These slides are from a presentation for the Nuffield Department of Primary Care Health Sciences. This covers reasons for capturing outcomes, challenges with commissioning and how our system of managing risks address these problems.
2. Inside Outcomes Purpose
• Improve Commissioning
• Use data to create social value
•Use data to describe the areas we live in
•Using communities to design services
• Relating what happens to policy
3. Challenges With Commissioning
• The difference between commissioning and procurement
• Co-designing services is more expensive
• What if the public don’t agree?
• Where do you get your data from?
4. What do we commission against?
• QOF Data
• Open Exeter Data
•National Data Sets (Atlas - life expectancy, infant mortality)
•Local data sets?
•Mosaic?
• Community data on service design?
5. There are two types of outcome
1. Systemic outcomes – e.g. A&E attendance, GP appointments
2. Individual outcomes – measurable improvement in health and
social situation
6.
7. Challenges with Measuring Outcomes
• Causality – Did you cause that outcome?
• Complexity – Especially in preventative services
• Why are you doing it?
• Contract specific outcomes can change a service
• Place vs service
8. Issues with data collected by services
• Compliance in collection
• Accuracy in collection
• Consistency in collection
• No closing the feedback loop
• Transactional in nature
9. All organisations collect four types of data
Demographics – Data that identifies individuals
Activity – What has happened to an individual
Outcome – What benefits, or disbenefits, have been received
Satisfaction – How happy the individual is
10. Name Definition Examples Strengths Weaknesses
Demographic data The identifying
factors for
individuals
Gender
Age
Ethnicity
Can help to measure how
representative of a community
a service is
Can be used as a comparator
for outcome data
On its own it is not very useful
data
Ease of collection can result in
excess collection
Data protection issues
Activity data A measurement of
the inputs provided
by a service
Number of people that have
used a service
Number of referrals (in and
out)
Number of sessions carried
out
Easy to measure
An important element in
calculating your costs
More of a measure of how
busy a service is rather than
how effective
Not a measure of quality
Outcome data A measurement of
the change in an
individual
Clients that have given up
smoking
Clients that have lost weight
Clients accessing entitled
range of benefits.
Much more focus on the
person receiving the service
A measure of the quality of the
service you provide
Can be used to compare with
other services
Can be hard to measure
Requires measurement at two
points
Satisfaction data Perception of the
intervention
Client satisfaction surveys Satisfaction is important in
assessing if people will return
to a service
Can be used as a basis adding
a personal element to
reporting
Inherently subjective
Not comparable inside an
organisation let alone with
other organisations
People liking a service doesn’t
mean it is a good service
11. Using Risk Maps – Managing Risk, Reducing Inequalities, Demonstrating Impact
• Performance management
• Consistent measurement
• Aggregated data for collective impact
• Auditable outcomes aligned to national frameworks
• Measures by risk and protective factor
• Evidence based
• Quality assured
12. Since 2012 there has been a push from Government to build
outcome frameworks
Services should be commissioned against outcome frameworks
Outcome frameworks often overlap
Outcome frameworks often contradict
Outcome frameworks
14. Data Collection is Structured to Match the Life Course
Starting
Well Data
Dictionary
Developing
Well Data
Dictionary
Working
Well Data
Dictionary
Living Well
Data
Dictionary
Ageing Well
Data
Dictionary
Diabetes
Data
Dictionary
But also service specific
Mental
Health Data
Dictionary
Supported
Housing Data
Dictionary
End of Life
Data
Dictionary
Domestic
Abuse Data
Dictionary
We have identified 93 common risks and issues.
Each has been defined and is monitored for any change in policy.
We are adding to this list all of the time.
15. Living Well Data Dictionary
Personal Circumstances:
• Domestic Abuse
• Homeless
• Temporary Accommodation
• Unsuitable Accommodation
• Vulnerable Adult
• Financial Hardship
• Social Isolation - Loneliness
• Environment - Noise
• Environment - Outdoor Spaces
Behaviour:
• Very Low Fruit & Vegetable Intake
• Low Fruit and Vegetable Intake
• Significant Fried and Processed Food
Intake
• Excessive Sugar
• Nutrition - Iron
• Physical Activity - Moderately
• Physical Activity - Inactive
• Alcohol Misuse
• Smoking
• Substance Misuse
Status:
• Weight - Overweight
• Weight – Obese
• Mental Health – Low Reported
Wellbeing
• Mental Health - Stress and Anxiety
• Sexual Health - Unwanted Pregnancy
• Sexual Health – Sexually Transmitted
Infections
• Pre - Diabetes: Non - Diabetic
• Screening - Increased Blood Pressure
• Screening - High Blood Pressure
16. Exampleple
Tracey:
● In debt
● Socially isolated
● Lives in a hostel
● Been to see GP 7 times in 3 months
● Stressed and anxious
● Attended A & E on two occasions with alcohol related
issues
● Smokes
● Misuses alcohol
● Poor diet
● No exercise
17.
18. Household
income is
>60% of UK
average
Reduce
households
where
neither
parent is in
work
Healthy Child
Programme
The family can
afford food
and clothing
items
Social Justice Outcomes
Framework
Department of Health
Department of Work and
Pensions
Financial Hardship
After
required fuel
costs the
family
remains
above the
poverty line
Improving
Outcomes
Supporting
Transparency
Reduce the
proportion of
those on
work-related
benefits
The number
of working
age adults
engaged in
work related
activity
19.
20.
21. Our Main Challenges
• Outcomes are not a priority
• Limited headspace to change
•Services focus on what they’re good at
•Measuring activity is easy
•Transactional data puts the focus on the service rather than community
• Who owns the outcomes?
•Risk aversion
22. Next Steps
•Developing an open standard
•Integration with other systems
•Promoting what can be measured
• Seeing data as an end in itself