Big Data

National Cancer Institute: Division of Cancer Control and Population Sciences What does it do?

- studies on patterns of care for persons with cancer before a cancer diagnosis, over the period of initial diagnosis and treatment, and during long-term follow-up

Big data in healthcare

- total of 2.5 quintillion terabytes of data were generated everyday in 2012 alone - healthcare has started focusing on big data - federal agencies, private firms, academia are increasingly investing in their analytic capabilities to help them make better sense of healthcare environment

Under HIPAA covered entities must follow a standardized Electronic Data Interchange (EDI) mechanism to submit and process insurance claims including:

1. **Healthcare Common Procedure Coding System (HCPCS)* to code medical procedures (chemotherapy) 2. *Common Procedural Terminology (CPT)* to code physicians procedures (office visits) 3. *International Classification of Diseases (ICD) 9* to code diagnosis and hospital procedures (surgery) 4. *National Drug Codes (NDC)* to code branded and generic medications

What does data show?

1. *Factual information* (as measurements of statistics) used as a basis for reasoning, discussion, or calculation 2. *Information output* by a sensitive device or organ that includes both useful and irrelevant or redundant information and must be processed to be meaningful 3. *Information* in numerical form that can be digitally transmitted or processed

Ways that we could eliminate waste in US healthcare

Decrease: - failures of care delivery - failures of care coordination - over treatment - administrative complexity - pricing failures - fraud and abuse

Strength & Limitations of Clinical Big Data

Strengths - broader array of data elements available Limitations - complex to query

Strength & weaknesses of Healthcare Utilization Big Data

Strengths - large sample size, standardized Weakness - timelineless - accessibility - data processing - quality

ISPOR*Digest of Databases goal:

To organize global healthcare databases into searchable models

as the US increases spending on health care what happens to life expectancy?

WE DO NOT see an increase

what is Big data?

a collection of large and complex datasets which are difficult to process using common database management tools or traditional data processing application

Top cause of drug-related adverse outcomes in pts 65+

agents effecting blood

What is ISPOR?

an available source of information for healthcare outcomes researchers, decision makers, pharmacists, clinicians and patients 407 databases form 47 coutnries

Top cause of drug-related adverse outcomes in pts 18-44

analgesics

Aim of big data: Research

generate study findings to aid clinical decision making and policy development

Top cause of drug-related adverse outcomes in pts 45-64

hormones

How does medicare penalize hospitals?

penalizes hospitals that have high rates of readmissions amount patents with: 1. Heart failure 2. Heart attack 3. pneumonia

Top cause of drug-related adverse outcomes in pts 17 and under

systemic agents

what is one of the driving forces to share utilize healthcare data?

to decrease costs and improve outcomes

Types of information in pharmacy data:

1. Group-level data 2. Service level data 3. Pharmacy-level data 4. Patient level data

National Center for Health Statistics consists of:

1. Medical Expenditure Panel Survey (MEPS) 2. Healthcare Cost and Utilization Project (HCUP) 3. National Health and Nutrition Examination Survey (NHANES) 4. National Nursing Home Survey (NNHS) 5. Longitudinal Studies of Aging (LSOA)

Types of Big Data in research

1. Registries (Surveillance epidemiology an End Results (SEER)- Cancer registry data that represents 25% of the US population with cancer 2. National Survey data (NHANES, BRFSS, MEPS) 3. Epidemiological and Economic data

where do you find Big data in healthcare?

1. Research 2. healthcare utilization data 3. Clinical data 4. patient generated data

Big data could save the healthcare industry up to $450 billion, but other things are important too, such as:

1. Right living: pts take more steps to improve their health 2. Right care: developing coordinated approach to care where all caregivers have access to same info 3. Right providers: professionals treating pts should have strong performance records 4. Right value: improving value while improving care quality 5. Right innovation: identifying new approaches to healthcare delivery

The 4 dimensions (V's) of Bid data

1. Velocity --> rapid processing 2. Volume --> size, storage 3. Variety -->structured/unstructured formats 4. Veracity--> accuracy

Data used in pharmacy for:

1. clinical 2. Administrative 3. Assessment

Hospitalizations- penalties for poor care

1. hospitalizations: more than 30% of the 2 trillion annual costs of healthcare in US, about 20% of all hospital admissions occur within 30 days of a previous discharge *potentially harmful and preventable*

Patient Big Data consists of:

1. personal Health Record 2. Online/Social Media 3. Patient Reported Outcomes 4. Retail Purchases 5. Devices/Electronic Monitors (data recorded via apps)

Overall goals of big data in healthcare

1. take advantage of the massive amounts of data and provide *right intervention to the right patient at the right time 2. personalized care 3. potentially benefit all the components of a healthcare system i.e. provider, payer, patient, management, government, society

About ___% of all hospital admissions occur within 30 days of a previous discharge

20%

Hospitalizations account for more than ___% of the 2 trillion annual costs of healthcare in US

30%

Pharmacy Quality Alliance (PQA) what is it?

A non-profit membership organization that: - promotes appropriate medication use and - aims at developing strategic plans for measuring and reporting performance

What does Healthcare Utilization Data consists of?

Administrative Claims data - medical services - pharmacy - lab and imaging

National Cancer Institute: Division of Cancer Control and Population Sciences what is it?

*SEER- Medicare Linked Dataset* - linkage of two large population-based sources of data that provide detailed information about Medicare beneficiaries with cancer

Clinical Big data captured via:

- Electronic Medical Records - Pharmacy Profiles - Registries

PQA-supported medication measure included in Medicare Part D plan ratings:

- High risk medications in the elderly - proportion of days covered (to measure medication adherence) - medication therapy management program completion rate for a CMR

example of healthcare utilization Data

- Medicare - Medicaid - Optum - MarketScan

Types of information in pharmacy data: Pharmacy-level data

- Pharmacy ID - name - address - phone #

Examples of Clinical Big Data

- Registry (SEER) - Cleveland Clinic Data

National Drug Code (NDC)

- a method to identify drugs during commercial distribution - a universal product identifier for human drugs

Data used in pharmacy for: Administrative

- billing (for pharmacy management, formulary management) - inventory (for pharmacy management)

what benefit does identifying patients at risk of readmission have?

- can guide *efficient resource utilization* and can potentially save millions of dollars each ear

Health Insurance portability and Accountability Act (HIPAA) of 1996

- developed national standards and safeguards for the use of electric healthcare information - covered entities must follow a standardized Electronic Data Interchange (EDI) mechanism to submit and process insurance claims

what does big data do?

- finds insights from complex, noisy, heterogenous, longitudinal and voluminous data - aims to answer questions that were previously unancered

Types of information in pharmacy data: Group-level data

- group ID - effective date - benefit design - name - address

Efficiently making predictions from such complex hospitalization data will require the development of:

- novel advanced analytical models

Types of information in pharmacy data: Patient level data

- patient ID - name - address - zip code - eligibility - phone #

Types of information in pharmacy data: Service level data

- patient ID - pharmacy ID - date of service - DOB - gender - Group ID - service - NDC - day's supply - charge

Data used in pharmacy for: Clinical

- patient profile (age, gender, race/ethnicity, socioeconomic status) - treatments (drugs, # of days supply, dose)

Data used in pharmacy for: Assessment

- performance measurement (could identify medication errors) - Clinical quality and safety (could identify medication errors)

NCD used in:

- product identification - inventory control - purchasing/billing - rebates

PatientsLikeMe

- started in 2006 - network with an online data sharing platform more than 200,000 patients and is tracking 1,500 diseases - Objective: given my status, what is the best outcome I can hope to achieve, and how do i get there? - people connect with others who have the same disease, track and share their own experiences, see what treatments have helped other patients like them, gain insights and identify any patterns