A Christmas guide to clinical coding

With the rollout of a new coding system in the UK under way, Richard Williams reveals a lighter side to recording clinical data If you didn’t find the section heading amusing you either have no sense of humour (SNOMED: 288801003—“Unable to use humour”), you don’t know any Christmas songs (SNOMED: 16170002—“Music blindness”), or, like most normal people, you don’t know much about clinical coding. In a nutshell, coding is the use of short alphanumeric codes to record symptoms, diagnoses, laboratory tests, procedures, and medicines in the electronic health record. For example, a GP in UK primary care might enter “C10F” to record a diagnosis of type 2 diabetes, “22A” along with a value and a unit to record a person’s weight, “di1m” to record a prescription of 300 mg soluble aspirin tablets, or “T550” for spacecraft launch pad accidents (more on this later). Typically, a clinician starts typing and the medical record software suggests appropriate codes from a dropdown list. Collectively, these codes are called either dictionaries (because each code has a definition) or terminologies. C10F, 229, and di1m are all examples of Read codes, which were created in the 1980s by James Read for use in primary care. Read codes have been the main clinical coding system in UK primary care since the mid-1990s.1 ### Accuracy and reliability Using clinical codes provides a level of standardisation that’s incredibly useful for continuity of care, monitoring safety, and improving quality. Scheduling an annual review for every patient with diabetes in a practice by searching through their typed notes would be hard enough given the numerous possible synonyms (“type 2” or “type II”; “diabetes” or “diabetes mellitus”; acronyms such as “T2DM”). However, imagine something more complex, such as finding all patients over a certain weight who had a particular medicine prescribed after new research had …