The role of data in health economics

Pedro Pita BarrosUniversidade Nova de Lisboa

1

Health Economics? (“data” you know well…)

• Covers:• Subjective and objective measures of valuation of health • Production of health: individual behaviors and health care + time• Demand and supply of health care services (matching and equilibrium)• Health insurance, public and private• Evaluation of health systems• Evaluation of products and services• Efficiency of organizations that provide health care• Ethics and access to health care

• Data matters, different data for different purposes

2

Aggregate decisions: how much to spend in health care to increase life expectancy?

Smalldata

Health exp/GDP

Mediumdata

Meso decisions: how to address a population’s health care needs in a region?

Unmet needs, main

diseases

Individual decisions: on treatments and efficient choices

EMR, hospitalrecords,

imaging, etc

Lots of data

Individual behavior and choices (why do we exercise? How to avoid start smoking?)

Everyday data, apps, smart

devices, internet searches on health

topics, etc.

No decisions: Biology of health | genetics – how health ”works”? (how does a cancer ”recruits” blood vessels?)

Big data Data deluge

1 human body, 50 trillion bacterial cells, 30 trillion

human cells

OECD health data

EU-SILC, SHAREsurveys

Providers of care & routine data

New sources ofdata collection

3

• What problems you can solve (in health economics), what data do you need, what data you can collect and use vary widely• Data and information created with such data only have economic

value (role) if decisions can be different because of it –• you may become aware or gain knowledge you did not have, but if all

decisions are the same, irrespective of data and information, then they play no role

4

What is health?

• WHO (most common definition): “A state of complete physical, mental and social well-being and not merely the absence of disease or infirmity”• Alternative one (Huber et al, 2011, BMJ): “The ability to adapt and to

self-manage, in the face of social, physical and emotional challenges”• The use of data (and data science) to help producing “health” seems

to be more helpful with the second definition – data that people collect, interpret and use to “adapt and to self—manage”, likely with the help of specialized / trained professionals (new profession: data scientists for health improvement?)

5

Most promising (new) uses of (new) data (?)

• Imaging and cancer growth – some limitations, including too many data – a challenge to train algorithms…• Prevention, behaviors and data collected from wearables and/or apps

– from the mostly harmless (exercise data…) to costly false positives (EKG with a smart watch)• Data from internet search interests – Joana Gonçalves-Sá and internet

searches to detect early seasonal influenza onset• What else?

6

What data (and digital) brings?

• Three alternatives to internal reorganization• Ability to have new services and products to produce health• Replace old services and products• Add to and co-exist with previous services and products

• More importantly, does it change (and how) the culture of organizations that provide health care?• Having better diagnostic tools improves ability to treat?• Having more and better data helps the patient to be involved in the decision?

7

Risks (?)

• Safety of data – not only privacy but inducing wrong decisions if data is altered on purpose – cybersecurity likely to become a new dimension in quality of care• Human – human interaction (emotions, compassion, empathy) - how

much can be replaced?• Data collection must have a purpose, be fit to the questions to be

addressed and not allow “gaming” of the health system – not just collecting because “it is there”• Unveil the “invisible” decisions – decisions and emotions – felt sick,

did not look for help – not registered unless he/she wants to

8

What do we (research group) currently use?

• General macro data (OECD, Health National Accounts, WHO, etc) –countries x years x variables • => question: what is a financially sustainable health system?

9

What do we (research group) currently use?

• Survey data – SHARE (30k+ each wave over 28 countries), Survey on access, National Health Survey, EU-SILC – capture the “invisible” and unmet needs • => question: what are the vulnerable populations and what unmet

needs for care exist?• => question: does supplementary health insurance improve your

health?

10

What do we (research group) currently use?

• Data from ACSS on hospital episodes (1M+ episodes each year), data on pharmacies’ transactions (3M+/month on a sample) • => question: do hospitals discharge too early (with health costs to

patients) when they have more demand? • Luis Filipe, “A bed constraint?”• => question: does the distance to the hospital matter for survival?• Eduardo Costa and Laman Orujova, “Time is on my side! Mortality

and distance to hospital care” (in Medium)

11

What do we (research group) currently use?

• Data from supermarkets sales on beverages• => question: is a tax good for your health? “Brown sugar, how come you taste so good? The impact of a soda tax on prices and consumption”, Judite Gonçalves & João Pereira dos Santos

12

https://www2.novasbe.unl.pt/health

We welcome collaborations putting together health economics and data

13