5/15/2015

Enlightiks Predictive Analytics Applied in Healthcare- Use Case

Dr. RUCHI DASS, CMO & Head Clinical Analytics,

Enlightiks Inc

Copyright © 2015 Enlightiks Inc., All Rights Reserved

1

1

Jyotika is 61 years old and lives in a small town in

Lucknow and is suffering from Diabetes for the past 14

years. She is coping well with fair HB1AC values and

lifestyle management. Recently, Jyotika felt that she is

not able to handle sunlight well and feels that her vision

is a bit blurred- she is speculating cataract and shows

up at Dr. Narula’s Clinic. Dr. Narula is the Chief

Ophthalmologist and has more than 35 years of

experience and currently runs a Polyclinic out of

Lucknow, UP.

Jyotika gets successfully operated for Cataract and is discharged the same day. Within 13 days

of discharge, Jyotika is admitted again but this time because of a Cardiac Disease. The family is

speculating if this has something to do with the recent cataract surgery. Dr Reddy, Head of

Department, Cardiology shows up and makes a comment to the family and Nurse- Well!

Oxidative damage to proteins in the human lens is believed to be important in the etiology of

age-related cataract. Because free radical-mediated oxidative damage to lipoproteins may

accelerate atherosclerosis, the authors hypothesized that the development of cataract might

be a marker for such damage and therefore might be associated with future risk of coronary

heart disease (CHD).

Mahesh, husband of Jyotika is not very happy with this. He believes that the doctors

demonstrated clear negligence when Jyotika was admitted last time and provided no warning

signs related to any possibility of a heart disease.

Is there a way that this story would have been different and more assuring for the patient?

Predictors like Physiological data (systolic blood pressure); Predictive Drugs usage and context

(drug use of alpha blockers, beta blockers, beta agonists), Previous disease history (especially

COPD, Lifestyle and environmental factors (age, smoking, and other coronary risk factors) can

help determine risk for the population well in advance with Predictive analytics in action.

Quick stats- In the United States, 1 in 4 women dies from heart disease. In fact, coronary heart

disease (CHD)—the most common type of heart disease—is the #1 killer of both men and

Copyright © 2015 Enlightiks Inc., All Rights Reserved

2

2

women in the United States. Also, a greater proportion of women (52 percent) than men (42

percent) with myocardial infarction die of sudden cardiac death before reaching the hospital.

“Advanced Predictive analytics companies like Enlightiks (based in India, US) have developed

interesting risk stratification modules that can crawl of millions of health records in a

population to understand patterns, trends, patient similarity metrics and natural language

processing system that also takes into context unstructured data like Physician notes and

discharge documents that are not often taken into account for analysis.”

Data goes through normalisation in a smart grid atmosphere and is then put into context to

further do analysis based on simple algorithms like Framingham’s scale coupled with patient

similarity study, isomorphism of patient graphs in a population, building cohorts and then

aligning the data to the outcomes. This provides more complete and accurate understanding

of each patient.

While studying the Prospective study of

cataract extraction and risk of coronary heart

disease in women 60,657 women aged 45--

63 years were followed. These women were

without known coronary disease, stroke, or

cancer in 1984.During 10 years of follow-up

(674,283 person-years), the authors of this

report (see source) documented 887 incident

cases of CHD and 2,322 deaths.

After adjustment for age, smoking, and other coronary risk factors, cataract extraction was

significantly associated with higher risk of CHD (relative risk (RR) = 1.88, 95% confidence

interval (CI): 1.41, 2.50) for total CHD, 2.44 (95% CI: 1.54, 3.89) for fatal CHD, and 1.63 (95%

CI: 1.14, 2.34) for nonfatal myocardial infarction).

The positive association between cataract extraction and total CHD was stronger among

women with a history of diabetes (RR = 2.80, 95% CI: 1.77, 4.42) than among those without

reported diabetes (RR = 1.51, 95 percent CI: 1.04, 2.18). In multivariate analyses, cataract

extraction was associated with significantly increased overall mortality (RR = 1.37, 95 percent

CI: 1.13, 1.66), which was entirely explained by the increased mortality from cardiovascular

disease (RR = 1.84, 95% CI: 1.29, 2.64).

Copyright © 2015 Enlightiks Inc., All Rights Reserved

3

3

THESE FINDINGS ARE COMPATIBLE WITH CURRENT HYPOTHESES RELATING OXIDATIVE DAMAGE AND TISSUE AGING TO

THE DEVELOPMENT OF CATARACT AND CHD.

Data mining and data analytics has been of immense importance to many different fields as

we witness the evolution of data sciences over recent years. Biostatistics and Medical

Informatics has proved to be the foundation of many modern biological theories and analysis

techniques. These are the fields which applies data mining practices along with statistical

models to discover hidden trends from data that comprises of biological experiments or

procedures on different entities.

In case of Jyotika if risk stratification was possible, it would have allowed advanced preventative

measures and would have also help bring down healthcare costs especially for a country like

India where 80% of healthcare expenditure is out of pocket. The claims and hospital admission

would have been carefully planned or avoidable. Also, such measures would have led to better

patient satisfaction.

Author: Dr. Ruchi Dass, CMO and Head Clinical Analytics, Enlightiks Inc. is an award winning Physician and

thought leader in the field of technology driven healthcare and has published several journals on User Driven

Healthcare. You can reach her at ruchi@enlightiks.com and follow @Enlightiks.

Copyright © 2015 Enlightiks Inc., All Rights Reserved

4

4

The Company: Enlightiks is a silicon valley based advanced analytics company with sole focus on healthcare industry. The company has created a cloud based platform, Querent™ that can be used by Providers, Health Plans, ACOs and Healthcare Consulting & Services organization for a variety of applications. Enlightiks delivers scalable, agile and decision-focused insights in the hands of Healthcare providers & decision makers via our proprietary, predictive clinical and computational platform – Querent™. Querent™ is an advanced analytics platform, which allows you to create adaptive visualization apps with built in clinical, cognitive and statistical models. Users can build apps or use the existing apps without having to know coding or statistical/analytical models. Querent™ is context aware and poses questions automatically based on the data, then proceeds to solve those questions. Querent™ can be used as an app on the cloud (currently on AWS, planned – GCE & Azure) or could be hosted on premise.

Copyright © 2015 Enlightiks Inc., All Rights Reserved

5

5

Credits:

1. Cardiac Data Mining (CDM); Organization and Predictive Analytics on Biomedical (Cardiac) Data. Musa M Bilal, Masood Hussain, Iqra Basharat, Mamuna Fatima, AIP Conference Proceedings 01/2013; 1559. DOI: 10.1063/1.4825018

2. A Predictive Model for Readmission of Patients with Congestive Heart Failure: A Multi-hospital Perspective- Eric Zheng

3. Prospective study of cataract extraction and risk of coronary heart disease in women. Hu FB1, Hankinson SE, Stampfer MJ, Manson JE, Colditz GA, Speizer FE, Hennekens CH, Willett WC.

4. nhlbi.nih.gov, How Does Heart Disease Affect Women? 5. http://www.cdc.gov/media/dpk/2013/images/vitalSigns/heart_disease/img14.jpg 6. Recent Insights in Coronary Artery Disease in Women- Biswakes Majumder, Azizul

Haque, Dipankar Ghosh Dastidar 7. Prevalence of risk factors for coronary artery disease in an urban Indian population-

2014- T Sekhri, R S Kanwar, R Wilfred, P Chugh, M Chhillar, R Aggarwal, Y K Sharma, J Sethi, J Sundriyal, K Bhadra, S Singh, N Rautela, Tek Chand, M Singh, S K Singh

8. Prevalence of coronary artery disease and coronary risk factors in Kerala, South India: A population survey – Design and methods- Geevar Zachariah,a S. Harikrishnan,b M.N. Krishnan,g,∗ P.P. Mohanan,c G. Sanjay,d K. Venugopal,e K.R. Thankappan,f and The Cardiological Society of India Kerala Chapter Coronary Artery Disease and Its Risk Factors Prevalence (CSI Kerala CRP) Study Investigators