Comparing the use of SNOMED CT and ICD10 for coding clinical conditions to implement laboratory guidelines

Laboratory medicine is responsible for an important part of hospital expenditure. Providing appropriate decision support to laboratory test requesters at the point of care is one of the main incentives for implementing laboratory guidelines, which can improve medical care. Laboratory guidelines developed by local experts in the Parisian region and two national guidelines for dyslipidemia were analyzed to extract test ordering recommendations. Clinical conditions which can be a trigger to order or not to order laboratory tests were extracted and mapped with ICD10 and SNOMED CT: 43.1% of clinical conditions were matched by ICD10 whereas SNOMED CT covered 80.1% of these conditions. For the non-mapped conditions, the main problem was found to be the ambiguity of the terms used in the guidelines. Ordinal characteristics of some clinical conditions and using terms more specific than SNOMED CT were other causes of mapping failure. Applying consistent and explicit concepts in the development of guidelines would lead to better implementation. By resolving the guideline ambiguity, SNOMED CT is a good choice and covers almost all of the clinical conditions in laboratory guidelines which are needed to implement in a Clinical Decision Support System.