Using Snoflake to access SNOMED CT

An Introduction to SNOMED CT

Click to Register

SNOMED Clinical Term (SNOMED CT ®) is a comprehensive and precise clinical reference terminology designed to make healthcare information useable and accessible. Global in scope SNOMED CT provides a common language of great depth that enables a consistent way of capturing, sharing and aggregating health data across clinical specialties and sites of care.

SNOMED CT is owned, maintained and distributed by the International Health Terminology Standards Development Organisation (IHTSDO ®). The IHTSDO is a not-for-profit association which is owned and governed by its national Members. SNOMED CT brings together the content and structure of SNOMED RT V1.0 with the NHS Clinical Terms Version 3 (Read Codes or CTV3). SNOMED CT delivers unmatched content and expressivity for clinical documentation and reporting, and provides a means of medical communication that spans languages, clinical specialties and geographic borders.

SNOMED CT contains concepts linked to clinical knowledge to enable unambiguous recording of data. The terminology's content also includes terms or synonyms relating to clinical concepts, each with their own unique identifier. Links, or semantic relationships, between clinical concepts provide formal definitions. The SNOMED CT structure makes possible interoperability across software applications that relate to diseases, treatments, etiologies, clinical findings, therapies, procedures and outcomes. The breadth and depth of the clinical terminology, as well as its computer-readable hierarchies, enable faster, reliable and consistent retrieval of robust clinical information. Concepts included in SNOMED CT with multiple levels of granularity including those illustrated below…

Snomed CT Concept types

This depth allows health care organisations to use their clinical information to their advantage. Users can record data just once, at the level of specificity they choose, then mine it repeatedly for decision support, statistical reporting, outcomes measurement, treatment guidelines and cost analysis. Cases can be retrieved based on numerous different criteria, including clinical and laboratory findings, causative agents, anatomical structure and therapeutic procedures.

SNOMED CT offers several structural advantages over Read Codes. A critical improvement for clinical decision support is the capacity to allow any coded disorder to 'belong to' more than one class of disorder. For example, in Read Version 2, “bacterial meningitis” is classed as an “inflammatory disease of the central nervous system”, but due to technical constraints within the Read Version 2 coding scheme, it cannot also be classed as a “bacterial disease”. For clinical decisions, the fact that “bacterial meningitis” is, in reality, both a nervous system disease and a bacterial disease cannot be ignored.

The 300,000+ concepts available when coding using SNOMED CT not only offer the facility to provide greater detail about a patient than was present in previous coding schemes, but also allow for additional context information to be coded and associated with the patient. For example, if a patient reports a family history of asthma, current clinical systems use varying methods to record this information. Some systems have the context of “family history” embedded into their structure and the asthma code is selected and added to this field; some use a specific code of family history of asthma; while others apply the context of family history to asthma using free text. Multiple methods of representing identical patient information, in particular free text, present immense challenges for clinical decision support.

A very important aspect of SNOMED CT is that its applicability is almost limitless. Clinicians and researchers can mine health information in ways that until the advent of SNOMED CT were not possible. SNOMED CT works through implementation in software applications, representing clinically relevant information in a reliable, reproducible manner is a natural and valid solution for a wide range of applications, some of which are listed below…

information could be in the used for the uses as listed

SNOMED CT can be seamlessly integrated with existing systems and although comprehensive on its own, it also cross-maps to other widely used legacy medical classifications, such as ICD-9-CM, ICD-O3, ICD-10, Laboratory LOINC and OPCS-4, making it easier to avoid duplicate data capture. In addition, SNOMED CT is aligned with numerous key health care standards such as HL7, DICOM, ANSI and ISO.

SNOMED CT also provides a framework to manage language dialects, clinically relevant subsets, qualifiers and extensions, such as the US and UK drug extensions of proprietary drug concepts and terms unique to these particular localities. This allows users to take advantage of its broad coverage while integrating their own specific codes, leading to easier adaptation to local needs.

The nature of SNOMED CT has, however, some practical problems regarding its use.

  1. Size. The English language version used by the NHS contains 400,000 concepts, around 1 million terms and 1.6 million relationships.
  2. Complexity. If SNOMED CT concept relationships followed the simple hierarchical parent-child model, where all concepts have a single parent and each parent can have multiple children, there would not be a problem, because its concepts could be logically envisaged using an Explorer. Although many SNOMED CT concept relationships are of the parent-child type, some are poly-hierarchical in that a concept may have multiple parents and multiple children.

Because of this property, an alternative visualisation model was required, and the Snoflake was specifically developed to provide the solution.


Share Snoflake: