Challenges and Opportunities for Data Reuse Ronald Cornet, PhD Dept. of Medical Informatics Academic Medical Center – University of Amsterdam.

Challenges and Opportunities forData Reuse

Ronald Cornet, PhD

Dept. of Medical Informatics

Academic Medical Center – University of Amsterdam

Overview

Why Data ReuseCategories of Data ReuseChallengesOpportunities

Data Reuse

Many data are collected in the care processMany data are needed e.g., for research

purposesTwo options:

Dedicated collection of purpose-specific data Re-using data from the care process

Dedicated Data Collection

Allows for optimally defined data setBut…May lead to redundant data collectionCan not always be performed (e.g.,

retrospectively)

Re-using Data from Care Process

Benefits from readily available dataBut…Data may be incompleteData may lack detailData may be biasedData may be incomparable

Seeking a balance

Data from the clinical care process often not fit for reuse

Dedicated data collection costly or impossible Recording “everything” about “everyone” is

impossible How to collect data in the primary care process

that can be reused with minimal drawbacks (e.g., bias, detail)?

Example – cost analysis

Cost analysis based on care recordsOnly recorded care is taken into account

for cost analysisE.g., during a busy night, a severely

traumatized patient enters emergency department, receives blood, resuscitation, and dies, is transported to morgue; no record created

Secondary Uses and Re-uses of Data

Protect and enhance public healthDevelop security and confidentiality

algorithms and test de-identification routines

Conduct researchCreate and maintain terminology and

representation formalisms

Secondary Uses and Re-uses (II)

Develop and apply decision support for health care providers

Support quality of patient careImprove patient safetyManage personal health

Secondary Uses and Re-uses (III)

Educate and credential healthcare providers and assess training activities

Analyze and Manage FinancesDetect fraud and illicit activityIdentify markets and promote sales

Examples

Public health

E.g., setting up disease-specific registries Cardiosurgery Renal Failure Intensive Care

Research

E.g., development of prognostic models

Examples

Decision support

E.g., develop algorithms, rules, and alertsQuality of patient care

Manage quality and outcomes» More on this tomorrow (Dr. De Keizer)

Manage staffing and resources Develop and assess quality indicators

Factors influencing Authorization for secondary Use (I)

Identification Status: Patient identifiable, de-identified, anonymized

Consent provided at the time of data collection Demographic representation, e.g., age, race,

gender Focus on a vulnerable population (e.g., prisoners) Original collector and aggregator of the data

(government, private)

Factors influencing Authorization for secondary Use (II)

Proposed secondary user of the data, e.g., government, academic institution

Funding source for secondary use Financial compensation to data collector for

providing data to a second party Beneficiary of secondary use, e.g., society,

researcher, academic institution Disclosure of secondary use, e.g., public

disclosure of results and/or methods

ERA-EDTA RegistryData about patients on End-Stage Renal

FailurePatient data recorded in renal centersData from centers to national registriesData from national registries to ERA-EDTAERA-EDTA shares outcomes with USRDS,

ANZDATA

Data Reuse Case StudyData Re-Re-Reuse Case Study

Challenges

Opportunities

Adjust data to the intended reuse

orAdjust the intended reuse to the data

Conclusion

Round-up

There are many kinds of data reuse

More information

[email protected]