System and method for detecting spatiotemporal clusters

• US 20040236604A1
• Filed: 12/22/2003
• Published: 11/25/2004
• Est. Priority Date: 12/20/2002
• Status: Abandoned Application

• Current Assignee: Cerner Innovation, Inc. (Oracle Corporation)
• Original Assignee: Cerner Innovation, Inc. (Oracle Corporation)
• Inventors: McNair, Douglas S.
• Application Number:
  US10/744,976
• Publication Number:
  US 20040236604A1
• Time in Patent Office: Days
• US Class Current: 705/2
• CPC Class Codes: G16H 50/80 for detecting, monitoring or modelling epidemics or pandemics, e.g. flu
• 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

Abstract

A method and system suitable for automated surveillance for disease detection and particularly for syndromic information represented in the transaction order records from clinical information systems of hospitals, clinics, and emergency rooms. Patient'"'"'s findings and conditions expressed as order annotations may be combined with evidence recorded by physicians in a syndromic classifier software application that yields daily counts of proband cases possibly denoting the occurrence of a bioterrorist-caused illness. Techniques from digital signal processing and statistical spatiotemporal image processing are combined in a method that allows for optimization of the parameters of the signal processing and statistical hypothesis testing, prediction for planning and decision-making, and inferencing. Once optimized, the method and system can achieve high-sensitivity high-specificity detection of true occurrences of syndromic illness such as bioterrorism, while avoiding false-alarm signals.

Claims

1. A method in a computing environment for disease surveillance and outbreak detection, the method comprising the steps of:

accessing population and location data for a plurality of geographies;

accessing clinical data originating from a plurality of reporting units associated with the geographies, said clinical data including a count of affected persons for a predetermined time period;

calculating exponential moving averages of standardized syndromic incidence ratios for each count of affected person from each reporting unit; and

generating kriging plots and Moran I statistics, wherein a p-value is computed for each spatiotemporal pattern.

2. The method of claim 1, wherein the geographies are counties.

3. The method of claim 1, wherein the predetermined time period is a day.

4. The method of claim 1, wherein the location data includes the longitude and latitude of the each geography.

5. The method of claim 1, further comprising the step of calculating the K-nearest neighbor based on the longitude and latitude of the geographies.

6. The method of claim 1, wherein the geographies are cities.

7. The method of claim 5, wherein the exponential moving averages are calculated for a period of days corresponding to an epidemiologically optimized value reflecting the statistical distribution of incubation periods of likely agents of bioterrorism.

8. The method of claim 5, further comprising the step of performing a Freemen Tukey transformation of the incidence ratio for each reporting unit.

9. The method of claim 8, further comprising the step of performing a logit transformation of the incidence ratio for each reporting unit.

10. The method of claim 1, further comprising the step of providing an alert if the p-level is greater than or equal to a predetermined constant

11. A method in a computing environment for effecting a controlled, recurring assessment of a spatiotemporal event incidence patterns on a state or national level, the method comprising the steps of:

accessing transmissions data received from a plurality of reporting units, said transmission data including proband counts for a predefined time period;

totalizing said proband counts;

recursively calculating exponential moving averages (EMAs) of standardized syndromic incidence ratios, for each reporting unit'"'"'s reported counts of proband cases;

generating kriging plots and Moran I statistics, wherein a p-value is computed for each spatiotemporal pattern, and

if the p-level is greater than or equal to a predetermined constant, initiating an alert or escalating vigilance in a particular geography.

12. The method of claim 11, further comprising the step of performing a Freeman-Tukey and logit transformation of each reporting unit'"'"'s standardized daily syndromic incidence ratio, to normalize the variance and to prevent excessive sensitivity to single-count reports from reporting units having small population.

13. The method of claim 11, wherein the predefined time period is a day.

14. A system in a computerized environment for disease surveillance and outbreak detection, the system comprising:

a first accessing component for accessing population and location data for a plurality of geographies;

a second accessing component for accessing clinical data originating from a plurality of reporting units associated with the geographies, said clinical data including a count of affected persons for a predetermined time period;

a calculating component for calculating exponential moving averages of standardized syndromic incidence ratios for each count of affected person from each reporting unit; and

a generating component for generating Moran I statistics, wherein a p-value is computed for each spatiotemporal pattern.