OPTIMIZING CLUSTER BASED COHORTS TO SUPPORT ADVANCED ANALYTICS

• US 20090240513A1
• Filed: 03/24/2008
• Published: 09/24/2009
• Est. Priority Date: 03/24/2008
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Kraemer, James R., Angell, Robert Lee, Friedlander, Robert R.
• Granted Patent: US 8,335,698 B2
• Time in Patent Office: Days
• US Class Current: 705/1.100
• CPC Class Codes: G06Q 10/10 Office automation; Time management

Abstract

A computer implemented method, apparatus, and computer program product is provided for selecting subjects for treatment study cohorts. A set of selected dimensions for optimizing selection of subjects for a treatment cohort group and a control cohort group associated with a treatment study is identified. Attribute data associated with subjects in the pool of available subjects is clustered at the atomic level to form clustered cohort data. A set of optimized subjects from a pool of available subjects is selected using the clustered cohort data and the set of selected dimensions. Subjects in the set of optimized subjects are optimized across the set of selected dimensions. Each subject in the set of optimized subjects is assigned to the treatment cohort group or the control cohort group.

Claims

1. A computer implemented method for selecting subjects for treatment study cohort groups, the computer implemented method comprising:

identifying a set of selected dimensions for optimizing selection of subjects for a treatment cohort group and a control cohort group associated with a treatment study;

clustering attribute data associated with subjects in a pool of available subjects at an atomic level to form clustered cohort data;

selecting a set of optimized subjects from the pool of available subjects using the clustered cohort data and the set of selected dimensions, wherein subjects in the set of optimized subjects are optimized across the set of selected dimensions; and

assigning each subject in the set of optimized subjects to the treatment cohort group or the control cohort group, wherein the each subject in the set of optimized subjects is assigned to the treatment cohort group or the control cohort group to minimize differences between the subjects in the treatment cohort group and the subjects in the control cohort group.

2. The computer implemented method of claim 1 wherein an objective function selects the subjects from the pool of available subjects to optimized the selection of subjects across the set of selected dimensions.

3. The computer implemented method of claim 1 wherein the treatment study comprises a plurality of treatments, wherein each treatment is associated with a treatment cohort group and a control cohort group, and further comprising:

assigning subjects in the pool of available subjects to each treatment cohort group and each control cohort group associated with each treatment in the plurality of treatments in accordance with a set of constraints for the each treatment in the plurality of treatments, and wherein the subjects are assigned to the each treatment cohort group and the each control cohort group to optimize the set of dimensions for the each treatment cohort group and the each control cohort group.

4. The computer implemented method of claim 1 further comprising:

identifying a set of minimum criteria and a set of constraints associated with each treatment in a plurality of treatments associated with the treatment study, wherein each treatment is associated with a treatment cohort group and a control cohort group;

identifying subjects in a plurality of potential subjects that satisfy the set if minimum criteria and the set of constraints to form the pool of available subjects;

analyzing attribute data associated with each subject in the pool of available subjects to form cohort data;

selecting a set of subjects for each treatment cohort group associated with the each treatment using the cohort data, wherein the set of subjects for the each treatment cohort group comprises subjects that satisfy the set of minimum criteria and the set of constraints, and wherein the each treatment is associated with a different set of constraints.

5. The computer implemented method of claim 1 wherein a given subject in the pool of available subjects assigned to the control cohort group remains available for assignment to other different control cohort groups, wherein a given subject is permitted to be a member of more than one control cohort group, and wherein any subject assigned to any control cohort group is unavailable for assignment to any treatment cohort group.

6. The computer implemented method of claim 1 further comprising:

responsive to a determination that a subject in the pool of available subjects is a desirable subject for both a first treatment and a second treatment associated with the treatment study, assigning the first subject to a control cohort group for the first treatment and a control cohort group for the second treatment, wherein the assignment of the subject to control cohort groups optimizes selection of subjects for a maximum number of control cohort groups; and

responsive to a determination that the subject is desirable for only the first treatment associated with the treatment study, assigning the subject to a treatment cohort group associated with the first treatment.

7. The computer implemented method of claim 1 wherein a given subject assigned to the treatment cohort group is unavailable for assignment to any other different treatment cohort group, and wherein the given subject assigned to the treatment cohort group is unavailable for assignment to any control cohort group.

8. The computer implemented method of claim 1 wherein the treatment cohort group comprises subjects that receive a given treatment and wherein the control cohort group comprises subjects that do not receive the given treatment.

9. The computer implemented method of claim 1 wherein the subjects in the set of optimized subjects comprises subjects having attributes that minimizes at least one dimension in the set of selected dimensions.

10. The computer implemented method of claim 1 wherein the subjects in the set of optimized subjects comprises subjects having attributes that maximizes at least one dimension in the set of selected dimensions.

11. A computer program product for selecting subjects for treatment study cohort groups, the computer program product comprising:

a computer-readable medium;

program code stored on the computer-readable medium for identifying a set of selected dimensions for optimizing selection of subjects for a treatment cohort group and a control cohort group associated with a treatment study;

program code stored on the computer-readable medium for clustering attribute data associated with subjects in a pool of available subjects at an atomic level to form clustered cohort data;

program code stored on the computer-readable medium for selecting a set of optimized subjects from the pool of available subjects using the clustered cohort data and the set of selected dimensions, wherein subjects in the set of optimized subjects are optimized across the set of selected dimensions; and

program code stored on the computer-readable medium for assigning each subject in the set of optimized subjects to the treatment cohort group or the control cohort group, wherein the each subject in the set of optimized subjects is assigned to the treatment cohort group or the control cohort group to minimize differences between the subjects in the treatment cohort group and the subjects in the control cohort group.

12. The computer program product of claim 11 wherein an objective function selects the subjects from the pool of available subjects to optimize the selection of subjects across the set of selected dimensions.

13. The computer program product of claim 11 wherein the treatment study comprises a plurality of treatments, wherein each treatment is associated with a treatment cohort group and a control cohort group, and further comprising:

program code stored on the computer-readable medium for assigning subjects in the pool of available subjects to each treatment cohort group and each control cohort group associated with each treatment in the plurality of treatments in accordance with a set of constraints for the each treatment in the plurality of treatments, and wherein the subjects are assigned to the each treatment cohort group and the each control cohort group to optimize the set of dimensions for the each treatment cohort group and the each control cohort group.

14. The computer program product of claim 11 further comprising:

program code stored on the computer-readable medium for identifying a set of minimum criteria and a set of constraints associated with each treatment in a plurality of treatments associated with the treatment study;

program code stored on the computer-readable medium for identifying subjects in a plurality of potential subjects that satisfy the set of minimum criteria and the set of constraints to form the pool of available subjects;

program code stored on the computer-readable medium for analyzing attribute data associated with each subject in the pool of available subjects to form cohort data;

program code stored on the computer-readable medium for selecting a set of subjects for each treatment cohort group associated with the each treatment using the cohort data, wherein the set of subjects for the each treatment cohort group comprises subjects that satisfy the set of minimum criteria and the set of constraints, and wherein the each treatment is associated with a different set of constraints.

15. The computer program product of claim 11 wherein a given subject assigned to the control cohort group remains available for assignment to other different control cohort groups, wherein a given subject is permitted to be a member of more than one control cohort group, wherein any subject assigned to any control cohort group is unavailable for assignment to any treatment cohort group, wherein a given subject assigned to the treatment cohort group is unavailable for assignment to any other different treatment cohort group, and wherein the given subject assigned to the treatment cohort group is unavailable for assignment to any control cohort group.

16. The computer program product of claim 11 further comprising:

program code stored on the computer-readable medium for assigning a subject to a control cohort group for a first treatment and a control cohort group for a second treatment in response to a determination that the subject is a desirable subject for both the first treatment and the second treatment associated with the treatment study, wherein the assignment of the subject to control cohort groups optimizes selection of subjects for a maximum number of control cohort groups; and

program code stored on the computer-readable medium for assigning the subject to a treatment cohort group associated with the first treatment in response to a determination that the subject is desirable for only the first treatment associated with the treatment study.

17. The computer program product of claim 11 wherein the subjects in the set of optimized subjects comprises subjects having attributes that minimizes at least one dimension in the set of selected dimensions.

18. The computer program product of claim 11 wherein the subjects in the set of optimized subjects comprises subjects having attributes that maximizes at least one dimension in the set of selected dimensions.

19. An apparatus comprising:

a bus system;

a communications system coupled to the bus system;

a memory connected to the bus system, wherein the memory includes computer usable program code; and

a processing unit coupled to the bus system, wherein the processing unit executes the computer-usable program code to identify a set of selected dimensions for optimizing selection of subjects for a treatment cohort group and a control cohort group associated with a treatment study;

cluster attribute data associated with subjects in a pool of available subjects at an atomic level to form clustered cohort data;

select a set of optimized subjects from the pool of available subjects using the clustered cohort data and the set of selected dimensions; and

assign each subject in the set of optimized subjects to the treatment cohort group or the control cohort group, wherein subjects in the set of optimized subjects are optimized across the set of selected dimensions, and wherein the each subject in the set of optimized subjects is assigned to the treatment cohort group or the control cohort group to minimize differences between the subjects in the treatment cohort group and the subjects in the control cohort group.

20. The apparatus of claim 19 wherein an objective function selects the subjects from the pool of available subjects to optimize the selection of subjects across the set of selected dimensions.

21. The apparatus of claim 19 wherein the treatment study comprises a plurality of treatments, wherein each treatment in the plurality of treatments comprises a treatment cohort group and a control cohort group, and wherein subjects in the pool of available subjects are assigned to treatment cohort groups and control cohort groups associated with each treatment in the plurality of treatments in accordance with a set of constraints for the each treatment in the plurality of treatments, and wherein the subjects are assigned to the each treatment cohort group and the each control cohort group to optimize the set of dimensions.

22. The apparatus of claim 19 wherein a given subject assigned to the control cohort group remains available for assignment to other different control cohort groups, wherein a given subject is permitted to be a member of more than one control cohort group, wherein any subject assigned to any control cohort group is unavailable for assignment to any treatment cohort group, wherein a given subject assigned to the treatment cohort group is unavailable for assignment to any other different treatment cohort group, and wherein the given subject assigned to the treatment cohort group is unavailable for assignment to any control cohort group.

23. The apparatus of claim 19 wherein the processor unit further executes the computer-usable program code to assign a subject to a control cohort group for a first treatment and a control cohort group for a second treatment in response to a determination that the subject is a desirable subject for both the first treatment and the second treatment associated with the treatment study, wherein the assignment of the subject to control cohort groups optimizes selection of subjects for a maximum number of control cohort groups;

and assign the subject to a treatment cohort group associated with the first treatment in response to a determination that the subject is desirable for only the first treatment associated with the treatment study.

24. The apparatus of claim 19 wherein the subjects in the set of optimized subjects comprises subjects having attributes that minimizes at least one dimension in the set of selected dimensions.

25. The apparatus of claim 19 wherein the subjects in the set of optimized subjects comprises subjects having attributes that maximizes at least one dimension in the set of selected dimensions.