SIMUL8 Healthcare: Cross-boundary, Cross-sector – using simulation

Cross-boundary, Cross-

sector – using simulation

Claire Cordeaux: Executive Director, Healthcare

SIMUL8 Corporation | SIMUL8.com | [email protected]

Agenda

• Healthcare policy agenda

• How simulation can help:• Prevention

• Screening for Hepatitis C

• Long Term Conditions

• Year of Care

• Schizophrenia

• Emergency Care Flow

• Community Services

• Questions

• Supporting people

outside hospital:

– Provides more accessible care

– Prevents exacerbation

– Saves unnecessary visits (and

expense)

– Speeds up hospital discharge

– Reduces Length of Stay

– Improves patient outcomes

Health Policy

Health Policy


• International studies

• But what does that mean for us?

• Hospital at home

• Intermediate care

• Early discharge

• Admission avoidance

• Transfer of care

• Telemedicine

The Evidence


Integrated Care

• Care organised around

the individual


Why simulation?

• A service and system redesign

• Understanding the impact of changing service

utilization on:

– Flow

– Cost

– Capacity/Resource

• No historic data

• Different impacts on organizations, costs and

patients


Health and Care System Flow

Lack of capacity?

Rural/urban

population?

Lack of access? Vulnerable

groups?

Not 24/7?

Click to edit Master title styleClick to edit Master title style


SIMULATION APPROACH:

PREVENTION


Case Study 1: Hepatitis C Screening –

preventing liver disease

• Impact of future demand if a new Hep C

service is delivered locally, increasing

patient attendance:– on future burden of disease

– projected treatment costs for the service


Simulating demand

Age-banded population projections

Age-banded disease

prevalenceDemand

749,805 X Hep C 0.45% = 2771(2.5% diagnosed, 55% not Genotype 1)


Rising demand

Simulated Demand: Catchment population


Diagnosis (Current State)


Burden of Disease

All patients undiagnosed or not cured go to

“warehouse” where disease progresses with

an annual cost


Treatment – 30% Dual, 70% PI


Costs

Step Unit CostInitial screening 6.3Further Screening 42.83Nurse Consultation 25 20Genotype test 52.61FBC 3 2.66HCVRNA and FBC 16.34HCVRNA 13.68

Consultant at initial treatment and 6 months SVR 55.98Warehouse disease progression p.a. 882Compensated Cirrhosis p.a. 1400Decompensated cirrhosis p.a. 11,218Carcinoma p.a. 9,996


Validation

Activity Data Source ScenarioGenerator result

Confirmed cases 56 Annual Hep Report 2012 p46

56

Commencing treatment

42 Annual Hep Report 2012 p44

42 (moderateand compensated cirrhosis)


Treated as current vs Treated in local

hospital– cost over 5 years

• Assumed 50% of currently treated patients

do not attend after initial appointment

• Costs reduce with increased diagnosis

• Costs include annual inflation


Patients cured and cost – 5 yearsBy:

• current state

• future state

• increasing diagnosis by 5%, 10% and 20%

Patients Cured by ScenarioIncreased diagnosis = increased patients cured

Cost per patient cured - decreases with increasing diagnosis


Annual Savings by Scenario with new

local provider

Driving business case for change




SEGMENTING POPULATIONS


Case Study 2: Long Term Conditions Year of

Care

Using risk stratification to identify and manage

patients with multiple conditions and test:

• What if they are proactively managed or

unmanaged?

• What if we applied an annual tariff?


Starting to simulate a new approach

Services “consumed”

Assessment of Need

Patients at Risk

Exacerbation


• No real correlation between risk score and level of need

But…

Assessment of Need

Patients at Risk



WHAT THE DATA IS

TELLING US

Kent whole population data

Over 30% of people over 75 years have multimorbidity

Multimorbidity is more common than single morbidity


The total health and social care cost is strongly related

to multimorbidity


The main contributors to total health & social care cost

are acute non-elective admissions


People with complex health & social care needs appear

to demonstrate a ‘crisis curve’


More community, mental health and social care

services are delivered to people following a ‘crisis’

than before the ‘crisis’


Some indications that an integrated care plan changes

the pattern of services delivered to people

BHR Costing Data

• Use local

data to test

assumptions

• Ability to

update and

review

Simulation

• Level of

acuity

• Increasing

numbers of

long term

conditions

Current Simulation

• Likelihood of patients accessing services by

changing state of patients (state transition)


• Patients in each “state” have

– A likelihood of accessing certain types of service

(Acute, Community, Mental Health, Social Care),

including accessing services more than once

• Costs associated with those services

How it works


Data builds an underlying discrete event

simulation model

• Number of

patients in

each “state”

by year

• Average cost

per patient

Results

• Cost by each area of service/organisation

Results

• Costs by state per year

• Average cost per patient

• Comparison with tariff

Results


Results Comparison

Trial results


• Known to integrated care team or not?

• Test against proposed tariff?

• Change variation in cost for services?

• Decrease transitions through states?

Scenarios


• Planning for demand

• Testing an improvement scenario

• Negotiation between healthcare providers

How is this helping?


Case Study 3: Improving mental health


Purpose of simulation

• To reduce pressure on acute mental health

services

• Test the impact of additional referrals to

– Early intervention services

– Rehabilitation

On costs and resource utilization


Results


Feedback from Janssen partners:

“encouraging”

• Trusts are interested and want to put their

own data in the model

• Generating engagement and discussion in

change

• Testing service redesign – business case

for change without jeopardising patient

• Efficient way of looking at the numbers

• Cost savings



SIMULATION APPROACH: WHOLE

SYSTEM FLOW


Case Study 4: Improving the emergency

care flow North Staffordshire

• Impact of increasing out of hospital services on

cost and capacity


• What does current unscheduled care flow

look like?

• What will it look like in 5 years taking into

account population change?

• What is the impact of increasing referrals to

domiciliary care direct from hospital?

Initially to answer following questions

Initial Model

Area NHS data

Scenario Generator

%

A+E 108,472125,302 (17,026 out-of-area)A&E out of area (5% S Staffs) 17,000

0.99864512

Total NEL Admissions 84,297 84,4701.00205227

Elective admissions 12,674 12,7101.00284046

Daycase 49,983 49,8950.9982394

Discharges to CommunityHospital

4560 4507

0.98837719

Discharge to social care teams (Stoke)

2183 2203

1.0091617

Discharges from Community Hospital

4347 4430

1.01909363

Intermediate Care (admission avoidance)

590 581

0.98474576

• Ran the model

through with

the received

population

data

• Set routing

percentages

so model

matches

activity data.

Baseline Results – 10 run trial

Item £ LOS

Hospital Bed £500 a day AMU/SAU/CDUInpatient

CommunityHospital Bed

£263 per day 21 days

Intermediate care £47 per hour 30 hours

A&E £105.5

Cost and Length of Stay Assumptions


With population increase

In 5 years

+ £11.3m (£1m domiciliary care)

(1% annual inflation)

Increase in A&E and

admissions over 9 years


Potential Domiciliary Care Scenario

• Average 6 week package for rehabilitation

• Other packages average 48 weeks

Scenario:

• Increase direct referrals from hospital – 30% of community

hospital referrals

• Average 2 additional days in hospital

• Referrals 10% to complex, 38% maintenance, 51% re-ablement


Cost per hour

Hours pw (normal)

LOSwks

Capacity (hrs pw)

Packagespw

Discharges to reablement from community 2.50% £20.98 11 6 1400 127

Discharges to reablement from acute 10.10% £20.98 11 6 1400 127

Discharges to maintenance care from community 4.50% £13.20 7 48 4100 586

Discharges to maintenance care from acute 7.60% £13.20 7 48 4100 586Discharge from reablement to maintenance 15% £13.20 7 48 4100 586

Discharge to complex £13.20 22 48 4100 186

All discharges from acute (stoke) 2183

All discharges from community (stoke) 876

Domiciliary Care Assumptions

• £2.6m savings overall

– Plus £4m social care

– Plus 1.3m additional LOS, max bed occupancy +

10, +1% utilization

– £7.6m savings community hospital, utilisation

reduced by 25%, max bed occupancy minus 90

Domiciliary care scenario results


Impacts

• Understanding the financial impacts

• Allows negotiation across providers and

between payers and providers


So let’s imagine…

• We have prevented disease

• We are managing long term conditions

effectively in the community

• Hospital admissions are reduced

• But how are community services coping?




OPERATIONAL IMPACT ON SERVICES


• What is the impact of improvement interventions

on a community team workload?

• For example: what is the impact of faster healing

wounds on workload (60%)?

– More time to care?

– More time to see other patients?

• Engaging with community team – what are the

pain points?

Case Study 5 - Impact on Community team

capacity


Community

Team

Patients

Daily allocation to

staff matching

patient need to

competencies

Referrals

Visits

Discharge or Death

Ageing Population

Clinical Assessment

Wound care only

Multi-morbidity

Not wound

care



Localising and testing improvements on

capacity and patient throughput.

• Change referrals, patient

types and priority

• Change visit times and

frequency

• Include travel times

• Test impact of

improvements by

condition on capacity and

patient wait times


Results – supporting evidence for

change


Final Thoughts

• Simulation supporting understanding of the

health and social care system

• Evidence:– To support cross-border/cross-sector dialogue

– To better understand the impact of change

– To inform financial and operational decision-making

Helping to define and solve problems



QUESTIONS?

SIMUL8 Healthcare: Cross-boundary, Cross-sector – using simulation

Healthcare

service simul8 corporation

healthcare simul8 corporation

individual simul8 corporation

pi simul8 corporation

patientssimul8 corporation

annual costsimul8 corporation

new hep c service

rising demand simulated