Table 1. Overall quality of mapping Read codes to ICD10-CM codes for a subset of 1,313 randomly selected Read codes from GP clinical data.
Number of codes (%) | Issues with mapping quality | Category* |
---|---|---|
796 (60.6%) | Perfect match: Read code is perfectly matched to ICD10-CM code | No issue |
413 (31.4%) | ICD10-CM ontology does not contain a specific code corresponding to a Read code. More general ICD10-CM term has to be used for mapping. (Fig 2B) | Imprecise |
36 (2.7%) | Some Read codes describe a group of conditions which can be explained by combinations of several ICD10-CM codes. (Fig 2A) | Multiple |
27 (2.0%) | Some Read codes refer to tumor morphology terms. Additional ontology which is specifically for oncology presentation (such as ICD-O-3) should be used to map Read code correctly. | Multiple |
16 (1.2%) | There is no ICD10-CM code which would match the Read code precisely. Therefore, the closest ICD10-CM term was selected. (Fig 2C) | Imprecise |
15 (1.1%) | Some Read codes refer to general conditions and were mapped to a block of ICD10-CM codes. (For example, Read code “[X]Diseases of inner ear” was mapped to ICD10-CM block “H80-H83 Diseases of inner ear.”) | Multiple |
10 (0.7%) | Some Read codes refer to general conditions and were mapped to specific ICD10-CM codes with “unspecified” or “uncomplicated” qualifiers. (For example, Read code “Rubella” was mapped to ICD10-CM code “B069 Rubella without complication.”) | Imprecise |
* “Multiple” refers to one-to-many mapping. “Imprecise” means that curators were unable to find precise one-to-one mapping.