Claims analytics engine

Abstract

Methods and systems for processing claims (e.g., healthcare insurance claims) are described. For example, prior to payment of an unpaid claim, a prediction is made as to whether or not an attribute specified in the claim is correct. Depending on the prediction results, the claim can be flagged for an audit. Feedback from the audit can be used to update the prediction models in order to refine the accuracy of those models.

What is claimed is: 
     
         1 . A computer-implemented method comprising:
 generating training data for training a predictive model to identify claims as likely erroneous or as not likely erroneous, the training data including, for each of a plurality of claims, (i) one or more intrinsic features that are derived from the claim itself, and (ii) a label indicating whether the claim is erroneous or not, the training data further including one or more extrinsic features that are not derived from the plurality of claims;   training the predictive model using the one or more intrinsic features, the one or more extrinsic features, and the labels included in the training data;   after training the predictive model, receiving a particular claim;   generating (i) one or more intrinsic features for the particular claim and (ii) one or more extrinsic features associated with the particular claim;   providing, to the predictive model, (i) the one or more intrinsic features for the particular claim and (ii) the one or more extrinsic features associated with the particular claim; and   obtaining, from the predictive model, an indication of whether the particular claim is likely erroneous or is not likely erroneous based on providing (i) the one or more intrinsic features for the particular claim and (ii) the one or more extrinsic features for the particular claim to the predictive model.   
     
     
         2 . The computer-implemented method of  claim 1 , further comprising submitting the particular claim to an audit process in response to obtaining an indication that the particular claim is likely erroneous. 
     
     
         3 . The computer-implemented method of  claim 2 , further comprising:
 receiving feedback associated with the particular claim from the audit process; and   updating the predictive model based on the received feedback.   
     
     
         4 . The computer-implemented method of  claim 2 , wherein submitting the particular claim to the audit process includes providing a description of why the particular claim is likely erroneous. 
     
     
         5 . The computer-implemented method of  claim 4 , wherein the description includes one or more potential errors in the particular claim. 
     
     
         6 . The computer-implemented method of  claim 1 , wherein the indication of whether the particular claim is likely erroneous or is not likely erroneous includes a probability score that indicates a probability that the claim is erroneous. 
     
     
         7 . The computer-implemented method of  claim 1 , wherein the training data for training the predictive model is generated from historical claim information. 
     
     
         8 . The computer-implemented method of  claim 1 , wherein each of the one or more extrinsic features included in the training data is associated with at least one claim from the plurality of claims. 
     
     
         9 . The computer-implemented method of  claim 1 , wherein each of the one or more extrinsic features included in the training data is associated with a patient included in at least one claim from the plurality of claims. 
     
     
         10 . A non-transitory, computer-readable medium storing instructions operable when executed to cause at least one processor to perform operations comprising:
 generating training data for training a predictive model to identify claims as likely erroneous or as not likely erroneous, the training data including, for each of a plurality of claims, (i) one or more intrinsic features that are derived from the claim itself, and (ii) a label indicating whether the claim is erroneous or not, the training data further including one or more extrinsic features that are not derived from the plurality of claims;   training the predictive model using the one or more intrinsic features, the one or more extrinsic features, and the labels included in the training data;   after training the predictive model, receiving a particular claim;   generating (i) one or more intrinsic features for the particular claim and (ii) one or more extrinsic features associated with the particular claim;   providing, to the predictive model, (i) the one or more intrinsic features for the particular claim and (ii) the one or more extrinsic features associated with the particular claim; and   obtaining, from the predictive model, an indication of whether the particular claim is likely erroneous or is not likely erroneous based on providing (i) the one or more intrinsic features for the particular claim and (ii) the one or more extrinsic features for the particular claim to the predictive model.   
     
     
         11 . The computer-readable medium of  claim 10 , the operations further comprising submitting the particular claim to an audit process in response to obtaining an indication that he particular claim is likely erroneous. 
     
     
         12 . The computer-readable medium of  claim 11 , the operations further comprising:
 receiving feedback associated with the particular claim from the audit process; and   updating the predictive model based on the received feedback.   
     
     
         13 . The computer-readable medium of  claim 11 , wherein submitting the particular claim to the audit process includes providing a description of why the particular claim is likely erroneous. 
     
     
         14 . The computer-readable medium of  claim 13 , wherein the description includes one or more potential errors in the particular claim. 
     
     
         15 . The computer-readable medium of  claim 10 , wherein the indication of whether the particular claim is likely erroneous or is not likely erroneous includes a probability score that indicates a probability that the claim is erroneous. 
     
     
         16 . The computer-readable medium of  claim 10 , wherein the training data for training the predictive model is generated from historical claim information. 
     
     
         17 . The computer-readable medium of  claim 10 , wherein each of the one or more extrinsic features included in the training data is associated with at least one claim from the plurality of claims. 
     
     
         18 . The computer-readable medium of  claim 10 , wherein each of the one or more extrinsic features included in the training data is associated with a patient included in at least one claim from the plurality of claims. 
     
     
         19 . A system comprising:
 memory for storing data; and   one or more processors operable to perform operations comprising:   generating training data for training a predictive model to identify claims as likely erroneous or as not likely erroneous, the training data including, for each of a plurality of claims, (i) one or more intrinsic features that are derived from the claim itself, and (ii) a label indicating whether the claim is erroneous or not, the training data further including one or more extrinsic features that are not derived from the plurality of claims;   training the predictive model using the one or more intrinsic features, the one or more extrinsic features, and the labels included in the training data;   after training the predictive model, receiving a particular claim;   generating (i) one or more intrinsic features for the particular claim and (ii) one or more extrinsic features associated with the particular claim;   providing, to the predictive model, (i) the one or more intrinsic features for the particular claim and (ii) the one or more extrinsic features associated with the particular claim; and   obtaining, from the predictive model, an indication of whether the particular claim is likely erroneous or is not likely erroneous based on providing (i) the one or more intrinsic features for the particular claim and (ii) the one or more extrinsic features for the particular claim to the predictive model.   
     
     
         20 . The system of  claim 19 , the operations further comprising submitting the particular claim to an audit process in response to obtaining an indication that the particular claim is likely erroneous.