Using machine learning to facilitate design and implementation of a clinical trial with a high likelihood of success

1. A computer-implemented method for monitoring clinical trial protocols, the computer-implemented method comprising:

receiving, by a computing platform, study design information for a clinical trial, the study design information including a set of parameters related to a success of the clinical trial and factors related to a relevance of the parameters;

providing, by the computing platform, plural sets of values for the set of parameters, wherein the plural sets of values are produced by varying values for the set of parameters, and wherein the values for the set of parameters are used to implement the clinical trial;

applying, by the computing platform, at least one trained machine learning model to the study design information including the provided sets of values to predict a score for each of a plurality of success scores including a predicted overall success score;

determining, by the computing platform, the provided set of values for the set of parameters that are associated with the at least one trained machine learning model predicting a best overall success score for the clinical trial; and

outputting, by the computing platform, the determined set of values for implementing the clinical trial.

2. The computer-implemented method of claim 1 , further comprising:

selecting a second plurality of sets of the provided values associated with next greatest predicted overall success scores; and

providing new variations of a set of values for the set of parameters based on the selected second plurality of sets of the provided values, and

applying the new variations of the set of values for the set of parameters to predict a score for the each of the plurality of success scores, wherein

the outputting, by the computing platform, further comprises outputting the set of values of the new variations associated with a best predicted overall success score.

3. The computer-implemented method of claim 1 , further comprising:

providing a user interface to allow a user to modify the study design information.

4. The computer-implemented method of claim 3 , further comprising:

responsive to the user modifying the study design information, applying, by the computing platform, the at least one trained machine learning model to the modified study design information to predict the score for the each of the plurality of success scores.

5. The computer-implemented method of claim 4 , further comprising:

determining weights and a bias adjustment applied by the at least one trained machine learning model to the set of parameters for predicting the score for the each of the plurality of success scores.

6. The computer-implemented method of claim 1 , wherein the factors related to the relevance of the parameters include a therapeutic area, a specific condition, a severity of progression of the condition, and demographics.

7. The computer-implemented method of claim 1 , wherein:

the set of parameters related to the success of the clinical trial include a total number of clinical trial visits, a duration of the clinical trial, an average duration of the clinical trial visits, an average frequency of the clinical trial visits over a given period of time, an average duration of each task, a number of data points collected during the clinical trial, and clinical trial risk factors, and

the plurality of success scores include a data quality score, a participant retention score, and a participant recruitment score.

8. A computing platform for monitoring clinical trial protocols, the computing platform comprising:

at least one processor; and

at least one memory connected with the at least one processor, wherein the at least one processor is configured to perform:

receiving study design information for a clinical trial, the study design information including a set of parameters related to a success of the clinical trial and factors related to a relevance of the parameters;

providing plural sets of values for the set of parameters, wherein the plural sets of values are produced by varying values for the set of parameters, and wherein the values for the set of parameters are used to implement the clinical trial;

applying at least one trained machine learning model to the study design information including the provided sets of values to predict a score for each of a plurality of success scores including a predicted overall success score;

determining the provided set of values for the set of parameters that are associated with the at least one trained machine learning model predicting a best overall success score for the clinical trial; and

outputting the determined set of values for implementing the clinical trial.

9. The computing platform of claim 8 , wherein the at least one processor is further configured to perform:

selecting a second plurality of sets of the provided values associated with next greatest predicted overall success scores;

providing new variations of a set of values for the set of parameters based on the selected second plurality of sets of the provided values, and

applying the new variations of the set of values for the set of parameters to predict a score for the each of the plurality of success factors, wherein

the outputting further comprises outputting the set of values of the new variations associated with a best predicted overall success score.

10. The computing platform of claim 8 , wherein the at least one processor is further configured to perform:

providing a user interface to allow a user to modify the study design information;

responsive to the user modifying the study design information, applying the at least one trained machine learning model to the modified study design information to predict the score for the each of the plurality of success scores; and

outputting the predicted score for the each of the plurality of success scores responsive to applying the at least one trained machine learning model to the modified design information.

11. The computing platform of claim 10 , wherein the at least one processor is further configured to perform:

determining weights and a bias adjustment applied by the at least one trained machine learning model to a set of parameters of the modified design information for predicting the score for the each of the plurality of success scores.

12. The computing platform of claim 8 , wherein the factors related to the relevance of the parameters include a therapeutic area, a specific condition, a severity of progression of the condition, and demographics.

13. The computing platform of claim 8 , wherein:

the set of parameters related to the success of the clinical trial include a total number of clinical trial visits, a duration of the clinical trial, an average duration of the clinical trial visits, an average frequency of the clinical trial visits over a given period of time, an average duration of each task, a number of data points collected during the clinical trial, and clinical trial risk factors.

14. The computing platform of claim 8 , wherein the at least one processor is further configured to perform:

calculating a per-visit task density score based on an average number of tasks per visit and a level intensity of each of the tasks, wherein a predicted participant retention rate score is based, at least partly, on the per-visit task density score, wherein the plurality of success scores includes the participant retention rate score.

15. The computing platform of claim 8 , wherein the at least one processor is further configured to perform:

calculating an average travel score for participants of the clinical trial based on travel cost, a distance to travel, travel time, and expected traffic level; and

determining the predicted overall success score based, at least partly, on the average travel score.

16. A computer program product for assessing clinical trial protocols, the computer program product comprising one or more computer readable storage media having program instructions collectively stored on the one or more computer readable storage media, the program instructions executable by at least one processor of a computing platform to cause the computing platform to:

receive study design information for a clinical trial, the study design information including a set of parameters related to a success of the clinical trial and factors related to a relevance of the parameters;

provide plural sets of values for the set of parameters, wherein the plural sets of values are produced by varying values for the set of parameters, and wherein the values for the set of parameters are used to implement the clinical trial;

apply at least one trained machine learning model to the study design information including the provided sets of values to predict a score for each of a plurality of success scores including a predicted overall success score;

determine the provided set of values for the set of parameters that are associated with the at least one trained machine learning model predicting a best overall success score for the clinical trial; and

output the determined set of values for implementing the clinical trial.

17. The computer program product of claim 16 , wherein the program instructions are further executable by the at least one processor of the computing platform to:

provide output to cause a display device to visually display graphical gauges indicating a corresponding score for a participant retention rate, a participant recruitment rate, a data quality score, and the predicted overall success score.

18. The computer program product of claim 16 , wherein the program instructions are further executable by the at least one processor of the computing platform to:

receive modified study design information from a client computing device, the modified study design information having been modified via a user interface of the client computing device;

responsive to the receiving of the modified study design information, apply the at least one trained machine learning model to the modified study design information to predict the score for the each of the plurality of success scores including the predicted overall success score;

responsive to determining that the predicted overall success score is less than a threshold value:

determine which one or more parameter values are negatively affecting the predicted overall success score, and

output information regarding the determined one or more parameter values; and

output the predicted score for the each of the plurality of success scores responsive to applying the at least one trained machine learning model to the modified study design information.

19. The computer program product of claim 16 , wherein the program instructions are further executable by the at least one processor of the computing platform to:

calculate a survey-related score based, at least partly, on measured participant response times, a frequency of participant reported outcomes, outcomes reported, and analysis of textual responses when available;

determine the predicted overall success score based, at least partly, on the survey-related score and the each score for the plurality of success scores; and

output the predicted overall success score and the each score for the plurality of success scores.

20. The computer program product of claim 16 , wherein the program instructions are further executable by the at least one processor of the computing platform to:

determine a number of queries received during the clinical trial, wherein

a predicted score for a data quality score is based, at least partly, on the number of queries received, wherein the plurality of success scores includes the data quality score.