Anonymous case study A
Auditing revenue loss across a group of veterinary clinics
Auditing veterinary misbilling to pinpoint problem areas and prioritize corrective action.
The directors of a veterinary group were concerned about misbilling and had reviewed a sample of recent customer invoices to find out the true extent of the problem. They found that a substantial proportion of customer bills included the main procedure but failed to charge for other products and services which their clinical protocols required.
Before taking this up with staff, they wanted to find out the full extent of the problem and understand exactly where these billing errors were occurring. Although they had good data skills in-house, they realized that the complexity of their billing records and the subtleties around their clinical protocols meant that they would need help carrying out a more comprehensive review.
Unpacking complex billing data
Using Sweetfish, we took a large sample of their historic billing data to analyze. They were using an older PMS, so our first challenge was to untangle the complexities of the data, both in terms of its structure and to take account of complex situations such as multiple invoices applying to a single hospital stay.
The next step was to analyze all the invoice data and find meaningful patterns amongst the products and services that are typically billed together. We combined machine learning techniques with the insights of our own veterinary team to develop a series of ‘rules’ which captured their clinical protocols in billing terms.
Calculating missed revenues
Sweetfish reviewed thousands of customer invoices and assessed each one against these rules. Wherever a rule was triggered by a missing product or service, Sweetfish calculated the amount of revenue that had been missed, based on their regular price list.
Because veterinary work is complex, there are often exceptions to established practices. So, for each rule we worked with the client to set a reasonable benchmark for compliance. Sweetfish used these benchmark levels to adjust the missed revenue figures to ensure they reflected real-world expectations.