Polyclinics Learning

1
Polyclinics Learning Development
26 June 2008

• Workshop 4

2
Polyclinics Learning Development
26 June 2008

• Financial and Commissioning Models

3
Why do we need this workshop?

• Polyclinics are a new type of service model in
  the NHS
• World Class Commissioning places new requirements
  on PCTs for information driven commissioning
• The tools and processes are currently not
  available to PCTs to provide this

4
Support is being provided by HfL to assist PCTs
with Commissioning

• HfL is providing tools to assist and support PCTs
  in answering key commissioning questions
• These models are being launched today
• Further detailed technical support will be
  provided in more technical workshops
• Models have been tested with some PCTs during
  development
• Models have been designed to meet local
  requirements by being flexible.

5
Six questions to answer when commissioning
Polyclinics

• What services should be commissioned in
  polyclinics?
• What levels of activity should be commissioned in
  polyclinics and how should these activities be
  defined and measured?
• What is a realistic price to pay for this
  activity at a polyclinic?
• Does the proposed polyclinic price represent
  value for money compared with existing provision?
• What is the overall financial impact of
  introducing polyclinics
• What is the overall cost of our commissioning
  plans and is this affordable?

6
Draft workshop agenda

• 1. Introduction 11.00 - 11.15
• 2. Service Lines
• Overview of Service Lines 11.15 11.30
• Small group discussion 11.30 12.00
• 3. The commissioning model
• Conceptual overview 12.00 12.15
• Group sessions case study 12.15 12.45
• Presentation from Tower Hamlets 12.45 13.00
• QA 13.00 13.15
• Lunch 13.15 14.00
• 4. Packages of care
• Definition 14.00 14.10
• Group discussions for a sample sub-service
  line 14.10 14.30
• 5. The financial model

7
What we will cover today
8
An approach to commissioning in polyclinics
Proposed solution for polyclinics
Current situation in non-acute care
Acute system
What is commissioned

• HRG
• Dominant FCE in spell

• Service lines/packages of care

• No agreed currency for community care

How much? (volume)

• Prior year activity

• Commissioning model

• Activity not always tracked

What is the cost of delivering the service?

• Reference cost Index (not seen by commissioner)

• Financial model

• Not broken down by activity

What price do I pay?

• PbR tariff

• Block contracts

• Financial model

9
Draft workshop agenda

• 1. Introduction 11.00 - 11.15
• 2. Service Lines
• Overview of Service Lines 11.15 11.30
• Small group discussion 11.30 12.00
• 3. The commissioning model
• Conceptual overview 12.00 12.15
• Group sessions case study 12.15 12.45
• Presentation from Tower Hamlets 12.45 13.00
• QA 13.00 13.15
• Lunch 13.15 14.00
• 4. Packages of care
• Definition 14.00 14.10
• Group discussions for a sample sub-service
  line 14.10 14.30
• 5. The financial model

10
Through this work we identified 12 service-lines
11
The service lines can be split into sub-service
lines
12
and further into packages of care
13
Example packages of care for planned care
8
New
New
New
New
Follow-up
New
Follow-up
Follow-up
New
Follow-up
Follow-up
New
Follow-up
Follow-up
New
New
Follow-up
Follow-up
New
Follow-up
New
Follow-up
New
Follow-up
Cardiac consultation
New
Oncology consultation
Follow-up
Neurology consultation
Rheumatology consultation
New
Follow-up
General surgery consultation
Endocrinology consultation
New
Gynaecology consultation
Dermatology consultation
Follow-up
Trauma orthopaedics consultation
New
Follow-up
Other surgery consultation
Other medicine consultation
ENT consultation
New
Follow-up
New
Respiratory medicine consultation
Sexual health
Ophthalmology consultation
Gastroenterology consultation
Follow-up
Urology consultation
General medicine consultation
New
Sexual health
Routine care (other)
Oral surgery consultation
Follow-up
Other consultation
Routine care (other)
Planned care
Procedures
Injury
Minor surgical procedures
Long-term conditions
Unplanned care
Joint-injections
Illness
14
Example packages of care for long-term conditions
8
TB
HIV
Musculo-skeletal
Learning disabilities
Severe
Other disabilities
Moderate
Mild
TB
HIV
Chronic kidney disease
Severe
Moderate
Mild
Asthma
Severe
Musculo-skeletal
Long-term neurological conditions
Learning disabilities
Other disabilities
Moderate
COPD
Mild
Pain Manage-ment
Severe
Chronic kidney disease
Diabetes
Moderate
Mild
Long-term neurological conditions
End of life care
Severe
Pain Management
Heart failure
Moderate
Mild
End of life care
Other
Coronary heart disease
Other
Mental health
Long-term conditions
Long-acting reversible contraception IUCD
fitting
Terminations
Depression/anxiety
Older people / Complex
Procedures
Vasectomies
Psychoses and other mental health disorders
15
Why use service lines?
Service lines
1

• allow us to define packages of care including
  interventions required in alignment with clinical
  evidence base
• facilitate clinical leadership and encourages
  integration of care
• act as the basis for measuring quality and
  productivity in a systematic way
• allow alignment of resources with patient need
• reflect a practical level of detail given the
  current environment of block budgets and limited
  payments-per-service

2
3
4
5
16
In Trusts, service lines lie between the
directorate (division) and HRG (activity) level
they are the natural business units

• Should follow natural structure of the
  organisation and be self-contained units
• Number of service lines needs to be tied to the
  natural structure and operations of the
  individual organisation there is no hard and
  fast rule
• All sub-parts of the service line should have
  aligned objectives and common performance
  measures
• Where possible service lines should be comparable
  in size and complexity

Source Monitor
17
Monitor have found that service line management
can also unlock the potential of hospital staff
From
To

• Its not my job

• I feel ownership of my service and can make
  improvements to how service is delivered.

Clinicians
Managers

• Its not my fault..

• I feel that I am a true thought partner for
  clinicians and help provide the management
  expertise to ensure decisions about the service
  are well informed

• We never have time to think about bigger
  strategic issues as we are involved in day to day
  operational details

• We have focused our agenda on those elements
  that we are uniquely positioned to contribute to

Boards/Strategy/ CEOs
We would like to achieve the same results in
Polyclinics
Source Monitor
18
In Germany, activity data is tracked and
aggregated up to service-lines for reporting
purposes
German Community Care System
Activity is grouped up into service-lines
Visits are the unit of measure used for codifying
and tracking activity
19
However, in the UK patient level and resource
group information is not currently widely
available so service lines are the natural
starting point
Polyclinic
Division
Service-line
Healthcare resource group (HRG)
Patient level
20
Service lines are the most practical solution to
delivering a patient focussed service given
limitations of information availability
Promotes
Hinders
Basis of splitting services
Chosen option
Factors for consideration when commissioning a
polyclinic
Individual Activities
Service lines
Resources
1
Defining packages of care commissioned
Commissioning along service lines enables the
polyclinic commissioning criteria to be met and
assessed
2
Clinician leadership and integration of care
3
Measuring quality and productivity
4
Alignment of resources with patient need
5
Reflect practical level of detail
21
To develop service-lines for non-acute activity
in London we used a combination of top-down and
bottom up analysis

• Interviewed PCTs and provider arms to determine
  how current activity is measured and tracked

• Worked with clinicians to review existing
  activity and classify it bottom-up

Service line split

• Examined current literature and published
  statistics to determine what data is available
  and what current groupings are used

• Tested and refined draft the service-line splits
  with commissioners, providers and clinicians in
  London, nationally and internationally

22
Polikum in Germany organise service in very
similar way to service lines proposed for
polyclinics
Activities
General practice services
Community services
Most outpatient appointments (including
antenatal/postnatal care)
Minor procedures
Urgent care
Diagnostics point of care pathology and
radiology
Interactive health information services,
including healthy living classes
Proactive management of long term conditions
Pharmacy
Other health professionals, e.g., optician,
dentist
Source Polykum, Germany
23
As do the family health centres of the Cleveland
Clinic in the US

• Primary Care
• Family Medicine
• Internal Medicine
• Obstetrics/Gynecology
• Maternal Fetal Medicine
• Pediatrics
• Surgery
• Colorectal Surgery
• General Surgery
• Outpatient Surgery Centre
• Plastic Surgery
• Vascular Surgery
• Community services
• Nutrition Services

• Diagnostics
• CT Scan
• DEXA (Bone Density Testing)
• Echocardiography (Echo, Stress Echo)
• Gastrointestinal Upper Lower GI
• General Radiology
• Genital/Urinary
• Mammography
• MRI
• Nuclear Medicine
• Total Body CT
• Ultrasound

• Specialty Care
• Allergy/Immunology
• Anesthesiology
• Audiology/Hearing Aid Services
• Breast Center
• Cardiology
• Dermatology
• Gastroenterology
• Infertility
• Medical Oncology/Hematology
• Neurology
• Orthopaedics
• Otolaryngology (Ear, Nose and Throat/ENT)
• Pain Management
• Pediatric Specialty
• CarePulmonology
• Rheumatology
• Urology

• Other support services
• Pharmacy
• Laboratory Services

Source Services Brochure for the Family Health
and Surgery Centre, Beachwood,
24
Review of service lines
25
Discussion groups for service lines

• Questions for discussion in groups
• What is your first impression of Service Lines?
• How do you think using Service Lines will help
  you commission polyclinics?

26
Place the 7 missing sub-service lines into the
appropriate service line
27
Service line exercise - Answers

• Sub-service line
• Smoking cessation
• Chronic kidney disease
• New dermatology consultation
• Minor injury
• Antenatal care
• Falls service
• Sickness note/ income benefits advice

• Service line
• Wellness
• Long term conditions
• Planned care
• Unscheduled care
• Reproductive health
• Older people/ Complex needs
• Social care

28
Draft workshop agenda

• 1. Introduction 11.00 - 11.15
• 2. Service Lines
• Overview of Service Lines 11.15 11.30
• Small group discussion 11.30 12.00
• 3. The commissioning model
• Conceptual overview 12.00 12.15
• Group sessions case study 12.15 12.45
• Presentation from Tower Hamlets 12.45 13.00
• QA 13.00 13.15
• Lunch 13.15 14.00
• 4. Packages of care
• Definition 14.00 14.10
• Group discussions for a sample sub-service
  line 14.10 14.30
• 5. The financial model

29
We have developed a tool which enables the
commissioner to calculate the level of health
services to be commissioned based on population
need
30
To understand patient need for a given health
activity, we must first understand what drives
the need
Activity associated with patient need
Driver of activity
Prevalence of diabetes
Diabetes services
Prevalence of incontinence
Continence service
Births
Post natal care
Smoking rates
Smoking cessation
31
QOF data cannot be used as a proxy for need, as
it frequently underestimates levels of disease
burden

• Obesity
• National prevalence rates 23 higher than TH QOF

• CHD
• National prevalence rates 12 higher than TH QOF

• Depression hypertension
• National prevalence rates 9 higher than TH QOF

Source Decision Resources PatientBase (2006)
Tower Hamlets QOF disease prevalence rates
2006-2007Humphreys, Improving the detection of
COPD, HSJ Apr 2008. Lister, Fats spreading as
quarter of Britons now register obese on the size
scale, The Times Dec 2005. 2003 Health Survey
for England.
32
The commissioning model allows calculation of
activity needed based on actual incidence rates
adjusted for local conditions
Incidence/ prevalence of sub-service line
activity driver
Expected patients for this driver

• Use local or national incidence data, or national
  data adjusted for local population
• Can have gt1 driver for any sub-service line

Expected patients in this sub-service line
Total activity needed
Segment share
Activity per patient (package of care)

• E.g., mild CHD sufferers as a proportion of total
  CHD sufferers

Population being modelled

• E.g., 10 visits per year

33
Example Severe Asthma
Driver is prevalence of asthma, which is 5.8
nationally
Incidence/ prevalence of sub-service line
activity driver
Expected patients for this driver
5.8
Expected patients in this sub-service line
2,900
1,450
Population being modelled
Total activity needed
Segment share
50,000
8,700
Activity per patient (package of care)
50
6
Represents average number of visits a patient
will have (more in next section)
50 of all asthma sufferers are considered to
have severe symptoms
34
Activity to be commissioned can be calculated in
three ways
Description
You would use this approach when

• Apply local or national prevalence rate data for
  the driver

• Good quality local or national data exists for
  the driver

• Use national data and amend for local
  demographics (e.g. age, ethnicity) and risk
  factors (e.g. obesity, smoking)

• Good quality national data exists, but does not
  reflect risk level of local population

• Use prior year activity data as a proxy for need

• No good quality data exists, or levels of
  commissioning will not be changed

This could be due to funding issues or due to
the driver in question not being a priority area
for the PCT
35
Case study exercise on commissioning for need

• For each case study discuss
• What are the drivers of activity
• Sources of information on the drivers
• What adjustments you might want to make for local
  calculations
• Feedback
• How did you find this?
• What parts would you find most difficult?

36
Case study Mild CHD
Incidence/ prevalence of sub-service line
activity driver
Expected patients for this driver
Expected patients in this sub-service line
Total activity needed
Activity per patient (package of care)
Segment share
Population being modelled
50,000
37
Case Study Minor Illness
Incidence/ prevalence of sub-service line
activity driver
Expected patients for this driver
Expected patients in this sub-service line
Total activity needed
Activity per patient (package of care)
Segment share
Population being modelled
50,000
38
Case Study Falls service
Incidence/ prevalence of sub-service line
activity driver
Expected patients for this driver
Expected patients in this sub-service line
Total activity needed
Activity per patient (package of care)
Segment share
Population being modelled
50,000
39
Case studies - CHD statistics
24.0
ILLUSTRATIVE
All data is for illustrative purposes only and
should not be used for decision-making
Prevalence of CHD,
The majority of my patients have relatively mild
symptoms. Last month, for instance, of the 20
patients I saw, 17 of them could be referred to a
specialist nurse without need for further
treatment Consultant
QOF
Patient base
Men
Women
HES
CHD patients should be monitored weekly in
community care. Following initial diagnosis by a
consultant, I would expect most service users to
receive up to 15 visits a year, although low
complexity cases may be managed in 35 visits per
year - Specialist nurse
40
Case studies - Minor illness statistics
23
ILLUSTRATIVE
All data is for illustrative purposes only and
should not be used for decision-making
I would say that on average, each person has a
minor illness 1.6 times a year - local GP
GP consultations by type,
Maternity/child care
Single visits
48
Long term conditions/repeat visits
Uk population 60 million Source National
Audit office Hospital activity statistics data
QResearch
41
Falls service information
ILLUSTRATIVE
All data is for illustrative purposes only and
should not be used for decision-making
Falls patients at local provider arm (population
110,000)
As an occupational therapist at the Falls
service, I see quite a broad range of patients.
About half of them are over 60, but many are
quite young. We see a lot of people with
disabilities and the service is exceptionally
valuable to them Sanjeev, OT
No. of patients
1987
97
2007
I recently helped my mother move to a bungalow
she had fallen several times in our 2-storey
housie and her recovery took 3 months. Between
physiotherapy, group rehabilitation sessions and
occupational therapy, I was driving her to the
hospital every week! Gemma, Falls patient
relative
42
PCT data
33
ILLUSTRATIVE
All data is for illustrative purposes only and
should not be used for decision-making
Local rate
National rate
Obesity,
Population 210,000
Demographics
Women
Men
Smoking rates,
Men
Women
43
Commissioning model exercise sample answers
Mild CHD
Minor illness
Falls service
Incidence/prevalence
To be decided by teams
160
Segment share
15
100
Activity per patient
4
1
Activity needed
Depends on incidence selected
80,000
44
NHS London Polyclinics Learning and
DevelopmentProgrammeLearning from Primary
Care Diagnostic in Tower Hamlets PCT
CONFIDENTIAL
26th June 2008
45
Service lines in Tower Hamlets? The Case for
Change

• History of very poor scores in patient
  satisfaction in all surveys Ipsos MORI, Picker
  diabetes
• High Bangladeshi population 33, of whom 30
  dissatisfied with NHS
• Worst scores in England on access 2007
• Not under-doctored overall 220 on performers
  list
• Access study showed many system inefficiencies
  and failure to use skill mix
• Health Inequalities work suggested high unmet
  need based on poor outcome data and very high
  SMRs
• History of siege mentality in practices 2004
  12,000 patient assignments, lists open but
  closed, population growth driving procurement of
  new practices, high patient turnover
• Current spend just under 10 total, ambition to
  move to top quartile - ? What investment and
  could primary care cope
• 80 90 patient experience in primary care not
  hospitals care closer to home could increase
  that BUT primary care must be fit for purpose

46
(No Transcript)
47
(No Transcript)
48
Can PCTs achieve great access when they have
challenging populations?
Some English PCTs have very different access
performance despite similar deprivation profiles
i
100
J
M
E
Q
S
G
L
O
48 hr access
F
I
80
H
C
R
D
P
K
B
N
A
T
60
20 most deprived PCTs
Index of multiple deprivation
A Sandwell B Camden C Heywood, Middleton
and Rochdale D Bradford and Airedale E Hartlep
ool F Southwark G Hull H Haringey I Birmin
gham EN J Salford K Newham L Nottingham
City M Middlesbrough N Islington O City and
Hackney P Heart of Birmingham Q Knowsley R M
anchester S Liverpool T Tower Hamlets
Source IMD Data 2004 taken from MORI Poll 2007
Access Mori poll, 2007
Some English PCTs have very different access
performance despite similar age profiles
ii
100
XXX versus other PCTs
48 hr access
80
60
Practice population under 5 or 65 and over
Source ADS 2006, reconciled to ONS mid-2005
estimates for LAs Access Mori poll, 2007
XXX has a higher proportion of one particular
ethnic minority than elsewhere in London. But
PCTS with smaller and similar ethnicity profiles
have very different access performance
iii
100
48 hr access
80
BROADER PICTURE
60
47
9
8
7
6
5
4
35
34
3
2
1
0
of population of this ethnicity
Source LHO-compiled, based on 2001 census
figures. 33 London PCTs included Access Mori
poll, 2007
49
What patients see at your practices
WORKING DRAFT
Information is more clear
Information is not clear
Not clear that practice is actually open
Door has an open sign
Language on the touch screen is not clear.
Welcome message is not displayed in alternative
languages
Touch screen is ineffectively placed as
reception is in patients eye-line when entering
practice, they head straight for the desk on
arrival
50
How often do patients see a GP?
1.0
How often do patients see their GP? Average
number of visits per year
Practice
Some views we have heard
O
More often than England average
M
Here in XXX our average visits per year are a
lot lower than in the rest of YYY The key
reason why access is poor is that our patient
come to us a lot more often than is normal for
London My patients come 13 times per year
L
T
J
Are these opinions in line with the facts?
C
P
H
D
Less often than England average
F
K
S
A
V
B
E
England average 3.25
49
Includes patients who visit 0 times in a
year Source Extracts from practice EMIS
systems, Q2 and Q3 2007 numbers extrapolated to
give per year numbers England average GP
consultation rate 2006 from Trends In
Consultation Rates 1995 to 2006, The Information
Centre team analysis
51
When can I see a GP?
WORKING DRAFT
GP availability on a typical Monday
A GP is available
A GP is not available
Practice
800
900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
Opens relatively late, shuts relatively early,
relatively short lunch
A
B
Impossible to see GP for 4 hrs in the middle of
the day
C
D
E
H
J
M
P
S
U
Note Timetable data reflects our current
understanding of what availability is like on an
average Monday, excludes extended hours provision
Source Team interviews during practice visits
52
How many appointments do you have?
2,000
WORKING DRAFT
These practices appear to have many more
appointments per 1000 patients than their peers.
Why might this be?
How many slots are offered per GP WTE per week?
O
How many slots are offered per 1,000 patients per
week?
L
P
D
48 Hour Access performance
54
O
J
10 fold diff in slot/WTE prob at least partly
driven by data quality issues at these practices
S
61
L
C
55
F
P
K
48
D
T
M
66
J
H
A
58
Overall, there is high variation in the number of
slots offered between practices. What could
account for this? How do you decide what the
right number is for your practice?
S
B
69
C
E
69
F
41
K
Text
Text
How many patients per GP WTE do practices have?
44
T
53
M
1,840
O
1,858
L
57
H
1,778
P
57
1,567
D
A
1,976
J
57
B
1,640
S
2,229
C
42
E
1,587
F
1,587
K
1,619
T
1,809
M
These practices appear to have many fewer
appointments per 1000 patients than their peers.
Why might this be?
2,238
H
1,530
A
1,908
B
2,000
E
Source Extracts from practice EMIS systems, Q2
and Q3 2007 PCT data MORI poll. Note
Practices excluded if one or more pieces of data
missing
53
Do your patients tend to come back often?
WORKING DRAFT
Example analysis for Practice O
48 of GPs time is spent with the 80 of
patients who have come 1 to 9 times so far this
year
36 of GPs time is spent with the 16 of
patients who have come 10 to 20 times so far this
year
16 of GPs time is spent with the 4 of patients
who have come more than 20 times so far this year
Amount of GPs time spent with this group of
patients
Number of times that patients have attended so
far in 2007
Source Practice EMIS extracts YTD 2007
54
Why Service Lines?

• Observation that clinicians spend a lot of time
  on tasks inappropriate for their skills (and
  costs)
• Ineffective and inefficient use of skill mix
  with more scope for delegation and substitution
• Many types of presentation share common
  requirements for clinical skills
• Change to units of clinical skill plus relevant
  admin rather than consultations or
  presentations
• A GP consultation of 10 minutes plus admin
  backupA Practice nurse consultation of 20
  minutesA HCA contact of 15 minutesGP Nurse
  HCA
• Care planning approach from Diabetes UK promotes
  self care and high quality consultation but may
  be less time 1 x 45 minute GP, 1 x 30 minute
  HCA instead of 6 x 10 minute Band 7 nurse 2 x
  10 min GP

55
Top tips for QOF/register use include the
following

• Compact, regularly updated manual for practices
• Strong challenge and validation of QOF by
  multidisciplinary team, particularly of high
  claimants
• Dedicated PCT resources for targeted performance
  development ,particularly for low claimants
• Calculation of actual versus expected numbers on
  registers, and systematic recovery plans
• Strict criteria for exclusions, with audit.
  Supported plans to reduce legitimate causes e.g.
  patient non-attendance.
• Calculate staffing requirement for registration
  and full annual patient review. Workforce plan
  for most cost-effective skill mix.
• Use integrated governance to drive for standards
  higher than QOF, and/or consider use of Local
  Enhanced Service agreement
• Promote systems of effective medicines management
  and patient adherence to therapy with support
  tailored to vulnerable groups
• To improve sustainability, carry out intelligence
  based campaigns to ensure most cost effective
  prescribing linked outcomes.
• Draw QOF and prescribing data together, with
  relevant hospital usage measures, on a
  benchmarked practice dashboard of inputs and
  outcomes

56
Demographics do not drive QOF performance
Example correlation of QOF DM17 vs deprivation
for English GP practices
Practices with challenging demographics do not
typically use exception reporting to improve QOF
scores (or at least, no more than other practices
do)
Source National QOF scores 2007, IMD
57
GP Practice performance on QOF for screening
varies dramatically and some practices exclude a
high percentage of their diabetics
1.5
QOF percent screened
Percent excluded
Adjusted QOF Screened
5.7
Practice 1
Practice 2
29.2
Practice 3
20.9
Practice 4
10.8
18.1
Practice 5
Practice 6
3.1
5.9
Practice 7
Practice 8
4.1
Practice 9
8.3
18.4
Practice 10
Practice 11
6.2
15.8
Practice 12
Practice 13
8.2
Practice 14
3.6
Practice 15
8.3
14.5
Practice 16
Practice 17
10.4
14.3
Practice 18
Practice 19
9.1
Practice 20
3.7
17.0
Practice 21
Practice 22
10.6
Practice 23
10.9
Practice 24
16.4
Practice 25
4.5
Practice 26
7.3
Practice 27
8.6
Practice 28
12.1
10.8
Practice 29
Practice 30
4.9
Practice 31
16.3
Practice 32
7.1
Practice 33
7.1
4.7
Practice 34
Practice 35
16.0
Practice 36
7.6
Practice 37
1.5
XX Avg
Eng Avg
XX Avg
Eng Avg
XX Avg
Source 2006-2007 QOF Scores and QMAS exclusion
reporting.
58
Unmet Need

• Estimated to be total of 12,000 diabetics in the
  borough

59
(No Transcript)
60
(No Transcript)
61
(No Transcript)
62
MODEL PREDICTS TWICE AS MANY PATIENTS AS
CURRENTLY REGISTERED
25.3

• The PCT currently treats 128,700 patients on QOF
  disease registers
• Model predicts 257,800 patients for these
  diseases

2007-2008 actual
Patients (thousands)
Model projected
Smoking
Hypertension
Obesity
Depression
Asthma
Diabetes
CHD
COPD
CKD
Mental health
Other
Other includes epilepsy, palliative care,
cancer, dementia, learning difficulties, heart
failure, atrial fibrillation, hypothyroidism and
stroke palliative care and learning difficulties
account for majority of difference between QOF
and models predictions Source QMAS disease
register 2007/08
63
MOST DISCREPANCIES BETWEEN MODEL AND QOF CAN BE
EXPLAINED
4,3
CONFIDENTIAL WORKING DRAFT
TH patients affected by each condition
Most likely reason for discrepancy
QOF (06-07)

• Model prevalence in line with other estimates
  (22 according to Health Survey for England)
• QOF is likely missing significant portion of
  obese pts

5,7
Obesity
Model
14,6

• QOF only includes pts w/ at least 3 abnormal BP
  readings, whereas model includes all pts with BP
  140/90 mm Hg

8,0
Hypertension
16,6
1,2

• Reason for discrepancy is unclear QOF and model
  measure pts 18 yrs old w/ stage 3-5 CKD

CKD
7,9

• QOF is limited to pts on diabetes and/or CHD
  register w/ depression in past 15 months model
  follows DSM-IV diagnosis criteria (pts w/ 5
  symptoms in 2 week period)

5,6
Depression
9,7
1,2

• QOF is symptom-based and likely detects only
  severe cases (model 0.93), whereas model
  includes moderate and severe cases, both of which
• COPD is also often misdiagnosed as asthma, due to
  similar symptoms

On average, model predicts 1.48 times more
patients than are currently captured by QOF
COPD
3,4

• QOF only records pts w/ history of coronary
  artery procedures and myocardial infarction, but
  does not include angina

1,9
CHD
3,7

• QOF reports only pts w/ EKG-confirmed diagnosis
  in past 12 months, whereas model includes all pts
  currently being treated w/ drugs for history of
  heart failure

0,5
Heart failure
1,6
0,2

• QOF relies on GP perception of impaired function
  and may thus report significantly fewer cases
  than the model, which includes a range of disease
  severity according to DSM-III

Dementia
1,0
4,5

• Model is in line with QOF
• Small discrepancies may arise if QOF missed some
  mild cases

Asthma
5,0

• Model is in line with QOF

4,4
Diabetes
4,3
64
MODEL PREDICTS STAFF AND COSTS REQUIRED TO SERVE
TH POPULATION, BASED ON DEMOGRAPHICS, DISEASE
PREVALENCE BEST CARE ASSUMPTIONS
180
CONFIDENTIAL WORKING DRAFT
9,140
0-4
9,280
0.1
0-4
0.1
15
0-4
15
14,110
5-14
14,720
0.1
5-14
0.1
5-14
24
25
58,780
15-39
64,980
1.4
15-39
2.0
1,112
15-39
1,811
23,190
40-64
31,040
3.3
40-64
6.8
40-64
1,060
2,941
8,030
65-84
7,220
10.5
65-84
18.1
65-84
1,179
1,829
1,300
85
670
85
10.8
16.9
85
195
158
2,610
0
Diabetes

• Best package of care
• Annual appts per case 4.3
• Appt length 20 min
• Staff mix
• GP 25
• Practice nurse 25
• Other 50

Staff assumptions Annual clinical hrs 820 GP
salary 290,000 NP salary 34,400
( 000)
9
Diabetes
Other diseases
9
Other diseases
619
310
GP
NP
Other practice staff
Other staff
NP
GP
Assumes 7 clinical sessions/week, 3
hours/session, 40 weeks per year Source Annual
appts per case 2003 UK doctor consultations by
selected ICD10 groups, Compendium of Health
Statistics (2007) appt length Penny Dash GP
salary, Laing Buisson NHS Financial Report 2007
(FY05/06)
65
AS WE REFINE OUR ASSUMPTIONS, THE RECOMMENDED
INVESTMENT MAY CHANGE
Modification to current calculations

• Changes to best care assumptions will likely
  increase the staff resources required to serve
  each patient

• Model includes many non-QOF diseases (not
  included in todays discussion), which will
  increase the number of patient visits

• We will do a thorough check to ensure model does
  not double count individuals with two related
  conditions that should be treated in a single
  appointment
• E.g., diabetes diabetic nephropathy,
  osteoarthritis osteoporosis

• We will check diseases with unexpectedly large
  prevalence rates to ensure only symptomatic cases
  requiring active treatment are included

66
Positive signs of impact following access work
Year-on-year change
2006/07 and 2007/08 48 hour MORI performance by
practice
06/07
07/08
lt70
70 to 80
gt80
Source MORI 06/07 and 07/08
67
We have segmented diabetic patients by attitude
to self-management and education to identify
appropriate interventions
Type of barrier to change/attending education
Proactive learners
Resigned

• Ive had diabetes for a while, Im not going
  to change my lifestyle any more
• Patient may have very poor knowledge about
  diabetes and control
• Reticent attitude means often ignore advice,
  dont seek out information
• Greatest challenge is to get patients interested
  and motivated to learn by tackling psychological
  barriers first

Willing to change

• I like getting education and managing my
  diabetes is very important to me
• Patient tends to have good control and good
  understanding of condition
• Proactively seeks out information
• Practical issues are main barrier to attendance
  at HAMLET

Not yet willing to change
Rational barriers
Emotional barriers
Psychological/ medical barriers
Confident with unknown knowledge gaps
Confused and anxious
Ignorantly unconcerned

• I know enough, Im not really interested in
  learning more
• Patient feels sufficiently informed although may
  lack clarity/information on issues
• Consequently often moderate to low proactivity
• Diabetes self-management competes with other
  demands in their lives

• I dont know what I can eat sometimes or what my
  test scores mean . . .
• Patients recognise that they lack knowledge
• However may not be very proactive in seeking out
  help/information
• Making access to education easier and clinician
  encouragement needed

• Diabetes is not so bad really, lots of people
  suffer from worse so I think Im doing ok
• Low level of knowledge, patient doesnt realise
  could be better
• Low proactivity
• Challenge is getting patient to appreciate
  seriousness of diabetes and self-management

66
Source Team analysis from patient and clinician
interviews
68
WCC Competency 4 - Collaborate with clinicians

• Clinical engagement remains critical for success
• Attract to change shouting does not work
• Identify champions and share
• Organisational development key teamwork and
  management very variable
• Preserve continuity of care and human scale

69
Evolution towards Polyclinics
70
(No Transcript)
71
Systematic long-term condition management that
enhances the patient experience within a
polyclinic setting

• Long-term condition self care comprises
• An entitlement to a care plan and navigator to
  aid in better self management
• More choice over nature of care and where it is
  delivered
• Services feel more integrated, personalised
  responsive

Greater choice of service provider
More information on managing condition
Entitlement to a care plan
More use of assistive technology
Navigator to help choose services
From the patients perspective
 

• The model of long-term condition care
• Empowered and informed patients - fully engaged
  and informed (Expert) patients, with continuous
  support and resources offered to patients where
  necessary
• Team based care - a team comprising of a diverse
  workforce, covering GPs, specialist nurses and
  specialists, as well as those with expertise in
  areas like nutrition, psychology, and behavior
  modification the team works in collaboration to
  improve the health of the patient
• Comprehensive clinical guidelines - treatment
  decisions based on explicit, proven guidelines
  integrated into the day-to-day practice of the
  primary care providers. Follow up as part of
  standard procedure
• Real time patient data - all clinicians have
  centralized, up-to-date information about the
  patients status. Information system tracks
  individual patients , as well as populations of
  patients and profile risk
• Effective risk modeling and patient selection -
  intensity of support matched to an assessment of
  need

72
Draft workshop agenda

• 1. Introduction 11.00 - 11.15
• 2. Service Lines
• Overview of Service Lines 11.15 11.30
• Small group discussion 11.30 12.00
• 3. The commissioning model
• Conceptual overview 12.00 12.15
• Group sessions case study 12.15 12.45
• Presentation from Tower Hamlets 12.45 13.00
• QA 13.00 13.15
• Lunch 13.15 14.00
• 4. Packages of care
• Definition 14.00 14.10
• Group discussions for a sample sub-service
  line 14.10 14.30
• 5. The financial model

73
Draft workshop agenda

• 1. Introduction 11.00 - 11.15
• 2. Service Lines
• Overview of Service Lines 11.15 11.30
• Small group discussion 11.30 12.00
• 3. The commissioning model
• Conceptual overview 12.00 12.15
• Group sessions case study 12.15 12.45
• Presentation from Tower Hamlets 12.45 13.00
• QA 13.00 13.15
• Lunch 13.15 14.00
• 4. Packages of care
• Definition 14.00 14.10
• Group discussions for a sample sub-service
  line 14.10 14.30
• 5. The financial model

74
What is a package of care?
ILLUSTRATIVE
Service
Example of care
3 visits per annum, 1 with a GP, 1 with practice
nurse, 1 with optometrist. Includes regular blood
tests and optometry tests
4 visits, 1 with a specialist nurse, 3 with
community nurses
5 visits from mid-wife at home
12 visits, 2 with GP, 10 as part of group session
taken by health trainer
75
In defining a package of care, it is important to
define all resource needs

• For each care pathway, define the types of visits
  first, follow-up, etc
• For each visit type, define the number of visits
  and average length of each visit
• Proportion of patients requiring each visit type

• For each type of visit, define the resource
  requirements in terms of
• Clinical staff
• Non-clinical staff
• Space
• Diagnostics
• Supplies
• Drugs

• Calculate the resource requirements for the
  average visit. This is achieved by applying a
  weighting to each resource based on the amount of
  time that is used as a proportion of the total
  package of care time

76
Example package of care Mild Diabetes
ILLUSTRATIVE
patients requiring visit,
No. of visits per patient
Time per visit, min
Clinical staff
Space
Visit types
Diagnostics
First visit
1
100
20
100 GP
100 consulting room
100 blood test
Follow-up visit
1
50
100 practice nurse
100 consulting room
100 blood test
20
Optometry visit
1
50
100 optometrist
100 consulting room
20
100 eye exam
Average resource requirements for a mild diabetes
visit (average resources used across the three
visits)

50 GP 25 practice nurse 25 optometrist
100 consulting room
20
75 blood test 25 eye exam
Number of visits per package of care
2
77
Why use packages of care?
Defining packages of care for patient services
has two key advantages

• Allows definition of package of care based on
  what high quality care is, defined by clinical
  evidence
• Allows tracking of quality of service by defining
  number and types of treatment, e.g.
• Visits with specialist clinicians
• Frequency and type of diagnostic test to be
  administered
• Ensures consistency of care across providers

• Once defined, packages of care enable a stronger
  link between price and commissioning

78
In reality, packages of care will vary at a
patient level according to patient needs
Example Mild Diabetes sub-service line
Average
Price or no. of visits
Because commissioning reflects population need,
the average package of care is used as a basis
for commissioning
Simple cases
Complex cases
Packages of care commissioned
79
Defining packages of care
Packages of care can be defined in a number of
different ways

• Based on clinical evidence base, e.g. NICE,
  NSFs, Cochrane reviews, map of medicine
• Using locally agreed pathways
• Interviewing staff at service providers to
  understand resource requirements

80
Group exercise for packages of care
The groups determine packages of care for 3
sub-service lines of increasing complexity Mild
CHD, Continence, Falls Service

• For each sub-service line, fill out the package
  of care template with
• Number of visits
• Staff involvement (clinical and non-clinical)
• Diagnostic requirements
• Space requirements
• Discuss
• How would you approach this in your PCT?

81
Blank template for calculating packages of care
patients requiring visit,
No. of visits per patient
Time per visit, min
Clinical staff
Space
Visit types
Diagnostics
82
Sample answers for packages of care Mild CHD
patients requiring visit,
No. of visits per patient
Time per visit, min
Clinical staff
Space
Visit types
Diagnostics
CHD package of care
5
Visits
Average resource requirements per visit
30
20 consultant 80 specialist nurse
60 clinic room 40 home
20 X-ray 20 ECG
83
Model answers for packages of care - Continence
patients requiring visit,
No. of visits per patient
Time per visit, min
Clinical staff
Space
Visit types
Diagnostics
First visit
1
100
60
100 specialist nurse
100 clinic room
100 bladder scan 100 Urine analysis
Continence package of care
5
Visits
80 specialist nurse 20 community nurse
Average resource requirements per visit
60
68 clinic room 32 home
100 Urine analysis 20 bladder scan 20 atrophic
stimulation
84
Model answers for packages of care Falls service
patients requiring visit,
No. of visits per patient
Time per visit, min
Clinical staff
Space
Visit types
Diagnostics
First visit physiotherapist
1
100
60
100 physio
100 clinic room
n/a
First visit OT
1
100
60
100 OT
100 clinic room
n/a
Follow-up physiotherapist
5
50
60
100 physio
100 clinic room
n/a
Follow-up OT group
5
50
60
100 OT
100 clinic room
n/a
Follow-up OT home
6
50
60
100 technician
100 home
n/a
Falls service package of care
10
Visits
Average resource requirements per visit
60
35 physio 40 OT 25 technician
75 clinic room 25 home
85
Draft workshop agenda

• 1. Introduction 11.00 - 11.15
• 2. Service Lines
• Overview of Service Lines 11.15 11.30
• Small group discussion 11.30 12.00
• 3. The commissioning model
• Conceptual overview 12.00 12.15
• Group sessions case study 12.15 12.45
• Presentation from Tower Hamlets 12.45 13.00
• QA 13.00 13.15
• Lunch 13.15 14.00
• 4. Packages of care
• Definition 14.00 14.10
• Group discussions for a sample sub-service
  line 14.10 14.30
• 5. The financial model

86
The commissioning model and the polyclinic model
are used together to commission a polyclinic
Commissioning model
Population
Needs
Activities
Polyclinic model

• PCT population size and growth
• Demographics
• Age
• Ethnicity
• Deprivation

• Identify drivers of each activity
• Apply prevalence of driver to derive local need
• If required, adjust figure for local variation in
  risk factors such as
• Smoking
• Obesity
• Child poverty etc.

• Activity by service line and sub-service line
• Projection over a number of years
• Package of care

Price
Output from Commissioning model

• PL
• Utilization
• Capex

The commissioning model defines activity
The polyclinic model calculates cost in order to
inform price
Input for Polyclinic model
Activities
Resources

• Staff
• Utilization
• Diagnostics
• Supplies
• Space
• Drugs

• Service line
• Package of care
• Number of visits

87
Polyclinic financial model
1
Polyclinic resources
3
Staff calcs
Staff needs
Commis-sioning model
Total staff costs
Utilisation
x
Staff costs
Activity level
or
Hard-coded inputs
8
4
Pricing model
Space calcs
Space needs
Resources needed
Total gross costs
Total space costs
Utilisation
x

• Packages of care
• CL Staff
• Non-CL staff
• Space
• Supplies and drugs
• Diagnostics (Dx)

Space costs
5
Diagnostics calcs

• Existing revenue sources
• Block contracts
• Price/unit

Dx needs
Pricing
Total Dx costs
Utilisation
x
Space costs
2
Polyclinic - costs
Staff costs
6
Indirect calcs

• New revenue sources
• Block contracts
• Price/unit
• Target return ()

Direct activity
Total indirect costs
Space costs
Indirect drivers
x
Indirect costs
Dx costs
7
Capex calcs
Indirect costs
Total capex costs / depre-ciation
Capex costs
88
Financial model - video
89
Group work for discussing the polyclinic model
At the tables discuss what you think will be the
most challenging parts of filling out the
polyclinic model
Questions to consider What did you think of
the model? Who in your organisation will take
ownership of the model and populate it? What
technical training will you need to complete the
model? What additional support will you need?
90
Draft workshop agenda

• 1. Introduction 11.00 - 11.15
• 2. Service Lines
• Overview of Service Lines 11.15 11.30
• Small group discussion 11.30 12.00
• 3. The commissioning model
• Conceptual overview 12.00 12.15
• Group sessions case study 12.15 12.45
• Presentation from Tower Hamlets 12.45 13.00
• QA 13.00 13.15
• Lunch 13.15 14.00
• 4. Packages of care
• Definition 14.00 14.10
• Group discussions for a sample sub-service
  line 14.10 14.30
• 5. The financial model

91
An approach to commissioning in polyclinics
Proposed solution for polyclinics
Acute system
What is commissioned

• HRG
• Dominant FCE in spell

• Service lines/packages of care

How much? (volume)

• Prior year activity

• Commissioning model

What is the cost of delivering the service?

• Reference cost Index (not seen by commissioner)

• Financial model

What price do I pay?

• PbR tariff

• Financial model

92
The two models are combined to calculate the
affordability of moving services to a polyclinic
Commissioning model
Activity
Financial model
Price

• Compare to
• current cost in other settings
• projected cost savings in acute care

93
Appendix
94
How would you account for GMS- or PMS-type
contracts in the financial model?
Commissioners can enter details of existing
contracts and the proportion of them that should
be allocated to each sub-service line. The model
then calculates the pre-existing funding flows
applicable to each sub-service line.
The value of the contract is entered
The activities covered by the contract are
attributed to the relevant sub-service lines
The funding from the contract feeds through to
the PL for each sub-service line
95
How do service lines map to the World Class
Commissioning specialties?
Non-acute service lines
WCC specialities

• Unscheduled care

• AE

• Planned care

• General surgery
• Trauma and orthopaedics
• General medicine
• Cardiac
• Other surgery
• Other medicine

Three service lines are split across a number of
specialties, and map only at the sub-service line
level

• Procedures

• General surgery
• Other medicine
• Gynaecology

• Long term conditions

• Cardiac
• Other medicine
• General medicine
• Other

• Mental health

• Mental health

• Older people/complex needs

• Elderly care

• Paediatrics

• Paediatrics

Most of the service lines map directly to the
world class commissioning specialties

• Maternity

• Maternity

• Wellness

• Other

• Social care

• Other

• Pharmacy

• Other

• Other

• Other

96
Updated polyclinic service lines proposal