Machine Learning Systems and Methods for Predicting Risk of Renal Function Decline

Abstract

Systems, methods and apparatuses are described herein that employ machine learning techniques to assess a likelihood or risk that one or more patients will experience an adverse outcome, such as a decline in renal function, within one or more timeframes. The embodiments may utilize patient data relating to demographics, vital signs, diagnoses, procedures, diagnostic tests, biomarker assays, genetic tests, behaviors, and/or patient symptoms, to determine risk information, such as important predictive features and patient risk scores. And the embodiments may automatically execute patient workflows, such as providing treatment recommendations to providers and/or patients, based on determined risk scores.

What is claimed is: 
     
         1 . A computer-implemented method of determining a risk of renal function decline for a patient comprising:
 analyzing, by a computer, input data received from one or more data sources to determine patient information associated with a patient, the patient information comprising:
 a plurality of demographics associated with the patient comprising:
 a demographic relating to an age of the patient; and 
 a demographic relating to a gender of the patient; 
 
 a plurality of lab tests associated with the patient, each lab test of the plurality of lab tests associated with lab test information comprising a lab test variable and a lab test value relating to the variable,
 wherein the plurality of lab tests comprises a first lab test associated with a first lab test variable relating to one of:
 tumor necrosis factor receptor-1 (“TNFR1”); 
 tumor necrosis factor receptor-2 (“TNFR2”); and 
 kidney injury molecule-1 (“KIM1”); 
 
 
   calculating, by the computer, a value for each of a plurality of features, based on the patient information, the plurality of features comprising:
 a plurality of demographic features, each demographic feature of the plurality of demographic features relating to at least one demographic of the plurality of demographics; and 
 a plurality of lab test features, each lab test feature of the plurality of lab test features relating to at least one lab test of the plurality of lab tests; 
   calculating, by the computer, a risk score for the patient, based on the values of the features, the risk score relating to a probability that the first patient will experience an outcome relating to a decline in renal function within a predetermined amount of time; and   outputting the risk score.   
     
     
         2 . A method according to  claim 1 , wherein the plurality of demographics further comprises a demographic relating to a race of the patient. 
     
     
         3 . A method according to  claim 1 , wherein:
 the first lab test variable associated with the first lab test relates to TNFR1; and   the plurality of lab tests further comprises:
 a second lab test associated with a second lab test variable relating to TNFR2; and 
 a third lab test associated with a third lab test variable relating to KIM1. 
   
     
     
         4 . A method according to  claim 3 , wherein the plurality of lab tests further comprises a fourth lab test associated with a fourth lab test variable relating to at least one of: estimated glomerular filtration rate (“eGFR”), urine albumin-creatinine-ratio (“UACR”), serum creatinine, blood urea nitrogen (“BUN”), serum sodium, serum potassium, serum chloride, serum bicarbonate, serum calcium, serum albumin, urine creatinine, urine albumin, urine microalbumin, urine protein, complete blood count (“CBC”), liver function, lipid profile, a coagulation panel, magnesium, phosphorus, brain natriuretic peptide (“BNP”), hemoglobin A1c (“HbA1c”), uric acid and endostatin. 
     
     
         5 . A method according to  claim 3 , wherein the fourth lab test variable relates to at least one of: eGFR, serum creatinine, BUN, serum bicarbonate, serum phosphorus, serum calcium, urine creatinine, urine albumin, urine microalbumin, urine protein, and UACR. 
     
     
         6 . A method according to  claim 1 , wherein the lab test information further comprises one or more of: a lab test identifier, a lab test date, a unit relating to the lab test value, a reference range of values, a sample type, facility identification information, provider information, radiological imaging data, and clinical notes. 
     
     
         7 . A method according to  claim 1 , wherein:
 the patient information further comprises one or more diagnoses associated with the patient, each diagnosis of the one or more diagnoses associated with diagnosis information comprising a diagnosis identifier; and   the plurality of features further comprises a plurality of diagnosis features, each diagnosis feature of the plurality of diagnosis features relating to at least one diagnosis of the one or more diagnoses.   
     
     
         8 . A method according to  claim 7 , wherein the one or more diagnoses comprises a first diagnosis associated with a kidney issue or a comorbidity. 
     
     
         9 . A method according to  claim 8 , wherein:
 the first diagnosis is associated with the kidney issue; and   the kidney issue relates to one of the group consisting of: polycystic kidney disease, renal agenesis, Alport Syndrome, rapidly progressive glomerulonephritis, focal segmental glomerulosclerosis, IgA nephropathy, membranous nephropathy, membranoproliferative glomerulopathy, mesangial proliferative glomerulopathy, minimal change disease, nephritis syndrome, nephrotic syndrome, nephrolithiasis, hypertensive nephropathy, analgesic nephropathy, diabetic nephropathy, lithium nephropathy, renal artery stenosis, Lupus nephritis, kidney myeloma, kidney amyloidosis, anti-glomerular basement disease, fatigue or weakness, edema, and proteinuria.   
     
     
         10 . A method according to  claim 8 , wherein
 the first diagnosis is associated with the comorbidity; and   the comorbidity relates to one of the group consisting of: alcohol abuse, anemia deficiency, rheumatoid arthritis, blood loss anemia, cardiac arrhythmia, congestive heart failure (“CHF”), chronic pulmonary disease (“CPD”), coagulopathy, acquired immunodeficiency syndrome (“AIDS”) or human immunodeficiency virus (“HIV”), depression, diabetes, drug abuse, hypertension, hypothyroidism, liver disease, lymphoma, a fluid or electrolyte disorder, metastatic cancer, a neurological disorder, obesity, paralysis, peripheral vascular disease, psychosis, and pulmonary circulation disorder.   
     
     
         11 . A method according to  claim 10 , wherein the plurality of diagnosis features further comprises a feature relating to a Charlson Comorbidity Index (“CCI”) score calculated for the first diagnosis. 
     
     
         12 . A method according to  claim 7 , wherein the diagnosis information further comprises one or more of: a diagnosis date, provider information, equipment information, clinical notes and vital signs information. 
     
     
         13 . A method according to  claim 1 , wherein
 the patient information further comprises one or more medications associated with the patient, each medication of the one or more medications associated with medication information comprising a medication identifier; and   the plurality of features further comprises a plurality of medication features, each medication feature of the plurality of medication features relating to at least one medication of the one or more medications.   
     
     
         14 . A method according to  claim 13 , wherein the medication information further comprises at least one of the group consisting of: a medication date, a medication type, a concentration, a quantity, an amount, date information, refill information, provider information, and clinical notes. 
     
     
         15 . A method according to  claim 14 , wherein the one or more medications comprises one or more of the group consisting of: an antibiotic medication; a non-steroidal anti-inflammatory drug (“NSAID”) medication; a beta-adrenergic receptor blocker medication; a dihydropyridine medication; an angiotensin II receptor blocker (“ARB”) medication; an angiotensin-converting enzyme (“ACE”) inhibitor medication; a sodium-glucose Cotransporter-2 (SGLT2) inhibitor medication; a Thiazide-class diuretic medication; a Loop-diuretic medication; and a HMG-CoA reductase inhibitor medication. 
     
     
         16 . A method according to  claim 1 , wherein the patient information comprises genetic information indicating that one or more risk variant alleles in an Apolipoprotein L1 gene (“APOL1”) of the patient are expressed. 
     
     
         17 . A method according to  claim 16 , wherein the plurality of features further comprises one or more features relating to the genetic information. 
     
     
         18 . A method according to  claim 1 , further comprising:
 determining that the risk score satisfies a workflow rule associated with a patient workflow; and   executing the patient workflow, based on said determining that the risk score satisfies said workflow rule.   
     
     
         19 . A method according to  claim 18 , wherein said executing the patient workflow comprises:
 determining a treatment recommendation for the patient, based on the risk score; and   transmitting a notification comprising the treatment recommendation to one or more recipients.   
     
     
         20 . A method according to  claim 1 , wherein the one or more data sources comprise at least one of the group consisting of: an electronic health records (“EHR”) system, a health facility system, an insurance system, a payment system, a user device, a medical device, a biometric device, and an engagement system.