How it works
Platocode CAC can be configured for any coding standard with a comprehensive index. This includes all variants of ICD-10, CPT, SNOMED, OPCS4.
Customers can create their own index entries for special facility-specific context. This could be for an entirely new set of codes, or to modify an existing code set - e.g. because of a locally named procedure variant.
There's no need to push hundreds of thousands or millions of documents through the engine to try to create statistical rules - because the binary pattern captures clinical context rather than multiple phrasing variations for the same concept.
The original Platocode CAC patent was for converting freeform text into a binary pattern using context resolution that the engine has been "taught" over 25 years. The Platocode engine now understands clinical context and builds binary patterns that retain the context no matter what phrasing or sequencing is used. This completely side-steps exponentially increasing rule libraries to cover all phrasing variants and also allows correlation of clinical contexts whether stated or not- such as Monteggia fracture. The binary specificity is far more granular than SNOMED which itself has an index, so can be delivered by the Platocode engine.
Multiple Code sets
Since there's a Platocode filter for each index, the same patient encounter can be coded to multiple standards by dropping through multple filters. This not only allows coding of (for example) SNOMED at the same time as ICD-10 for DRG calculation, it also allows customers to minimize service consumption. E.g. an Australian facility may prefer to use its own ICD-10 procedure codes and only apply CAC for diagnosis, while a US physician practice may request only E/M for one set of episodes but ICD-10 and CPT for another.
How the filter is used
First, health encounter documentation is converted into binary patterns that retain all the clinical context. The resulting patterns are dropped through the binary filter created from the coding index. The filter is sequenced with highest specificity first, so the most specific code always is preferred and "other and unspecidied" codes are automatic defaults if nothing else matches. This simple idea means that accuracy is the same for frequent phrasing as for phrasing never encountered before.
When a coding standard update is published, Platocode CAC simply rebuilds its binary pattern based on the new index and the engine is ready to go. There's no customer burden involving document submission to re-train the engine or hard-coded rules.
Integration with Codectomy
If you have Codectomy, adding the CAC service creates a new CAC panel option. When selected, the panel contains all the diagnosis and procedure codes that the Platocode engine was able to extract from supplied documentation. When the coder selects a code, the CAC panel shows the document source/s for that code. This allows a coder to quickly assess whether a CAC code should be dragged or copied across to the Episode ICD or other code panels.