SlideShare a Scribd company logo
1 of 44
The Tripadvisor approach for health?
Using patients' online descriptions of their care to
understand healthcare quality

Felix Greaves
Imperial College London
@felixgreaves
England is proud of the National Health Service
But it is not without problems
Quality can be variable, and we often fail to hear the patient perspective
At the same time, we love rating things on the internet
We now do it for our experiences of healthcare too
The number of patients describing their care online is
increasing
USA

UK

Number of hospital ratings

25000

Gao et al., JMIR, 2012

20000
15000
10000
5000
0

Greaves et al., JMIR, 2012
There is resistance from many clinicians to the idea of
being rated online:
Arguments for and against online rating
Why it’s bad

Why it’s good

1. Strain doctor-patient
1.Doctors can often be poor
relationships
judges of their patients’
experience
2. Reviews may be malicious
or fake
2.Feedback changes doctors’
performance
3. Selection bias by those
leaving reviews
3.People are using the
Internet to voice opinions,
4. Lack of meaningful data on
so why not capture this
technical quality of health
information in a useful form
care
Research Question

Is there a relationship between ratings online
and ‘hard’ measures of quality?
The opportunity for a natural experiment

NHS Choices website allows patients to review all NHS services in England
Method:
Obtained national review data from the Department of
Health
We thought we could compare ratings with some
traditional measures of quality

Vs.
Average scores for 10,000 ratings
Mean rating
(out of 5)

Score

Range

The environment where I
was treated was…

3.6

2.6-5.0

I was treated with dignity
and respect by the hospital
staff…
I was involved with
decisions about my care…
The hospital staff worked
well together…

4.0

2.7-5.0

3.8

2.4-5.0

4.1

2.9-5.0

Greaves et al. BMJ Quality and Safety, 2012
Average scores for 10,000 ratings
Score
The environment where I
was treated was…

Mean rating
(out of 5)

3.6

Range

2.6-5.0

I was treated with dignity
4.0
2.7-5.0
and respect by The majority of ratings are positive
• the hospital
staff…
• 67% would recommend to a friend
I was involved with
3.8
2.4-5.0
decisions about my care…
The hospital staff worked
4.1
2.9-5.0
well together…
Greaves et al. BMJ Quality and Safety, 2012
Ratings compared to patient experience surveys
NHS Choices
Measure

Survey question:

Spearman
Rho

p value

Proportion of
patients
recommending

“Overall, how would you rate the
quality of care you received”

0.40

<0.001

Rating of being
treated with dignity
and respect

“Overall, did you feel you were
treated with dignity and respect
while in hospital?”

0.33

<0.001

Rating of staff
working together

“How well would rate how well the
doctors and nurse worked
together?”

0.32

<0.001

Rating of cleanliness “How clean was the hospital ward or 0.48
room you were in?”

<0.001

Greaves et al. BMJ Quality and Safety, 2012
Ratings compared to patient experience surveys
NHS Choices
Measure
Proportion of
patients
recommending

Survey question:
“Overall, how would you rate the
quality of care you received”

Spearman
Rho
0.40

There is a moderate, highly
Rating of being
“Overall, did you feel you were
0.33
significant association between
treated with dignity treated with dignity and respect
ratings online and large
and respect
while in hospital?”
surveys of patient experience
Rating of staff
working together

“How well would rate how well the
doctors and nurse worked
together?”

p value
<0.001

<0.001

0.32

<0.001

Rating of cleanliness “How clean was the hospital ward or 0.48
room you were in?”

<0.001

Greaves et al. BMJ Quality and Safety, 2012
Ratings compared to outcomes
NHS Choices Measure

Spearman
Rho

Other variable

Proportion of patients Hospital Standardised Mortality
recommending
Ratio

p value

-0.20

0.01

Proportion of patients Standardised morality rate for high -0.22
recommending
risk conditions

0.01

Proportion of patients Standardised morality rate among -0.00
recommending
surgical inpatients with serious
treatable complications
Proportion of patients Emergency readmission rate
-0.31
recommending
within 28 days

0.99

Patient perception of
cleanliness

<0.001

Rate of MRSA bacteraemia (per
1,000 bed days)

-0.30

<0.001

Greaves et al. Arch Int Med, 2012
Ratings compared to outcomes
NHS Choices Measure

Spearman
Rho

Other variable

Proportion of patients Hospital Standardised Mortality
recommending
Ratio

p value

-0.20

0.01

Proportion of patients Standardised morality rate for high -0.22
recommending There is a weak, significant
risk conditions

0.01

association between ratings
Proportion of patients Standardised morality rate among -0.00
recommending online and mortality rates
surgical inpatients with serious

0.99

treatable complications
Proportion of patients Emergency readmission rate
recommending
within 28 days

-0.31

<0.001

-0.30

<0.001

Patient perception of
cleanliness

Rate of MRSA bacteraemia (per
1,000 bed days)

Greaves et al. Arch Int Med, 2012
A study in the US found
the same results
comparing Yelp reviews
with the HCAHPS survey
A wisdom in the crowd of patients?

Sir Francis Galton won ‘guess-the-weight-of-the-bull’ competitions
by asking lots of local farmers for their best guess, and then taking
the average – he gave us the idea of a ‘wisdom of crowds’
Comparison of the NHS Inpatient Survey and ratings on the NHS
Choices website
NHS Inpatient Survey NHS Choices ratings
Mechanism

Paper-based survey

Ratings left on a website

Number of
responses
Selection

69,000 per year

5,000 per year
Self-selecting; patients are
not solicited

Proportion
positive

Random; patients receive a
survey requesting
completion after leaving
hospital
79% rated their overall care
as excellent or very good

Cost

Likely more expensive

Likely less expensive

67% would recommend to a
friend
Comparison of the NHS Inpatient Survey and ratings on the NHS
Choices website
NHS Inpatient Survey NHS Choices ratings
Mechanism
Number of
responses
Selection

Proportion
positive
Cost

Paper-based survey

Ratings left on a website

There were 10,000 hospital ratings inper year
69,000 per year
5,000 the
UK over 2 years
Random; patients receive a Self-selecting; patients are
Over the same time
not were
survey requesting period there solicited
29,118,009 hospital
completion after leaving admissions
hospital
79% rated hospital admissions are rated
0.04% of their overall care 67% would recommend to a
as excellent or very good
friend
Likely more expensive

Likely less expensive
The stereotypical reviewer?
Where are people
rating their care?
Associations between whether a practice is rated
with population and practice characteristics
Independent
variable

Z statistic

p value

Practice
population size
IMD score of
patients
Population
density

15.38

<0.001

-7.82

<0.001

6.72

<0.001

Singlehander
Proportion of
population aged
over 65 years
Proportion of
population who
are white
Type of contract
Training practice

-4.50
-3.88

<0.001
<0.001

-1.58

0.11

-0.71
0.35

0.48
0.73
Associations between whether a practice is rated
with population and practice characteristics
Independent
variable

Z statistic

p value

15.38
<0.001
Practice
population size
-7.82
<0.001
IMD score of
• Practices serving younger people
patients
<0.001
Population more6.72
likely to be rated
density

are

• Practices serving less deprived people

Singlehander
-4.50
<0.001
-3.88
Proportion are more likely to be rated <0.001
of
population aged
• Practices in urban areas are more
over 65 years
0.11
Proportion likely -1.58be rated
of
to
population who
are white
Type of contract -0.71
0.48
Training practice 0.35
0.73
Cloud of patient experience ?
A problem

Free Text
A new trend: ‘Big Data’ and social media analytics
Sentiment analysis
Another natural experiment

We can compare patients’ free text descriptions of care with their own quantitative ratings
What we did: Machine learning
• You need to teach an algorithm how to recognise particular
words and phrases
• We used all comments and ratings from 3 years as a training
set (13,802 comments)
We tried to predict patients ratings of their care from their
comments in 2010
• Whether the patient would recommend the hospital or not
• Whether the hospital was clean or not
• Whether the patient was treated with dignity or not
Used open source Weka software
Sentiment analysis can be tricky
Simple punctuation contains little information

(
But when paired with other punctuation, it’s very
important

:(
And small changes can have big effects on sentiment

:)
Some words the algorithm thought were important

Overall
• rude
• excellent
• hours
• pain
• communication

Cleanliness
• dirty
• floor
• filthy
• bed
• blood

Dignity
• rude
• told
• thank
• friendly
• attitude
How good is it at predicting ratings?

Question we are predicting
the answer to

Total number Prediction
Kappa P value
of comments accuracy

Would you recommend the
hospital?

6412

88.7

0.75

<0.0001

How clean was the hospital
room or ward that you were in?

6139

81.2

0.40

<0.0001

Overall, did you feel you were
treated with respect and dignity
while you were in the hospital?

6239

83.6

0.56

<0.0001
How does this compare to patient surveys?

Patient Survey
Question

Machine learning
prediction

Spearman
rho

Overall, how would you
rate the care you
received?

Whether the patient would
0.46
recommend the hospital

Probability

P<0.001

Overall, did you feel you
were treated with respect Whether the patient was
treated with dignity
and dignity while you
were in the hospital?

0.50

P<0.001

In your opinion, how
clean was the hospital
room or ward that you
were in?

0.37

P<0.001

Standard of cleanliness
Limitations

• Selection bias
• Sarcasm / Irony
• Culturally specific
 Phrases like ‘cup of tea’ important,
but effect depends on context
The UK quality regulator now
includes online reviews and
comments in its performance
measurement framework
The NHS has started publically reporting social media sentiment
Conclusions
Online rating and reviewing are on the rise

Evidence of some wisdom in the crowd of patients

Substantial professional resistance to the idea

Selection bias in those commenting online

The ‘cloud of patient experience’ is a potentially valuable
source of information to understand patient views
Thanks to my co-authors, and funders

•
•
•
•
•
•
•
•
•

Christopher Millett
Ara Darzi
Dominic King
Henry Lee
Utz Pape
Liam Donaldson
Azeem Majeed
Robert Wachter
Daniel Ramirez-Cano

More Related Content

What's hot

CLIENT SATISFACTION SURVEY 2016
CLIENT SATISFACTION SURVEY 2016CLIENT SATISFACTION SURVEY 2016
CLIENT SATISFACTION SURVEY 2016Nsubuga Nicholas
 
IUHA ED Patient Satisfaction Quality Improvement Presentation
IUHA ED Patient Satisfaction Quality Improvement PresentationIUHA ED Patient Satisfaction Quality Improvement Presentation
IUHA ED Patient Satisfaction Quality Improvement PresentationMarissa Wuethrich
 
QMAC Patient Involvement
QMAC Patient InvolvementQMAC Patient Involvement
QMAC Patient InvolvementAdam Thompson
 
Patient Activation: Where Do I Start?
Patient Activation: Where Do I Start?   Patient Activation: Where Do I Start?
Patient Activation: Where Do I Start? EngagingPatients
 
iWGC symposium 2016 slide deck
iWGC symposium 2016 slide deck iWGC symposium 2016 slide deck
iWGC symposium 2016 slide deck Emily Fovargue
 
Multisource feedback & its utility
Multisource feedback & its utilityMultisource feedback & its utility
Multisource feedback & its utilityIAMRAreval2015
 
6. Advanced Access and Predictive Analytics
6. Advanced Access and Predictive Analytics6. Advanced Access and Predictive Analytics
6. Advanced Access and Predictive AnalyticsMichele Molden
 
Prudent healthcare and patient activation (1)
Prudent healthcare and patient activation (1)Prudent healthcare and patient activation (1)
Prudent healthcare and patient activation (1)Andrew Rix
 
3 Sisters Brochure
3 Sisters Brochure3 Sisters Brochure
3 Sisters BrochureBarry Duncan
 
How to double client outcomes in 18 seconds (Lambert, 2014)
How to double client outcomes in 18 seconds (Lambert, 2014)How to double client outcomes in 18 seconds (Lambert, 2014)
How to double client outcomes in 18 seconds (Lambert, 2014)Scott Miller
 
Engaging with Clinicians by Creating Highly Adoptable Improvement: Relevance ...
Engaging with Clinicians by Creating Highly Adoptable Improvement: Relevance ...Engaging with Clinicians by Creating Highly Adoptable Improvement: Relevance ...
Engaging with Clinicians by Creating Highly Adoptable Improvement: Relevance ...Canadian Patient Safety Institute
 
Patient & Family Advisory Councils: the Business Case for Starting a PFAC & P...
Patient & Family Advisory Councils: the Business Case for Starting a PFAC & P...Patient & Family Advisory Councils: the Business Case for Starting a PFAC & P...
Patient & Family Advisory Councils: the Business Case for Starting a PFAC & P...EngagingPatients
 
10 Must Know Techniques for Managing Physician Relations in Today's Digital W...
10 Must Know Techniques for Managing Physician Relations in Today's Digital W...10 Must Know Techniques for Managing Physician Relations in Today's Digital W...
10 Must Know Techniques for Managing Physician Relations in Today's Digital W...Endeavor Management
 
Behavioral Health Outcomes
Behavioral Health OutcomesBehavioral Health Outcomes
Behavioral Health Outcomesstclairer
 

What's hot (20)

CLIENT SATISFACTION SURVEY 2016
CLIENT SATISFACTION SURVEY 2016CLIENT SATISFACTION SURVEY 2016
CLIENT SATISFACTION SURVEY 2016
 
IUHA ED Patient Satisfaction Quality Improvement Presentation
IUHA ED Patient Satisfaction Quality Improvement PresentationIUHA ED Patient Satisfaction Quality Improvement Presentation
IUHA ED Patient Satisfaction Quality Improvement Presentation
 
Aware - Just Clinical
Aware - Just ClinicalAware - Just Clinical
Aware - Just Clinical
 
Canadian MedRec Quality Audit Month 2015 - Results
Canadian MedRec Quality Audit Month 2015 - ResultsCanadian MedRec Quality Audit Month 2015 - Results
Canadian MedRec Quality Audit Month 2015 - Results
 
QMAC Patient Involvement
QMAC Patient InvolvementQMAC Patient Involvement
QMAC Patient Involvement
 
The Canterbury Initiative
The Canterbury InitiativeThe Canterbury Initiative
The Canterbury Initiative
 
Patient Activation: Where Do I Start?
Patient Activation: Where Do I Start?   Patient Activation: Where Do I Start?
Patient Activation: Where Do I Start?
 
iWGC symposium 2016 slide deck
iWGC symposium 2016 slide deck iWGC symposium 2016 slide deck
iWGC symposium 2016 slide deck
 
Underrepresented in Emergency Medicine: Studying the Medical Education Pipeli...
Underrepresented in Emergency Medicine: Studying the Medical Education Pipeli...Underrepresented in Emergency Medicine: Studying the Medical Education Pipeli...
Underrepresented in Emergency Medicine: Studying the Medical Education Pipeli...
 
Multisource feedback & its utility
Multisource feedback & its utilityMultisource feedback & its utility
Multisource feedback & its utility
 
6. Advanced Access and Predictive Analytics
6. Advanced Access and Predictive Analytics6. Advanced Access and Predictive Analytics
6. Advanced Access and Predictive Analytics
 
Robert _highly_organized_primary_care_2
Robert  _highly_organized_primary_care_2Robert  _highly_organized_primary_care_2
Robert _highly_organized_primary_care_2
 
Prudent healthcare and patient activation (1)
Prudent healthcare and patient activation (1)Prudent healthcare and patient activation (1)
Prudent healthcare and patient activation (1)
 
3 Sisters Brochure
3 Sisters Brochure3 Sisters Brochure
3 Sisters Brochure
 
How to double client outcomes in 18 seconds (Lambert, 2014)
How to double client outcomes in 18 seconds (Lambert, 2014)How to double client outcomes in 18 seconds (Lambert, 2014)
How to double client outcomes in 18 seconds (Lambert, 2014)
 
Engaging with Clinicians by Creating Highly Adoptable Improvement: Relevance ...
Engaging with Clinicians by Creating Highly Adoptable Improvement: Relevance ...Engaging with Clinicians by Creating Highly Adoptable Improvement: Relevance ...
Engaging with Clinicians by Creating Highly Adoptable Improvement: Relevance ...
 
Systematic Screening: Applying lessons learned from an international best pra...
Systematic Screening: Applying lessons learned from an international best pra...Systematic Screening: Applying lessons learned from an international best pra...
Systematic Screening: Applying lessons learned from an international best pra...
 
Patient & Family Advisory Councils: the Business Case for Starting a PFAC & P...
Patient & Family Advisory Councils: the Business Case for Starting a PFAC & P...Patient & Family Advisory Councils: the Business Case for Starting a PFAC & P...
Patient & Family Advisory Councils: the Business Case for Starting a PFAC & P...
 
10 Must Know Techniques for Managing Physician Relations in Today's Digital W...
10 Must Know Techniques for Managing Physician Relations in Today's Digital W...10 Must Know Techniques for Managing Physician Relations in Today's Digital W...
10 Must Know Techniques for Managing Physician Relations in Today's Digital W...
 
Behavioral Health Outcomes
Behavioral Health OutcomesBehavioral Health Outcomes
Behavioral Health Outcomes
 

Viewers also liked

5 day training schedule
5 day training schedule5 day training schedule
5 day training scheduleJames Gonzalez
 
チーム子育てプロジェクト企画書(概要版)
チーム子育てプロジェクト企画書(概要版)チーム子育てプロジェクト企画書(概要版)
チーム子育てプロジェクト企画書(概要版)mh0306052
 
La standards chart powerpoint
La standards chart powerpointLa standards chart powerpoint
La standards chart powerpointcatherinereed00
 
Danny -10 things you need to know about vietnam
Danny -10 things you need to know about vietnamDanny -10 things you need to know about vietnam
Danny -10 things you need to know about vietnamPhúc Đức
 
El cuerpo humano
El cuerpo humanoEl cuerpo humano
El cuerpo humanoRosabelori
 

Viewers also liked (7)

5 day training schedule
5 day training schedule5 day training schedule
5 day training schedule
 
チーム子育てプロジェクト企画書(概要版)
チーム子育てプロジェクト企画書(概要版)チーム子育てプロジェクト企画書(概要版)
チーム子育てプロジェクト企画書(概要版)
 
La standards chart powerpoint
La standards chart powerpointLa standards chart powerpoint
La standards chart powerpoint
 
Danny -10 things you need to know about vietnam
Danny -10 things you need to know about vietnamDanny -10 things you need to know about vietnam
Danny -10 things you need to know about vietnam
 
Insurance1.2 uchart
Insurance1.2 uchartInsurance1.2 uchart
Insurance1.2 uchart
 
Seminar
SeminarSeminar
Seminar
 
El cuerpo humano
El cuerpo humanoEl cuerpo humano
El cuerpo humano
 

Similar to The 'Tripadvisor' Approach to Health

A study on health care services at rajarajeshwari
A study on health care services at rajarajeshwariA study on health care services at rajarajeshwari
A study on health care services at rajarajeshwarineelaiahgari madhavi
 
TitlePATIENTS SAISFACTION ABOUT PATIENTS REFER.docx
TitlePATIENTS SAISFACTION ABOUT PATIENTS REFER.docxTitlePATIENTS SAISFACTION ABOUT PATIENTS REFER.docx
TitlePATIENTS SAISFACTION ABOUT PATIENTS REFER.docxjuliennehar
 
Question of Quality Conference 2016 - Quality at UCLH
Question of Quality Conference 2016 - Quality at UCLHQuestion of Quality Conference 2016 - Quality at UCLH
Question of Quality Conference 2016 - Quality at UCLHHCA Healthcare UK
 
Clinician Satisfaction Before and After Transition from a Basic to a Comprehe...
Clinician Satisfaction Before and After Transition from a Basic to a Comprehe...Clinician Satisfaction Before and After Transition from a Basic to a Comprehe...
Clinician Satisfaction Before and After Transition from a Basic to a Comprehe...Allison McCoy
 
David Prior: driving improvements in the quality of care across the system
David Prior: driving improvements in the quality of care across the systemDavid Prior: driving improvements in the quality of care across the system
David Prior: driving improvements in the quality of care across the systemThe King's Fund
 
Deborah Schaler - EHPR 2014 - Patient Feedback Research
Deborah Schaler - EHPR 2014 - Patient Feedback ResearchDeborah Schaler - EHPR 2014 - Patient Feedback Research
Deborah Schaler - EHPR 2014 - Patient Feedback ResearchDeborah Schaler
 
SIP-Feedback system College
SIP-Feedback system CollegeSIP-Feedback system College
SIP-Feedback system CollegeShreekanth Dangi
 
Patient’s experience, improve the quality health3
Patient’s experience, improve the quality health3Patient’s experience, improve the quality health3
Patient’s experience, improve the quality health3zsaddique
 
The art of the possible will
The art of the possible   willThe art of the possible   will
The art of the possible willhowardcooper
 
Realizing the Promise of Patient-Reported Outcomes Measures
Realizing the Promise of Patient-Reported Outcomes MeasuresRealizing the Promise of Patient-Reported Outcomes Measures
Realizing the Promise of Patient-Reported Outcomes MeasuresHealth Catalyst
 
Improving patients' experiences
Improving patients' experiencesImproving patients' experiences
Improving patients' experiencesminu2
 
Clinical audit for the enlightened ian callanan hslg conference 2013
Clinical audit for the enlightened ian callanan hslg conference 2013Clinical audit for the enlightened ian callanan hslg conference 2013
Clinical audit for the enlightened ian callanan hslg conference 2013hslgcommittee
 
How many more staff do you need to improve the quality of care?
How many more staff do you need to improve the quality of care?How many more staff do you need to improve the quality of care?
How many more staff do you need to improve the quality of care?mckesson
 
Embrace Transparency to Transform the Experience HMPS 2022.pptx
Embrace Transparency to Transform the Experience HMPS 2022.pptxEmbrace Transparency to Transform the Experience HMPS 2022.pptx
Embrace Transparency to Transform the Experience HMPS 2022.pptxSuzanne Hendery
 
Suzanne Hendery Embrace Transparency to Transform the Experience-Suzanne Hend...
Suzanne Hendery Embrace Transparency to Transform the Experience-Suzanne Hend...Suzanne Hendery Embrace Transparency to Transform the Experience-Suzanne Hend...
Suzanne Hendery Embrace Transparency to Transform the Experience-Suzanne Hend...Suzanne Hendery
 
2014 Physician Compensation and Employment Report
2014 Physician Compensation and Employment Report2014 Physician Compensation and Employment Report
2014 Physician Compensation and Employment ReportMeaghan O'Neil
 
A comparative study on patients’ satisfaction in health care service
A comparative study on patients’ satisfaction in health care serviceA comparative study on patients’ satisfaction in health care service
A comparative study on patients’ satisfaction in health care serviceAlexander Decker
 
How to Define Effective and Efficient Real World Trials
How to Define Effective and Efficient Real World TrialsHow to Define Effective and Efficient Real World Trials
How to Define Effective and Efficient Real World TrialsTodd Berner MD
 

Similar to The 'Tripadvisor' Approach to Health (20)

A study on health care services at rajarajeshwari
A study on health care services at rajarajeshwariA study on health care services at rajarajeshwari
A study on health care services at rajarajeshwari
 
TitlePATIENTS SAISFACTION ABOUT PATIENTS REFER.docx
TitlePATIENTS SAISFACTION ABOUT PATIENTS REFER.docxTitlePATIENTS SAISFACTION ABOUT PATIENTS REFER.docx
TitlePATIENTS SAISFACTION ABOUT PATIENTS REFER.docx
 
Question of Quality Conference 2016 - Quality at UCLH
Question of Quality Conference 2016 - Quality at UCLHQuestion of Quality Conference 2016 - Quality at UCLH
Question of Quality Conference 2016 - Quality at UCLH
 
Clinician Satisfaction Before and After Transition from a Basic to a Comprehe...
Clinician Satisfaction Before and After Transition from a Basic to a Comprehe...Clinician Satisfaction Before and After Transition from a Basic to a Comprehe...
Clinician Satisfaction Before and After Transition from a Basic to a Comprehe...
 
David Prior: driving improvements in the quality of care across the system
David Prior: driving improvements in the quality of care across the systemDavid Prior: driving improvements in the quality of care across the system
David Prior: driving improvements in the quality of care across the system
 
Deborah Schaler - EHPR 2014 - Patient Feedback Research
Deborah Schaler - EHPR 2014 - Patient Feedback ResearchDeborah Schaler - EHPR 2014 - Patient Feedback Research
Deborah Schaler - EHPR 2014 - Patient Feedback Research
 
SIP-Feedback system College
SIP-Feedback system CollegeSIP-Feedback system College
SIP-Feedback system College
 
Patient’s experience, improve the quality health3
Patient’s experience, improve the quality health3Patient’s experience, improve the quality health3
Patient’s experience, improve the quality health3
 
The art of the possible will
The art of the possible   willThe art of the possible   will
The art of the possible will
 
Realizing the Promise of Patient-Reported Outcomes Measures
Realizing the Promise of Patient-Reported Outcomes MeasuresRealizing the Promise of Patient-Reported Outcomes Measures
Realizing the Promise of Patient-Reported Outcomes Measures
 
Improving patients' experiences
Improving patients' experiencesImproving patients' experiences
Improving patients' experiences
 
Clinical audit for the enlightened ian callanan hslg conference 2013
Clinical audit for the enlightened ian callanan hslg conference 2013Clinical audit for the enlightened ian callanan hslg conference 2013
Clinical audit for the enlightened ian callanan hslg conference 2013
 
healthcare quality and safety
 healthcare quality and  safety healthcare quality and  safety
healthcare quality and safety
 
How many more staff do you need to improve the quality of care?
How many more staff do you need to improve the quality of care?How many more staff do you need to improve the quality of care?
How many more staff do you need to improve the quality of care?
 
Embrace Transparency to Transform the Experience HMPS 2022.pptx
Embrace Transparency to Transform the Experience HMPS 2022.pptxEmbrace Transparency to Transform the Experience HMPS 2022.pptx
Embrace Transparency to Transform the Experience HMPS 2022.pptx
 
Suzanne Hendery Embrace Transparency to Transform the Experience-Suzanne Hend...
Suzanne Hendery Embrace Transparency to Transform the Experience-Suzanne Hend...Suzanne Hendery Embrace Transparency to Transform the Experience-Suzanne Hend...
Suzanne Hendery Embrace Transparency to Transform the Experience-Suzanne Hend...
 
2014 Physician Compensation and Employment Report
2014 Physician Compensation and Employment Report2014 Physician Compensation and Employment Report
2014 Physician Compensation and Employment Report
 
Patient Experience June 2015
Patient Experience June 2015Patient Experience June 2015
Patient Experience June 2015
 
A comparative study on patients’ satisfaction in health care service
A comparative study on patients’ satisfaction in health care serviceA comparative study on patients’ satisfaction in health care service
A comparative study on patients’ satisfaction in health care service
 
How to Define Effective and Efficient Real World Trials
How to Define Effective and Efficient Real World TrialsHow to Define Effective and Efficient Real World Trials
How to Define Effective and Efficient Real World Trials
 

Recently uploaded

Night 7k to 12k Navi Mumbai Call Girl Photo 👉 BOOK NOW 9833363713 👈 ♀️ night ...
Night 7k to 12k Navi Mumbai Call Girl Photo 👉 BOOK NOW 9833363713 👈 ♀️ night ...Night 7k to 12k Navi Mumbai Call Girl Photo 👉 BOOK NOW 9833363713 👈 ♀️ night ...
Night 7k to 12k Navi Mumbai Call Girl Photo 👉 BOOK NOW 9833363713 👈 ♀️ night ...aartirawatdelhi
 
Lucknow Call girls - 8800925952 - 24x7 service with hotel room
Lucknow Call girls - 8800925952 - 24x7 service with hotel roomLucknow Call girls - 8800925952 - 24x7 service with hotel room
Lucknow Call girls - 8800925952 - 24x7 service with hotel roomdiscovermytutordmt
 
The Most Attractive Hyderabad Call Girls Kothapet 𖠋 6297143586 𖠋 Will You Mis...
The Most Attractive Hyderabad Call Girls Kothapet 𖠋 6297143586 𖠋 Will You Mis...The Most Attractive Hyderabad Call Girls Kothapet 𖠋 6297143586 𖠋 Will You Mis...
The Most Attractive Hyderabad Call Girls Kothapet 𖠋 6297143586 𖠋 Will You Mis...chandars293
 
Vip Call Girls Anna Salai Chennai 👉 8250192130 ❣️💯 Top Class Girls Available
Vip Call Girls Anna Salai Chennai 👉 8250192130 ❣️💯 Top Class Girls AvailableVip Call Girls Anna Salai Chennai 👉 8250192130 ❣️💯 Top Class Girls Available
Vip Call Girls Anna Salai Chennai 👉 8250192130 ❣️💯 Top Class Girls AvailableNehru place Escorts
 
Book Paid Powai Call Girls Mumbai 𖠋 9930245274 𖠋Low Budget Full Independent H...
Book Paid Powai Call Girls Mumbai 𖠋 9930245274 𖠋Low Budget Full Independent H...Book Paid Powai Call Girls Mumbai 𖠋 9930245274 𖠋Low Budget Full Independent H...
Book Paid Powai Call Girls Mumbai 𖠋 9930245274 𖠋Low Budget Full Independent H...Call Girls in Nagpur High Profile
 
Call Girls Service Surat Samaira ❤️🍑 8250192130 👄 Independent Escort Service ...
Call Girls Service Surat Samaira ❤️🍑 8250192130 👄 Independent Escort Service ...Call Girls Service Surat Samaira ❤️🍑 8250192130 👄 Independent Escort Service ...
Call Girls Service Surat Samaira ❤️🍑 8250192130 👄 Independent Escort Service ...CALL GIRLS
 
Russian Escorts Girls Nehru Place ZINATHI 🔝9711199012 ☪ 24/7 Call Girls Delhi
Russian Escorts Girls  Nehru Place ZINATHI 🔝9711199012 ☪ 24/7 Call Girls DelhiRussian Escorts Girls  Nehru Place ZINATHI 🔝9711199012 ☪ 24/7 Call Girls Delhi
Russian Escorts Girls Nehru Place ZINATHI 🔝9711199012 ☪ 24/7 Call Girls DelhiAlinaDevecerski
 
Best Rate (Hyderabad) Call Girls Jahanuma ⟟ 8250192130 ⟟ High Class Call Girl...
Best Rate (Hyderabad) Call Girls Jahanuma ⟟ 8250192130 ⟟ High Class Call Girl...Best Rate (Hyderabad) Call Girls Jahanuma ⟟ 8250192130 ⟟ High Class Call Girl...
Best Rate (Hyderabad) Call Girls Jahanuma ⟟ 8250192130 ⟟ High Class Call Girl...astropune
 
Call Girls Ludhiana Just Call 9907093804 Top Class Call Girl Service Available
Call Girls Ludhiana Just Call 9907093804 Top Class Call Girl Service AvailableCall Girls Ludhiana Just Call 9907093804 Top Class Call Girl Service Available
Call Girls Ludhiana Just Call 9907093804 Top Class Call Girl Service AvailableDipal Arora
 
Premium Call Girls Cottonpet Whatsapp 7001035870 Independent Escort Service
Premium Call Girls Cottonpet Whatsapp 7001035870 Independent Escort ServicePremium Call Girls Cottonpet Whatsapp 7001035870 Independent Escort Service
Premium Call Girls Cottonpet Whatsapp 7001035870 Independent Escort Servicevidya singh
 
Call Girls Bhubaneswar Just Call 9907093804 Top Class Call Girl Service Avail...
Call Girls Bhubaneswar Just Call 9907093804 Top Class Call Girl Service Avail...Call Girls Bhubaneswar Just Call 9907093804 Top Class Call Girl Service Avail...
Call Girls Bhubaneswar Just Call 9907093804 Top Class Call Girl Service Avail...Dipal Arora
 
(Rocky) Jaipur Call Girl - 9521753030 Escorts Service 50% Off with Cash ON De...
(Rocky) Jaipur Call Girl - 9521753030 Escorts Service 50% Off with Cash ON De...(Rocky) Jaipur Call Girl - 9521753030 Escorts Service 50% Off with Cash ON De...
(Rocky) Jaipur Call Girl - 9521753030 Escorts Service 50% Off with Cash ON De...indiancallgirl4rent
 
Artifacts in Nuclear Medicine with Identifying and resolving artifacts.
Artifacts in Nuclear Medicine with Identifying and resolving artifacts.Artifacts in Nuclear Medicine with Identifying and resolving artifacts.
Artifacts in Nuclear Medicine with Identifying and resolving artifacts.MiadAlsulami
 
VIP Call Girls Indore Kirti 💚😋 9256729539 🚀 Indore Escorts
VIP Call Girls Indore Kirti 💚😋  9256729539 🚀 Indore EscortsVIP Call Girls Indore Kirti 💚😋  9256729539 🚀 Indore Escorts
VIP Call Girls Indore Kirti 💚😋 9256729539 🚀 Indore Escortsaditipandeya
 
Call Girls Horamavu WhatsApp Number 7001035870 Meeting With Bangalore Escorts
Call Girls Horamavu WhatsApp Number 7001035870 Meeting With Bangalore EscortsCall Girls Horamavu WhatsApp Number 7001035870 Meeting With Bangalore Escorts
Call Girls Horamavu WhatsApp Number 7001035870 Meeting With Bangalore Escortsvidya singh
 
Low Rate Call Girls Kochi Anika 8250192130 Independent Escort Service Kochi
Low Rate Call Girls Kochi Anika 8250192130 Independent Escort Service KochiLow Rate Call Girls Kochi Anika 8250192130 Independent Escort Service Kochi
Low Rate Call Girls Kochi Anika 8250192130 Independent Escort Service KochiSuhani Kapoor
 
Chandrapur Call girls 8617370543 Provides all area service COD available
Chandrapur Call girls 8617370543 Provides all area service COD availableChandrapur Call girls 8617370543 Provides all area service COD available
Chandrapur Call girls 8617370543 Provides all area service COD availableDipal Arora
 
All Time Service Available Call Girls Marine Drive 📳 9820252231 For 18+ VIP C...
All Time Service Available Call Girls Marine Drive 📳 9820252231 For 18+ VIP C...All Time Service Available Call Girls Marine Drive 📳 9820252231 For 18+ VIP C...
All Time Service Available Call Girls Marine Drive 📳 9820252231 For 18+ VIP C...Arohi Goyal
 
Call Girls Darjeeling Just Call 9907093804 Top Class Call Girl Service Available
Call Girls Darjeeling Just Call 9907093804 Top Class Call Girl Service AvailableCall Girls Darjeeling Just Call 9907093804 Top Class Call Girl Service Available
Call Girls Darjeeling Just Call 9907093804 Top Class Call Girl Service AvailableDipal Arora
 

Recently uploaded (20)

Night 7k to 12k Navi Mumbai Call Girl Photo 👉 BOOK NOW 9833363713 👈 ♀️ night ...
Night 7k to 12k Navi Mumbai Call Girl Photo 👉 BOOK NOW 9833363713 👈 ♀️ night ...Night 7k to 12k Navi Mumbai Call Girl Photo 👉 BOOK NOW 9833363713 👈 ♀️ night ...
Night 7k to 12k Navi Mumbai Call Girl Photo 👉 BOOK NOW 9833363713 👈 ♀️ night ...
 
Lucknow Call girls - 8800925952 - 24x7 service with hotel room
Lucknow Call girls - 8800925952 - 24x7 service with hotel roomLucknow Call girls - 8800925952 - 24x7 service with hotel room
Lucknow Call girls - 8800925952 - 24x7 service with hotel room
 
The Most Attractive Hyderabad Call Girls Kothapet 𖠋 6297143586 𖠋 Will You Mis...
The Most Attractive Hyderabad Call Girls Kothapet 𖠋 6297143586 𖠋 Will You Mis...The Most Attractive Hyderabad Call Girls Kothapet 𖠋 6297143586 𖠋 Will You Mis...
The Most Attractive Hyderabad Call Girls Kothapet 𖠋 6297143586 𖠋 Will You Mis...
 
Vip Call Girls Anna Salai Chennai 👉 8250192130 ❣️💯 Top Class Girls Available
Vip Call Girls Anna Salai Chennai 👉 8250192130 ❣️💯 Top Class Girls AvailableVip Call Girls Anna Salai Chennai 👉 8250192130 ❣️💯 Top Class Girls Available
Vip Call Girls Anna Salai Chennai 👉 8250192130 ❣️💯 Top Class Girls Available
 
Book Paid Powai Call Girls Mumbai 𖠋 9930245274 𖠋Low Budget Full Independent H...
Book Paid Powai Call Girls Mumbai 𖠋 9930245274 𖠋Low Budget Full Independent H...Book Paid Powai Call Girls Mumbai 𖠋 9930245274 𖠋Low Budget Full Independent H...
Book Paid Powai Call Girls Mumbai 𖠋 9930245274 𖠋Low Budget Full Independent H...
 
Escort Service Call Girls In Sarita Vihar,, 99530°56974 Delhi NCR
Escort Service Call Girls In Sarita Vihar,, 99530°56974 Delhi NCREscort Service Call Girls In Sarita Vihar,, 99530°56974 Delhi NCR
Escort Service Call Girls In Sarita Vihar,, 99530°56974 Delhi NCR
 
Call Girls Service Surat Samaira ❤️🍑 8250192130 👄 Independent Escort Service ...
Call Girls Service Surat Samaira ❤️🍑 8250192130 👄 Independent Escort Service ...Call Girls Service Surat Samaira ❤️🍑 8250192130 👄 Independent Escort Service ...
Call Girls Service Surat Samaira ❤️🍑 8250192130 👄 Independent Escort Service ...
 
Russian Escorts Girls Nehru Place ZINATHI 🔝9711199012 ☪ 24/7 Call Girls Delhi
Russian Escorts Girls  Nehru Place ZINATHI 🔝9711199012 ☪ 24/7 Call Girls DelhiRussian Escorts Girls  Nehru Place ZINATHI 🔝9711199012 ☪ 24/7 Call Girls Delhi
Russian Escorts Girls Nehru Place ZINATHI 🔝9711199012 ☪ 24/7 Call Girls Delhi
 
Best Rate (Hyderabad) Call Girls Jahanuma ⟟ 8250192130 ⟟ High Class Call Girl...
Best Rate (Hyderabad) Call Girls Jahanuma ⟟ 8250192130 ⟟ High Class Call Girl...Best Rate (Hyderabad) Call Girls Jahanuma ⟟ 8250192130 ⟟ High Class Call Girl...
Best Rate (Hyderabad) Call Girls Jahanuma ⟟ 8250192130 ⟟ High Class Call Girl...
 
Call Girls Ludhiana Just Call 9907093804 Top Class Call Girl Service Available
Call Girls Ludhiana Just Call 9907093804 Top Class Call Girl Service AvailableCall Girls Ludhiana Just Call 9907093804 Top Class Call Girl Service Available
Call Girls Ludhiana Just Call 9907093804 Top Class Call Girl Service Available
 
Premium Call Girls Cottonpet Whatsapp 7001035870 Independent Escort Service
Premium Call Girls Cottonpet Whatsapp 7001035870 Independent Escort ServicePremium Call Girls Cottonpet Whatsapp 7001035870 Independent Escort Service
Premium Call Girls Cottonpet Whatsapp 7001035870 Independent Escort Service
 
Call Girls Bhubaneswar Just Call 9907093804 Top Class Call Girl Service Avail...
Call Girls Bhubaneswar Just Call 9907093804 Top Class Call Girl Service Avail...Call Girls Bhubaneswar Just Call 9907093804 Top Class Call Girl Service Avail...
Call Girls Bhubaneswar Just Call 9907093804 Top Class Call Girl Service Avail...
 
(Rocky) Jaipur Call Girl - 9521753030 Escorts Service 50% Off with Cash ON De...
(Rocky) Jaipur Call Girl - 9521753030 Escorts Service 50% Off with Cash ON De...(Rocky) Jaipur Call Girl - 9521753030 Escorts Service 50% Off with Cash ON De...
(Rocky) Jaipur Call Girl - 9521753030 Escorts Service 50% Off with Cash ON De...
 
Artifacts in Nuclear Medicine with Identifying and resolving artifacts.
Artifacts in Nuclear Medicine with Identifying and resolving artifacts.Artifacts in Nuclear Medicine with Identifying and resolving artifacts.
Artifacts in Nuclear Medicine with Identifying and resolving artifacts.
 
VIP Call Girls Indore Kirti 💚😋 9256729539 🚀 Indore Escorts
VIP Call Girls Indore Kirti 💚😋  9256729539 🚀 Indore EscortsVIP Call Girls Indore Kirti 💚😋  9256729539 🚀 Indore Escorts
VIP Call Girls Indore Kirti 💚😋 9256729539 🚀 Indore Escorts
 
Call Girls Horamavu WhatsApp Number 7001035870 Meeting With Bangalore Escorts
Call Girls Horamavu WhatsApp Number 7001035870 Meeting With Bangalore EscortsCall Girls Horamavu WhatsApp Number 7001035870 Meeting With Bangalore Escorts
Call Girls Horamavu WhatsApp Number 7001035870 Meeting With Bangalore Escorts
 
Low Rate Call Girls Kochi Anika 8250192130 Independent Escort Service Kochi
Low Rate Call Girls Kochi Anika 8250192130 Independent Escort Service KochiLow Rate Call Girls Kochi Anika 8250192130 Independent Escort Service Kochi
Low Rate Call Girls Kochi Anika 8250192130 Independent Escort Service Kochi
 
Chandrapur Call girls 8617370543 Provides all area service COD available
Chandrapur Call girls 8617370543 Provides all area service COD availableChandrapur Call girls 8617370543 Provides all area service COD available
Chandrapur Call girls 8617370543 Provides all area service COD available
 
All Time Service Available Call Girls Marine Drive 📳 9820252231 For 18+ VIP C...
All Time Service Available Call Girls Marine Drive 📳 9820252231 For 18+ VIP C...All Time Service Available Call Girls Marine Drive 📳 9820252231 For 18+ VIP C...
All Time Service Available Call Girls Marine Drive 📳 9820252231 For 18+ VIP C...
 
Call Girls Darjeeling Just Call 9907093804 Top Class Call Girl Service Available
Call Girls Darjeeling Just Call 9907093804 Top Class Call Girl Service AvailableCall Girls Darjeeling Just Call 9907093804 Top Class Call Girl Service Available
Call Girls Darjeeling Just Call 9907093804 Top Class Call Girl Service Available
 

The 'Tripadvisor' Approach to Health

  • 1. The Tripadvisor approach for health? Using patients' online descriptions of their care to understand healthcare quality Felix Greaves Imperial College London @felixgreaves
  • 2. England is proud of the National Health Service
  • 3. But it is not without problems Quality can be variable, and we often fail to hear the patient perspective
  • 4. At the same time, we love rating things on the internet
  • 5. We now do it for our experiences of healthcare too
  • 6. The number of patients describing their care online is increasing USA UK Number of hospital ratings 25000 Gao et al., JMIR, 2012 20000 15000 10000 5000 0 Greaves et al., JMIR, 2012
  • 7. There is resistance from many clinicians to the idea of being rated online:
  • 8. Arguments for and against online rating Why it’s bad Why it’s good 1. Strain doctor-patient 1.Doctors can often be poor relationships judges of their patients’ experience 2. Reviews may be malicious or fake 2.Feedback changes doctors’ performance 3. Selection bias by those leaving reviews 3.People are using the Internet to voice opinions, 4. Lack of meaningful data on so why not capture this technical quality of health information in a useful form care
  • 9. Research Question Is there a relationship between ratings online and ‘hard’ measures of quality?
  • 10. The opportunity for a natural experiment NHS Choices website allows patients to review all NHS services in England
  • 11. Method: Obtained national review data from the Department of Health
  • 12. We thought we could compare ratings with some traditional measures of quality Vs.
  • 13. Average scores for 10,000 ratings Mean rating (out of 5) Score Range The environment where I was treated was… 3.6 2.6-5.0 I was treated with dignity and respect by the hospital staff… I was involved with decisions about my care… The hospital staff worked well together… 4.0 2.7-5.0 3.8 2.4-5.0 4.1 2.9-5.0 Greaves et al. BMJ Quality and Safety, 2012
  • 14. Average scores for 10,000 ratings Score The environment where I was treated was… Mean rating (out of 5) 3.6 Range 2.6-5.0 I was treated with dignity 4.0 2.7-5.0 and respect by The majority of ratings are positive • the hospital staff… • 67% would recommend to a friend I was involved with 3.8 2.4-5.0 decisions about my care… The hospital staff worked 4.1 2.9-5.0 well together… Greaves et al. BMJ Quality and Safety, 2012
  • 15. Ratings compared to patient experience surveys NHS Choices Measure Survey question: Spearman Rho p value Proportion of patients recommending “Overall, how would you rate the quality of care you received” 0.40 <0.001 Rating of being treated with dignity and respect “Overall, did you feel you were treated with dignity and respect while in hospital?” 0.33 <0.001 Rating of staff working together “How well would rate how well the doctors and nurse worked together?” 0.32 <0.001 Rating of cleanliness “How clean was the hospital ward or 0.48 room you were in?” <0.001 Greaves et al. BMJ Quality and Safety, 2012
  • 16. Ratings compared to patient experience surveys NHS Choices Measure Proportion of patients recommending Survey question: “Overall, how would you rate the quality of care you received” Spearman Rho 0.40 There is a moderate, highly Rating of being “Overall, did you feel you were 0.33 significant association between treated with dignity treated with dignity and respect ratings online and large and respect while in hospital?” surveys of patient experience Rating of staff working together “How well would rate how well the doctors and nurse worked together?” p value <0.001 <0.001 0.32 <0.001 Rating of cleanliness “How clean was the hospital ward or 0.48 room you were in?” <0.001 Greaves et al. BMJ Quality and Safety, 2012
  • 17. Ratings compared to outcomes NHS Choices Measure Spearman Rho Other variable Proportion of patients Hospital Standardised Mortality recommending Ratio p value -0.20 0.01 Proportion of patients Standardised morality rate for high -0.22 recommending risk conditions 0.01 Proportion of patients Standardised morality rate among -0.00 recommending surgical inpatients with serious treatable complications Proportion of patients Emergency readmission rate -0.31 recommending within 28 days 0.99 Patient perception of cleanliness <0.001 Rate of MRSA bacteraemia (per 1,000 bed days) -0.30 <0.001 Greaves et al. Arch Int Med, 2012
  • 18. Ratings compared to outcomes NHS Choices Measure Spearman Rho Other variable Proportion of patients Hospital Standardised Mortality recommending Ratio p value -0.20 0.01 Proportion of patients Standardised morality rate for high -0.22 recommending There is a weak, significant risk conditions 0.01 association between ratings Proportion of patients Standardised morality rate among -0.00 recommending online and mortality rates surgical inpatients with serious 0.99 treatable complications Proportion of patients Emergency readmission rate recommending within 28 days -0.31 <0.001 -0.30 <0.001 Patient perception of cleanliness Rate of MRSA bacteraemia (per 1,000 bed days) Greaves et al. Arch Int Med, 2012
  • 19. A study in the US found the same results comparing Yelp reviews with the HCAHPS survey
  • 20. A wisdom in the crowd of patients? Sir Francis Galton won ‘guess-the-weight-of-the-bull’ competitions by asking lots of local farmers for their best guess, and then taking the average – he gave us the idea of a ‘wisdom of crowds’
  • 21. Comparison of the NHS Inpatient Survey and ratings on the NHS Choices website NHS Inpatient Survey NHS Choices ratings Mechanism Paper-based survey Ratings left on a website Number of responses Selection 69,000 per year 5,000 per year Self-selecting; patients are not solicited Proportion positive Random; patients receive a survey requesting completion after leaving hospital 79% rated their overall care as excellent or very good Cost Likely more expensive Likely less expensive 67% would recommend to a friend
  • 22. Comparison of the NHS Inpatient Survey and ratings on the NHS Choices website NHS Inpatient Survey NHS Choices ratings Mechanism Number of responses Selection Proportion positive Cost Paper-based survey Ratings left on a website There were 10,000 hospital ratings inper year 69,000 per year 5,000 the UK over 2 years Random; patients receive a Self-selecting; patients are Over the same time not were survey requesting period there solicited 29,118,009 hospital completion after leaving admissions hospital 79% rated hospital admissions are rated 0.04% of their overall care 67% would recommend to a as excellent or very good friend Likely more expensive Likely less expensive
  • 24. Where are people rating their care?
  • 25. Associations between whether a practice is rated with population and practice characteristics Independent variable Z statistic p value Practice population size IMD score of patients Population density 15.38 <0.001 -7.82 <0.001 6.72 <0.001 Singlehander Proportion of population aged over 65 years Proportion of population who are white Type of contract Training practice -4.50 -3.88 <0.001 <0.001 -1.58 0.11 -0.71 0.35 0.48 0.73
  • 26. Associations between whether a practice is rated with population and practice characteristics Independent variable Z statistic p value 15.38 <0.001 Practice population size -7.82 <0.001 IMD score of • Practices serving younger people patients <0.001 Population more6.72 likely to be rated density are • Practices serving less deprived people Singlehander -4.50 <0.001 -3.88 Proportion are more likely to be rated <0.001 of population aged • Practices in urban areas are more over 65 years 0.11 Proportion likely -1.58be rated of to population who are white Type of contract -0.71 0.48 Training practice 0.35 0.73
  • 27. Cloud of patient experience ?
  • 29. A new trend: ‘Big Data’ and social media analytics
  • 31. Another natural experiment We can compare patients’ free text descriptions of care with their own quantitative ratings
  • 32. What we did: Machine learning • You need to teach an algorithm how to recognise particular words and phrases • We used all comments and ratings from 3 years as a training set (13,802 comments) We tried to predict patients ratings of their care from their comments in 2010 • Whether the patient would recommend the hospital or not • Whether the hospital was clean or not • Whether the patient was treated with dignity or not Used open source Weka software
  • 34. Simple punctuation contains little information (
  • 35. But when paired with other punctuation, it’s very important :(
  • 36. And small changes can have big effects on sentiment :)
  • 37. Some words the algorithm thought were important Overall • rude • excellent • hours • pain • communication Cleanliness • dirty • floor • filthy • bed • blood Dignity • rude • told • thank • friendly • attitude
  • 38. How good is it at predicting ratings? Question we are predicting the answer to Total number Prediction Kappa P value of comments accuracy Would you recommend the hospital? 6412 88.7 0.75 <0.0001 How clean was the hospital room or ward that you were in? 6139 81.2 0.40 <0.0001 Overall, did you feel you were treated with respect and dignity while you were in the hospital? 6239 83.6 0.56 <0.0001
  • 39. How does this compare to patient surveys? Patient Survey Question Machine learning prediction Spearman rho Overall, how would you rate the care you received? Whether the patient would 0.46 recommend the hospital Probability P<0.001 Overall, did you feel you were treated with respect Whether the patient was treated with dignity and dignity while you were in the hospital? 0.50 P<0.001 In your opinion, how clean was the hospital room or ward that you were in? 0.37 P<0.001 Standard of cleanliness
  • 40. Limitations • Selection bias • Sarcasm / Irony • Culturally specific  Phrases like ‘cup of tea’ important, but effect depends on context
  • 41. The UK quality regulator now includes online reviews and comments in its performance measurement framework
  • 42. The NHS has started publically reporting social media sentiment
  • 43. Conclusions Online rating and reviewing are on the rise Evidence of some wisdom in the crowd of patients Substantial professional resistance to the idea Selection bias in those commenting online The ‘cloud of patient experience’ is a potentially valuable source of information to understand patient views
  • 44. Thanks to my co-authors, and funders • • • • • • • • • Christopher Millett Ara Darzi Dominic King Henry Lee Utz Pape Liam Donaldson Azeem Majeed Robert Wachter Daniel Ramirez-Cano