Identifying risk of a medical event

• US 20080009684A1
• Filed: 05/31/2007
• Published: 01/10/2008
• Est. Priority Date: 05/31/2006
• Status: Active Grant

• Current Assignee: University of Rochester
• Original Assignee: University of Rochester
• Inventors: Ryan, Daniel, Corsetti, James, Sparks, Charles, Moss, Arthur
• Granted Patent: US 8,979,753 B2
• Time in Patent Office: Days
• US Class Current: 600/300
• CPC Class Codes: A61B 5/00 Measuring for diagnostic purposes radiation diagnosis A61B6/00; diagnosis by ultrasonic, sonic or infrasonic waves A61B8/00; Identification of persons
• CPC Class Codes: A61B 5/14532 for measuring glucose, e.g. by tissue impedance measurement
• CPC Class Codes: A61B 5/7271 Specific aspects of physiological measurement analysis A61B5/316, A61B5/349 and A61B5/374 take precedence
• CPC Class Codes: A61B 5/7275 Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor
• CPC Class Codes: G01N 2800/04 Endocrine or metabolic disorders
• CPC Class Codes: G01N 2800/60 Complex ways of combining multiple protein biomarkers for diagnosis
• CPC Class Codes: G06T 17/00 Three dimensional [3D] modelling, e.g. data description of 3D objects
• CPC Class Codes: G06T 2215/16 Using real world measurements to influence rendering
• CPC Class Codes: G16H 15/00 ICT specially adapted for medical reports, e.g. generation or transmission thereof
• 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: Y10S 128/92 Computer assisted medical diagnostics
• CPC Class Codes: Y10S 128/923 by comparison of patient data to other data
• CPC Class Codes: Y10S 707/99937 Sorting
• CPC Class Codes: Y10S 707/9994 Distributed or remote access

Abstract

Described is a method of determining the relative risk of an outcome based on an analysis of multiple risk factors. A graphical method is used to take values corresponding to risk parameters and an event outcome to produce a smoothed surface map representing relative risk over an entire space defined by n risk factors. Applying a query data point to the surface map permits the determination of the estimated outcome probability for the query data point, based on its location on the surface map. The method then reports a relative risk or other probability measure associated with the query data point. Also described is a method of analysis in which subpopulations previously identified as high-risk can be further analyzed with respect to risk posed by additional factors.

Claims

1. A method for identifying risk of a medical outcome based on an analysis of multiple risk factors, comprising:

providing a reference database comprising a plurality of reference data points, each reference data point comprising a value for each of n risk-factor parameters, wherein n≧

2;

wherein each reference data point further comprises a corresponding outcome value; and

wherein the outcome value is determined by the prior occurrence or non-occurrence of a medical outcome;

mapping each reference data point as a map point in (n+1)-dimensional space, thereby producing a reference map;

wherein the position of each map point in n dimensions is determined by the values of the associated reference data point'"'"'s n risk-factor parameters; and

wherein the position of each map point in the (n+1)th dimension is determined by the outcome value for the associated reference data point;

applying an interpolation algorithm to the reference map to produce an interpolated (n+1)-dimensional map;

wherein the interpolated (n+1)-dimensional map comprises a surface that defines a function that maps values of the n risk-factor parameters to predicted outcome values; and

wherein each of the predicted outcome values represents at least one of a relative risk and a probability of occurrence of the medical outcome;

mapping a query data point, comprising values for each of the n risk-factor parameters, onto the interpolated (n+1)-dimensional map;

wherein the location of the query data point on the interpolated (n+1)-dimensional map is determined, at least in part, by the query data point'"'"'s values of the n risk-factor parameters;

determining a predicted outcome value for the query data point by mapping, with the function, the query data point'"'"'s values of the n risk-factor parameters to a point in the surface, the point in the surface being indicative of the predicted outcome value for the query data point; and

outputting at least one of the query data point'"'"'s predicted outcome value, its associated relative risk, and its associated probability, to an output device.

2. The method of claim 1, wherein n=2.

3. The method of claim 1, wherein at least one of the risk-factor parameters comprises a continuous range of values.

4. The method of claim 1, wherein at least one of the risk-factor parameters comprises a range of discrete values.

5. The method of claim 1, wherein at least one of the risk-factor parameters comprises a binary range of values.

6. The method of claim 1, wherein at least one of the risk-factor parameters comprises a presence or absence of a risk factor, a pre-existing condition, or a history of a pre-existing condition.

7. The method of claim 1, wherein the outcome comprises a coronary event in a patient.

8. The method of claim 7, wherein the coronary event comprises at least one of a malignant arrhythmia, a myocardial infarction, and an episode of unstable angina.

9. The method of claim 1, wherein a relative risk represented by the query data point'"'"'s predicted outcome value comprises a range of values of relative risk.

10. The method of claim 1, wherein a probability represented by the query data point'"'"'s predicted outcome value comprises a range of values of probability.

11. The method of claim 1, wherein the surface comprises points representing values or functions of probabilities of occurrence of the medical outcome.

12. The method of claim 1, further comprising outputting a graphical representation of the interpolated (n+1)-dimensional map to an output device.

13. The method of claim 1, wherein the output device comprises at least one of a numerical display, a graphical display, an electronic monitor, paper, an LCD, and a hand-held display.

14. The method of claim 1, wherein at least one of the n risk-factor parameters is selected from the group consisting of blood levels of apoA1, apoB, total serum cholesterol, HDL, cholesterol, triglyceride, glucose, insulin, plasminogen activator inhibitor-1, lipoprotein (a), C-reactive protein, von Willibrand factor, von Willibrand factor antigen, fibrinogen, D-dimer, Factor VII, Factor VIIa, and lipoprotein-associated phospholipase A₂.

15. The method of claim 1, wherein at least one of the n risk-factor parameters is selected from the group consisting of an LDL peak particle diameter, a history of type II diabetes, a history of a pre-diabetic syndrome, a history of metabolic syndrome, a polymorphism in NAD(P)H oxidase, a polymorphism in Glycoprotein-Ibα

, a history of obesity, a body mass index, and a history of a previous coronary event.

16. The method of claim 15, wherein the coronary event comprises at least one of a malignant arrhythmia, a myocardial infarction, and an episode of unstable angina.

17. The method of claim 1, wherein two of the risk-factor parameters comprise a genetic polymorphism in NAD(P)H oxidase and a serum HDL level.

18. The method of claim 17, wherein an additional risk-factor parameter comprises at least one of a C-reactive protein level and a cholesterol level.

19. The method of claim 17, wherein the genetic polymorphism comprises a polymorphism in the p22phox subunit of the NAD(P)H oxidase.

20. The method of claim 19, wherein the genetic polymorphism comprises a C or T nucleotide at position +242 of the NAD(P)H oxidase gene p22phox subunit.

21. The method of claim 20, wherein the polymorphism further comprises a homozygosity for a C nucleotide at position +242 of the NAD(P)H oxidase gene p22phox subunit.

22. The method of claim 17, wherein the HDL is H_2a. HDL.

23. The method of claim 17, wherein the HDL is an HDL sub-fraction comprising particles corresponding in size to H_2a HDL particles.

24. The method of claim 1, wherein two of the risk-factor parameters comprise two of the following:

a serum HDL level, a CRP level, and a total serum cholesterol level.

25. The method of claim 1, wherein one of the risk-factor parameters comprises a history of a prior myocardial infarction.

26. The method of claim 1, wherein two of the risk-factor parameters comprise a body fluid glucose level and a serum triglyceride level.

27. The method of claim 26, wherein the body fluid comprises at least one of serum and cerebrospinal fluid.

28. The method of claim 1, wherein two of the risk-factor parameters comprises a body mass index and a fibrinogen level.

29. The method of claim 1, wherein one of the risk-factor parameters comprises a plasminogen activator inhibitor level.

30. The method of claim 29, wherein plasminogen activator inhibitor comprises plasminogen activator inhibitor-1 (PAI-1).

31. The method of claim 1, wherein one of the risk-factor parameters comprises a history of metabolic syndrome.

32. The method of claim 1, wherein one of the risk-factor parameters comprises a polymorphism in Glycoprotein-Ibα

such that there is a threonine or a methionine residue at amino acid residue 145 of Glycoprotein-Ibα

.

33. The method of claim 32, wherein the polymorphism comprises having at least one Glycoprotein-Ibα

allele that encodes a Met at amino acid residue 145.

34. The method of claim 33, wherein one of the risk-factor parameters comprises a history of diabetes.

35. The method of claim 33, wherein one of the risk-factor parameters comprises a triglyceride level.

36. The method of claim 33, wherein one of the risk-factor parameters comprises a lipoprotein-associated phospholipase A₂ level.

37. A method, of identifying or estimating risk of a medical outcome in a subpopulation of patients, comprising:

providing a first plurality of reference data points representing a population of patients, each reference data point comprising a value for each of n risk-factor parameters, wherein n≧

2;

wherein each reference data point further comprises a corresponding outcome value; and

wherein the outcome value is determined by the prior occurrence or non-occurrence of a medical outcome;

mapping each reference data point as a map point in (n+1)-dimensional space, thereby producing a reference map;

wherein the position of each map point in n dimensions is determined by the values of the associated reference point'"'"'s n risk-factor parameters; and

wherein the position of each map point in the (n+1)th dimension is determined by the outcome value for the associated reference data point;

applying an interpolation algorithm to the reference map to produce an interpolated (n+1)-dimensional map;

wherein the interpolated (n+1)-dimensional map comprises a surface that defines a function that maps values for the n risk-factor parameters to predicted outcome values; and

wherein each of the predicted outcome values represents at least one of a relative risk and a probability of occurrence of the medical outcome;

locating on the interpolated (n+1)-dimensional map a first subpopulation of reference data points corresponding to a first subpopulation of patients within the population of patients, the first subpopulation of reference data points representing an increased or decreased risk of the medical outcome relative to another subpopulation of patients within the population;

providing a plurality of subpopulation data points corresponding to the first subpopulation of patients, each subpopulation data point comprising a value for each of m risk-factor parameters, wherein m≧

2;

wherein each subpopulation data point further comprises a corresponding subpopulation outcome value; and

wherein the subpopulation outcome value is determined by the prior occurrence or non-occurrence of the medical outcome;

mapping each subpopulation data point as a map point in (m+1)-dimensional space, thereby producing a second reference map;

wherein the position of each map point in m dimensions is determined by the values of the associated subpopulation data point'"'"'s m risk-factor parameters; and

wherein the position of each map point in the (m+1)th dimension is determined by the subpopulation outcome value for the associated subpopulation data point;

applying an interpolation algorithm to the second reference map to produce an interpolated (m+1)-dimensional map;

wherein the interpolated (m+1)-dimensional map comprises a second surface that defines a second function that maps values for the m risk-factor parameters to subpopulation predicted outcome values; and

wherein each of the subpopulation predicted outcome values represents at least one of a subpopulation relative risk and a subpopulation probability of occurrence of the medical outcome; and

outputting at least one of a subpopulation predicted outcome value, a subpopulation relative risk, and a subpopulation probability of occurrence of the medical outcome, to an output device.