Computing platform for establishing referrals
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for a computing platform that identifies information about a trial program, where the information is related to healthcare data included in datasets, and identifies an investigator based on the information about the trial program. A data analytics model of the platform generates an initial provider score for each provider in a group of providers based on analysis of the information. The analyzed information of the datasets includes healthcare data describing interactions between patients and providers in the group and criteria for the trial program. The platform provides a request to a subset of providers using the initial provider scores. The request is an invitation to for each provider to join a referral network. The platform uses the request to establish referral connections between the trial investigator and a provider in the subset.
1. A computer system-implemented method comprising:
generating, by at least some of one or more processors of the computer system, machine learning inferences that represent data connections from prior clinical trial programs, wherein the machine learning inferences are generated based on iterative analysis of new or modified information about the prior clinical trial programs;
using, by at least some of the one or more processors of the computer system, the generated inferences to enhance a predictive analytics model, wherein the predictive analytics model is configured to predict prospective patient referral activity of a provider based on a weighting of the data connections represented by the machine learning inferences, and wherein enhancing the predictive analytics model comprises modifying the weighting of the data connections based on the new or modified information about the prior clinical trial programs;
generating, by at least some of the one or more processors of the computer system, a score from the enhanced predictive analytics model that represents a likelihood that the provider will refer candidates to an investigator that conducts a clinical trial program;
based on a comparison of the generated score to a threshold value, generating, by at least some of the one or more processors of the computer system, a referral connection between the investigator and the provider for providing a patient referral to the investigator; and
providing, by at least some of the one or more processors of the computer system, a communication indicative of the patient referral between the provider and the investigator.
2. The computer system-implemented method of claim 1 , further comprising:
generating, by at least some of the one or more processors of the computer system, a second score from the enhanced predictive analytics model that represents a decreased likelihood that a different provider will refer the candidates to the investigator that conducts the clinical trial program.
3. The computer system-implemented method of claim 1 , wherein the generating the score from the enhanced predictive analytics model that represents the likelihood that the provider will refer the candidates further comprises:
generating, by at least some of the one or more processors of the computer system, the score from the enhanced predictive analytics model that represents an increased likelihood that the provider will identify and refer the candidates to the investigator that conducts the clinical trial.
4. The computer system-implemented method of claim 1 , wherein generating the score is based on at least one of:
a quantity of patients managed by the provider that are eligible for participation in the clinical trial;
a proximity of the provider relative to one or more of multiple investigators conducting the clinical trial; or
a quantity of patients managed by the provider that are shared with one or more of the multiple investigators conducting the clinical trial.
5. The computer system-implemented method of claim 1 ,
wherein the data connections from the prior clinical trial programs comprises:
healthcare data describing interactions between patients and different providers; and
a set of criteria for the trial program.
6. The computer system-implemented method of claim 5 , wherein the set of criteria for the trial program comprises:
inclusion criteria for identifying the candidates for participating in the trial program; and
exclusion criteria for identifying other candidate to be excluded from participating in the trial program.
7. The computer system-implemented method of claim 1 , wherein the predictive analytics model is trained using at least a neural network implemented on one or more hardware circuits.
8. The computer system-implemented method of claim 1 , further comprising:
generating, by at least some of the one or more processors of the computer system, a listing of different providers that are identified in a referral network;
ranking, by at least some of the one or more processors of the computer system, each provider of the different providers relative to one another in the listing based on an initial provider score for each provider in the listing; and
determining, by at least some of the one or more processors of the computer system, a subset of providers based on a ranking order for each provider in the listing.
9. The computer system-implemented method of claim 8 , wherein the initial provider score is generated by the predictive analytics model before the generated inferences are used to enhance the predictive analytics model.
10. A system, comprising:
one or more processors; and
one or more non-transitory machine-readable storage devices storing instructions that are executable by the one or more processing devices to cause performance of operations comprising:
generating, by at least some of the one or more processors of the system, machine learning inferences that represent data connections from prior clinical trial programs, wherein the machine learning inferences are generated based on iterative analysis of new or modified information about the prior clinical trial programs;
using, by at least some of the one or more processors of the system, the generated inferences to enhance a predictive analytics model, wherein the predictive analytics model is configured to predict prospective patient referral activity of a provider based on a weighting of the data connections represented by the machine learning inferences, and wherein enhancing the predictive analytics model comprises modifying the weighting of the data connections based on the new or modified information about the prior clinical trial programs;
generating, by at least some of the one or more processors of the system, a score from the enhanced predictive analytics model that represents a likelihood that the provider will refer candidates to an investigator that conducts a clinical trial program;
based on a comparison of the generated score to a threshold value, generating, by at least some of the one or more processors of the system, a referral connection between the investigator and the provider for providing a patient referral to the investigator; and
providing, by at least some of the one or more processors of the system, a communication indicative of the patient referral between the provider and the investigator.
11. The system of claim 10 , further comprising:
generating, by at least some of the one or more processors of the system, a second score from the enhanced predictive analytics model that represents a decreased likelihood that a different provider will refer the candidates to the investigator that conducts the clinical trial program.
12. The system of claim 10 , wherein the generating the score from the enhanced predictive analytics model that represents the likelihood that the provider will refer the candidates further comprises:
generating, by at least some of the one or more processors of the system, the score from the enhanced predictive analytics model that represents an increased likelihood that the provider will identify and refer the candidates to the investigator that conducts the clinical trial.
13. The system of claim 10 , wherein generating the score is based on at least one of:
a quantity of patients managed by the provider that are eligible for participation in the clinical trial;
a proximity of the provider relative to one or more of multiple investigators conducting the clinical trial; or
a quantity of patients managed by the provider that are shared with one or more of the multiple investigators conducting the clinical trial.
14. The system of claim 10 ,
wherein the data connections from the prior clinical trial programs comprises:
healthcare data describing interactions between patients and different providers; and
a set of criteria for the trial program.
15. The system of claim 14 , wherein the set of criteria for the trial program comprises:
inclusion criteria for identifying the candidates for participating in the trial program; and
exclusion criteria for identifying other candidate to be excluded from participating in the trial program.
16. The system of claim 10 , wherein the predictive analytics model is trained using at least a neural network implemented on one or more hardware circuits.
17. The system of claim 10 , further comprising:
generating, by at least some of the one or more processors of the system, a listing of different providers that are identified in a referral network;
ranking, by at least some of the one or more processors of the system, each provider of the different providers relative to one another in the listing based on an initial provider score for each provider in the listing; and
determining, by at least some of the one or more processors of the system, a subset of providers based on a ranking order for each provider in the listing.
18. The system of claim 17 , wherein the initial provider score is generated by the predictive analytics model before the generated inferences are used to enhance the predictive analytics model.
19. One or more non-transitory machine-readable storage devices storing instructions that are executable by one or more processors to cause performance of operations comprising:
generating, by at least some of the one or more processors, machine learning inferences that represent data connections from prior clinical trial programs, wherein the machine learning inferences are generated based on iterative analysis of new or modified information about the prior clinical trial programs;
using, by at least some of the one or more processors, the generated inferences to enhance a predictive analytics model, wherein the predictive analytics model is configured to predict prospective patient referral activity of a provider based on a weighting of the data connections represented by the machine learning inferences, and wherein enhancing the predictive analytics model comprises modifying the weighting of the data connections based on the new or modified information about the prior clinical trial programs;
generating, by at least some of the one or more processors, a score from the enhanced predictive analytics model that represents a likelihood that the provider will refer candidates to an investigator that conducts a clinical trial program;
based on a comparison of the generated score to a threshold value, generating, by at least some of the one or more processors, a referral connection between the investigator and the provider for providing a patient referral to the investigator; and
providing, by at least some of the one or more processors, a communication indicative of the patient referral between the provider and the investigator.
20. The machine-readable storage device of 19 , further comprising:
generating, by at least some of the one or more processors, a second score from the enhanced predictive analytics model that represents a decreased likelihood that a different provider will refer the candidates to the investigator that conducts the clinical trial program.