Illogical placing of codes within a clinical classification.

Readers may have noticed problems on practice computer systems when searching for codes associated with chronic kidney disease (CKD). This is due to anomalies in the Read Code System. The Read Code System is based on a branching tree system with progressive refining of accuracy of terms with each level of the tree and addition of a further digit to the code. The first digit is either a letter or a number. Codes starting with a letter signify a diagnosis or disease, codes starting with a number signify process of medicine. So, for example, a code starting with the letter K indicates a genitourinary systemdisease – such as ‘nephritic syndrome’. On the other hand, a code starting with a number 1 would indicate a history or symptom – such as ‘frequency of micturition’. Unfortunately the CKD codes have been located in the Read Code System under the section starting with the number 1 (history/symptoms) – codes start with 1Z1. This means that computer systems will file these codes not under diagnoses but under symptoms and any search for renal failure under code K05 will miss any patient with a CKD code. Further problems are caused because the CKD codes do not follow the normal Read method of progressive refinement with each extra digit. For example ‘CKD 3A with proteinuria’ would by Read convention be a childcode of ‘CKD 3A’. This is not the case, they are all simply children codes of chronic renal failure symptom 1Z1. This explainswhy, whenCKD3A is entered into the computer, its logical child codes ‘CKD3A with proteinuria’ and ‘CKD3A without proteinuria’ are not displayed. The long promised SNOMED CT code system is due for release soon. I wait to see if this sort of problem will be solved.