Projecting time-to-pay of healthcare invoices

A method for forecasting when clinical invoices will be paid, the system comprising:

• accessing paid billing records that comprise clinical procedure information received over a computer network from an electronic health record;
• generating a plurality of payment probability curves using a survival analysis that uses the paid billing records as input, wherein the plurality of the payment probability curves comprises individual payment probability curves for different types of payers, and wherein the plurality of the payment probability curves comprises individual payment probability curves for different billed amounts, and wherein an individual payment probability curve in the plurality of payment probability curves is not able to accurately predict future payments;
• determining that a subset of payment probability curves from the plurality of payment probability curves are similar;
• generating a composite payment probability curve by clustering the subset of payment probability curves into the composite payment probability curve;
• associating the composite payment probability curve with a first type of payer associated with an individual payment probability curve from the subset of payment probability curves used to generate the composite payment probability curve;
• and
• generating a cash flow forecast to be received from a payer of the first type using the composite payment probability curve and accounts receivable for the payer.

A method for forecasting when clinical invoices will be paid, the method comprising:

• accessing paid billing records that comprise clinical factor information received over a computer network from an electronic health record;
• generating a plurality of payment probability curves using a survival analysis that uses the paid billing records as input, wherein the plurality of the payment probability curves comprises individual payment probability curves for different combinations of clinical factors, wherein the first combination of clinical factors comprises a payer class and a clinical procedure type, and wherein an individual payment probability curve in the plurality of payment probability curves is not able to accurately predict future payments;
• determining that a subset of payment probability curves from the plurality of payment probability curves are similar;
• generating a composite payment probability curve by clustering the subset of payment probability curves into the composite payment probability curve;
• associating the composite payment probability curve with a first combination of clinical factors associated with an individual payment probability curve from the subset of payment probability curves used to generate the composite payment probability curve;
• calculating an estimated probability of payment for an invoice having the first combination of clinical factors using a time the invoice has been unpaid and the composite payment probability curve;
• determining that the estimated probability of payment exceeds a threshold;
• and
• designating the invoice for a collection effort.