Methodologies for CRNs: Can Statisticians, Epidemiologists, and Machine Learners Play in the Same Sand Box? Perspectives from OHDSI

Size: px

Start display at page:

Download "Methodologies for CRNs: Can Statisticians, Epidemiologists, and Machine Learners Play in the Same Sand Box? Perspectives from OHDSI"

Megan Thompson
5 years ago
Views:

1 Methodologies for CRNs: Can Statisticians, Epidemiologists, and Machine Learners Play in the Same Sand Box? Perspectives from OHDSI Patrick Ryan, PhD Janssen Research and Development 2 October 2015

2 Introducing OHDSI The Observational Health Data Sciences and Informatics (OHDSI) program is a multistakeholder, interdisciplinary collaborative to create open-source solutions that bring out the value of observational health data through large-scale analytics OHDSI has established an international network of researchers and observational health databases with a central coordinating center housed at Columbia University

OHDSI: a global community OHDSI Collaborators: >100 researchers in academia, industry and government Statisticians, Epidemiologists, Informaticists, Machine Learners,

3 OHDSI: a global community OHDSI Collaborators: >100 researchers in academia, industry and government Statisticians, Epidemiologists, Informaticists, Machine Learners, Clinical Scientists all playing in the same sandbox >10 countries OHDSI Distributed Data Network: >40 databases standardized to OMOP common data model >500 million patients

4 OHDSI ongoing collaborative activities Methodological research Open-source analytics development Clinical applications Observational data management Data quality assessment Common Data Model evaluation ATHENA for standardized vocabularies WhiteRabbit for CDM ETL Usagi for vocabulary mapping HERMES for vocabulary exploration ACHILLES for database profiling Clinical characterization Phenotype evaluation CIRCE for cohort definition CALYPSO for feasibility assessment HERACLES for cohort characterization Chronic disease therapy pathways Population-level estimation Empirical calibration LAERTES for evidence synthesis CohortMethod SelfControlledCaseSeries SelfControlledCohort TemporalPatternDiscovery HOMER for causality assessment Patient-level prediction Evaluation framework and benchmarking PatientLevelPrediction APHRODITE for predictive phenotyping PENELOPE for patient-centered product labeling

expansion based on the experiences of the OMOP community who bring on a

5 OMOP CDMv2 Evolution of the OMOP Common data model OMOP CDM now Version 5, following multiple iterations of implementation, testing, modifications, and expansion based on the experiences of the OMOP community who bring on a growing landscape of research use cases. OMOP CDMv4 OMOP CDMv5 Page 5

Standardized clinical data Drug safety surveillance Device safety surveillance Vaccine safety surveillance Comparative effectiveness One model, multiple use cases Health economics Quality of care

6 Standardized clinical data Drug safety surveillance Device safety surveillance Vaccine safety surveillance Comparative effectiveness One model, multiple use cases Health economics Quality of care Clinical research Person Observation_period Specimen Death Standardized health system data Location Care_site Provider Standardized meta-data CDM_source Concept Visit_occurrence Procedure_occurrence Drug_exposure Device_exposure Condition_occurrence Measurement Note Observation Fact_relationship Payer_plan_period Procedure_cost Drug_era Visit_cost Drug_cost Device_cost Cohort Cohort_attribute Condition_era Dose_era Standardized health economics Standardized derived elements Vocabulary Domain Concept_class Concept_relationship Relationship Concept_synonym Concept_ancestor Source_to_concept_map Drug_strength Cohort_definition Attribute_definition Standardized vocabularies

7 Patients with total knee replacement Administrative claims: Truven MarketScan CCAE Electronic health records: UK CPRD Hospital billing records: Premier

9 Evidence OHDSI seeks to generate from Clinical characterization: observational data Natural history: Who are the patients who have exposure to cranial perforaters? Quality improvement: what proportion of patients who have exposure to cranial perforaters experience seizure? Population-level estimation Safety surveillance: Do cranial perforaters cause seizure? Comparative effectiveness: Do cranial perforaters cause seizure more than non-perforated craniectomy? Patient-level prediction Precision medicine: Given everything you know about me and my medical history, if I have exposure to a cranial perforator, what is the chance that I am going to have seizure in the next year? Disease interception: Given everything you know about me, what is the chance I will develop the need for a cranial perforater?

10 Cranial perforator in the OHDSI vocabulary as a device!...but none of our source data map to it

11 Causal inference problem: Exposure? Covariates?? Outcome Problem: The data I have doesn t allow me to answer the question I want Stop and do nothing Wait to get new data Change your question

12 Opportunities for changing the question Clinical characterization: Natural history: Who are the patients who have exposure to cranial perforaters a craniectomy with burr hole procedure? Quality improvement: what proportion of patients who have exposure to cranial perforaters a craniectomy with burr hole procedure experience seizure? Population-level estimation Safety surveillance: Does cranial perforaters a craniectomy with burr hole procedure cause seizure? Comparative effectiveness: Do cranial perforaters a craniectomy with burr hole procedure cause seizure more than non-perforated craniectomy? Patient-level prediction Precision medicine: Given everything you know about me and my medical history, if I have exposure to a cranial perforaters a craniectomy with burr hole procedure what is the chance that I am going to have seizure in the next year? Disease interception: Given everything you know about me, what is the chance I will develop the need for a cranial perforaters a craniectomy with burr hole procedure?

13 Feasibility assessment: cohort definition

14 Cohort characterization

15 Cohort exploration

16 What can we learn: Scenario 1 : mature technology Clinical characterization of indicated population and users of mature technology Population-level estimation of new users of mature technology vs. alternative treatments or non-exposure Patient-level prediction to summary past experience of patients like me

17 What can we learn: Scenario 2: New technology Clinical characterization of indicated population and users of current treatments PLUS new technology Population-level estimation of new users of new technology vs. alternative treatments or non-exposure Patient-level prediction to summary past experience of patients like me

18 Concluding thoughts We need to do more to understand what we can learn from the data we ve already got (changing the question as necessary) and get more data to address what we know we re missing Product lifecycle continuum (new -> mature product) can fit with same evidence generation continuum: clinical characterization population-level estimation patient-level prediction.. same analytics but different use cases The entire evidence lifecycle (methods research open-source analytics development clinical application) needs to be cultivated to support robust product surveillance and evaluation

19 Join the journey Interested in OHDSI? Questions or comments? Contact: 19

Applying Hill's criteria as a framework for causal inference in observational data

Applying Hill's criteria as a framework for causal inference in observational data Patrick Ryan, PhD Janssen Research and Development Columbia University Medical Center 10 June 2015 Perspectives on the