The SHARPn phenotyping funnel. Phenotype specific patient cohorts

Transcription

1 The SHARPn phenotyping funnel Mayo Clinic EHR data QDMs CEMs Intermountain EHR data DRLs Phenotype specific patient cohorts [Welch et al., JBI 2012; 45(4):763-71] 2012 MFMER slide-1

2 Algorithm Development Process - Modified Phenotype Algorithm Semi-Automatic Execution Standardized and structured representation of phenotype definition criteria Use the NQF Quality Data Model (QDM) Rules Evaluation Visualization Conversion of structured phenotype criteria into executable queries Use JBoss Drools (DRLs) Standardized representation of clinical data Create new and re-use existing clinical element models (CEMs) Transform Transform Data Mappings NLP, SQL [Welch et al., JBI 2012; 45(4):763-71] 2012 MFMER slide-2

3 Clinical Element Models Higher-Order Structured Representations [Stan Huff, IHC] 2012 MFMER slide-3

4 CEMs available for patient demographics, medications, lab measurements, procedures etc. [Stan Huff, IHC]

5 SHARPn data normalization pipeline - I 2012 MFMER slide-5

6 SHARPn data normalization pipeline - II CEM MySQL database with standardized and normalized patient data [Welch et al., 2012 JBI 2012; MFMER 45(4):763-71] slide-6

7 Algorithm Development Process - Modified Semi-Automatic Execution Standardized and structured representation of phenotype definition criteria Use the NQF Quality Data Model (QDM) Rules Phenotype Algorithm Transform Transform Visualization Data Evaluation Standardized representation of clinical data Create new and re-use existing clinical element models (CEMs) Mappings NLP, SQL [Welch et al., JBI 2012; 45(4):763-71] 2012 MFMER slide-7

8 Example algorithm: Hypothyroidism 2012 MFMER slide-8

9 NQF Quality Data Model (QDM) Standard of the National Quality Forum (NQF) A structure and grammar to represent quality measures and phenotype definitions in a standardized format Groups of codes in a code set (ICD-9, etc.) "Diagnosis, Active: steroid induced diabetes" using "steroid induced diabetes Value Set GROUPING ( ) Supports temporality & sequences AND: "Procedure, Performed: eye exam" > 1 year(s) starts before or during "Measurement end date" Implemented as a set of XML schemas Links to standardized terminologies (ICD-9, ICD-10, SNOMED-CT, CPT-4, LOINC, RxNorm etc.) 2012 MFMER slide-9

10 Example: Diabetes & Lipid Mgmt. - I 2012 MFMER slide-10

11 Example: Diabetes & Lipid Mgmt. - II Human readable HTML 2012 MFMER slide-11

12 Example: Diabetes & Lipid Mgmt. - III 2012 MFMER slide-12

13 Example: Diabetes & Lipid Mgmt. - IV 2012 MFMER slide-13

14 Example: Diabetes & Lipid Mgmt. - V Computer readable XML (based on HL7 RIM semantics) 2012 MFMER slide-14

15 Algorithm Development Process - Modified Phenotype Algorithm Semi-Automatic Execution Standardized and structured representation of phenotype definition criteria Use the NQF Quality Data Model (QDM) Rules Evaluation Visualization Conversion of structured phenotype criteria into executable queries Use JBoss Drools (DRLs) Standardized representation of clinical data Create new and re-use existing clinical element models (CEMs) Transform Transform Data Mappings NLP, SQL [Welch et al., JBI 2012; 45(4):763-71] 2012 MFMER slide-15

16 JBoss Drools rules management system Represents knowledge with declarative production rules Origins in artificial intelligence expert systems Simple when <pattern> then <action> rules specified in text files Separation of data and logic into separate components Forward chaining inference model (Rete algorithm) Domain specific languages (DSL) 2012 MFMER slide-16

17 Example Drools rule {Rule Name} rule Glucose <= 40, Insulin On when {binding} {Java Class} {Class Getter Method} $msg : GlucoseMsg(glucoseFinding <= 40, currentinsulindrip > 0 ) then {Class Setter Method} glucoseprotocolresult.setinstruction(glucoseinstructions GLUCOSE_LESS_THAN_40_INSULIN_ON_MSG); end Parameter {Java Class} 2012 MFMER slide-17

18 [Li et al., AMIA 2012; (Epub ahead of print)] 2012 MFMER slide-18

19 2012 MFMER slide-19 The executable Drools workflow [Li et al., AMIA 2012; (Epub ahead of print)]

20 [Endle et al., AMIA 2012; (Epub ahead of print)]

21 1. Converts QDM to Drools 2. Rule execution by querying the CEM database 3. Generate summary reports

22

23

24