Does mapping to SNOMED CT improve precision of subjective clinical evaluations ?

14
Does mapping to SNOMED CT improve precision of subjective clinical evaluations? SHI2010 presentation Kirstine Rosenbeck Gøeg & Anne Randorff Rasmussen

description

Does mapping to SNOMED CT improve precision of subjective clinical evaluations ?. SHI2010 presentation Kirstine Rosenbeck Gøeg & Anne Randorff Rasmussen. Semantic of clinical documentation. Introduction. Ambiguous documentation. “Impairment of vision”. - PowerPoint PPT Presentation

Transcript of Does mapping to SNOMED CT improve precision of subjective clinical evaluations ?

Page 1: Does mapping to SNOMED CT improve precision of subjective clinical evaluations ?

Does mapping to SNOMED CT improve precision of subjective

clinical evaluations?

SHI2010 presentationKirstine Rosenbeck Gøeg & Anne Randorff Rasmussen

Page 2: Does mapping to SNOMED CT improve precision of subjective clinical evaluations ?

INTRODUCTION

Semantic of clinical documentation

Page 3: Does mapping to SNOMED CT improve precision of subjective clinical evaluations ?

Ambiguous documentation

“Impairment of vision”

Worsening of something chronic

First diagnose

Impairment causing change of occupation

Page 4: Does mapping to SNOMED CT improve precision of subjective clinical evaluations ?

A problem?

• Interview 2009 “Vi har nogle begreber som vi har brugt i årevis, eksempelvis tilladelig blødning..... nogen der siger, at det er 100ml, mens andre siger at det er det i hvert fald ikke. Så der opdager vi, at vi ikke har den samme opfattelse. Så der må vi blive enige om at tilladelig blødning er under 100ml, hvis moren ellers har det godt, blodtrykket holder og hun ikke besvimer. Der var så meget mere, der skulle knyttes til sådan et begreb....”

• Rector 1999 “Common usage may differ from the expectations of information retrieval, e.g. it may seem odd to say that ‘The fingernail is a part of the arm’, but it would be an error not to retrieve ‘evulsion of nail’ under ‘trauma to upper extremity’.”

Interview 2009: http://projekter.aau.dk/projekter/files/17651356/speciale-rapportTilUdskrift.pdfRector 1999: Why is terminology so hard? Method Inform Med 1999; 38: 239–52

Page 5: Does mapping to SNOMED CT improve precision of subjective clinical evaluations ?

A problem?

• Increasingly possible to– Exchanging information to

perform shared care– Use data for statistical

purposes

• Characterised by– No input validation– Most likely to stay in one

organisation and being used only for patient treatment purposes

Semantic precision!

Page 6: Does mapping to SNOMED CT improve precision of subjective clinical evaluations ?

METHOD

Mapping subjective evaluations to SNOMED CT

Page 7: Does mapping to SNOMED CT improve precision of subjective clinical evaluations ?

Method• Two cases

– Tolerable bleeding/ Tilladelig blødning

– Eating too little/ Småtspisende

• Charaterised by– Real life examples of clinical

evaluations– From spring 2009– Expressed difficulties in

mapping from paper documentation to electronic documentation

• Three approaches to improve precision using SNOMED CT– Mapping to existing pre-

coordinated SNOMED CT terms consisting of a finding and a qualifier. Map the evaluation as is, without interpretation.

– Mapping by analysing the clinical situation/context in which the evaluation is obtained.

– Mapping by nesting the subjective evaluation with the rationale behind the evaluation.

Page 8: Does mapping to SNOMED CT improve precision of subjective clinical evaluations ?

RESULTS

SNOMED CT mappings

Page 9: Does mapping to SNOMED CT improve precision of subjective clinical evaluations ?

Existing pre-coordinated terms

Hemorrhage

Actute Massive Subacute ChronicDiffuse

Bleeding(finding)Associated morphology

Finding of vaginal bleeding

Normal menstual Fresh Scanty Moderate Profuse Abnormal

Bleeding(finding)

Finding of quantity of eating

Binging Excessive Fasting Overeating

Finding of eating pattern

Picky eater Greediness

Page 10: Does mapping to SNOMED CT improve precision of subjective clinical evaluations ?

Adding the context of the finding

Finding: Tolerable bleeding

Context: During labour

Maternal blood loos within normal

limits

SNOMED CTmapping

Finding: Eating too little

Context: Screening, some time duration

Inadequate dietary caloric intake,

continual

Inadequate dietary caloric intake,

continual

Page 11: Does mapping to SNOMED CT improve precision of subjective clinical evaluations ?

Nesting observation and evaluation

Page 12: Does mapping to SNOMED CT improve precision of subjective clinical evaluations ?

DISCUSSION

Choosing between mappings

Page 13: Does mapping to SNOMED CT improve precision of subjective clinical evaluations ?

Discussion

• A study with only two clinical expressions – pros and cons

• Choosing between the three approaches?– Mapping directly is not recommended based on

the results of this study – Considering the clinical situation and

consequences of different mappings– Asking clinical experts

• Can levels of precision be pre-specified?

Page 14: Does mapping to SNOMED CT improve precision of subjective clinical evaluations ?

• Correspondence:Kirstine Hjære RosenbeckPhD Fellow, Aalborg UniversityDepartment of Health Science and Technology

E-mail: [email protected]:+45 9940 8835