Early detection of disease outbreak using electronic patient data to reduce public health threat from bio-terrorism

• US 8,392,152 B2
• Filed: 08/13/2008
• Issued: 03/05/2013
• Est. Priority Date: 12/14/2001
• Status: Active Grant

• Current Assignee: Cerner Innovation, Inc. (Oracle Corporation)
• Original Assignee: Siemens Medical Solutions USA Incorporated (Siemens AG)
• Inventors: Rao, R. Bharat
• Primary Examiner(s): Rivas, Omar Fernandez
• Assistant Examiner(s): LUU, CUONG V
• Application Number:
  US12/190,675
• Publication Number:
  US 20090076851A1
• Time in Patent Office: 1,665 Days
• Field of Search: 703/2, 703/11, 705/2, 705/3, 600/300
• US Class Current: 703/2
• CPC Class Codes: G06Q 30/0202 Market predictions or forecasting for commercial activities
• CPC Class Codes: G16H 10/60 for patient-specific data, e.g. for electronic patient records
• CPC Class Codes: G16H 50/50 for simulation or modelling of medical disorders
• CPC Class Codes: G16H 50/80 for detecting, monitoring or modelling epidemics or pandemics, e.g. flu
• CPC Class Codes: G16H 70/60 relating to pathologies
• CPC Class Codes: G16Z 99/00 Subject matter not provided for in other main groups of this subclass
• CPC Class Codes: Y02A 90/10 Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
• CPC Class Codes: Y10S 707/99939 Privileged access
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing

Abstract

A method is provided for automatically identifying a disease outbreak indicative of a potential bio-terror attack. Patient records are mined from structured and unstructured clinical sources. The patient records are then analyzed by correlating selected patient data contained in the patient records with disease indicia for each of a plurality of diseases. A probability of a disease outbreak is estimated at least in part based on these correlations. Suspicion may also be indicated if anomalous disease clusters are found. If any of the estimated probabilities exceeds a threshold value, a disease outbreak alert is outputted. The disease indicia may be defined by disease progression models, which may be stored in a disease knowledge base.

Claims

1. A method for automated identification of disease outbreak, the method comprising:

obtaining patient records for a plurality of patients, said records including structured and unstructured data;

for each patient of the plurality of patients, mining data from the unstructured data of the patient record using domain-specific knowledge contained in a knowledge base to produce one or more probabilistic assertions;

for each patient of the plurality of patients, correlating the structured data and the probabilistic assertions with disease indicia from at least one disease to determine a correlation value between the patient and each of the at least one disease;

outputting an alert when at least one correlation value exceeds a first predetermined threshold;

calculating a joint probability that at least one patient of the plurality of patients has the at least one disease based on each of the correlation values and a measure of geographic proximity between each of the plurality of patients; and

outputting an alert when the calculated joint probability exceeds a second predetermined threshold.

2. The method of claim 1, wherein the unstructured data of the patient records includes information regarding patient symptoms.

3. The method of claim 2, wherein the information regarding patient symptoms is derived at least in part from treatment notes.

4. The method of claim 1, wherein the disease indicia are specified by disease progression models stored in a disease knowledge base.

5. The method of claim 1, wherein the outputted alert includes information regarding suspect patients.

6. The method of claim 5, wherein the outputted information about the suspect patients preserves the anonymity of the suspect patients.

7. The method of claim 1, wherein the structured data includes at least one database.

8. The method of claim 7, wherein the at least one database contains one or more of laboratory information, prescription information, test result information or combinations thereof.

9. The method of claim 1, wherein the unstructured data comprises information in text format.

10. The method of claim 9, wherein the information in text format includes one or more of treatment notes, admission slips, reports or combinations thereof.

11. The method of claim 1, wherein the disease indicia includes expected symptoms.

12. The method of claim 1, wherein at least some of the disease indicia are temporally defined.

13. The method of claim 1, wherein the obtained patient records include information gathered from different hospital departments.

14. The method of claim 1, wherein the outputted alert includes a request for additional information.

15. The method of claim 14, wherein the request for additional information includes a request to verify the existence of a specified symptom.

16. The method of claim 14, wherein the request for additional information includes a request to perform a specified task.

17. The method of claim 1, wherein each of the first and second predetermined thresholds depends upon the at least one disease.

18. The method of claim 1, wherein each of the first and second predetermined thresholds depends upon a terror threat level.

19. The method of claim 1, wherein each of the first and second predetermined thresholds depends upon previous false alerts.

20. The method of claim 1, wherein the correlating is performed at a central location and the patient records are provided using a networked hospital information system.

21. The method of claim 1, wherein the outputted alert is sent to a monitoring facility.

22. The method of claim 1, wherein the outputted alert is sent to a government agency.

23. The method of claim 1, where the disease indicia are determined with respect to previous patients whose records are available.

24. The method of claim 1, wherein the alert is outputted even though patient disease progressions do not follow expected disease progressions.

25. The method of claim 24, wherein the disease indicia are extracted from a set of previous patient records.

33. The method of claim 1 wherein the indicia are provided by the Centers for Disease Control and Prevention or another government agency.

35. The method of claim 1 wherein the patient records are obtained from health care facilities or hospitals.

26. A program storage device readable by a machine, the program storage device tangibly embodying a program of instructions executable on the machine to perform a method for automated identification of disease outbreaks, the method comprising:

obtaining patient records for a plurality of patients, said records including structured and unstructured data;

for each patient of the plurality of patients, mining data from the unstructured data of the patient record using domain-specific knowledge contained in a knowledge base to produce one or more probabilistic assertions;

for each patient of the plurality of patients, correlating the structured data and the probabilistic assertions with disease indicia from at least one disease to determine a correlation value between the patient and each of the at least one disease;

outputting an alert when at least one correlation value exceeds a first predetermined threshold;

calculating a joint probability that at least one patient of the plurality of patients has the at least one disease based on each of the correlation values and a measure of geographic proximity between each of the plurality of patients; and

outputting an alert when the calculated joint probability exceeds a second predetermined threshold.

27. A method for automated identification of disease outbreak, the method comprising:

data mining of unstructured data from selected patient records using domain-specific knowledge contained in a knowledge base;

determining a correlation value between the selected patients and indicia of a disease based on the data mined from the unstructured data;

determining a joint probability that at least one patient of the selected patients has the disease based on each of the correlation values;

indicating a disease outbreak when the joint probability exceeds a threshold value; and

outputting a disease outbreak alert for the disease outbreak.

28. The method of claim 27 wherein the selected patient records includes information regarding patient symptoms.

29. The method of claim 27 wherein the mining and determining comprises:

extracting information from a plurality of clinical data sources to create a set of probabilistic assertions;

combining the set of probabilistic assertions to create one or more unified probabilistic assertion; and

inferring patient symptoms from the one or more unified probabilistic assertion.

30. The method of claim 27 wherein the disease indicia includes disease progression models.

31. The method of claim 27, wherein the threshold value depends upon the disease.

32. The method of claim 27, wherein the threshold value depends upon a terror threat level.

34. The method of claim 27 wherein the indicia are provided by the Centers for Disease Control and Prevention or another government agency.

36. The method of claim 27 wherein the selected patient data is obtained from health care facilities or hospitals.

37. A method for automated identification of disease outbreak, the method comprising:

obtaining patient records for a plurality of patients, said records including structured and unstructured data;

for each patient of the plurality of patients, mining data from the unstructured data of the patient record using domain-specific knowledge contained in a knowledge base to produce one or more probabilistic assertions;

for each patient of the plurality of patients, correlating the structured data and the probabilistic assertions with disease early-stage indicia and disease late-stage indicia from at least one disease to determine an whether there is at least a partial match between the early-stage indicia of each patient and the disease and to determine whether there is a match between the late-stage indicia of each patient and the disease; and

outputting an alert when the early-stage indicia for a particular patient only partially matches the disease and the late-stage indicia for the particular patient matches the disease.