Decision analysis system and method for evaluating patient candidacy for a therapeutic procedure

• US 6,533,724 B2
• Filed: 04/26/2001
• Issued: 03/18/2003
• Est. Priority Date: 04/26/2001
• Status: Expired due to Fees

• Current Assignee: Abiomed, Inc. (Johnson & Johnson)
• Original Assignee: Abiomed, Inc. (Johnson & Johnson)
• Inventors: McNair, Douglas S.
• Primary Examiner(s): Winakur, Eric F.
• Assistant Examiner(s): McCrosky, David J.
• Application Number:
  US09/843,243
• Publication Number:
  US 20020193667A1
• Time in Patent Office: 691 Days
• Field of Search: 600/300, 128/898, 128/897, 702/181
• US Class Current: 600/300
• CPC Class Codes: G16H 15/00 ICT specially adapted for medical reports, e.g. generation or transmission thereof
• CPC Class Codes: G16H 20/00 ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: G16H 50/30 for calculating health indices; for individual health risk assessment
• CPC Class Codes: G16H 50/50 for simulation or modelling of medical disorders

Abstract

A decision analysis method for evaluating patient candidacy for a therapeutic procedure is provided. The method includes predicting, in accordance with a predetermined quantitative predicative scoring system, the probability of an adverse effect on such patient absent such procedure where less than all of data elements used in the scoring system may be available. The method is used to predict a probability of death of a patient within a predetermined period of time and includes predicting, in accordance with a predetermined quantitative predicative scoring system. The probability of death for such patient within the predetermined period of time being determined from medical data elements of the patient where less than all of such medical data elements may be available for use by the scoring system. A method is also provided for generating a scoring system to predict a probability of death of a patient within a predetermined period of time.

Claims

1. A method for predicting a healthcare outcome for a patient, comprising:

(a) collecting data values representing a plurality of medical data elements from a plurality of past patients, the medical data elements being likely to be predictors for the healthcare outcome;

(b) grouping the medical data elements into a plurality of composite variables, each composite variable including a plurality of medical data elements;

(c) dichotomizing the medical data elements based on threshold values and assigning points to each medical data element based on the dichotomization;

(d) totaling the dichotomized points within each composite variable for each past patient;

(e) correlating composite variable point totals to healthcare outcome for each past patient to establish a predictive model for the healthcare outcome;

(f) applying the predictive model to the patient to predict the healthcare outcome for the patient.

2. The method of claim 1, wherein the plurality of composite variables comprises at least three composite variables.

3. The method of claim 2, wherein, in addition to at least three composite variables, patient age is collected and maintained as a separate medical data element.

4. The method of claim 1, wherein the medical data elements are grouped into composite variables after determining a significance value for each medical data element, wherein the significance value indicates the significance of the medical data element to the healthcare outcome being predicted, and the plurality of groups each include a plurality of medical data elements having similar significance values.

5. The method of claim 1, wherein the composite variables are dichotomized based on threshold values.

6. The method of claim 5, wherein the composite variables are dichotomized into binary form.

7. The method of claim 1, wherein the medical data elements include only data elements that are available during a predetermined period of time after a patient has been admitted to a clinical care unit.

8. The method of claim 7, wherein the time period is 24 hours.

9. The method of claim 1, wherein the healthcare outcome predicted is mortality within a specified period of time.

10. The method of claim 9, wherein the specified period of time is 30 days.

11. The method of claim 9, wherein the healthcare outcome predicted is mortality without any therapeutic procedure.

12. The method of claim 9, wherein the healthcare outcome predicted is mortality where given a specific therapeutic procedure.

13. The method of claim 12, wherein the predictive model is applied to predict mortality with and without a specific therapeutic procedure for determining whether the specific therapeutic procedure will be effective.

14. The method of claim 13, wherein the predictive model is applied to determine whether the patient is a candidate for a more invasive treatment than the specific therapeutic treatment.

15. The method of claim 14, wherein the more invasive treatment is the implantation of an artificial heart.

16. The method of claim 12, wherein the specific therapeutic procedure is the application of a ventricular assist device.

17. The method of claim 1, further comprising validating the predictive model before applying it to a patient to predict a healthcare outcome.

18. The method of claim 17, wherein validating the predictive model comprises applying the predictive model to a validation cohort of past patients and comparing healthcare outcomes predicted by the predictive model to historical healthcare outcomes.

19. The method of claim 18, wherein the comparison is made by calculating a c-index value.

20. The method of claim 18, further comprising revising threshold values in the predictive model to improve its predictive accuracy.

21. The method of claim 1, wherein the model can be accurately applied to a patient to predict a healthcare outcome where data values are not available for the patient for each of the plurality of medical data elements.

22. The method of claim 21, wherein each composite variable includes a plurality of medical data elements for which data values are available.

23. A method for developing a predictive model for determining a healthcare outcome for a patient, the predictive model operating on a plurality of medical data elements and being capable of predicting a healthcare outcome for the patient when data values for less than all of the plurality of medical data elements are available, the method comprising:

(a) collecting historical patient data values for the plurality of medical data elements;

(b) transforming the data values into dichotomized values;

(c) total the dichotomized values into a plurality of composite variable values; and

(d) correlating the composite variable values to historical healthcare outcomes so that the predictive model can predict a healthcare outcome when data values are not available for all of the medical data elements.

24. The method of claim 23, wherein the plurality of composite variables comprises at least three composite variables.

25. The method of claim 24, wherein, in addition to at least three composite variables, patient age is collected and maintained as a separate medical data element.

26. The method of claim 25, wherein the composite variables are dichotomized into binary form.

27. The method of claim 23, wherein the dichotomized values are totaled into a plurality of composite variables by grouping the medical data elements into composite variables after determining a significance value for each medical data element, wherein the significance value indicates the significance of the medical data element to the healthcare outcome being predicted, and the plurality of groups each include a plurality of medical data elements having similar significance values.

28. The method of claim 23, wherein the data values are transformed into dichotomized values based on threshold values established for the medical data elements.

29. The method of claim 23, wherein the composite variables are dichotomized based on threshold values.

30. The method of claim 23, wherein the medical data elements include only data elements that are available during a predetermined period of time after a patient has been admitted to a clinical care unit.

31. The method of claim 30, wherein the time period is 24 hours.

32. The method of claim 23, wherein the healthcare outcome predicted is mortality within a specified period of time.

33. The method of claim 32, wherein the specified period of time is 30 days.

34. The method of claim 32, wherein the healthcare outcome predicted is mortality without any therapeutic procedure.

35. The method of claim 32, wherein the healthcare outcome predicted is mortality where given a specific therapeutic procedure.

36. The method of claim 35, wherein the specific therapeutic procedure is the application of a ventricular assist device.

37. The method of claim 23, further comprising validating the predictive model before applying it to a patient to predict a healthcare outcome.

38. The method of claim 37, wherein validating the predictive model comprises applying the predictive model to a validation cohort of past patients and comparing healthcare outcomes predicted by the predictive model to historical healthcare outcomes.

39. The method of claim 38, wherein the comparison is made by calculating a c-index value.

40. The method of claim 38, further comprising revising threshold values in the predictive model to improve its predictive accuracy.

41. A method for generating a scoring system to predict a probability of death of a patient within a predetermined period of time, comprising:

obtaining medical information comprising a plurality of medical data elements from each one of a plurality of patients, the medical information including only those data elements available during a predetermined period of time after each of the patients has been admitted to a critical care unit;

determining a significance value for each of the medical data elements, the significance value providing an indication of the significance of the medical data elements on any one of the patients dying within the predetermined period of time;

establishing from the determined significance values a plurality of groups of the medical data elements, the medical data elements in each one of the groups having significance values within a corresponding predetermined significance value range;

dichotomizing each medical data element with a corresponding inclusion/exclusion threshold level;

assigning a point value to each one of the dichotomized medical data elements;

for each one of the patients, calculating a total of the point values for each one of the groups including in the total only the medical elements determined by the last-mentioned dichotomization to be included within the total;

establishing a group threshold level for each one of the groups;

producing a binary output for each one of the groups in accordance with a comparison between the total of the points determined for each one of the groups and the corresponding group threshold; and

processing the binary outputs from the plurality of groups to provide a score corresponding to each one of the binary outputs indicating a probability of death within the predetermined period of time.