The client operates ER (Emergency Department) across the US. The revenue per year is $22M with an average of 18000 visits per month. They have a collective total of 45 billing/collections staff, both onshore and offshore. In client incurred revenue loss due to inaccurate charging and coding processes for emergency department (ED) patients. The inconsistencies and reduced quality of the processes not only hindered the performance but also lead to denial of claims. The deployment AI-based tool for predicting the erroneously coded chart and generating audit samples led to improvement in accuracy, consistency in coding, and error prediction. It also enhanced revenue, while also lowering costs, reducing the number of failed claims and improving coding accuracy.

• Inconsistency in coding performance and drop in quality
• Reduced acuity and revenue per encounter
• Denied claim due to charge and coding issues
• Timely filing denials due to delay in receiving documentation and missing charges
• High rate of eligibility related denials and incomplete demographic information
• The non-tracking of several Underpayment issues
• Unidentified compliance risks

Based on the assessment, the Apexon team recommended the following technology initiatives to transform the reconciliation and quality procedures.

• Usage of technology such as Assisted coding and error prediction tool to improve accuracy and consistency in coding
• Implementing high-level reporting solution to identify any anomaly and address the same
• 100% eligibility verification before submitting a claim to the payer
• Change in the workflow to improve the overall process efficiency
• Trained the team on coding standards and implementing process guidelines to handle regular coding denials

A strategic approach towards quality led to the creation of an AI-based automation tool. This resulted in accurate and timely predicting of the erroneously coded chart and generating audit samples which were crucial in improving the performance and revenue. It delivered the following.

1. It helped maintain manually coded/ auto coded sheet based on daily coded data and corresponding audits
2. Performed as an error prediction tool with customizable algorithm for predicting error codes and the likely reason for the inaccuracy
3. It helped in prioritizing error charts with maximum error propensity across different critical to quality fields
4. It helped in generating reports and monitoring quality with an eye on improving the coding process