DETECTING MISSING HCPCS CODES THROUGH MACHINE LEARNING APPROACH

Abstract

Currently hospitalization systems in the US are facing huge losses around 5 billion a year due to the problem of missing HCPCS codes in the insurance claim. The insurance companies do no compensate the amount for the missing codes in the bill list. The missing HCPCS codes problem occur usually in the case of outpatients. There are only one or two operators to process the bill for the patients, which is done manually and chances for the operators for missing of some codes are high. Therefore, we can see that on a daily basis; if such codes are missing largely the losses annually accumulate to a high amount. For incurring these losses, the hospitals hire accountants, who solve the problem by analysing the missing codes and recover some losses. Unfortunately, the hospitals have to pay a huge amount to these accountants too. So the problem of revenue still sustains. We in this project have developed a model for the hospital systems, which can help in finding the missing codes and account for the losses being faced. We have used Market Basket Analysis on FP Growth algorithm to detect the missing HCPCS codes in the bills. This model can solve the problem completely and help the hospitals to avoid such high losses.