Identify Databases International Society for Pharmacoeconomics and Outcomes Research, 17th Annual Meeting, June 2-6, 2012 Washington, DC, USA 19

20 2. & 3. Ascertain Interest and Survey General database attribute checklist General information: Description and type of database (in patient, out patient, claims data, other) Database content: Patient demographics (address, age, gender, race) Outcome data (clinical endpoints, mortality, lab reports, other) Treatment data (prescriptions, surgires, other) Resource utilization (admissions, referrals, doctor visits, other) Technical details: Questions determined according to evidence gaps Coding system (for diagnosis, prescriptions, and surgical procedures) Data linkage potential (unique identifiers, legal aspects) Access scheme ( open for public, for fees, ethical approval obligation to publish) Contact details International Society for Pharmacoeconomics and Outcomes Research, 17th Annual Meeting, June 2-6, 2012 Washington, DC, USA 20

21 4. Score & Weight Example Data element Availability of information in the RWD sources Amount of data elements collected in routine clinical practice Demographic data and medical history capture (age, gender, co-morbidities, CIRS) High High Disease characteristics capture (date of first diagnosis, etc) Medium High Survival status (date of death & cause of death) Medium High Treatment capture (name, treatment start date and treatment stop date, MIPI score, safety) Low Medium Assessment of response to treatment, date of disease progression and/or date of treatment failure) Low Low Medical service utilization (hospitalization and length of stay, types and number of lab test) Low Medium PRO (e.g. quality of life instruments) Low Low 21

22 5. Analyze Suitability for Research Question Assessment of databases for the capture of four major data groups: Demographic data Treatment data Clinical outcomes Detailed diagnosis data groups including patient demographics 3 data groups including patient demographics Can be considered for data linkage 28 International Society for Pharmacoeconomics and Outcomes Research, 17th Annual Meeting, June 2-6, 2012 Washington, DC, USA 4 data groups Minimum requirements for comparative effectiveness research 22

23 Systematic Approach How to use evidentiary needs to guide optimal choice of real-world data sources Systematically determine evidentiary needs Define research question Prepare protocol synopsis & data elements Determine primary, secondary or hybrid research approach Systematically evaluate realworld data sources Finalize protocol & study plan 7 8 Conduct research Disseminate evidence 23

24 Determine Primary, Secondary or Hybrid Research Approach Data elements of interest to answer the research question? Data elements collected routinely? Available in electronic databases? Available in localized clinic records? Data elements not collected routinely Secondary Research Database analysis Secondary Research Patient chart review Primary Research Prospective data collection 24

25 Systematic Approach How to use evidentiary needs to guide optimal choice of real-world data sources Systematically determine evidentiary needs Define research question Prepare protocol synopsis & data elements Determine primary, secondary or hybrid research approach Systematically evaluate realworld data sources Finalize protocol & study plan 7 8 Conduct research Disseminate evidence 25

26 Summary Real-world evidence provides valuable evidence to support healthcare stakeholder decision makers In a complex and evolving stakeholder and real-world data landscape it can be challenging to determine what evidence is needed and how to gather that evidence A systematic approach to understand evidentiary needs and real-world data options ensures that we always use the right data for the right questions 26

27 Thank You! 27

28 Contact Information David Thompson, PhD SVP & Head of EMR Data & Analytics Real-World & Late Phase Research Quintiles Louise Parmenter, PhD Global Head of Operations Epidemiology & Outcomes Research Real-World & Late Phase Research Quintiles T: E: T: + 44 (0) E: louise.parmenter@quintiles.com 28