Reducing Avoidable AdmissionsJoint Strategic Needs Assessment

Andrew PulfordSenior Public Health Research Officer

BACKGROUND

Trends in emergency admissions to hospitalLocal policy context

Trends in emergency admissions

• Steady rise in the number of emergency inpatient admissions to hospital over past 30 years - major source of pressure for the NHS

• Reasons for this increase are complex and include an ageing population and changing social factors increasing the demand for formal care

• A proportion of emergency admissions will be completely appropriate

• Growing evidence that a significant proportion of patients treated in A&E are not there because it is the best place for them to be treated

• It is important that we begin to understand what proportion of all emergency admissions could have been treated more appropriately elsewhere, and what that care could look like

Local policy context

• Shifting the Balance of Care

• Health & Social Care Integration

• Discussions between Public Health and Health and Social Care Partnerships (HSCPs) identified an understanding of emergency admissions as a top priority for the partnerships

• JSNA requested by a group representing the three HSCPs to aid strategic planning

Local policy context

• Analysis of emergency admission data, combined with review of the available evidence base, can help to focus or target interventions aimed at providing evidence informed integrated care

• Provide planners within HSCPs with an enhanced understanding of what proportion of all emergency admissions could have been treated more appropriately elsewhere, and what that care could look like

METHODS

Defining avoidable admissionsJSNA methodology

Defining avoidable admissions

• Primary analysis focused on Ambulatory Care Sensitive Conditions (ACSC) as identified by Purdy et al (2009)

– Incorporates 35 categories of conditions for which admission could be avoided by interventions in primary, community or social care

– Should not be considered as “inappropriate admissions” but rather one that could have been avoided if health and social care services were configured in a different way

– ACSC coding was based on the ICD-10 code in the primary diagnosis position (i.e. main presenting condition)

Defining avoidable admissions

• Secondary analysis focused on intentional and unintentional injuries

– ICD-10 codes for classification of the cause of injury, poisoning and other adverse effects

– These are designed to provide supplementary information to a primary diagnosis and as such are coded within diagnosis positions 2-6

JSNA methodology• Acute hospital discharge data (SMR01) extracted

– Continuous inpatient stays beginning/ending 2013/14– Ayrshire and Arran residents– Coded as an emergency admissions– Analysed by age, deprivation, gender and HSCP area

• Rates calculated either as age standardised to the 1976 European standard population or as age specific rates for five year age bands

– Allow comparison between groups and/or over time whilst discounting any difference in population age structure

JSNA methodology

• Two literature reviews were undertaken around reducing avoidable emergency admissions:

– review of systematic reviews of models of care which show potential for reducing avoidable emergency admissions

– review of current evidence base for condition-specific interventions with the potential to reduce avoidable admissions for the top five ACSCs in Ayrshire and Arran

ANALYSIS OF EMERGENCY ADMISSIONS

Ambulatory care sensitive conditionsInjuries

Summary of emergency admissions involving Ayrshire & Arran residents during 2013/14

All emergency admissions

Avoidable admissions -

ACSC

Avoidable admissions -

injuries

Number of admissions 48,378 17,621 7,559

% of emergency admissions 36% 16%

Number of patients 33,025 13,767 6,654

Continuous inpatient stays resulting from an emergency admission with primary diagnosis classed as ACSC 2013/14; by sex and HSCP area

Ayrshire & Arran

East Ayrshire North Ayrshire

South Ayrshire

Ayrshire & Arran

East Ayrshire North Ayrshire

South Ayrshire

Males Females

-

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

Eurp

oean

age

-sta

ndar

dise

rat

e pe

r 10

0,00

0 po

pula

tion

South Ayrshire's age standardised overall ACSC rates are lower than East and NorthNorth Ayrshire females’ age standardised ACSC rates higher than female A&A rate

Continuous inpatient stays resulting from an emergency admission with primary diagnosis classed as ACSC 2013/14; by age and HSCP area

0-4 5-9 10-14

15-19

20-24

25-29

30-34

35-39

40-44

45-49

50-54

55-59

60-64

65-69

70-74

75-79

80-84

85+0

5,000

10,000

15,000

20,000

25,000

East AyrshireNorth AyrshireSouth AyrshireAyrshire & Arran

Age

spec

ific

rate

s pe

r 10

0,00

0 po

pula

tion

Rates are high among 0-4 years before dropping sharply at 5-9 yearsSteady rise until around 70 years followed by sharper increase in age specific rates

Continuous inpatient stays resulting from an emergency admission with primary diagnosis classed as ACSC 2013/14; by SIMD 2012

quintile

SIMD 1 (most deprived)

SIMD 2 SIMD 3 SIMD 4 SIMD 5 (least deprived)

0

1,000

2,000

3,000

4,000

5,000

6,000

Euro

pean

age

-sta

ndar

dise

d ra

tes

per

100,

000

popu

latio

n

Ratio SIMD 1 : SIMD 5 (Ayrshire & Ar-ran)

All ages 2.26:10-19 years 1.43:120-64 years 2.96:165+ 2.06:1

Clear deprivation gradient observed (note partly explained by inclusion of emergency admissions in Health domain of SIMD)

Difference between most and least deprived quintiles greatest among working age adults

Top five ACSC categories (1st position) of emergency admissions in Ayrshire and Arran residents during 2013/14, all ages*

Frequency % of all emergency admissions

% of all ACSC conditions

Angina 4,174 8.6% 24%

Urinary Tract Infection (UTI)/pyelonephritis

1,759 3.6% 10%

Chronic Obstructive Pulmonary Disease (COPD)

1,652 3.4% 9%

Dehydration / gastroenteritis

1,295 3.0% 7%

Influenza/pneumonia 1,092 2.3% 6%

European age standardised rates per 100,000 population of continuous inpatient stays resulting from an emergency admission with primary

diagnosis classed as ACSC 2013/14; by top 5 categories and HSCP area

Ayrs

hire

& A

rran

East

Ayr

shire

Nort

h Ay

rshi

re

Sout

h Ay

rshi

re

Ayrs

hire

& A

rran

East

Ayr

shire

Nort

h Ay

rshi

re

Sout

h Ay

rshi

re

Ayrs

hire

& A

rran

East

Ayr

shire

Nort

h Ay

rshi

re

Sout

h Ay

rshi

re

Ayrs

hire

& A

rran

East

Ayr

shire

Nort

h Ay

rshi

re

Sout

h Ay

rshi

re

Ayrs

hire

& A

rran

East

Ayr

shire

Nort

h Ay

rshi

re

Sout

h Ay

rshi

re

Angina COPD Dehydration/Gastroen-teritis

Influenza/Pneumonia UTI/Pyelonephritis

-

200

400

600

800

1,000

1,200

Euro

pean

age

-sta

ndar

dise

d ra

te p

er 1

00,0

00 p

opul

a-tio

n

Ratio SIMD 1 : SIMD 5 (Ayrshire & Arran)

Angina 2.33:1COPD 9.37:1Dehydration/Gastroenteritis 1.58:1Influenza/Pneumonia 2.50:1UTI/Pyelonephritis 1.56:1

Continuous inpatient stays resulting from an emergency admission classed as unintentional or intentional injury (position 2-6) 2013/14

0-4 5-9 10-14

15-19

20-24

25-29

30-34

35-39

40-44

45-49

50-54

55-59

60-64

65-69

70-74

75-79

80-84

85+ -

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000 Unintentional injury

Intentional injury

Age bands

Age

spe

cific

rate

s pe

r 100

,000

pop

ulati

on

Continuous inpatient stays resulting from an emergency admission classed as unintentional or intentional injury (position 2-6) 2013/14;

by HSCP area

Ayrshire & Arran

East Ayrshire North Ayrshire

South Ayrshire

Ayrshire & Arran

East Ayrshire North Ayrshire

South Ayrshire

Unintentional injury Intentional injury

-

200

400

600

800

1,000

1,200

1,400

Euro

pean

age

-sta

ndar

dise

d ra

te p

er 1

00,0

00 p

opul

a-tio

n

Ratio SIMD 1 : SIMD 5 (Ayrshire & Arran)

Unintentional 1.98:1Intentional 3.83:1

SUMMARY OF THE CURRENT EVIDENCE BASE

Models of careCondition specific interventions

Evidence base – some caveats• The models vary considerably in design

(apples V pears)

• Evaluation studies also varied in: design, duration, target population, local v regional, etc

• Difficult to identify the elements of complex models that may/not be effective

• Effective models might be lost in the overview of a review

• Models might not reduce admissions but may have other desirable outcomes, both clinical and patient/carer-centred

• Evidence of effectiveness can take time to establish – often looking at a lack of evidence rather that evidence that model not effective

• Strong association between avoidable admissions and measures of deprivation, so interventions must reflect this

• As most avoidable admissions are due to a range of factors, no single model or intervention will be effective in reducing admission rates, therefore a whole-systems approach will be required

• There is a clear need to develop robust evaluation, both local and national if possible, when introducing any new models of care without a robust evidence base

We recommend ensuring adequate funding and resource is in place for the following interventions aimed at reducing avoidable hospital

admissions:

• Interventions at A&E: review by senior clinician and GP-led assessment units for urgent referrals from community GPs

• Integrated Clinical Care programmes for heart failure, COPD, asthma and diabetes

• Exercise-based rehabilitation for CHD

• Case management for heart failure

• Home visits (plus telephone support) for heart failure patients; pregnant women with hypertension and/or diabetes, and mental health patients

• Self-management, including practitioner review, in asthma and COPD patients

• Specialist clinics for heart failure patients

• Assertive Community Treatment for mental health patients

• Managed Clinical Networks (MCN) in patients with angina and diabetes

• Tele-related health care in older people and in people with heart failure, CHD, hypertension and diabetes

For reducing avoidable admissions for COPD exacerbations, we recommend ensuring adequate resource and funding is in place for:

• Smoking cessation to be offered to all patients with COPD

• The step-wise approach to drug therapy as outlined in the NICE Guideline for COPD

• Pulmonary rehabilitation for all patients with moderate to severe COPD

• Influenza vaccination for patients with COPD

Evidence to date is inconclusive with respect to two models of care which are relevant to HSCPs:

• Hospital at home:

– Non significant increase in admissions compared to inpatient hospital care

– Varying degrees of success in reducing admissions or readmissions for specific patients and particular conditions

– Important to remember that hospital at home may be achieving other important patient outcomes

• Integrated care plans - horizontal integration between Health & Social Care:

– 16 heterogeneous pilots of health and social care integration initiatives in England did not provide evidence of reduced admissions

– May not have been realistic to expect such outcomes to emerge in short term

– Strong evidence that ICP within health systems, i.e. vertical integration between primary and secondary care, can reduce hospitalizations in patients with chronic conditions

– Vertical ICP involves combinations of MDTs for disease management, specialist clinics, out-patient support, plus patient education & self-management

CONCLUSION

JSNA provides planners within Health and Social Care Partnerships with an enhanced understanding of:

– The current state of emergency admissions in Ayrshire and Arran

– What proportion could have been avoided if primary, community or social care systems were configured differently

– Effective models of care for reducing avoidable admissions

• The scope of this report has been limited to these three areas

• Must not lose sight of the important role of disease prevention

• There is a preventable element to each of the top reasons for avoidable admission

• We have also illustrated the proportion of potentially avoidable emergency admissions due to injuries - clear role for more upstream prevention