330 likes | 361 Views
Representing Results of Genetic Tests Using LOINC. Stanley M. Huff, M.D. Intermountain Health Care coshuff@ihc.com. Outline. The data exchange environment HL7 basics LOINC basics LOINC names for results of genetic tests. Clinical Integration. Sunquest Lab. AGFA Radiology.
E N D
Representing Results of Genetic Tests Using LOINC Stanley M. Huff, M.D.Intermountain Health Care coshuff@ihc.com
Outline • The data exchange environment • HL7 basics • LOINC basics • LOINC names for results of genetic tests
Clinical Integration Sunquest Lab AGFA Radiology Tamtron Anatomic Pathology McKesson Pharmacy ARUP Blood Bank MIMIR Blood Gas Machines Dictaphone Varis Oncology MRS Mammography Logicare ER Computrition Dietary HELP (Inpatient HIS) 3M IDX (Outpatient) IDX Systems HDM & Medrec 3M ADT, Orders, Results, Billing Registration, Scheduling ADT, Billing Health Data Dictionary 3M DataGate Interface Engine STC Tuxedo ADT, Orders, Results, Billing ADT, Billing, Case Mix Billing & Financial IHC ADT, Results, Orders Registration, Scheduling Tuxedo DataStage CIS EMMI/EMR Database (HEMS) 3M Clinical Workstation 3M Tuxedo Data Warehouse IHC
Why coded data? • You can’t do synchronous decision support on free text • Misspellings • Homographs (homonyms) • Multiple terminologies • Outcomes • Knowledge discovery • Knowledge sharing (Arden, GLIF) • Flexibility in reports • Data re-use • Maintenance of the knowledge base
Typical Clinical Data Flow Standard Interfaces EMR Interface Engine
HL7 Lab Result Message MSH|^~\&|OADD||DADD||19941122100053||ORU^M01| PID|||661041||GARDNER^REED^M| PV1||I|E7^703^^LDS| OBR||^A000520|LYTES^Serum Electrolytes| OBX|1|NM|NAS^Serum Sodium|1|138|mmol/L| OBX|2|NM|K^Serum Potassium|1|3.2|mmol/L| OBX|3|NM|CL^Serum Chloride|1|114|mmol/L| OBX|4|NM|CO2^Serum CO2|1|24|mmol/L|
OBX: the flexible segment A code that identifies the datatype of OBX-5 OBX-5: Data Status A code that identifies the units of numerical data in OBX-5 Other data fields include: date of observation, identity of provider giving observation, normal ranges, abnormal flags OBX||NM|11289-6^^LN||38|C^^ISO+|||||F A code that identifies the data in OBX-5(Temp Reading)
OBX: with a coded value A code that identifies the datatype as a coded element The code is from LOINC The code is from SNOMED A code that identifies the data in OBX-5(Fetal Gender) OBX-5: Data A code for Female OBX||CE|11882-8^Fetal Gender^LN||T-D0AA0^Female^SMI|
So we are all using HL7, what is the problem? • Site 1: OBX|1|CE|ABO^ABO GROUP||O^Type O| • Site 2: OBX|1|CE|BLDTYP^ABO GROUP||TYPEO^Type O| • Site 3: OBX|1|CE|ABOTYPE^ABO GROUP||OPOS^Type O|
The goal • Site 1: OBX|1|CE|883-9^ABO GROUP||F-D1250^Group O| • Site 2: OBX|1|CE|883-9^ABO GROUP||F-D1250^Group O| • Site 3: OBX|1|CE|883-9^ABO GROUP||F-D1250^Group O| Agree on a universal coding system for clinical observations.
LOINC Objective Create universal observation identifiers (test codes) for use in data exchange standards that follow a name-value pair strategy (HL7, DICOM, ASTM, CEN, etc.).
Important facts • LOINC 2.08 • 24477 Laboratory LOINC codes • 5775 Clinical LOINC codes • LOINC is free for use • LOINC list servers at www.hl7.org • Download files and tools from Regenstrief Web Site • www.regenstrief.org/loinc/loinc.htm • RELMA – a tool for mapping local codes to LOINC
Laboratory LOINC Subject Areas • Chemistry • Urinalysis • Toxicology • Hematology • Microbiology • Antibiotic Susceptibilties • Immunology/Serology • Genetic testing
General Form of LOINC Names LOINC codes are created systematically using a six axis model <component> : <property> : <timing> : <system> : <scale> : <method> 2947-0 SODIUM :SCNC :PT :BLD :QN The first 5 parts are mandatory, but method is optional.
Objective of GeneTests-LOINC Collaboration • Initiated summer 2001 (Peter Tarczy-Hornoch MD) • Goal of the LOINC Committee: to develop formal names and codes for identifying clinical reports, observations, and measurements, including laboratory tests, to develop compatible data sets for electronic storage, transmission and display • Goal of GeneTests: Catalog and characterize available genetic testing and laboratories offering these tests • Goal of Collaboration: • Increase granularity of descriptive test data in GeneTests • Provide data necessary for generation of LOINC codes for available clinical genetic tests currently not in LOINC • Automatic linkage (WWW) between relevant LOINC and GeneTests database entries
Collaboration steps • Development by GeneTests of custom API to access selected GeneTests database content for LOINC (done) • Analysis by LOINC of reports and preliminary identification of additional data needed for LOINC names and codes (done) • GeneTests pilot poll of clinical labs (in process) • Commercial (n=5) & academic (n=5) • Request for additional information • Representative sample reports for all tests offered by lab • Both normal and range of abnormal sample reports • Scrubbed of patient identifying data • Joint development of formal methodology nomenclature • Final identification (joint) of data to request of labs • GeneTests solicitation of data from all labs • Joint establishment of auto-linkage • LOINC use of API & data to generate new codes/names
Nomenclature for Locating Genetic Defects Three types of nomenclatures for identifying the location of a genetic defect Designation Explanation p c g Identify the defect by codon by counting the amino acids in the protein produced by the gene counting the first amino acid.Identify the defect by counting nucleotides from the messenger RNA used to produce the protein with intron excluded. These will produce numbers 3x as large as those in the first method. Identify the defect by counting from the first nucleotide in the DNA as it exists as a gene natively in the chromosome with introns included.
List of single letter amino acid codes Amino Acid Code Amino Acid Code Alanine A Leucine L Arginine R Lysine K Asparagine N Methionine M Aspartic acid D Phenylalanine F Cysteine C Proline P Glutamic acid E Serine S Glutamine Q Threonine T Glycine G Tryptophan W Histidine H Tyrosine Y Isoleucine I Valine V
Specific disease gene mutations The pattern for the first part of the LOINC name is: <gene name> GENE. <mutation nomenclature>. <mutation and its location>
Factor V Leiden mutation • F5 GENE.P.R506Q • "F5" identifies the gene • "GENE" is a fixed part • "P" identifies the kind of mutation nomenclature (protein) • “R506Q” indicates that the amino acid arginine (R) is replaced by glutamine (Q) at codon #506.
Fully specified names (specific disease genes) • F5 GENE.P.R506Q:ARB:PT:BLD/TISS:ORD:MOLGEN • Synonyms = Factor V Leiden, Factor V resistance, APC resistance gene • HFE GENE.P.C282Y:ARB:PT:BLD/TISS:ORD:MOLGEN • Synonyms = HLA-H gene, hemochromatosis gene • CFTR GENE.P. F508 DEL :ARB:PT:BLD/TISS:ORD:MOLGEN • Synonyms = Cystic Fibrosis Transmembrane Regulator • Scale is ORD (ordinal), possible answers are: • no mutation found • heterozygous mutation (the mutation found in one gene) • homozygous mutation (the mutation was found in both genes in the gene pair)
Reporting of multiple Allels • APO E Allel 1:PRID:PT:BLD/TISS:NOM:MOLGEN • Synonyms: APO E Alzheimer's risk • Scale is now NOM (nominal) • Answers = E1, E2, E3, or E4 • APO E Allel 2:PRID:PT:BLD/TISS:NOM:MOLGEN • Answers = E1, E2, E3, or E4
Gene Mutation Analysis • CFTR MUTATION ANALYSIS :IMP:PT:BLD/TISS:NOM:MOLGEN • Synonyms = Cystic fibrosis transmembrane regulator • Answers: Identifiable Mutation, Not Identifiable Mutation • BRCA1 MUTATION ANALYSIS :IMP:PT:BLD/TISS:NOM:MOLGEN • Synonyms = breast cancer risk gene • Answers: Identifiable Mutation, Not Identifiable Mutation • CFTR MUTATIONS TESTED FOR: PRID:PT;BLD/TISS:NOM:MOLGEN • The answers could include: • Delta F508, G542X, R553X, W1282X, N1303K
Trinucleotide repeats • FRAXE GENE.CGG REPEATS: ARB:PT:BLD/TISS:ORD:MOLGEN • Synonym = FRAGILE X SYNDROME • HD GENE.CGG REPEATS :ARB:PT:BLD/TISS:ORD:MOLGEN • Synonym = HUNTINGTON DISEASE, IT15, HD, HUNTINGTON CHOREA • SPINOCEREBELLAR ATAXIA GENES.CAG REPEATS :ARB:PT:BLD/TISS:ORD:MOLGEN • DMPK GENE.CTG REPEATS :ARB:PT:BLD/TISS:ORD:MOLGEN • Synonym = MYOTONIC DYSTROPHY • Scale is ORD with usual answers of: • not expanded • indeterminate • expanded
Gene Rearrangements (hematopathology) • TCRB GENE REARRANGEMENTS :ARB :PT:BLD/TISS:ORD:MOLGEN • Synonym = T cell receptor beta chain • TCRD GENE REARRANGEMENTS :ARB :PT:BLD/TISS:ORD:MOLGEN • Synonym = T cell receptor delta chain • TCRG GENE REARRANGEMENTS :ARB :PT:BLD/TISS:ORD:MOLGEN • Synonym = T cell receptor gamma chain • Results would be reported as: • clonal, not clonal
Translocations • Pattern: T(<breakpoint gene 1>,<breakpoint gene 2>) (<gene1>,<gene2>):gene translocation • T(9,22) (ABL1,BCR) GENE TRANSLOCATION :ARB:PT:BLD/TISS:ORD:MOLGEN • Synonyms = Philadelphia chromosome, BCR1, chronic myeloid leukemia, CML • T(14,18) (IGH,BCL2) GENE TRANSLOCATION :ARB:PT:BLD/TISS:ORD:MOLGEN • Synonyms = Follicular B cell lymphoma, oncogene B-cell leukemia 2, CLL, chronic lymphatic leukemia, follicular lymphoma • T(15,17) (PML,RAR) GENE TRANSLOCATION :ARB:PT:BLD/TISS:ORD:MOLGEN • Synonyms = RAR, promyelocytic leukemia, myelogenous, retinoic acid receptor, acute promyelocytic leukemia, APL • Answers?
Fraction of cells that have the Rearrangement • CELLS.T(9,22).(ABL1,BCR)/CELLS TOTAL :NRF:PT:BLD/TISS:QN:MOLGEN • “/” represents division • Scale is QN (quantitative) • Property is NRF (number fraction)
Other genetic tests (specifics not shown) • Identity testing using DNA probes • Copy number of gene (N-Myc gene, Growth control gene) • Gene loss
Laboratory LOINC Committee Ray Aller, John Baenziger, Pamela D. Banning, Jim Bristol, Tom Burgess, Jim Case, Linda Charles, Jim Cimino , Diane Dwyer, Arden Forrey, Andy Gajda, Norbert Goldfield, Brian Griffin, Ed Hazell, Gil Hill, Stan Huff, Kathy Hutchins, Kathy Kammerer, Dennis Leavelle, Diane Leland, Pat Maloney, Doug Martin, Clem McDonald, Bill Meilahn, Karen Sieber, Frank Stalling, John Stelling, Bill Thurston, Dan Yokota
Clinical LOINC Committee James Barthel , (Dean Bidgood), Bruce Bray, (Bill Francis), Alan Golichowski, Karl Hammermeister, James Campbell, Sue Bakken, Pat Wilson, Stan Huff, Clem McDonald, Dan Pollock, (Angelo Rossi Mori), Jeff Suico, Anders Thurin, Wayne Tracy, Barry Gordon, Warren Williams, Pavla Frazier, William Karitis, Shaun Shakib, Jim Cimino, James Campbell, Susan Matney
Literature References • Huff SM, Rocha RA, McDonald CJ, De Moor GJE, etal. Development of the LOINC (Logical Observation Identifier Names and Codes) Vocabulary. Journal of American Medical Informatics Association, 1998, 5:276-292. • Dolin RH, Huff SM, Rocha RA, Spackman KA, Campbell, KE. Evaluation of a “Lexically Assign, Logically Refine”Strategy for Semi-Automated Integration of Overlapping Terminologies. Journal of American Medical Informatics Association, 1998, 5:203-213. • Rocha RA, Huff SM. Coupling Vocabularies and Data Structures: Lessons from LOINC. Journal of American Medical Informatics Association, AMIA Annual Fall Symposium Supplement, 1996, 90-4. • Forrey AW, McDonald CJ, DeMoor G, Huff SM , Leavelle D, Leland Fiers DT, Charles L, Griffin B, Stalling F, Tullis A, Hutchins K, Baenziger J. Logical Observation Identifier Names and Codes (LOINC) Database: A public use set of codes and names for electronic reporting of clinical laboratory test results. Clinical Chemistry, 1995. • Bakken S, Cimino JJ, Haskell R, Kukafka R, Matsumoto C, Chan GK, Huff SM. Evaluation of the Clinical LOINC (Logical Observation Identifier Names and Codes) Semantic Structure as a Terminology Model for Standardized Assessment Measures.Journal of the American Medical Informatics Association, 2000, 7:529-538.