Theory & Reality Ateev Mehrotra Harvard Medical School Theory - - PowerPoint PPT Presentation

Price Shopping for Healthcare: Theory & Reality

Ateev Mehrotra

Harvard Medical School

Theory

2

“When consumers apply pressure on an industry, whether it’s retailing or banking, cars or computers, it in invaria iably ly pr produce duces s a surge ge of in innovation ion that in increa eases es pr produ ductiv ctivit ity, , red educes es pr pric ices es, , impr improves es quality, and expands choices.”

• - Regina Herzinger, Harvard Business School,

Market-Driven Health Care

3

Patients are engines of change

Freedom of choice Bear greater fraction of costs

• f their care

Transparency of costs and quality Judicious use of low-value care Preferential selection of low- cost, high-quality providers Market forces will drive providers to improve value

Lots of money to be saved if patients switched to lower-priced providers?

Yearly y spendi nding ng How much ch would d be saved ed La Labs 270 13 136 Imagi aging ng 436 436 254 254 Durabl able e Medical cal Equi quipmen pment 61 37 Tot

• tal

al 767 427

58% savings!

• Model scenario if patients who received care which cost

above the median price switched to median price facility in their community

• Focus on commodity services

5

Joining CDHP has at most modest effect on prices paid

6

 Among 8 out of 9 services prices paid by CDHP did

not differ (Sood, 2013)

 No impact on prices (Brot-Goldberg, NBER)  Entry into CDHP did have modest effect of

switching from brand name to generics (Huckfeldt, NBER)

Relatively small differences also in self-reported price shopping

Deductible During your last health care encounter: None $1-1250 $1251+ Know your costs before? 86 61 48 Actively seek cost information? 13 11 15 Consider other providers? 7 10 10 Consider other providers AND compare costs? 1 3 4 Sinaiko, JAMA IM, 2015

Is the problem lack of price data?

8

Enthusiasm for Price Transparency

 Over half the states have passed laws requiring

either payers or providers to disclose pricing information to patients

 Component of president’s health care agenda

9

10

11

12

Study design

 Intervention group: two large companies offered

Truven Cost Calculator

 Control group: similar employees of other

companies not offered tool

 Outcomes

• Change in outpatient spending
• Switching from hospital outpatient setting

13

What Did We Find?

 Findings

• No decrease in overall spending
• No shift in care from hospital outpatient setting
• No decrease in spending among those with higher

deductibles

 Concerns

• Tool is not well designed
• Tool was poorly marketed
• Too broad of a focus on outpatient spending

14

2nd Study

15  Intervention group: all CalPERS individuals in a PPO

in California

 Control group: all non-CalPERS California Anthem

members in PPO plan with a deductible ($250-750)

 To address differences between the intervention and

control population we used propensity score weighting

 Outcome: Smaller set of services for which people

use the price transparency tool

Impact of Offering Transparency Tool

728 788 756 812 600 650 700 750 800 850 900

Year before Year after Spending on “Shoppable Services”

Control ntrol Offere ered d price ce transp nspare renc ncy y tool

• l

16

Price Paid by Those Who Searched

$0 $200 $400 $600 $800 $1,000 $1,200 $1,400

E&M Lab Imaging

Non- searchers Searchers Non- searchers Searchers Non- searchers Searchers

• 14%

17

Fraction of services received preceded by a search

18

All employees

• ffered tool

Among those who signed up

Labs

0.3% 2.2%

Office visits

1.0% 7.5%

Imaging

1.0% 7.0%

Summary

19  Theory is that combination of higher deductibles and

price data will drive consumers to lower-priced providers

 Evidence is that neither high deductibles alone or

deductibles + price transparency has had meaningful effect on prices

 Why?

• Is theory wrong?
• Other barriers exist that limit patients from being

price shoppers?

• Ateev Mehrotra
• mehrotra@hcp.med.harvard.edu

21

22

Relatively Few Offered the Tools Signed Up

 Truven tool

• 10% of households signed up for tool
• But only 2% used it over time

 Castlight tool

• 24% of households signed up for tool
• 12% used it for a price search
• But only 4% had sustained use (3x use with 2x separated by

90 days)

• 60% of use was in first four months of aggressive marketing

23

Why Isn’t Price Transparency Decreasing Spending?

 Small fraction of people are signing up for tool  Even among those who sign up, few use the tool

before seeking care

 When they do use the tool, for most services

searchers do not choose a lower cost provider

24

But why?

25

To answer this question

26  National survey of adults who received care in last

year

 Interviews with employees offered a price

transparency tool

Not an explanation: Patients Do Not Care About Prices

 Tell us they are very interested

• 3/4ths said that out-of-pocket costs were very important

when they chose a provider

 “It’s just like going to get a car: [if] people are out

looking around, trying to get the best price, [dealers] are going to drop the price for you because they want your business. I don’t think health care will be any different.”

27

Not an explanation: Belief Higher Prices = Higher Quality

 Only 20% said that it was likely or very likely costs

were related to quality

 “I don’t think that it hurts to get an idea [of price] if

you’re going to get…comparable quality care at a location that’s gonna cost you a little bit less.”

28

Not an explanation: Just increase deductibles

2 4 6 8 10 12 14 16 18 $0 $1-500 $501-1250 $1251-2500 $2501+

Fracti tion

• n of people

ple who actively ely se searched hed out ut price ce data

29

Awareness & Salience

• Do not know where to find data
• Hard to access
• Not sure if the data they see on these websites is

relevant to their care

30

Complexity of Billing

• Strange codes
• CPT/HCPCS/ICD10
• Division of costs
• Single ED visit could mean many bills
• Facility, Lab, radiologist, ED physician,
• Post-hoc nature of billing
• I got some stupid GI bug…in May. I was in the ER twice, and…missed 6
• r 7 days of work. And holy crap, that bill is insane. It would have been

cheaper to have been admitted. Seriously.

• All of a sudden, [my husband] took an insulin shot in the hospital and it

cost me $500. If I had known that, I would have brought the insulin myself and given it to him. So all of a sudden I get this bill for $500 that I have to pay, and there's no way out of it.

Obtuse Nature of Health Care Billing

Most popular aspect of Castlight were looking up bills and benefits

33

Use of tool % Research benefit-design 16 Look up prior claims 14 Finding a provider/facility 11

Deductibles only reward shopping for lower-cost care

Percent Searches by Price Category % Searches $2500+ 37% $1250-$2500 16% $500-$1250 15% $100-$500 19% $0-$100 12%

I was looking into [the price of arthroscopy on tool]…but it didn’t matter because I had already exceeded my deductible for the year, so really it wasn’t out-of-pocket at that point

34

Other factors more important than cost

• Rely

ly on referral: erral: We pretty much go with the recommendation of whatever the physician says…I’ve counted on my primary physician to make the quality check”

• Litt

ttle choice ce: : Here in Redding we only have like two places in town that you can actually get an MRI, so, we don’t have a whole lot of choice

• Loyalty

yalty to PCP/ P/spe specialis cialist: For example our pediatrician, his quality is amazing. We would not, I can’t imagine searching [for] our pediatrician [based on price]…I mean, there’s three doctors I can think of that I would not go elsewhere for

35

Tools themselves

• Most said they were satisfied with

Castlight tool

• Yet low net promoter score and few

used it more than once.

How to move forward?

37

Key takeaways

• Great potential for savings by shifting to

lower-cost providers

• Price shopping still more theory vs. reality
• Transparency is fundamental yet little

evidence that current efforts are increasing shopping and empowering members

• Some potential ideas on how to move

forward

38

Why aren’t people price shopping?

 Current benefit design makes it irrelevant for a

large fraction of care

 “Search friction”

• Price data hard to find and understand
• Complexity of the billing system
• Tools themselves

 Limited circumstances to price shop

• Not everything is “shoppable”
• No alternative choices
• Do not want to disrupt relationships
• Physician recommendation is key

39

Involving Physicians

 Physicians care because their patients care  Physicians care because key way to decrease

spending is to refer to lower-cost providers

 But physicians do not have this information easily

accessible

 Make it easier for them to give a price quote

Focus on select clinical areas

42

 Lab tests  Imaging  Acute care

Profile physician groups and individual PCPs using price-indices

43