08 - BD2K Think Tank on EHR data methodologies in Clinical Research

(1)

BD2K Think Tank on EHR Data

Methodologies in Clinical Research

(2)

1. Clinical questions

o Are allergies associated with

higher risk of uveitis in kids

with Juvenile Arthritis?

2. Insights from the data

o Learning off-label use of drugs

3. Phenotypic profiling

o Finding the rate of

incontinence after prostate

surgery

4. Predictive models

o Who is going to be depressed?

• EHR as the major data source

• Dealing with imperfect EHR data as they are

• Wide range of clinical research studies

(3)

Terms, Concepts

Counts

Variables

Any allergy med

?

Profiling risk factors for chronic uveitis

(4)

Learning off-label use (drug-indication)

Classifier for “used-to-treat” relationship

EMR Prior knowledge

• Remove known uses • Remove similar indications

• Supported in FAERS • Supported in MEDLINE • Plausibility

• Not an adverse event

₃₂

₃₅

17

27

R is k H igh L o w Cost Low High Prioritize 407 well-supported usages 2,362,950 drug-indication pairs

EMR Prior knowledge

+

-T

1,325 67

F

_{424 6,968}

Classifier for

“used-to-treat”

relationship

Example features:

• Similarity with on-label uses • Fraction of uses that are

approved Example features: • Co-mentions • Drug-first fraction 7,112 positive and 27,938 negative training examples 1,749 positive and 7,035 negative test examples 8,861 positive examples 34,973 negative examples Evaluate in test set Medi-span, Drugbank Clinical Notes

PPV

0.952 Spec

0.990 Recall

0.758 F1

0.844

(5)

From CDW to patient feature matrix

Beta blocker

Diseases Procedures Drugs Devices

A

D

C

B

IC

14 0

(6)

PAST MEDICAL/SURGICAL HISTORY: Positive for atrial fibrillation. The patient had AVR 6 years ago. Peripheral arterial disease with hypertension, peripheral neuropathy, atherosclerosis, hemorrhoids, proctitis, CABG, and cholecystectomy.

FAMILY HISTORY: Positive for atherosclerosis, hypertension, autoimmune diseases in the family.

REVIEW OF SYSTEMS: Weight loss of 25 pounds within the last 6 months, shortness of breath, constipation, bleeding from hemorrhoids, increased frequency of urination, muscle aches, dizziness and faintness, focal weakness and numbness in both legs, knees and feet.

LABORATORY DATA AND RADIOLOGICAL RESULTS: The patient had a chest x-ray, which showed cardiomegaly with atherosclerotic heart disease, pleural thickening and small pleural effusion, a left costophrenic angle which has not changed when compared to prior examination, COPD pattern. The patient also had a head CT, which showed atrophy with old ischemic changes. No acute intracranial findings.

DISCHARGE DIAGNOSIS: Syncope. DISCHARGE MEDICATIONS: The patient was discharged on the following medications; Cardizem 90 mg p.o. thrice daily, digoxin 0.125 mg p.o. once daily, allopurinol 100 mg two times daily, Coumadin 4 mg p.o. q.h.s., and Remeron 15 mg p.o. q.h.s.

past medical surgical history: positive for atrial fibrillation. patient avr years ago. peripheral arterial disease hypertension, peripheral neuropathy, atherosclerosis, hemorrhoids, proctitis, cabg, and cholecystectomy.

family history: positive for

atherosclerosis, hypertension, autoimmune diseases family.

review of systems: weight loss pounds months, shortness of breath, constipation, bleeding hemorrhoids, increased frequency of urination, muscle aches, dizziness and faintness, focal weakness and numbness both legs, knees and feet. laboratory data and results: patient chest x-ray, which cardiomegaly atherosclerotic heart disease, pleural thickening and small pleural effusion, left costophrenic angle which not changed compared prior examination, copd pattern. patient head ct, which atrophy old ischemic . no acute intracranial findings.

discharge diagnosis: syncope.

discharge medications: patient following medications; cardizem mg . . daily, digoxin 0.125 mg . . once daily, allopurinol 100 mg two times daily, coumadin mg . . . . ., and remeron mg . . . . .

Input Text Annotations

True Internal Representation

(with some keys shown for illustration)

53 8 44 5 8539 8 3 16 (colon) 996 3 1363 13 (atrial fibrillation) 1 19 (period) 8 7 5087 12 129 6 158 6 1 3

16091 3 (peripheral arterial disease) 254 33 2 12 4624 2 2 21 (comma) 6198 2 (atherosclerosis) 2 15 (comma) 2835 2 2 11 10647 2 (proctitis) 2 9 2026 2 (cabg) 2 4 11 2 1907 4 (cholecystectomy) 1 15 (period) ... Reconstructed Representation PAST MEDICAL/SURGICAL HISTORY: Positive for

atrial fibrillation. The patient had AVR 6 years ago. Peripheral arterial disease with hypertension, peripheral neuropathy, atherosclerosis, hemorrhoids, proctitis, CABG, and cholecystectomy.

FAMILY HISTORY: Positive for atherosclerosis, hypertension, autoimmune diseases in the family.

REVIEW OF SYSTEMS: Weight loss of 25 pounds within the last 6 months, shortness of breath, constipation, bleeding from hemorrhoids, increased frequency of urination, muscle aches, dizziness and faintness, focal weakness and numbness in both legs, knees and feet.

LABORATORY DATA AND RADIOLOGICAL RESULTS: The patient had a chest x-ray, which showed cardiomegaly with atherosclerotic heart disease, pleural thickening and small pleural effusion, a left costophrenic angle which has not changed when compared to prior examination, COPD pattern. The patient also had a head CT, which showed atrophy with old ischemic changes. No acute intracranial findings.

DISCHARGE DIAGNOSIS: Syncope. DISCHARGE MEDICATIONS: The patient was discharged on the following medications; Cardizem 90 mg p.o. thrice daily, digoxin 0.125 mg p.o. once daily, allopurinol 100 mg two times daily, Coumadin 4 mg p.o. q.h.s., and Remeron 15 mg p.o. q.h.s.

past medical surgical history: positive for atrial fibrillation. patient avr years ago. peripheral arterial disease hypertension, peripheral neuropathy, atherosclerosis, hemorrhoids, proctitis, cabg, and cholecystectomy.

family history: positive for

atherosclerosis, hypertension, autoimmune diseases family.

review of systems: weight loss pounds months, shortness of breath, constipation, bleeding hemorrhoids, increased frequency of urination, muscle aches, dizziness and faintness, focal weakness and numbness both legs, knees and feet.

laboratory data and results: patient chest x-ray, which cardiomegaly atherosclerotic heart disease, pleural thickening and small pleural effusion, left costophrenic angle which not changed compared prior examination, copd pattern. patient head ct, which atrophy old

ischemic . no acute intracranial

findings.

discharge diagnosis: syncope.

discharge medications: patient following medications; cardizem mg . . daily, digoxin 0.125 mg . . once daily, allopurinol 100 mg two times daily, coumadin mg . . . . ., and remeron mg . . . . .

Input Text Annotations

True Internal Representation

(with some keys shown for illustration)

53 8 44 5 8539 8 3 16 (colon) 996 3 1363 13 (atrial fibrillation) 1 19 (period) 8 7 5087 12 129 6 158 6 1 3

16091 3 (peripheral arterial disease) 254 33 2 12 4624 2 2 21 (comma) 6198 2 (atherosclerosis) 2 15 (comma) 2835 2 2 11 10647 2 (proctitis) 2 9 2026 2 (cabg) 2 4 11 2 1907 4 (cholecystectomy) 1 15 (period) ... Reconstructed Representation Diseases Procedures Drugs BioPortal – knowledge graph

Devices

Term recognizer

Event Definition

Juvenile Idiopathic Arthritis

Juvenile spondyloarthropathy

:

Uveitits

Iridocylitis

1

2

3 http://clever.stanford.edu

Present, positive

mentions, about

the patient

(7)

(8)

B – Cleaning Concept mappings

Adrenal insufficiency hypoadrenalism syndrome Addison’s disease

SNOWMED CT

MeSH

MedDRA

tid

str

freq suppress source

8829

adrenal insufficiency

4861 (null)

(null)

95136

adrenal insufficiency, primary

7 (null)

(null)

138843

adrenal gland insufficiency

1 TRUE

auto

tid

str

freq suppress source

36442

addison disease

180 (null)

(null)

24680

addison's disease

511 (null)

(null)

52232

addisonian crisis

61 (null)

(null)

86884

addisons disease

10 (null)

(null)

tid

str

freq suppress source

46107

hypoadrenalism

89 (null)

(null)

2662032

hypoadrenalism (disorder)

0 (null)

(null)

909629

hypoadrenalism with achalasia

0 (null)

(null)

869335

hypoadrenalism, primary

0 (null)

(null)

(9)

myocardial infarction C0027051 myocardial infarction 1597 myocardial infarction 1377

myocardial infarction C0027051 myocardial infarction 1597 heart attack 426

myocardial infarction C0027051 myocardial infarction 1597 acute myocardial infarction 131

myocardial infarction C0027051 myocardial infarction 1597 myocardial infarction (mi) 129

myocardial infarction C0027051 myocardial infarction 1597 non-st elevation myocardial infarction 88

myocardial infarction C0027051 myocardial infarction 1597 inferior myocardial infarction 77

myocardial infarction C0027051 myocardial infarction 1597 nstemi 71

myocardial infarction C0027051 myocardial infarction 1597 non-q wave myocardial infarction 49

myocardial infarction C0027051 myocardial infarction 1597 anterior myocardial infarction 46

myocardial infarction C0027051 myocardial infarction 1597 myocardial infarctions 42

myocardial infarction C0027051 myocardial infarction 1597 inferior wall myocardial infarction 32

myocardial infarction C0027051 myocardial infarction 1597 old myocardial infarction 28

myocardial infarction C0027051 myocardial infarction 1597 cardiogenic shock 27

myocardial infarction C0155626 acute myocardial infarction 142 acute myocardial infarction 142

myocardial infarction C0155626 acute myocardial infarction 142 acute myocardial infarction, unspecified site 1

myocardial infarction C0036980 shock, cardiogenic 79 cardiogenic shock 79

myocardial infarction C0542269 non-q wave myocardial infarction 51 non-q wave myocardial infarction 51

myocardial infarction C0155668 old myocardial infarction 29 old myocardial infarction 29

myocardial infarction C0340324 silent myocardial infarction 25 silent myocardial infarction 25

myocardial infarction C0340321 old inferior myocardial infarction 22 old inferior myocardial infarction 22

myocardial infarction C1273976 first myocardial infarction 15 first myocardial infarction 15

myocardial infarction C0589368 postoperative myocardial infarction 14 postoperative myocardial infarction 14

myocardial infarction C2349195 acute anterior wall myocardial infarction 10 acute anterior myocardial infarction 10

myocardial infarction C0264700 acute inferior myocardial infarction 8 acute inferior myocardial infarction 8

myocardial infarction C0340320 old anterior myocardial infarction 6 old anterior myocardial infarction 6

myocardial infarction C0264699 acute anteroseptal myocardial infarction 2 acute anteroseptal myocardial infarction 2

myocardial infarction C0348593 subsequent myocardial infarction 2 subsequent myocardial infarction 2

myocardial infarction C0155627 acute myocardial infarction of anterolateral wall 1 acute anterolateral myocardial infarction 1

myocardial infarction C0340308 acute myocardial infarction of inferolateral wall 1 acute inferolateral myocardial infarction 1

myocardial infarction C0340318 acute posterior myocardial infarction 1 acute posterior myocardial infarction 1

(10)

Look at the data

• 59 note types (out of 8,132) cover 7.6 mil.

notes (out of 9.5 mil)

• 0.7% types cover 80% of the data

• 2.2% types (182) cover 90% of the data

• 5% of notes, cover 95% of the data.

• K = 95.

(11)

(12)

Phenotyping trade off

Acc

PPV Time

0.98 0.96 1900

Acc

PPV Time

(13)