Using routinely collected data Using routinely collected data t fi t ht fi t hto figure out where to figure out where 

the NHS is going wrongthe NHS is going wrongthe NHS is going wrong.the NHS is going wrong.

Helene Irvine, Consultant in Public Health MedicineJohn Gomez, Senior Information Analyst

Presented to the Scottish Faculty of Public Health Conference Dunblane Hydro 27 November 2016Conference, Dunblane Hydro, 27 November 2016. 

Full version, complete with embedded notes and 2 pageembedded notes and 2 page 

summary, originally presented at the y, g y pAnnual Conference of the Deep End j h ld bProject, held on 24 November 2015 at the Erskine Hotel is available at:the Erskine Hotel, is available at:

http://www.gla.ac.uk/media/media_443697_en.pdf

NRAC and CSR work revealed two major challenges in GG&C that are linked

4

Financial over‐parity for HCHS due to

challenges in GG&C  that are linked

Over‐subscription and for HCHS due to 

rising relative over‐provision of 

ELECTIVE CARE to

‘failing performance’ of UNSCHEDULED CARE, uptake dominated by th d i d d thELECTIVE CARE to 

the affluent and relatively healthy

Loss of the gatekeeper function

the deprived and thosein greater immediate 

need

Health neurosis Uncontrolled access to 2y care

The weakening of general practice, dating back to the 

Screening (programmes &non evidence based ad hoc); excessive and 

Weakening of district nursing; investment in CH(C)P based staff well placed to identify unmet need but p , g

1980s (Read Lewis Ritchie’s 2003 paper), which was further undermined by the new GP 

conflicting health promotion; the internet; health consumerism; supply of highly 

insufficiently skilled to deal with it; Unscheduled Care Programme and HEAT target; increased A&E medical staffing; cutbacks to social work 

contract in 2004.specialised health care; disease‐specific NGOs; etc

budgets and community based CoE; inadequate provision of respite, etc

All admissions (all types, all specs) : GG&C/RoS quotients of age standardised ratesbetween 2001/2 and 2010/11, by gender and SIMD quintile, n=315,553 and

/

6

836,010 adms in GG&C and RoS, respectively, in 2010/11, shown as a graph.Source: SMR01 dataset downloaded by J Gomez.1.4 1 4

1.301.31.34

1.3

1.4

1.21.2

1.08

1 0

1.1

Q11.05

1.1

Q1

0.9

1.0

Q2

Q30.9

1.0Q2

Q3

0.8

Q4

Q5 0.8

Q4

Q5

Females Males

Age standardised rates of elective admission (oncology) for GG&C andRest of Scotland females, by selected SIMD quintile, n=11,432 admissions

9

Rest of Scotland females, by selected SIMD quintile, n 11,432 admissionsfor GG&C females. Source: SMR01 dataset downloaded by J Gomez.

25

18 2

y = 0.4986x + 14.35R² = 0.5605

20

25

Evidence of Inverse care 

16.3

18.2

15

20law at local level!

11.6

12.7

10 GG&C SIMD Q1

5

GG&C SIMD Q1

GG&C SIMD Q5

RoS SIMD Q1

0

RoS SIMD Q1

RoS SIMD Q5

Li (GG&CLinear (GG&C SIMD Q5)

My current assumptionsMore likely Regression model ‐ acute sector 10

•The NRAC formula is reasonably correct and was more clever than I thoughtat predicting our need‐related costs. It deliberately avoids perverseincentives and thereby ignores rising rates of activity that are unrelated to

MLC indexEquation that best fits the data:incentives and thereby ignores rising rates of activity that are unrelated to

need.

•The increasing financial bail‐out (over‐parity) is not legitimately ours and

Dependant variable (y)

Actual activity

e.g. Predicted y=bx+ a

probably does belong to NHS Lothian and other boards that are under‐parity.

•The high and rising rates of admission are not entirely related to socialdeprivation; an increasing proportion of our admissions are for affluent

Actual activity multiplied by national costing fees/predicted cost for that age

COSdeprivation; an increasing proportion of our admissions are for affluent

residents and therefore must contain a considerable proportion ofinappropriate or unnecessary admissions. Inverse care law is a problem for us.

•The high and rising rates of admission are indeed driving the over parity in

cost for that age sex distribution for each IDZ GG&C is

inadvertently flattening the slope

T

•The high and rising rates of admission are indeed driving the over‐parity inthat this excessive workload costs money to deliver and is not recognised bythe NRAC formula even though the latter is a utilisation‐based formula.Unit of analysis = IDZ 

n=~1300, ~300 in GG&C

flattening the slope of the line via the mismatch of need and provision

•Supply is a major determinant of activity as demonstrated by the fact thatthe national dataset doesn’t fit unless considerable numbers of supply factorsare included in the regression modelling. NEED

and provision.

Independent variables (x):  1)  Prevalence of SRLLTI (Census 2001); 2)  SMR All cause <75 yrs for each IDZ.

Number of emergency admissions (all specs, all ages, all stays) atGG&C sites 1995/6 ‐ 2014/15 Source: SMR01 data from J Gomez

13

GG&C sites, 1995/6 ‐ 2014/15. Source: SMR01 data from J Gomez.

170,000 5) UCCP, intro of 4 hr A&E target ↑A&E Counting 

160,000 1) New GP Contract2) New Hospital Consultant 

Contract

A&E target, ↑A&E consultants.

6) Funding starts to transfer from general prac ce→ CH

gGRI AAU stays

140,000

150,000 3) Loss of GP incentive to do OOH work

4) Commencement of transfer of LHCC functions to CHP

prac ce → CH services

111 NHS24

130,0007) CHPs have 

completely

WI, RAH

GRI

Intro of

110,000

120,000GEMS Co‐op in GG 

NHS24

completely replaced LHCCs

8) Council tax freeze  (SW)

9) ↓ District

Intro of AAUs

LHCCs

Change Fund

100,000

9) ↓ District Nurses

Percentage of total national territorial board NHS funding spenton general practice vs community services 2001 2013 Source:

19

on general practice vs community services, 2001‐2013. Source:ISD Scotland website funding data.

18%

12.6%

17.0%

14%

16%

46.1% rise over entire period34 9% i i 2006/7

11.6%

10%

12%34.9% rise since 2006/7

8.13%9.30% 7.66%

6%

8%5.7% decline over entire period17 6% d li i 2005/6

2%

4%

6%LES budgets devolved to health boards 2007New GP 

contract 2004

17.6% decline since 2005/6

0%

2%

GP

CommunityCommunity services

Actual vs real terms expenditure (after SG GDP deflator applied) onl ti i S tl d b t 2001/2 d 2012/3 S

20

general practice in Scotland between 2001/2 and 2012/3. Source:ISD website funding and SG provided deflators at the time of 2013.

£834,554,000 

£756,016,000 800,000 

900,000 

£701,009,000 

500 000

600,000 

700,000 

300 000

400,000 

500,000 

Real terms expenditure A 9% decline in real terms between 

100 000

200,000 

300,000  (£000s)

Actual

2005/6 and 2012/3.

‐

100,000  Actual expenditure (£000s)

HCHS Medical staff (all grades), All GPs (all grades), All GPs in 2013assuming 8 and 9 sessions per WTE: numbers of WTE per annum

21

assuming 8 and 9 sessions per WTE: numbers of WTE per annumemployed in Scotland. Source: ISD Scotland manpower and survey data.

14 00011,485.0 

12,000

14,000

All HCHS medical staff

New GP and Hospital Doctor Contracts 2004

7,159.2 

9,261.6 

8,000

10,000staff

'All GPs' estimate 8

Contracts 2004

4,140.1 4 196 93 697 2 4,073.8 

6,000

estimate 8 sessions/WTE

'All GPs' estimate 9 4,196.9 3,910.1 

3,781.9 

3,697.2 

2,000

4,000 sessions/WTE

All GPs WTE

0

Manpower ratio: Numbers of WTE HCHS medical consultant staff: Numbersof GP principals and salaried assuming 8 sessions per WTE in 2009 and 2013,

23

p p g p ,by year, for selected health boards, RoS, Scotland and England. Source: ISDmanpower data and Centre for Workforce Intelligence for English data.

1.6

1.8

1.42

1.241 22

1.4

1.22

1

1.2

GG and GG&C

0.750.73

0 6

0.8GG&C adjTaysideLothianG i

0.4

0.6 GrampianScotlandRoSL k hiLanarkshireEngland

GP principal in general practice, headcount: Crude rate per 10,000 pop’n, 2004onwards, n=789 in GG&C and 2,933 in the RoS by 2014, using GROS denominator;

25

, , y , g ;and 3 WTE estimates from 2005 ISD Return and 2009 and 2013 WorkforceSurveys for GG/GG&C and RoS (8 sessions/WTE). Source: ISD website.

7.006 266 64

7

8

6.26 6.226.64

6.215.87

5

6

3

4 RoS (headcount)Merger of Greater Glasgow with Clyde

1

2

3GG/GG&C (headcount)

RoS WTE (estimates)11% and 12%  reduction in WTE over 8 years, RoS and GG/GG&C

0

1

GG/GG&C WTE (estimates)

District Nurses: Crude rate of WTE provision per 10,000, forGG GG&C A&C RoS 2000 to 2013 Source: ISD Website on Nursing

31

GG, GG&C, A&C, RoS, 2000 to 2013. Source: ISD Website on Nursingand Midwifery stats, based on data submitted by health boards.

4 55 0

3 5

4.0

4.5

4.0

4.5

5.0

Greater Glasgow 

No data available

2 5

3.0

3.5

3.0

3.5

4.0and GG&C

Agenda For Change 2007

1 5

2.0

2.5

2.0

2.5 A&CMerger of GG and 

Clyde 2006

Change 2007

0 5

1.0

1.5

1.0

1.5

RoS

0.0

0.5

0.0

0.5RoS

Per capita expenditure in real terms on social services for the 65+ inGlasgow C based on LFR03 Returns compared to the percentage of the

34

Glasgow C based on LFR03 Returns compared to the percentage of theGlasgow City 65+ population that was 85+ based on GROS mid yearestimates. Source: LFR03 Returns published by SG and NRS MYEs.

£3,057 12.5%14%£3,500

£2,639

11.9%10%

12%

£2,500

£3,000Per capita expenditure on social services for 

6%

8%

£1,500

£2,000 the 65+

% of the 65+

2%

4%

£500

£1,000% of the 65+ pop'n that was 85+

0%£0

Live births in Scotland, 1900‐2010 and Percentage of GG&C populationthat was 90+ years of age over time 2000 2014 Source: NRS

35

0.8%n=8,200 in 2013

4 hr A&E 

that was 90+ years of age over time, 2000‐2014. Source: NRS.

136 546

160,000 

0.6%

0.7%2013compliance 

starts to collapse in GG&C

136,546 

120,000 

140,000 

0.5%

0.6%

n=6,204 i 2009

100,000 

0.3%

0.4%in 2009

60,000 

80,000 

0.2% Intro of AAUs20 000

40,000 

0.0%

0.1%AAUs

‐

20,000 

00 0 0 0 40 0 60 70 80 90 00 0

190

191

192

193

194

195

196

197

198

199

200

201

Tracing back problems described in Auditor General’s Report to General Practice40

47

Thank you for listening!Thank you for listening!

Any comments or questions?Any comments or questions?

hirvine@nhs nethirvine@nhs.net