managing complexity towards submission
play

Managing complexity towards submission Phuse US Connect 2019-02-27 - PDF document

3/15/19 Managing complexity towards submission Phuse US Connect 2019-02-27 Johannes Ulander 2019-02-27 Outline Managing complexity Challenges in study setup and reporting ... and submitting Biomedical Concepts: What?


  1. 3/15/19 Managing complexity towards submission Phuse US Connect 2019-02-27 Johannes Ulander 2019-02-27 Outline • Managing complexity • Challenges in study setup … • … and reporting • ... and submitting • Biomedical Concepts: What? Why? • Summary Suite Picture by @cyclingalps, www.cycling-challenge.com 1

  2. 3/15/19 Overview A3 - A suite of tools • An MDR at the center providing a single source of knowledge • Terminology • Biomedical Concepts (BCs) • Forms • SDTM domains • Provides an API to other tools • A study workbench tool • To construct clinical studies • And automate define creation • A tool to generate SDTM datasets (planned) • Then expand across lifecycle • Also, not shown, an experimental tool linking healthcare and clinical research prototyping the SDTM auto generation 3 Study Setup Translating study specifications to reality Is the blood pressure to be measured Which units to use – STANDING, SITTING Which tests are Which SDTM domains are do we allow for or SUPINE – what being measured needed for this study – do multiple units or just should be on the CRF? in BLOOD, we need a new or where one? Which ones are Do we need a new SERUM, or does this information applicable for the CRF? PLASMA belong in SDTM domains? test? Do we need a new annotation? Protocol Measurements V1 V2 V3 V4 Demographics x Vital signs x x CT Scan x Dosing x x x EQ-5D x x Lipids x x Haemoglobin x x Adverse Events x x x x EDC, Labs etc. protocol -> data collection -> reporting 4 2

  3. 3/15/19 Study Setup Standards … but not connected Standard Library of Mappings CRFs and aCRFs Silo Silo Silo Protocol Measurements V1 V2 V3 V4 Demographics x Vital signs x x CT Scan x Dosing x x x EQ-5D x x Lipids x x Haemoglobin x x Adverse Events x x x x EDC, Labs etc. protocol -> data collection -> reporting 5 No link to Study Setup terminology Standards … but not connected No link to terminology Standard Library of Mappings CRFs and aCRFs Silo Silo No link to Silo annotation Protocol Measurements V1 V2 V3 V4 Demographics x Vital signs x x CT Scan x Dosing x x x EQ-5D x x Lipids x x Haemoglobin x x Adverse Events x x x x EDC, Labs etc. protocol -> data collection -> reporting 6 3

  4. 3/15/19 Study Setup Challenge – maintaining transformations and traceability effectively Protocol CRF SDTM Measurements V1 V2 V3 V4 Demographics x Transformation Transformation Temperature Vital signs x x ### C or F CT Scan x Dosing x x x Diastolic Traceability Traceability EQ-5D x x ### mmHg Lipids x x Systolic Haemoglobin x x ### mmHg Adverse Events x x x x Biomedical Concepts (BCs) break down the “silos” and facilitate a consistent way of collecting the data across studies 7 What is a Biomedical Concept? Clinical view • A collection of variables Systolic Blood Pressure • A real-world entity Value Test Result mmHg Units • Something a physician would Standing recognize Sitting Position Supine Arm Location Left Right Laterality 8 4

  5. 3/15/19 What is a Biomedical Concept? Data Manager / Programmer view • Some more detail Systolic Blood Pressure C25298 Value • Linked to the CDISC terminology Format = 3.0 Test Result mmHg C49670 Units • Consistent information with views Standing tailored to user C62166 Sitting C62122 Position Supine C62167 • Present in a way that is meaningful to Arm C32141 Location the consumer Left C25229 Right Laterality C25228 9 Biomedical Concepts Implementation 1 0 1 0 5

  6. 3/15/19 Refer domains to Biomedical Concepts Where does the collected information belong? • Multiple domains can refer to the same BC. • The context of the study will then determine which domain is used • IMPORTANT: The domain refers to the BC. They BC DOES NOT refer to the domain. Can be used with any appropriate SDTM domain of the correct class 11 Biomedical Concepts Domain relationship • We can automatically add these “mappings” because the BC is linked to a domain --TEST / --TESTCD Systolic Blood Pressure --ORRES C25298 Value Format = 3.0 • In reality linked to a class hence the “--” --ORRESU Test rather than “VS” Result mmHg C49670 Units Standing --POS C62166 Sitting C62122 Position Supine C62167 Arm C32141 Location Left --LOC C25229 Right Laterality C25228 --LAT 12 6

  7. 3/15/19 Biomedical Concepts Domain relationship We use a framework (BRIDG) and templates • to ensure consistency of approach 13 Why bother? Consistency across the life-cycle • BC is a standardised definition that can be used for various purpose and facilitates faster and more flexible study setup CRF/aCRF Biomedical Therapeutic Protocol Areas Concepts Sharing data collection definitions 14 7

  8. 3/15/19 Standardisation - From Forms to BCs Flexible form creation • Ensures we have compliance with the CDISC standards 15 Demo 1 6 8

  9. 3/15/19 Standardisation - From Forms to BCs Flexible form creation The flexibility is achieved on 3 levels: 1. The user can decide which BCs to add to a Vital Signs form, i.e. which questions are included in the vital sign page 2. The user can enable/disable data elements in a BC, e.g. decide not to use the baseline flag question if not relevant, i.e. what the level of response is needed for the question 3. The user can decide if questions and/or responses are optional/mandatory 17 Upstream Standardisation BCs in Protocols Assessments BC Screening Baseline Treatment EndOfTrial Demographics Age x Assessments Race Screening Baseline Treatment EndOfTrial Date of Birth Demographics x Sex Vital signs Height x x Vital signs x x Weight Heart Rate CT Scan x Temperature Dosing x x x Systolic Blood EQ-5D x x Pressure Diastolic Blood Lipids x x Pressure Haemoglobin x x CT Scan CT scan timing x Adverse Events x x x x Dosing Number of dosing x x x units EQ-5D Mobility x x Lipids LDL x x HDL Total Cholesterol Haemoglobin Haemoglobin x x Adverse Events x x x x Adverse Events 18 9

  10. 3/15/19 Upstream Standardisation BCs in Protocols Assessments BC Screening Baseline Treatment EndOfTrial Demographics Age x Race Date of Birth Sex Vital signs Height x x Weight Heart Rate Temperature Systolic Blood Pressure Diastolic Blood Pressure CT Scan x CT scan timing • Forms containing these BCs will Dosing Number of dosing x x x then be listed and the CRF designer units EQ-5D Mobility x x would then pick the appropriate Lipids LDL x x CRF. HDL Total Cholesterol Haemoglobin Haemoglobin x x • CRF reviewed early by the study Adverse Events x x x x Adverse Events group allowing for adjustments to the protocol if needed. 19 Sharing BCs Common definitions based on templates • BCs should be made based on templates to ensure consistency. • Contributors could be both pharmaceutical companies, CROs as well as regulators. • Sharing the definition of a BC does not disclose any intellectual property or any sensitive data that needs de-identification 20 10

  11. 3/15/19 Sharing BCs Common definitions based on templates • Distribute the information via SAS tables, linked data, spreadsheets (MS Excel), database loads etc. • Share with CDISC Export Export Export 21 BCs For Defining Therapeutic Areas (TAs) Current metadata for TAs – a BC template Doesn’t define the possible/required tests for the TA Domain Question Text Prompt SDTM or BRIDG Definition Codelist CRF Core Data CDASH Completion Type Variable Instructions Name VS What is the <Test> VSTEST PerformedActivity.Per Verbatim VSTEST Record the Highly Char vital sign test formedObservation > name of the name of the Recommended name? PlannedActivity > test or vital sign test StudyActivity > examination if not pre- DefinedActivity.Define used to obtain printed on the dObservation.nameC the CRF. ode.CD.originalText measurement or finding. Doesn’t define the {VSTEST} possible/required (See Section 2.2.) units for the test VS What was the <Units> VSORRESU PerformedObservatio Original units VSRESU 'Record or Recommended Char unit of the n > in which the select the unit /Conditional measurement? PerformedObservatio data were of measure nResult.value.ANY=> collected. associated PQ.translation.PQR.c {VSRESU} with the test, ode (See Section if not pre- 2.2.) printed on the CRF.' 22 11

  12. 3/15/19 BCs For Defining Therapeutic Areas (TAs) TAs as a set of BCs • TAs could be defined as a set of BCs • Also a set of relationships with target domains • Those domain relationships might change in TAs! 23 Demo 2 4 12

  13. 3/15/19 Demo 2 5 Summary S-Cubed A3 suite • Solid foundation based around the MDR • The old world can co-exist with BC’s • Web-based tools • Allow for other tools to access exports or APIs • Deployment should not require large-scale IT projects such that all sponsors can benefit from high-quality definitions. 2 6 13

  14. 3/15/19 Thank you for listening www.s-cubed-global.com CONTACT: Johannes Ulander ju@s-cubed.dk 14

Recommend


More recommend