Self-improving method of using online communities to predict health-related outcomes

Abstract

The invention is directed, in part, to method of using self-reported health data in online communities to predict significant health events in life-changing illnesses to improve the lives of individuals and to improve patient self-management. The invention provides a method for providing real-time personalized medical predictions for an individual patient. The method includes: providing a database containing patient information for a plurality of other patients including one or more attributes for each patient in the database; constructing a model of a disease based on disease progressions for the plurality of patients; receiving a request from the individual patient, the patient associated with one or more attributes; and making a real-time prediction for the individual patient based on the mode and the individual patient's attributes.

1 - 20 . (canceled) 
     
     
         21 . A computer implemented method for providing medical predictions for an individual patient in a community of patients, the method comprising:
 providing a server coupled, via a network, to a plurality of computers, each computer having a graphical user interface and being associated with a particular patient wherein each computer includes a processor configured with executable instructions to allow each patient of the community of patients to input self-reported information relating to one or more medical condition attributes of the particular patient without being responsive to a series of pre-programmed questions, the server being configured to:   receive the self-reported patient information from each of the plurality of computers;   store the self-reported patient information in a database;   construct a model of a disease based on disease progressions wherein a disease progression is based on a patient's disease and self-reported patient information stored in the database.   
     
     
         22 . The computer implemented method as recited in claim  1 , wherein the server is further configured to:
 receive a request from an individual patient from the community of patients; and   determine a real-time prediction concerning the effect of an intervention for the individual patient based on the model and individual patient's attributes and analyzing an effect of the intervention by obtaining a difference between or comparing the outcome of a disease progress with and without an intervention.   
     
     
         23 . The method of claim  1 , wherein the one or more attributes includes at least one selected from the group consisting of: age, race, ethnicity, gender, height, weight, body mass index (BMI), body volume index (BVI), genotype, phenotype, severity of the disease, progression rate of the disease, measures of functional ability, quality of life, interventions, and remedies. 
     
     
         24 . The method of claim  1 , wherein the disease includes at least one selected from the group consisting of: neurological diseases, Amytrophric Lateral Sclerosis (ALS), Multiple Sclerosis (MS), Parkinson's Disease, Human Immunodeficiency Virus (HIV), Acquired Immune Deficiency Syndrome (AIDS), depression, mood disorders, cancer, blood cancer, fibromyalgia, epilepsy, post traumatic stress disorder, traumatic brain injury, cardiovascular disease, osteoporosis, chronic obstructive pulmonary disease, arthritis, allergies, autoimmune diseases, and lupus. 
     
     
         25 . The method of claim  1 , wherein the model is based on data for a subset of the plurality of patients and the server is further configured to process a request from the patient to modify a composition of the subset of the plurality of patients. 
     
     
         26 . The method of claim  2 , wherein the server is further configured to calculate a confidence interval for the prediction, which includes selecting a set of reported data points from the plurality of other patients. 
     
     
         27 . The method of claim  6 , wherein for each of the reported data points in the set, the server is configured to:
 obtain a data set for a corresponding other patient to the reported data point;   calculate a predicted value with the data set and the model;   calculate an error between the predicted value and the reported data point;   determine a distribution of the errors; and   calculate a confidence interval from the distribution.   
     
     
         28 . The method of claim  2 , wherein the difference is measured for a plurality of individual patients. 
     
     
         29 . The method of claim  2 , wherein the difference is compared to the distribution of error. 
     
     
         30 . The method of claim  2 , wherein the difference is compared to the confidence interval for the model. 
     
     
         31 . The method of claim  10 , wherein the server is further configured to identify one or more of the differences that exceed the confidence interval for the model. 
     
     
         32 . The method of claim  6 , wherein the confidence interval is calculated with a chi-square test. 
     
     
         33 . The method of claim  6 , wherein the confidence interval is calculated from a measure of variance of the individual patient's attributes. 
     
     
         34 . The method of claim  6 , wherein the confidence interval is calculated by comparing the individual patient's attributes to a model fit for the individual patient using the model. 
     
     
         35 . A non-transitory and tangible computer-readable medium whose contents cause a computer to perform a computer-implemented method for providing medical predictions for an individual patient in a community of patients, the method comprising:
 receiving, by a server, self-reported patient information from a plurality of computers wherein each computer has a graphical user interface associated with a particular patient with each computer including a processor configured with executable instructions to allow each patient of the community of patients to input self-reported information relating to one or more medical condition attributes of the particular patient;   store, by the server, the self-reported patient information in a database;   construct, by the server, a model of a disease based on disease progressions wherein a disease progression is based on a patient's disease and self-reported patient information stored in the database;   receive, by the server, a request from an individual patient from the community of patients; and   determine, by the server, a real-time prediction concerning the effect of an intervention for the individual patient based on the model and individual patient's attributes and analyzing an effect of the intervention by obtaining a difference between or comparing the outcome of a disease progress with and without an intervention.   
     
     
         36 . The non-transitory and tangible computer-readable medium as recited in claim  15 , wherein each patient of the community of patients inputs the self-reported information relating to one or more medical condition attributes of the particular patient without being responsive to a series of pre-programmed questions. 
     
     
         37 . The non-transitory and tangible computer-readable medium as recited in claim  15 , wherein the model is based on data for a subset of the plurality of patients and the server is further configured to process a request from the patient to modify a composition of the subset of the plurality of patients. 
     
     
         38 . The non-transitory and tangible computer-readable medium as recited in claim  15 , wherein the server further calculates a confidence interval for the prediction, which includes selecting a set of reported data points from the plurality of other patients. 
     
     
         39 . The non-transitory and tangible computer-readable medium as recited in claim  15 , wherein the difference is measured for a plurality of individual patients. 
     
     
         40 . The non-transitory and tangible computer-readable medium as recited in claim  15 , wherein the difference is compared to the distribution of error.