Powerful analysis, influencing decisions 22 April, 2015 Commissioning Analysis and Intelligence Team Andrew Jackson Overview of Tools Analysis to support.

Powerful analysis, influencing decisions18 April 2023

Commissioning Analysis and Intelligence Team

Andrew Jackson

Overview of ToolsAnalysis to support Commissioning

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

218 April 2023 2

Quantity of beer

benefit

Quantity of beer

benefit

Why Variation Matters - Application to beer(with thanks to Angela Bate, Newcastle University, Institute of Health and Society!)

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

318 April 2023 3

Application to beer

benefitbenefit

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

418 April 2023 4

Diminishing Marginal Returns

benefit

Quantity of beer

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

518 April 2023 5

Quantity of beer

benefit

Negative Marginal Returns

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

618 April 2023 6

Marginal Analysis

Quantity of beer

benefit Marginal benefit

Quantity of beer – or £ invested

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

718 April 2023 7

What We Do

• Annual Data Collection each summer from 152 PCTs for 23 PB Categories, plus Sub-Categories.

– England level, programme budgeting expenditure published each autumn;– Benchmarked expenditure data at PCT level published each autumn on DH website;– One page linked spend and outcome data published via SPOT tool;– NCHOD site links expenditure data with:

Outcome data – e.g. disease mortality, 75 >YLL, 30 day > hospital mortality; QOF data – e.g. disease prevalence, disease incidence, blood / cholesterol control; HES data – e.g. admissions, alos, beddays, day case rate, FHS Prescribing – expenditure and volume.

• For PBC activity, using programme budgeting codes we are able to use NHS Comparators to;

– Provide a selection (elective / non – elective admissions, plus FHS prescribing) of programme budgeting data more quickly, and more frequently (every quarter);

– Data are available at England, SHA, PCT, and individual practice level.

• Developing additional supporting tools;

– Links variation in programme budgeting category with high volume HRGs within the programme – IVET tool available now;– Analysis underway on high growth admissions to question whether clinical thresholds are changing, work being undertaken in partnership with NICE;– Analysis scheduled (and resourced) to take place over the summer analyse PROMS data, but from a commissioner perspective to highlight variation in

initial health status i.e. the threshold, and in due course, final health status / improvement in health status.

• Dissemination to the NHS taken forward jointly via;

– The World Class Commissioning Programme;– QIPP ‘Right Care’ programme.

• wcc.networks.nhs.uk/healthinvestment

– Provides links to all material, plus on line training.

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

818 April 2023 8

2008/9 Programme Level Expenditure2007/ 08£000s

2008/ 09£000s

Difference£000s % change

05 Mental Health Disorders 10,278,503 10,415,840 137,337 1%

10 Problems of Circulation 7,227,743 7,420,201 192,458 3%

02 Cancers and Tumours 4,964,282 5,134,948 170,666 3%

11 Problems of the Respiratory System 3,802,489 4,247,325 444,836 12%

15 Problems of Musculo Skeletal System 4,085,030 4,212,469 127,440 3%

13 Problems of Gastro Intestinal System 4,099,120 4,097,920 1,200- 0%

17 Problems of Genito Urinary System 3,645,977 4,000,641 354,664 10%

07 Neurological 3,438,256 3,683,873 245,616 7%

16 Problems due to Trauma and Injuries 3,075,116 3,299,792 224,676 7%

22 Social Care Needs 2,069,407 3,156,039 1,086,632 53%

18 Maternity and Reproductive Health 2,950,693 3,100,821 150,128 5%

12 Dental Problems 3,017,291 3,087,416 70,125 2%

06 Problems of Learning Disability 2,856,102 2,916,182 60,080 2%

04 Endocrine, Nutritional and Metabolic 2,428,797 2,526,152 97,355 4%

21 Healthy Individuals 1,729,476 1,908,832 179,357 10%

14 Problems of the Skin 1,698,307 1,794,226 95,920 6%

08 Problems of Vision 1,598,679 1,664,102 65,423 4%

01 Infectious Diseases 1,333,222 1,410,980 77,757 6%

03 Disorders of Blood 1,237,154 1,253,786 16,632 1%

19 Conditions of Neonates 955,303 1,101,470 146,167 15%

20 Adverse effects and poisoning 833,801 955,442 121,641 15%

09 Problems of Hearing 421,965 417,167 4,798- -1%

23 Other 25,436,712 24,836,323 600,389- -2%

Total Gross Expenditure 93,183,426 96,641,949 3,458,523 3.6%

Programme Budgeting Category

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

918 April 2023

Expenditure Variation of Cancer Source – PB Spreadsheet

Powerful analysis, influencing decisions

Expenditure (£000s) per 100,000 selected population

-

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Rank

Expenditure per 100,000 selected population

Expenditure per 100,000 selected population for PCTs within selected cluster

Expenditure for selected PCT

02 Cancers and Tumours

Q39 Bristol Teaching PCT (5QJ )

EXPENDITURE ON OWN POPULATION

UNIFIED WEIGHTED POPULATION

1 SUPER (7 groups)

2007-08

Selected PCT

Selected expenditure

Selected programme

Selected cluster level

Selected population

Selected year

Exp

end

itu

re (

£000

s)

per

100

,000

sel

ecte

d

po

pu

lati

on

(Use the drop down boxes on the right to change the variables)CALCULATE

NOW

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

1018 April 2023

Spend v Outcomes – National ComparisonSource – APHO Spot Tool

Outcomes and expenditure relative to other PCTs in England

Inf

Canc,Skin,TraumaBlood,Gastro,Pois,Hlth

End

MH

LDNeuro

Vision

Hear,SocCirc

Resp,GUDent

Musc,Neo

Mat

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5

Spend Z Score

Hea

lth

Ou

tco

me

Z S

core

Lower Spend, Better Outcome

Lower Spend,Worse Outcome

Higher Spend,Worse Outcome

Higher Spend,Better Outcome

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

1118 April 2023

Spend v Outcomes – Cluster ComparisonSource – APHO SPOT tool.

Powerful analysis, influencing decisions

Arrows show

movement from

national to cluster picture

Ranking of outcomes and expenditure relative to ONS Cluster group

Inf,Trauma

Canc Blood,Hlth

End

MH

LD Neuro

Vision

Hear,Skin,Soc

Circ

Resp

Dent

GastroMusc

GU

Mat Neo

Pois

0

3

5

8

10

035810Spend Ranking

Hea

lth

Ou

tco

me

/ N

eed

Ran

kin

gLower SpendBetter Outcome

Higher SpendBetter Outcome

Lower SpendWorse Outcome

Higher SpendWorse Outcome

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

1218 April 2023 12

Programme Budget Atlas – Scatterplot shows CVD Expenditure and Mortality data from previous slides on the same graph – BEN PCT (highlighted by blue dot) have low spend (vertical axis) and high mortality (horizontal axis) – PCTs can use the chart to identify PCTs

with good outcomes (and can filter by SHA and ONS similar PCT cluster)

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

1318 April 2023 13

Programme Budget Atlas – CVD Non-Elective Hospital Admissions per 100,000 population (weighted for age, sex and need) – Darker areas represent higher number of admissions – BEN PCT has a high number of Non-Elective Admissions

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

1418 April 2023 14

Programme Budget Atlas – CVD Non-Elective Average Length of Stay per spell in hospital – Darker areas represent higher LOS – BEN PCT has a low average LOS for Non-Elective Admissions

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

1518 April 2023 15

Programme Budget Atlas – CVD Emergency Readmissions to hospital within 28 days of discharge – Darker areas represent higher number of readmissions – BEN PCT has a high number of CVD Emergency Readmissions

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

1618 April 2023

NHS Comparators – Range of Activity and Expenditure Data (age and sex standardised)

Powerful analysis, influencing decisions

Ealing PCTEstimate of the

difference between actual and expected

spend (based on national average

expenditure applied to own population)

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

1718 April 2023

NHS Comparators – Admissions data can be broken down by Programme Budget category

Powerful analysis, influencing decisions

Ealing PCT

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

1818 April 2023

NHS Comparators – Variation between GP Practices in a PCT

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

1918 April 2023

NHS Comparators – Time series data – Outpatient First Attendances

18% growth in Outpatient

First Attendances

in 2008/9

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

2018 April 2023

NHS Comparators – Reported vs Expected Prevalence – Indication of unmet need

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

2118 April 2023

NHS Comparators - Disease (or Programme Budgeting) Level Expenditure on Admissions

This table is available at SHA, PCT, and Practice level, on a quarterly or annual basis.

The table shows for inpatient admissions in each disease area, actual spend, expected based on national averages, and expenditure differences in absolute and % terms.

Because expenditure is calculated as activity multiplied by tariff, the table is less useful where tariff is less well developed – e.g. Mental Health.

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

2218 April 2023

NHS Comparators - Disease (or Programme Budgeting) Expenditure on Prescribing

This table is available at SHA, PCT, and Practice level, on a quarterly or annual basis.

The table shows for each disease area for FHS prescribing, actual spend, expected based on national averages, and expenditure differences in absolute and % terms.

Data are only available on NHS net – hence thus are not accessible by pharmaceutical industry!

Powerful analysis, influencing decisions18 April 2023

MapAdditional view

Google copyright© Google

Highest 20%

Lowest 20%

Selected areahighlighted

No gradients within cells

Google copyright

Powerful analysis, influencing decisions18 April 2023

Map

Google copyright

Highest 20%

Lowest 20%

Map centraliseswhen selected area

when selected

Powerful analysis, influencing decisions18 April 2023

Map

On pin click, show practice

name

Google copyright

Highest 20%

Lowest 20%

Garforth practice

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

2618 April 2023

E03 - AgeSexNeeds Expenditure Rate per 1000 population for PCTs

0

1,000

2,000

3,000

4,000

5,000

6,000

1 11 21 31 41 51 61 71 81 91 101 111 121 131 141 151

PCT

Ag

eS

ex

Ne

ed

s E

xp

en

dit

ure

Ra

te (

£

pe

r 1

00

0 p

op

ula

tio

n)

Cardiac valve procedures – Inpatient expenditure rate, 2008/9(weighted for age, sex and need - per 1,000 population)

Source: DH CAI (using HES)

London

Some overlap with circulatory

96% of expenditure in analysis

Cardiac valve procedures cost £12,343 each.

There is a 5-fold variation in expenditure between PCTs(adjusting for age, sex and need).

The coefficient of variation is 26.5%.(This takes into account all PCTs, not just the top and bottom PCTs.)

The potential savings are £19M(if PCTs with rates higher than the median reduced to this level).

Relatively high unwarranted variation so large potential savings.

Small drops in activity give large savings

Top 30 PCTs(Lowest Rates)Next 31 PCTs

Next 30 PCTs

Next 31 PCTs

Bottom 30 PCTs (Highest Rates)

Top 30 PCTs(Lowest Rates)Next 31 PCTs

Next 30 PCTs

Next 31 PCTs

Bottom 30 PCTs (Highest Rates)

Top 30 PCTs(Lowest Rates)Next 31 PCTs

Next 30 PCTs

Next 31 PCTs

Bottom 30 PCTs (Highest Rates)

Total Inpatient Expenditure (£M)

Potential Saving using 50th percentile (£M)

Potential Saving as % of Total Inpatient Expenditure

148 19 12.9%

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

2718 April 2023

Coronary bypass – Inpatient expenditure rate, 2008/9(weighted for age, sex and need - per 1,000 population)

Source: DH CAI (using HES)

Total Inpatient Expenditure (£M)

Potential Saving using 50th percentile (£M)

Potential Saving as % of Total Inpatient Expenditure

154 14 9.0%

London

Coronary bypasses cost £8,660 each.

There is a 9-fold variation in expenditure between PCTs(adjusting for age, sex and need).

The coefficient of variation is 27.3%.(This takes into account all PCTs, not just the top and bottom PCTs.)

The potential savings are £14M(if PCTs with rates higher than the median reduced to this level).

E04 - AgeSexNeeds Expenditure Rate per 1000 population for PCTs

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

1 11 21 31 41 51 61 71 81 91 101 111 121 131 141 151

PCT

Ag

eS

ex

Ne

ed

s E

xp

en

dit

ure

Ra

te (

£

pe

r 1

00

0 p

op

ula

tio

n)

Top 30 PCTs(Lowest Rates)

Next 31 PCTs

Next 30 PCTs

Next 31 PCTs

Bottom 30 PCTs (Highest Rates)

Top 30 PCTs(Lowest Rates)

Next 31 PCTs

Next 30 PCTs

Next 31 PCTs

Bottom 30 PCTs (Highest Rates)

Top 30 PCTs(Lowest Rates)

Next 31 PCTs

Next 30 PCTs

Next 31 PCTs

Bottom 30 PCTs (Highest Rates)

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

2818 April 2023

Non-transient stroke or cerebrovascular accident >69 – Inpatient expenditure rate, 2008/9

(weighted for age, sex and need - per 1,000 population) Source: DH CAI (using HES)

Total Inpatient Expenditure (£M)

Potential Saving using 50th percentile (£M)

Potential Saving as % of Total Inpatient Expenditure

171 14 8.0%

London

Non-transient stroke or cerebrovascular accidents >69 cost £4,096 each.

There is a 5-fold variation in expenditure between PCTs(adjusting for age, sex and need).

The coefficient of variation is 18.7%.(This takes into account all PCTs, not just the top and bottom PCTs.)

The potential savings are £14M(if PCTs with rates higher than the median reduced to this level).

A22 - AgeSexNeeds Expenditure Rate per 1000 population for PCTs

0

1,000

2,000

3,000

4,000

5,000

6,000

1 11 21 31 41 51 61 71 81 91 101 111 121 131 141 151

PCT

Ag

eS

ex

Ne

ed

s E

xp

en

dit

ure

Ra

te (

£

pe

r 1

00

0 p

op

ula

tio

n)

Top 30 PCTs(Lowest Rates)

Next 31 PCTs

Next 30 PCTs

Next 31 PCTs

Bottom 30 PCTs (Highest Rates)

Top 30 PCTs(Lowest Rates)

Next 31 PCTs

Next 30 PCTs

Next 31 PCTs

Bottom 30 PCTs (Highest Rates)

Top 30 PCTs(Lowest Rates)

Next 31 PCTs

Next 30 PCTs

Next 31 PCTs

Bottom 30 PCTs (Highest Rates)

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

2918 April 2023

Of this £12 million, over £2 million a year can be saved if they reduce Primary Knee Replacements to the national average rateSource – IVET tool.

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

3018 April 2023

The benchmarking tool allows PCTs to select a high cost disease or procedure, choose a benchmark level (e.g. median, lowest 10%) and potential savings are displayed.Source IVET tool.

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

3118 April 2023

Improvements to Programme Budgeting Data (1)

• Move to a Commissioner-based return (trial-run for 2009/10 collection and replace Reference Cost based return for 2010/11 collection)

• This will improve the accuracy of the data in future– Based on actual PCT spend rather than Provider’s cost (e.g. PbR tariff)

– Commissioners have greater incentive to ensure data is as accurate as possible

• This will improve the timeliness of data in future– Benchmarking tools will be available to PCTs sooner

– Potential for in-year monitoring and benchmarking

• This will also reduce the overall burden on NHS in future

• We are also working to improve the allocation of Outpatients and A&E expenditure to Programme Budget categories

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

3218 April 2023

Improvements to Programme Budgeting Data (2)

• We also propose to develop a template that provides a more detailed breakdown of Programme Budget data

• The intention is to provide PCTs with the opportunity to compare spend in different areas within a disease category (e.g. Primary Care, Secondary Care, Community Care)

• The long term aim is for the template to be based on patient pathways (the template will initially include prevention spend for each category)

• The basis for all our developments is to provide information to PCTs in the most useful way to allow them to make informed health investment decisions

• We have a data quality group to advise on these developments, but are always seeking feedback from the NHS – if anyone has any comments on this and/or would like to be involved in developing the collection please let us know

Co

mm

issi

on

ing

An

aly

sis

an

d In

telli

ge

nce

Te

am

3318 April 2023

Useful Links

nww.nhscomparators.nhs.uk

wcc.networks.nhs.uk/healthinvestment