Systems and methods for medical data processing and analysis

• US 10,446,264 B2
• Filed: 01/30/2014
• Issued: 10/15/2019
• Est. Priority Date: 01/30/2013
• Status: Active Grant

• Current Assignee: Attomarker Limited
• Original Assignee: Attomarker Limited
• Inventors: Shaw, Andrew Mark
• Primary Examiner(s): Lin, Jerry
• Application Number:
  US14/764,557
• Publication Number:
  US 20150363560A1
• Time in Patent Office: 2,084 Days
• Field of Search: 702 19
• CPC Class Codes: G16B 20/00 ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations
• 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

Abstract

We describe a method of predicting the response of a patient to a medical procedure, the method comprising: inputting acute phase response (APR) biomarker data defining a level of an acute phase response (APR) biomarker in said patient at a succession of biomarker measurement times following said medical procedure, said APR biomarker data defining a biomarker time course representing an evolution over time of said acute phase response; and processing said APR biomarker data to determine a derivative with respect to time of said time course from said APR biomarker data to provide APR time series data; determining a prediction of the response of said patient to said medical procedure from said APR time series data.

Claims

1. A method for treating a patient following a surgical procedure, the method comprising:

inputting acute phase response (APR) biomarker data defining a level of an acute phase response (APR) biomarker in said patient at a succession of biomarker measurement times following said medical procedure, said APR biomarker data defining a biomarker time course representing an evolution over time of said acute phase response;

processing said APR biomarker data to determine a derivative with respect to time of said time course from said APR biomarker data to provide APR time series data, wherein said processing of said APR biomarker data to determine said derivative with respect to time of said time course further comprises correcting for a clearance rate of said biomarker in said patient;

identifying one or both of a maximum and a minimum in said APR time series data;

determining a value for an alert state variable responsive to said APR time series data; and

if the alert state variable has a first value then treating the patient for infection.

2. A method as claimed in claim 1 further comprising one or both of disregarding a data value and flagging an error, where said determined derivative has a value defining a rate of fall of said level of said APR biomarker greater than said clearance rate.

3. A method as claimed in claim 1 wherein said determining of said value for said alert state variable comprises identifying one or more of the following conditions:

(i) a falling level of said APR time series data after an initial delay time;

(ii) a point of inflection in said APR time series data;

(iii) a minimum in said APR time series data;

(iv) a rise in said APR time series data in response to a said biomarker measurement; and

(v) a fall in the APR time series data in response to a said biomarker measurement;

and, responsive to said identifying, adjusting said value for said alert state variable.

4. A method as claimed in claim 1 wherein determining if the alert state variable has a first value comprises determining, given said value of said alert state variable, a probability of one of a set of possible patient responses, the set of possible responses including infection.

5. A method as claimed in claim 4 wherein determining if the alert state variable has a first value comprises determining the probability P(response|alert state variable) of a said response given said value of said alert state variable, comprises determining as a function of time:

P(alert state variable|response)·

P(alert state variable)

where P(alert state variable|response) is a probability of said alert state variable given said response, and P(alert state variable) is a prior probability of said alert state variable.

6. A method as recited in claim 5 wherein one or more terms of P(alert state variable|response)·

P(alert state variable) is dependent upon a comorbidity of said patient.

7. A method as recited in claim 1 wherein said processing of said APR biomarker data comprises evaluating:

8. A method as recited in claim 1 further comprising performing an assay to determine said level of said CRP.

9. A physical data carrier carrying computer program code to implement the method of claim 1.

10. A method as claimed in claim 1 further comprising:

if the alert state variable has a second value then identifying the patient as suitable for discharge.

11. A method as claimed in claim 1 wherein said APR biomarker comprises CRP (C-Reactive Protein) and said APR biomarker data defines a level of CRP in the patient.

12. A medical system for treating a patient following a surgical procedure, the system comprising:

working memory;

program memory;

a processor coupled to said working memory and to said program memory;

an input to receive APR biomarker data comprising data representing a level of an acute phase response (APR) biomarker in said patient at a succession of biomarker measurement times following said medical procedure, said APR biomarker data defining a biomarker time course representing an evolution over time of said acute phase response, said APR biomarker comprising CRP (C-Reactive Protein) and said APR biomarker data defining a level of CRP in the patient; and

wherein said program memory stores processor control code to;

process said APR biomarker data to determine a derivative with respect to time of said time course from said APR biomarker data to provide APR time series data, wherein said processing of said APR biomarker data to determine said derivative with respect to time of said time course further comprises correcting for a clearance rate of CRP in said patient;

identify one or both of a maximum and a minimum in said APR time series data;

determine a value for an alert state variable responsive to said APR time series data; and

if the alert state variable has a first value then identify the patient for treatment for infection.