Envisioning enhanced primary care in Singapore: a group model building approach 2 nd Asia-Pacific Region System Dynamics Conference John P. Ansah, PhD Assistant Professor Program in Health Services and Systems Research Duke-NUS Medical School, Singapore 1
Why is chronic care an issue? The chronic care problem is an episodic acute disease oriented system meeting a rapidly expanding long-lived population People are living longer with chronic conditions Many chronic conditions are unmanaged Unmanaged chronic conditions become complex This dynamic exacerbates system stress use of acute services (triply undesirable) Inefficiencies due to service mismatch: PCPs making referrals to specialists, specialists trying to provide primary care 2
Proportion with Multiple Chronic Conditions People are living longer with chronic conditions 3
Service mismatch: Many beds filled with people not needing acute care Based on reviewing case notes of Geriatrics ward in an acute hospital in Singapore 4
Enhancing Chronic Care in Singapore An effective primary care system is key to successful chronic disease management Coordination of health needs across the various medical and health-related social services Accessible, affordable and patient-centric Education and illness prevention However, Singapore s health system poses unique challenges Multiple stakeholders with competing interests Constraints on number and training of providers Limited interface between public and private sector health services 5
Primary Care in Singapore Polyclinics 18 polyclinics Attends to 45% of chronic patients Payment is combination of out-ofpocket, Medisave payment and subsidies Provides comprehensive range of services outpatient medical care, health screening, education and vaccinations, x-ray and laboratory services Patients assigned any available doctor from a common group of medical officers and family physician Private GP Clinics ~1,500 private GP clinics Attends to 55% of chronic patients Payment is usually on a fee-forservice basis and Medisave payment and subsidies not always available Usually do not possess on-site facilities Service provision is heterogenous some offer aesthetic medicine Patient sees the same solo physician 6
Primary Care Roundtable 2015 50 stakeholders with interests in chronic care GPs, polyclinic doctors, medical educators, representatives from hospitals and ministries, as well as health services researchers Group model building (GMB) exercise Presentations of perspectives Preliminary model Insights Subsequent integration of results into a more detailed model, capturing key dynamics 7
Presentation of perspectives Public primary care providers Bailing water from a leaky boat Private primary care providers Many willing but what s the business case? Government Ready to support major innovation International experts Enhanced chronic care works but the devil s in the details 8
Consensus: The quadruple aim A health system consists of multiple sectors: acute and chronic, generalist and specialist, public and private, and so on. This manifold enterprise is intended to produce a mix of services that maximizes the health of the population (effectiveness), while maintaining sustainable costs (efficiency) and a high level of patient and provider satisfaction. An optimal mix of these features is the so-called quadruple aim of health care. 9
Consensus: The level of analysis = needs segments A framework for planning, implementing, and evaluating health and health-related social services to optimize population health Objective of our efforts = meeting the needs Need = health and social features that increase risk for bad outcomes, and risk can be alleviated by providing specific services 10
Segments and typical needs I. Healthy, at risk II. III. Stable chronic conditions(s) Complex chronic conditions I. Long course of decline II. Limited reserve and serious exacerbations Physician-level services & procedures Patient self-management & education Home services (non-medical) Befriending services Care coordination Medication adherence Caregiver support & education Day care Skilled nursing services Monitoring of symptoms, signs, and biomarkers and prompt followup Hospice care (palliative care) 11
Healthy, at risk Number of doctors Frequency of seeing the same doctor Length of consultation Stable chronic condition Service gap Doctorpatient relationship Model of care efficiency Visit waiting time Building relationships Clinic volume IV Provider work life Population health I Population health Complicated chronic condition III Percapita cost Change in positive patient experience II Positive patient experience Attractiveness of clinic Out-of-pocket cost 12
time to adjust relationship initial doctor patient relationship relative frequency of seeing same doctor mortality rate healthy births relative doctor patient relationship doctor patient relationship average doctor patient relationship building relationship relative length of consultation frequency of seeing same doctor Translating the conceptual model into a dynamic simulation model initial regression rate regression rate mortality rate stable mortality healthy mortality stable regression proportion of population with complex condition healthy, at risk incidence stable chronic condition progression complicated chronic condition <total population> mortality complicated incidence rate average service gap progression rate mortality rate complicated cost per person relative service gap service gap total cost of service provision total cost of care waiting time relative waiting time change in satisfaction clinic volume time to change patient satisfaction unit cost of care cost of hospitalization max time per doctor initial patient satisfaction patient satisfaction hospitalization rate healthy cost per hospitalization length of consultation estimated doctors average patient satisfaction uptake rate of services relative patient satisfaction out of pocket cost hospitalization rate stable chronic hospitalization rate complicated number of doctors affordability total attractiveness attractiveness of clinic cost per doctor total labor cost 13
Interaction between care venues time to adjust relationship initial doctor patient relationship initial regression rate regression rate mortality rate stable mortality rate healthy mortality healthy mortality stable regression initial regression rate births mortality rate healthy healthy, at risk births incidence rate incidence mortality healthy mortality rate stable stable chronic condition mortality stable progression regression complicated regression chronic condition rate proportion of population with complex condition mortality complicated healthy, at risk average service gap incidence complicated chronic condition incidence rate mortality rate mortality proportion complicated of complicated population with <total complex condition population> total cost of cost per mortality care rate person complicated <total population> relative doctor patient relationship service gap cost per person waiting time relative doctor patient relationship max time per doctor total cost of service provision total cost of care clinic volume max estimated time per doctors patient satisfaction change average in patient satisfaction satisfaction time to change patient satisfaction doctor patient relative relationship length of consultation average length doctor of patient consultation relationship uptake rate of services relative frequency initial doctor of time to adjust seeing same patient doctor relationship relationship building relationship relative length of consultation number of doctors length of consultation average patient satisfaction estimated total attractiveness doctors out of pocket cost hospitalization unit cost of rate healthy carehospitalization rate stable chronic cost per hospitalization hospitalization hospitalization rate ratehealthy hospitalization rate cost of complicated stable chronic hospitalization cost per hospitalization hospitalization rate complicated relative frequency of seeing same doctor number of doctors service gap waiting time average stable chronic service gap relative waiting time clinic volume uptake rate of condition attractiveness of relative service relative patient services clinic gap initial patient satisfaction satisfaction relative cost per progression waiting time doctor rate relative service affordabilityrelative patient progression gap initial patient satisfaction satisfaction patient satisfaction out of pocket change in progression cost total labor satisfaction rate cost affordability total cost of service provision time to change patient satisfaction unit cost of care cost of hospitalization doctor patient relationship average doctor patient relationship building relationship frequency of seeing same doctor Enhanced primary care frequency of seeing same doctor total attractiveness attractiveness of clinic cost per doctor total labor cost Normal Primary care 14
Policy experimentation Equilibrium Effectiveness of normal and enhanced primary care Out-of-pocket costs Number of doctors Effective enhanced primary care Service gap in enhanced primary care reduced from 0.5 to 0.1 at time 5 Service gap in normal primary care remained at 0.5 15
Policy experimentation Reduced out-of-pocket costs: enhanced primary care Service gap in enhanced primary care reduced from 0.5 to 0.1 at time 5 Service gap in normal primary care remained at 0.5 Out-of-pocket costs for enhanced primary care set at half that of normal care Proactive increase of enhanced primary care providers Service gap in enhanced primary care reduced from 0.5 to 0.1 at time 5 Service gap in normal primary care remained at 0.5 Out-of-pocket costs for enhanced primary care set at half that of normal care Supply of enhanced primary care providers (number of doctors) proactively increased in response to demand 16
$/person/year Dimensionless Dimensionless Dimensionless Effective Reduced Proactive Equilibrium: enhanced out-of-pocket increase of primary enhanced All costs key care: for policy primary The effectiveness variables care: providers: Reduced were of receiving kept out-of-pocket Reduced constant care out-of-pocket at costs an for enhanced costs for enhanced primary care primary venue care, was and assumed proactive to be increase significantly in enhanced higher primary to that of normal care providers primary care 0.16 Proportion of population with complex condition 2.5 Average patient satisfaction 0.12 2.1 0.08 1.7 1.3 0.04 0.9 0 1 6 11 16 21 26 31 36 41 46 51 56 61 0.5 1 6 11 16 21 26 31 36 41 46 51 56 61 Equilibrium Low out-of-pocket cost Effective enhanced PC Proactive increase in providers Equilibrium Low out-of-pocket cost Effective enhanced PC Proactive increase in providers 500 Cost per person 1.6 Average doctor patient relationship 475 1.4 450 425 400 375 1.2 1 0.8 350 1 6 11 16 21 26 31 36 41 46 51 56 61 0.6 1 6 11 16 21 26 31 36 41 46 51 56 61 Equilibrium Low out-of-pocket cost Effective enhanced PC Proactive increase in providers Equilibrium Low out-of-pocket cost Effective enhanced PC Proactive increase in providers
Insights GMB facilitates the identification of important system interactions A major potential policy option is to empanel patients in high risk segments under contract (e.g., in the UK mode of commissioning ) The private sector has capacity and willingness; need a business case Enhancing chronic care requires enhancing the capacity and capability of primary care Finding the optimal mix of quadruple aims To work we must consider the broader health care ecosystem 18
How does an enhanced chronic care sector impact the ecosystem? Investment: Staff Facilities Subsidies Training Rules of engagement: Empanelment Payment rules Regulations ENHANCED CHRONIC CARE One point of contact Coordination High capability Means: public-private partnership Staff & patient satisfaction Improved care Shift to private sector Reduced SOC & hospital use 19
Future Work What are the met and unmet needs? Identifying needs by segment, not just by diseases Understanding the dynamics of needs over time What makes different modes of care more or less attractive To providers To patients Engage patient groups: not involved in GMB Simulating current and potential policies Evaluating different policies in silico Performing sensitivity analysis: assessing confidence and further data needs Use models to track, reassess, improve iteratively 20
Thank You