23
The Application of Fair Shares for All to Family Health Service Expenditure National Resource Allocation Committee August 2005

NRAC Presentation August 2005 v0.1

Embed Size (px)

Citation preview

Page 1: NRAC Presentation August 2005 v0.1

The Application of Fair Shares for All to Family Health Service Expenditure

National Resource Allocation Committee August 2005

Page 2: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

Agenda

• Background to the review

• Conceptual approach

• Utilisation data versus the epidemiological approach

• Cross boundary flow

• Summary

Page 3: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

Aim

• To present work on extending the Fair Shares For All (FSFA) methodology into Family Health Services (FHS) on behalf of the FHS Advisory Group

• To provide a broad conceptual overview of our approach

• To highlight one or two key issues for discussion

• To identify areas for further detailed discussion with NRAC given the quantum of research

Page 4: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

Background

Deloitte was commissioned in September 2004 to develop a series of needs based formulae for the allocation of three FHS budgets1:

Pharmaceutical Services (PS) £117m

Primary and Community Dental Services

General Dental Service (GDS) £199m

Community Dental Service (CDS) £ 32m

General Ophthalmic Services (GOS) £ 36m

Total (2003/04) £384m

1 The research was based on a specification from September 2002

Page 5: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

Current Financial Arrangements

The FHS budget is currently allocated to each NHS Board based on historical expenditure (the CDS budget is allocated as part of the community formula and the GMS budget has a dedicated formula).

These were not covered in the original Fair Shares for All review because it was considered impractical within the timescales:

‘there was a consensus that these smaller elements…should be examined at a later date in the medium term future, once the methods for

distributing larger budgets had been decided’

The aim of this work is to ensure that resources are distributed equitably across Scotland reflecting the population’s relative need for resources rather than reflecting the current pattern and location of services.

Page 6: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

Legislative & Policy Context

There has been a substantial volume of policy and legislative change within FHS over the past few years:

• The development of a new community pharmacy contract (Pharmaceutical Care Services). This will place an increased emphasis on pharmaceutical care in addition to dispensing. This may influence patterns of service utilisation over time and have implications for the development of a PCS formula.

• The extension of free sight tests and dental check up to all as outlined in the Smoking, Health and Social Care Bill 2004. These changes will influence the range of people eligible for NHS treatment and thus need for NHS resources.

Whilst the need for FHS resources should not be influenced by policy decisions per se, the development of the formulae will need to take cognisance of these changes and be flexible

Page 7: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

What’s different about FHS?

• There is a mixed economy of provision with a high proportion of services provided by private sector contractors. The public sector provides a safety net function (particularly in rural areas)

• Services are not universally free for all with co-payment for certain services

• There is a large private sector market

• A high proportion of contractors income is from non-NHS sources (pharmacy c10%, dentistry c50%, ophthalmic >50%)

• Patients have a greater choice and more importantly ease of choice regarding provider

• Contractors do not have a defined/registered population (likewise NHS Boards have a responsibility for services not a defined resident population)

Page 8: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

Conceptual Framework

We adopted standard capitation techniques based on the FSFA approach. However, we tailored this approach depending upon the characteristics of each service and data availability. We applied five adjustments:

• the population share of the NHS Board

• the age and sex characteristics of the population

• the morbidity and life circumstances of the population

• an adjustment for unavoidable costs to reflect the additional cost of providing services in remote areas

• an adjustment for the extent of cross boundary flow

However, we did not necessarily apply each adjustment to each component of each formulae (as discussed below).

Page 9: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

Pharmaceutical Services

Controlled Drug PS

Age/Sex

General PS

Unavoidable Costs

Cross Boundary Flow

Total Population Total Population

Age/Sex Controlled Drugs only

Prescribing Formula MLC

Controlled Drug PS

Age/Sex

General PS

Unavoidable Costs

Cross Boundary Flow

Total Population Total Population

Age/Sex Controlled Drugs only

Prescribing Formula MLC

This was applied across both sub-programmes

The Group recognised the limitations of this approach and that new data may soon be available

Due to different need profile and fee structure we separated this expenditure

Page 10: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

General Ophthalmic Services

Vouchers, Repairs & Replacements

Age/Sex Adjustment

Sight Tests

Remoteness

Cross Boundary Flow

Total Population Eligible Population

Age Adjustment

High Need Group Adjustment No MLC

Vouchers, Repairs & Replacements

Age/Sex Adjustment

Sight Tests

Remoteness

Cross Boundary Flow

Total Population Eligible Population

Age Adjustment

High Need Group Adjustment No MLC

The Sight Test programme will soon be free for all

We incorporate no explicit MLC adjustment

Defined using eligibility criteria and presented separately as expenditure is based on ‘take up rates’’

Page 11: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

Primary and Community Dentistry

Adult Dental Health

Total Population

Child Dental Health

Remoteness

Cross Boundary Flow

Child Orthodontic Edentate Dentate

Total Population

Edentate Population

Dentate Population

Age/Sex

MLC

Age/Sex

No MLC No MLC

No Age/Sex

MLC

Age/Sex

Adult Dental Health

Total Population

Child Dental Health

Remoteness

Cross Boundary Flow

Child Orthodontic Edentate Dentate

Total Population

Edentate Population

Dentate Population

Age/Sex

MLC

Age/Sex

No MLC No MLC

No Age/Sex

MLC

Age/Sex

Orthodontic need was specified separately as no MLC was applied

Edentate and dentate populations have very different needs, we used survey data to determine the size of each population

Separate Adult and Child programmes were used mainly to reflect different fee structures

Page 12: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

The Data Mountain

We blended mountains of administrative, survey and expenditure data:

• Administrative Data: we used whole data extracts from MIDAS (GDS system), OPTIX (GOS system), SMR13 (CDS system) and the Prescribing Information System

• CHI Matched Data: original research by ISD Scotland enabled the matching of the CHI register to OPTIX and MIDAS (4m plus observations)

• Epidemiological Data: DHSRU released data on three national child dental health surveys (22,000 observations)

• Survey Data: we used data from the Scottish Health Survey, the UK Adult Dental Health Survey and Health Education Population Survey

• Benefit/Credit Data: 3 of the first 10 FoI requests to DWP related to this study

• Other Data: population, expenditure, contractor locations, census, earnings surveys, clinical literature etc

Page 13: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

A Hybrid Approach

We took a pragmatic view of the best way of developing a needs based formula using both utilisation and epidemiological data when appropriate. In general we preferred to use utilisation data unless it was considered ‘too problematic’:

• we used utilisation data to generate the age and sex adjustment in all three formulae. We assumed that on average the pattern of resource use across age reflects the relative need for resources.

• we used epidemiological data for the construction of the MLC adjustment in the dental formula only. We assume that the gradient of resource need across deprivation was proportional to dental health need.

Page 14: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

Pitfalls of Utilisation Data

The FHS Advisory Group was concerned regarding the use of data from the GDS for constructing a capitation formula (MIDAS):

• Treatment claims are influenced by fee structures

• Administrative datasets are not comprehensive

• No patient location identifiers were available

• Co-payment for services may deter certain groups from accessing services

The use of dental services was not considered an adequate approximation for the need for resources. For example, people in more deprived areas use dental services the least despite all measures of poor dental health being higher in deprived areas (the ‘inverse care law’). No amount of ‘unmet needing’ would address this fundamental issue2.

2 These drawbacks applied to the development of an age and sex adjustment but particularly to the development of an MLC adjustment.

Page 15: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

Determinants of Dental Visits

0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75

Edentate

Brush teeth less thanonce per day

Own House

Elderly Living Alone

Area Level EducationDeprivation

Odds Ratio

If you live in a more ‘educationally’ deprived area you are less likely to visit your dentist

If you have no teeth you have poor dental health but less need to visit the dentist

Behavioural variables are important determinants

Page 16: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

The Epidemiological Approach

• We used a standard measure of dental health (a count of decayed, missing and filled teeth) as a proxy for dental health

• We used data from 3 child dental health surveys to develop an age and sex standardised dmft score for each small area in Scotland (22,000 observations)

• We regressed this on a series of area level explanatory variables including supply variables and socio-economic variables

• A number of the Scottish Index of Multiple Deprivation (SIMD) domains were significantly related to dental health, particularly the education deprivation domain (note the NRRA index was not a good explanatory variable)

• We used this model to calculated the MLC adjustment for both Child and Adult sub-programmes

Page 17: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

Dental Health and Deprivation

0

50

100

150

200

250

-2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5

Education Domain Score

Sta

nda

rdis

ed

dmft

Sco

re

Page 18: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

Issues with the Approach

Problems

• there may not be a proportionate relationship between epidemiological indicators of need and the need for resources, for example, an individual with a dmft score of two does not necessarily need twice the resources as the individual with a score of one

• those with a dmft score of zero still need resources for preventative care

Positives

• the metric is easy to understand and has a clear link to dental health

• there are fewer issues with ‘tails of the distribution’ (i.e. the influence of unmet need or private practice)

• the relationship looks about right

Page 19: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

Cross Boundary Flow

• NHS Boards have a statutory responsibility to pay for services provided in their area, not for services only to their resident population

• Patients have a greater freedom of movement which makes it difficult to assess whether an area is relatively under or over provided with services

• For example, a city centre may appear to be over funded relative to need because patients may ‘commute’ across boundaries to visit a contractor

• The extent of cross boundary flow has never been examined prior to this study due to data constraints

• The results are very interesting and the interpretation challenging

Page 20: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

Cross Boundary Flow

• These data illustrate the fluidity of the FHS market, for example:

– the median number of ‘feeder’ areas for each dental list is 27

– the median number of ‘feeder’ GPs for a community pharmacist is 78

– one dental list contained patients from 275 different areas

– one high street pharmacist in a city centre saw patients from over 700 GP practices (note there are only 1,000 in Scotland)

• We could only match data from OPTIX (GOS) to NHS Board of treatment so are unable to provide a similar profile for each optician location

• The adjustment is not based on the simple percentage of the population crossing boundaries (i.e. inflow – outflow), but takes into account the need profile of the small area from which patients originate

Page 21: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

Cross Boundary Flow Adjustment

-15%

-10%

-5%

0%

5%

10%

15%A

A

Bor AC

Fife GG

Hig

h

Lan

Gra

mp

Ork

Loth

Tay FV WI

DG

Sh

et

Cro

ss B

ound

ary

Flo

w A

dju

stm

ent

GDS GOS PS

Page 22: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

Cross Boundary Flow

• The results illustrate the extent of cross boundary flow for the first time

• It could be argued that the adjustment is not reflecting need but supply restrictions in certain areas (particularly in the dental formula)

• On the other hand the flow could illustrate patient preferences for treatment locations or simply optimal locations from a retailing perspective (for example, in the GOS a high proportion of fees are from non-NHS sources)

• Without this information it is difficult to assess the relative provision of services in each NHS Board

Page 23: NRAC Presentation August 2005 v0.1

©2005 Deloitte MCS Limited. Private and Confidential

Summary

• As with any research we identify almost as many questions has we answer:– How should the cross boundary flow adjustment be interpreted?

– Is utilisation data robust enough for generating the age and sex adjustment?

– Does resource use increase in proportion to dental health?

– Is the unavoidable cost adjustment reinforcing traditional service delivery patterns?

• We consider that we have taken these issues at least ‘90% of the way’

• What happens next?