METHODS FOR DETERMINING RISK AND TREATING DISEASES AND CONDITIONS THAT CORRELATE TO WEATHER DATA

• US 20170308672A1
• Filed: 09/04/2015
• Published: 10/26/2017
• Est. Priority Date: 09/05/2014
• Status: Active Grant

• Current Assignee: ERREx, Inc., University of Cincinnati
• Original Assignee: University of Cincinnati
• Inventors: Martin, Vincent, Levin, Linda, Simmons, Mark Russell, Nicholson, Robert Allen, Peterlin, Albert
• Granted Patent: US 10,665,349 B2
• Time in Patent Office: Days
• CPC Class Codes: G01W 1/02 Instruments for indicating weather conditions by measuring two or more variables, e.g. humidity, pressure, temperature, cloud cover or wind speed G01W1/10 takes precedence
• CPC Class Codes: G06F 17/11 for solving equations , e.g. nonlinear equations, general mathematical optimization problems optimization specially adapted for a specific administrative, business or logistic context G06Q10/04
• CPC Class Codes: G06F 2111/10 Numerical modelling
• CPC Class Codes: G06F 30/20 Design optimisation, verification or simulation optimisation, verification or simulation of circuit designs G06F30/30
• CPC Class Codes: G16H 40/20 for the management or administration of healthcare resources or facilities, e.g. managing hospital staff or surgery rooms
• 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
• CPC Class Codes: G16H 50/70 for mining of medical data, e.g. analysing previous cases of other patients
• 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

Methods and model equations are provided for predicting a risk of a subject who experiences adverse medical events associated with the weather, of experiencing a new-onset event. Methods include a) identifying a climate region of interest; b) collecting daily mean barometric pressure (BP) data for a time frame and dividing the days of the time frame into at least upper, middle and lower quantile BP days to identify the upper quantile BP days.

Claims

1. A method for generating a model set of equations for predicting a risk of a subject who experiences adverse medical events associated with the weather, of experiencing a new-onset event (NOE), the method comprising:

a) identifying a climate region of interest;

b) collecting daily mean barometric pressure (BP) data for a time frame and dividing the days of the time frame into at least upper, middle and lower quantile BP days to identify the upper quantile BP days;

c) collecting daily NOE data for a subject cohort consisting of subjects known to suffer from the adverse medical event for the time frame and calculating a daily incident rate (IR) of NOEs for each day of the time frame and dividing the days of the time frame into at least upper, middle and lower quantile IR-NOE days to identify the upper quantile IR-NOE (UQ-IR-NOE) days;

d) determining a relevant number of seasons based on an association between the upper IR-NOE quantile days identified in c) and the upper BP quantile days identified in b);

e) collecting hourly weather data for the time frame for a number of weather parameters and determining a set of weather variables;

f) employing a generalized linear regression analysis to generate a rank for each weather variable as a predictor of the UQ-IR-NOE days for each relevant season, for each BP quantile;

g) identifying a predictive equation using a forward stepwise approach,

wherein the model comprises a set of one or more equations for predicting the risk of a subject experiencing a new-onset event (NOE) at the completion of step (g).

2. The method according to claim 1, wherein the number of relevant seasons is 2-4 and the time frame is one year.

3. The method of claim 1, wherein the generalized linear regression analysis of step (f) comprises a generalized mixed analysis or a generalized estimating equation (GEE) analysis;

4. The method of claim 3, wherein the predictor of step (f) is continuous.

5. The method of claim 4, wherein the generalized linear regression analysis of step (f) comprises GEE and the forward stepwise approach of step (g) comprises the following steps:

i) identifying a best predictive single variable equation based on p-value and quasi-likelihood independence criterion (QIC) fit of the first-ranked weather variable in each season, for each BP quantile;

ii) adding the next-ranked weather variable to the identified equation from g) in each season, for each BP and determine if fit improves;

iii) repeating step (ii) until addition of the next-ranked weather variable to the model fails to improve fit.

6. The method according to claim 4, wherein the generalized linear regression analysis of step (f) comprises a generalized mixed regression analysis and the forward stepwise approach of step (g) comprises the following steps:

i) identifying a best predictive single variable equation based on Akaike information criterion (AIC) fit of the first-ranked variable in each season, for each BP quantile;

ii) adding the next-ranked weather variable to the identified equation from (g) in each season, for each BP and determine if fit improves;

iii) repeating step (ii) until addition of the next-ranked variable to the model fails to improve fit.

7. The method according to 1, wherein the number of seasons determined in step (d) is based on a regression analysis of the UQ-BP and the UQ-IR-NOE, said regression analysis comprising one or both of leverage point and outlier detection diagnostics.

8. The method according to claim 7, wherein a daily incident rate is equal to the number of subjects experiencing a NOE divided by the total number of eligible subjects;

“

eligible subject”

being defined as a subject who has not experienced an event in the preceding 24 hours.

9. The method according to claim 1, wherein the weather variables are based on weather parameters selected from the group consisting of barometric pressure (BP), wind speed (WS), wind direction (WD), dry bulb temperature (DBT), lightening activity (LA), precipitation (P) and relative humidity (RH).

10. The method according to claim 9, wherein the weather variables comprise a daily mean and multiple differential maximums, minimums and means for each selected parameter.

11. The method according to claim 10, wherein the multiple differentials are selected from a 24 hour differential, a 12 hour differential, a 6 hour differential a 3 hour differential, and combinations thereof for each selected parameter.

12. The method according to claim 8, wherein collected data is collected from multiple geographic locations within the climate region.

13. The method according to claim 12, wherein the climate region is classified as Cfa in the Kö

ppen-Geiger climate region classification system and the relevant seasons comprise F/W/S (9/21-6/20) and Summer (6/21-9/20).

14. The method according to claim 13, wherein the weather associated adverse event is new onset of a migraine headache (NOE=NOH), the quantile is a tertile and the model set of equations includes the following weather variables:

A) BP 24 hour differential mean;

B) WS 24 hour differential maximum;

C) DBT 24 hour differential mean;

D) DBT 6 hour differential mean;

E) BP 24 hour differential maximum;

F) DBT daily mean;

G) BP daily mean; and

H) RH 24 hour differential mean.

15. A model set of equations generated according to the method of claim 12.

16. A model set of equations generated according to the method of claim 14, wherein R=risk of a given day being a UT-IR-NOH day, the model set of equations comprising:

[season=W,BP tertile=low]R=e^[β

₀^+β

₁^A+β

₂^B]+N+ε

or 1=e[^β

₀+β

₁A+β

₂^A(exp)2+β

₃B+β

₄^B(exp)2]+N+ε

 

 

1.

[season=W,BP tertile=mid]R=e[β

₀^+β

₁^A+β

₂^B]+N+ε

 

 

2.

[season=W,BP tertile=high]R=e^[β

₀^+β

₁^A+β

₂^C]+N+ε

 

 

3.

[season=S,BP tertile=low]R=e^[β

₀^+β

₁^E]+N+ε

 

 

4.

[season=S,BP tertile=mid]R=e^[β

₀^+β

₁^G+β

₂^F]+N+ε

 

 

5.

[season=S,BP tertile=high]R=e^[β

₀^+β

₁^H]+N+ε

 

 

6.

wherein N is the number of subjects in the cohort eligible to have an NOH on a given day and is the denominator in the IR-NOH calculation, and ε

is an error term of GEE regression modeling.

17. A method of assessing a risk of a subject predisposed to experiencing weather-associated adverse events for experiencing a new onset event on a given day, the method comprising:

a) determining a climate region associated with the geographical location of the subject;

b) determining the relevant season in which the given day falls;

c) determining the projected mean BP for the day and identifying it as an upper, middle or lower quantile BP day;

d) selecting an equation from the model set of equations according to claim 15 specific to the determined climate region, the determined relevant season and the determined BP quantile; and

e) entering weather variable data for the given day into the selected equation to yield an assessment of the risk.

18. The method according to claim 17, wherein the weather-associated adverse event is associated with a condition selected from asthma, emphysema, depression, cardiovascular disease, arthritis, artherosclerosis, and diabetes.

19. The method according to claim 18, wherein the weather-associated adverse event is associated with artherosclerosis and comprises heart attack or stroke.

20. A method of assessing a risk of a subject residing in climate region Cfa and predisposed to experiencing weather-associated adverse events for experiencing a new onset migraine headache on a given day, the method comprising:

a) determining the relevant season in which the given day falls;

b) determining whether the given day is an upper, middle or lower BP tertile day;

c) selecting an equation from the model set of equations according to claim 16 specific to the determined relevant season and the determined BP tertile;

d) entering weather variable data for the given day into the selected equation to yield an assessment of the risk.

21. The method according to claim 1, wherein the subject cohort is further controlled according to factors known to influence frequency of a medical condition precipitating the adverse event associated with the weather.

22. The method according to claim 21, wherein the factors are selected from race, gender, age, socio-economic status and combinations thereof.

23. A method for reducing the risk of a subject predisposed to experiencing weather-associated migraine headaches for experiencing a new onset migraine headache on a given day, the method comprising:

assessing the risk according to claim 20;

treating the patient prophylactically to mitigate or avoid the migraine headache if the day;

is assessed as more likely than not to be a UT-IR-NOH day.

25. An article of manufacture comprising computer-readable code for implementing the method according to claim 16.

26. The article of manufacture according to claim 25 comprising a mobile application software product.

24. A method for increasing efficiency in a hospital staffing and resource commitment by predicting high admission days, a “

high”

admission day being defined as a day falling in an upper quantile of the hospital'"'"'s daily admissions for a year, the method comprising;

1) generating a model set of equations for predicting a risk of subjects who suffer from weather-associated medical conditions of being admitted to the hospital, wherein “

generating”

comprises the steps of;

a) identifying the climate region in which the hospital is located;

b) collecting daily mean barometric pressure (BP) data for the year and dividing the mean BP data into upper, lower and middle quantiles;

c) collecting daily hospital admissions data for a subject cohort for the year, said subject cohort consisting of subjects known to suffer from a weather-associated medical condition and who have been admitted to a hospital at least once previously due to experiencing an adverse event associated with the condition, to calculate a daily admission rate (AR) for each day of the year and to determine an upper quantile of days associated with the AR;

d) determining a number of relevant seasons based on regression analysis of the upper BP quantile days and the upper AR quantile days;

e) collecting weather parameter data across the year;

f) employing GEE modeling to generate a rank for each weather variable as a continuous predictor of the upper quantile AR days for each relevant season, for each BP quantile;

g) identifying a best predictive single variable equation based on p-value and QIC fit of the first-ranked weather variable in each relevant season, for each BP quantile;

h) adding the next-ranked weather variable to the identified equation from g) in each season, for each BP and determining if fit improves;

i) repeating step h) until addition of the next-ranked weather variable fails to improve fit, wherein the model comprises a set of equations for predicting a risk of subjects who suffer from weather-associated medical conditions of being admitted to the hospital at the completion of step i);

2) employing the model to determine which days are likely to be upper quantile admission rate (UQ-AR) days; and

3) staffing the hospital and committing resources to the hospital on the basis of the determination in step

     2).