Automating ICD10 coding and claims payments processing

Automating ICD10 coding and claims payments processing

Challenge 1

 Challenge

  • Medical coding of diagnosis for pathology data is done manually
  • Accuracy of ICD10 coding process affects payor reimbursement rates
  • Medical coding staff can be fully utilized in higher-value activities

Solution-2

Solution

  • Using Microsoft Azure services, Neal Analytics developed an ensemble of candidate machine learning models capable of classification of diagnosis from text
  • Neal Analytics integrated the models with diagnosis text, clinical history, code set, and historical labeled data allowing coding to be addressed automatically
  • Neal Analytics used text vectorization and augmented feature engineering to translate and process natural language, creating a friendly user interface that requires no training to use

Result

  • Increased ICD10 coding accuracy from manual
  • Automated Claims submission with ICD10, resulting in better reimbursement and payor reporting