Dynamic Modelling to Support Collaborative Planning and Decision Making

Case Studies October 2012

David Rees Founding Partner Synergia Ltd

Long-Term Planning in Local Government (2011/12)  (Work  Conducted  for  District  Council)  

David  Rees,  Synergia,  Auckland,  New  Zealand  

While  they  had  detailed  and  robust  financial  planning  underpinning  their  Long  Term  Plan,  a  District  Council  found  it  difficult  to  respond  quickly  to  requests  for  alteraIon  to  the  plan.  What  would  be  the  consequences,  for  example,  of  shiLing  a  major  capital  project  back  by  two  years?    What  would  be  the  consequences  of  adding  or  deleIng  any  of  the  projects  currently  in  the  plan?    While  their  financial  models  were  detailed  and  robust,  they  were  unable  to  answer  quesIons  such  as  these  in  an  easy,  flexible  and  speedy  way.     The  development  of  a  dynamic  simulaIon  model,  

calibrated  with  their  own  financial  model,    provided  them  with  a  tool  that  enabled  them  to  conduct  mulIple  ‘what-­‐if’  scenarios.    The  model  ‘dashboard’  allows  them  to  quickly  modify  assumpIons  in  their  LTP  and  see  the  consequences  for  revenues,  expenditures  and  their  overall  financial  posiIon  over  the  lifeIme  of  the  LTP.    

Using Systems Modelling to Integrate Multiple Workstreams within Energy Sustainability Research (2011)  (Work  Conducted  for  University  of  Otago  Energy  Research  Centre)  

David  Rees,  Synergia,  Auckland,  New  Zealand  

Faced  with  data  coming  from  mulIple  research  streams  within  the  mulI-­‐disciplinary  research  team,  the  research  centre  wanted  ways  of  integraIng  their  findings.  The  purpose  of  the  modelling  was  to  disIll  the  key  findings  from  the  different  research  streams  and  any  uncover  issues  that  may  have  emerged  during  the  research  process.    Phase  II  of  that  research  project  is  now  underway  and  over  the  next  four  years  we  will  be  working  with  the  research  team,  using  dynamic  modelling  to  integrate  the  research  workstreams,  and  use  the  simulaIon  capabilites  to  explore  future  scenario  arising  out  of  the  research.  

Phase II (2012 – 2016) Energy Culture II Energy sustainability in households, transport and SMEs Renewable Energy & the Smart Grid Exploring the supply and demand dynamics in a future based on extensive use of renewable energy sources

Regional Transport in Canterbury: Health Impact Analysis (2010)  (Work  Conducted  for  Environment  Canterbury)  

David  Rees,  Synergia,  Auckland,  New  Zealand  Dr.  Adrian  Field,  Synergia,  Auckland,  New  Zealand  

In  October  2009  Environment  Canterbury  iniIated  a  Health  Impact  Assessment  (HIA)  of  its  Regional  Land  Transport  Strategy.  The  aim  of  the  HIA  was  to  assess  the  links  between  transport  planning,  health  determinants,  and  health  outcomes  for  the  Canterbury  RLTS.      This  simulaIon  model  supported  the  HIA  by  exploring  the  links  between  transport  planning  and  health  outcomes  that  were  idenIfied  in  the  iniIal  scoping  workshop.    The  HIA  idenIfied  some  of  the  linkages,  such  as  those  between  safety  and  cycle  use  and  focused  its  analysis  on  three  key  areas;  safety,  mode  choice  and  healthier  environments.    The  aim  of  the  simulaIon  model  was  to  help  inform  policy  by  quanIfying  some  of  the  key  linkages  and  the  size  and  Iming  of  potenIal  health  impacts  resulIng  from  policy  opIons  being  considered  in  the  RLTS.  

Op$ons  for  Demen$a  Care  (2010/11)  (Work  Conducted  for  Health  Workforce  New  Zealand)  

David  Rees,  Synergia,  Auckland,  New  Zealand  Geoff  McDonnell,  AdapIve  Care  Systems,  University  of  NSW  Dr.  Ray  Naden,  Clinical  Director,  Synergia    

In  work  we  undertook  for  Health  Workforce  New  Zealand,  Synergia  explored  the  opportuniIes  for  improving  care  for  people  with  moderate  demenIa  in  the  home  and  community  secngs,  and  the  potenIal  impact  this  may  have  upon  admissions  to  aged  residenIal  care  (ARC).    The  report  provided  an  overview  of  the  modelling  used  to  explore  the  dynamics  of  home-­‐based  care  –  specifically  carer  stress  –  and  its  impact  upon  reducing  admissions  to  ARC.  The  report  then  provided  a  descripIon  of  the  models  of  care  required  to  bring  that  reducIon  about.    Because  demenIa  is  an  area  in  which  there  is  a  paucity  of  data,  our  modelling  had  to  bring  together  informaIon  from  a  number  of  sources.  Furthermore,  it  had  to  allow  a  range  of  scenarios  to  be  run  under  a  range  of  different  assumpIons.  The  model  allows  stakeholders  to  obtain  a  richer  understanding  of  what  the  future  possibiliIes  are,  the  constraints  upon  those  possibiliIes,  and  the  variables  that  have  an  impact  upon  determining  which  scenario  is  more  likely  to  come  to  pass.  

A Population-Based Approach  to  Planning  Mental  Health  Services  in  Primary  Care  (2010)  ((Work  Conducted  for  Health  Research  Council)  

David  Rees,  Synergia,  Auckland,  New  Zealand  Philip  Gandar,  Synergia,  Auckland,  New  Zealand  

MildSymptoms

ModerateSymptoms

SevereSymptomsbecoming

moderatebecoming

severe

need for MHservices

PMHCinterventions

SMHCinterventions

recoveringmoderate

recoveringsevere

recoveringmild

adequacy ofprovider resources

quality ofcare

prevention &management of risk

factors

investing inservice

improvement capability ofresources

amount ofresources

investing in riskmanagement

fundsavailable

investing in socialdeterminants

SocialStrength

change insocial strength

model of carerequirements

servicedemand

accesslevels

serviceprovision

<fundsavailable>

model ofcare

average level offunctioning

NoSignificantSymptoms developing

symptoms

developing moderatesymptoms

developing severesymptoms

individualattributes

RequiringSecondary

Careenteringsecondary care

discharging fromsecondary care

The  issues  that  any  region  faces  in  planning  Primary  Mental  Health  Care  (PMHC)  are  varied  and  complex.  There  is  no  one  soluIon  that  can  be  applied  across  the  country,  and  because  of  this  it  is  important  that  planners  in  each  region  know  their  own  populaIon  and  its  needs,  and  the  characterisIcs  of  the  people  and  resources  who  can  respond  to  them.  

This  model  is  designed  to  help  facilitate  conversaIons  about  PMHC  in  local  regions,  so  that  they  can  design  soluIons  that  best  fit  their  parIcular  circumstances.  It  takes  a  systems  approach  because  we  know  that  any  soluIon  that  does  help  improve  mental  health  services  will  be  required  to  address  many  issues.  IsolaIng  a  single  issue  simply  will  not  work.  To  facilitate  the  conversaIons  we  have  designed  a  model  of  the  key  elements  within  PMHC  and  how  those  elements  link  together.  The  model  is  based  on  our  conversaIons  with  planners  and  providers  within  each  DHB  and  focuses  on  key  themes  that  are  common  across  all.    

Review  of  Aged-­‐Care  Workforce  (2010)  (Work  Conducted  for  Health  Workforce  New  Zealand)  

David  Rees,  Synergia,  Auckland,  New  Zealand  Geoff  McDonnell,  AdapIve  Care  Systems,  University  of  NSW  Dr.  Ray  Naden,  Clinical  Director,  Synergia      

Trainees Workforce

Older People Receiving

Care

Service Configuration

A  System  Dynamics  (SD)  Model  was  designed  to  provide  a  framework  for  meeIng  the  challenge  of  developing  and  managing  the  future  aged-­‐care  workforce.  It  did  so  by  describing  the  dynamic  relaIonships  between  older  people  in  need  of  health  care  services,  the  services  that  have  been  established  to  respond  to  those  needs  and  the  workforce  that  exists  within  each  service.    Central  to  the  model  is  the  key  quesIon;  “What  is  the  workload  that  the  workforce  has  to  undertake?”  Furthermore,  the  model  highlights  that  workload  is  a  funcIon  of  those  receiving  care  and  the  configuraIon  of  the  services  designed  to  provide  that  care.    In  addiIon,  the  configuraIon  of  the  services  is  a  funcIon  of  the  work  needed  to  be  done  and  the  workforce  able  to  undertake  it.    As  a  consequence,  discussions  about  future  workforce  requirements  has  to  be  based  on  an  understanding  of  the  dynamic  interplay  between  each  of  the  three  elements.    The  need  for  care  was  modelled  by  using  funcIonal  impairment  as  the  key  modifiable  factor.  The  data  for  calculaIng  this  was  taken  from  the  Department  of  StaIsIcs  and  from  the  Australian  Bureau  of  StaIsIcs  survey  of  disability,  ageing  and  carers,  which  was  calibrated  for  the  New  Zealand  populaIon.  This  survey  (which  is  a  self  assessment)  provided  the  best  available  data  on  the  likely  levels  of  funcIonal  impairment  (disability)  in  the  populaIon.  FuncIonal  impairment  was  defined  as  any  limitaIon,  restricIon  or  impairment,  (physical  or  cogniIve)  which  has  lasted  or  is  likely  to  last  for  at  least  6  months  and  restricts  everyday  acIviIes.    Model  projecIons  indicate  that  those  65+  with  severe  funcIonal  impairment  will  rise  from  127,874  in  2010  to  207,409  by  2026.    Research  indicates  that  the  rates  at  which  people  develop  funcIonal  impairment  could  be  reduced  by  as  much  as  30%.  If  this  did  occur  the  numbers  of  people  with  severe  funcIonal  impairment  would  rise  to  175,178,  by  2026;  a  reducIon  of  43,000  when  compared  with  the  baseline.    

Exploring the Impact of Adherence to Asthma Medication on Healthcare Utilisation (2010)  (Work  Conducted  for  private  healthcare  provider)  

David  Rees,  Synergia,  Auckland,  New  Zealand  

Recently  a  private  healthcare  provider  completed  a  trial  of  a  medicaIon  adherence  programme,  which  involved  targeted  text  messaging  designed  to  change  percepIons  and  improve  adherence  to  asthma  preventer  medicaIon.  The  results  were  impressive,  showing  a  39%  increase  in  adherence,  versus  the  baseline,  aLer  6  months.    The  quesIon  that  this  raised  for  the  Company  was  whether  or  not  this  improvement  could  have  significant  enough  impacts  upon  healthcare  uIlisaIon  to  jusIfy  further  investments  in  the  programme.  Of  special  interest  was  whether  or  not  the  impact  upon  healthcare  uIlisaIon  could  be  significant  enough  to  interest  Pharmac  in  supporIng  the  programme.    To  help  answer  this  Synergia  was  commissioned  to  develop  a  dynamic  simulaIon  model  that  could  explore  the  impact  of  increased  adherence,  generated  by  programme,  on  healthcare  uIlisaIon.    This  would  then  enable  the  Company  to  make  a  more  rigorous  assessment  of  its  commercial  viability  in  the  New  Zealand  market.    

A  Whole  of  System  Approach  to  Compare  Op$ons  for    CVD  Interven$ons  in  Coun$es  Manukau,  New  Zealand  (2009)  (Australia  New  Zealand  Journal  Of  Public  Health.  (2012)  Volume  65,  Issue  3.)  

SmokingPrevalence

ObesityPrevalence

Secondhandsmoke

Poor dietfraction

Inadequatephysical activity

fraction

Stressedfraction

Diagnosisand control

First-time CVevent and death

rates

Use of smoking quitproducts and

services

Use of mental healthservices by stressed

Sources ofstress

Use of weight lossservices by obese

Use ofprimary care

Particulate airpollution

Use of qualityprimary care

Tobacco taxes andsales/marketing

regulations

Smoking bans atwork and public

places

Quality of primarycare provision

Anti-smokingsocial marketing

Highcholesterol

High bloodpressure

Diabetes

UncontrolledChronic Disorder

Prevalences

Non-CVDPopn

Post-CVDPopnFirst-time

events survived

Recurrent CVevent and death

rates

CV events anddeaths

Peopleturning 35

Non-CVD Popndeaths

Post-CVD Popndeaths

Timothy  Kenealy,  SecIon  of  Integrated  Care,  South  Auckland  Clinical  School,  University  of  Auckland,  New  Zealand  David  Rees,  Synergia,  Auckland,  New  Zealand  Nicolese  Sheridan,  SecIon  of  Integrated  Care,  South  Auckland  Clinical  School,  University  of  Auckland,  New  Zealand    Allan  Moffis,  Director  of  Primary  Care,  CounIes  Manukau  District  Health  Board,  New  Zealand  Sarah  Tibby,  Programme  Manager,  Long  term  CondiIons,  CounIes  Manukau  District  Health  Board,  New  Zealand  Jack  Homer,  Homer  ConsulIng,  Voorhees,  New  Jersey,  United  States.    

Objec$ve  To  assess  the  usefulness,  to  planning  and  funding  decision  makers,  of  a  naIonal  and  a  local  System  Dynamics  model  of  cardiovascular  disease.  Methods  In  an  iteraIve  process,  an  exisIng  naIonal  model,  based  on  earlier  work  by  Jack  Homer,  was  populated  with  local  data  and  was  presented  to  stakeholders,  in  CounIes  Manukau,  New  Zealand.  They  explored  the  plausibility,  usefulness  and  implicaIons  of  the  model.  Data  were  collected  from  30  people  using  quesIonnaires,  and  from  field  notes  and  interviews,  both  of  which  were  themaIcally  analysed.  Results  PotenIal  users  readily  understood  the  model  and  acIvely  engaged  in  discussing  it.  None  disputed  the  overall  model  structure,  but  most  wanted  extensions  to  the  model  to  elaborate  areas  of  specific  interest  to  them.    Local  data  made  lisle  qualitaIve  difference  to  data  interpretaIon  but  was  nevertheless  considered  to  be  a  necessary  step  to  support  confident  local  decisions.    Conclusion  Some  limitaIons  to  the  model  and  its  use  were  recognised,  but  users  could  allow  for  these  and  sIll  derive  use  from  the  model  to  qualitaIvely  compare  decision  opIons.  Implica$ons  The  System  Dynamics  modelling  process  is  useful  in  complex  systems  and  is  likely  to  become  established  as  part  of  the  rouInely  used  suite  of  tools  used  to  support  complex  decisions  in  CounIes  Manukau  District  Health  Board.  Keywords  Cardiovascular  diseases,  system  dynamics,  populaIon  health,  decision  making,  health  care  quality  access  and  evaluaIon