Modelling as an innovation tool Modelling as an innovation tool - PowerPoint PPT Presentation

Modelling as an innovation tool Modelling as an innovation tool Digital methods for better decisions Digital methods for better decisions Ian Gibson and Peter Bodon of Digital methods for better decisions

Modelling as an innovation tool Digital methods for better decisions Health

Modelling as an innovation tool Digital methods for better decisions Limits of current methods: Oversimplification Health is a complex system that is dynamic, Benchmarks, rules of thumb interdependent, variable and resource – relevance with rapid change??? dependent. - Use of average Fragmented consideration of Health relies on communication and facility design information flows. organisational change information and communications technology Lean methods Quality management Models of care – limited consideration of resources, variability.

Modelling as an innovation tool Digital methods for better decisions “We can't solve problems by using the same kind of thinking we used when we created them.” - Albert Einstein

Modelling as an innovation tool Digital methods for better decisions Leading US Nurses concluded from a study of nursing practices in wards that: “A holistic approach is needed whereby people, process, and technology come together harmoniously in a physical space to produce the maximum medical-surgical unit efficiency.” (Reference: A 36-Hospital Time and Motion Study: How Do Medical-Surgical Nurses Spend Their Time? The Permanente Journal/ Summer 2008/ Volume 12 No. 3)

Modelling as an innovation tool Digital methods for better decisions Approach is a design method using: Systems thinking Data analytics Computer simulation Project management Change management Facilitation Enables Thorough understanding of current system and likely future demands. Testing of the options – “what if scenarios” using same criteria as finished facility – before design is complete

Modelling as an innovation tool Digital methods for better decisions Computer simulation Computer simulation uses mathematics and logic to construct virtual models of real systems. The models enable understanding of the system and experiments to develop optimised systems. “Computer simulation is this mind -blowing Car crash performance advance that will increase the rate of innovation.” - Bill Gates

Modelling as an innovation tool Digital methods for better decisions “Computer simulation is this mind - blowing advance that will increase the rate of innovation.” - Bill Gates

Modelling as an innovation tool Digital methods for better decisions References on systems engineering approaches in healthcare Building a Better Delivery System: a New Engineering/Health Care Partnership. Proctor, P., et al., et al. Washington D.C. : National Academy Press, 2005. President’s Council of Advisors on Science and Technology. Report to the President Better Health Care and Lower Costs: Accelerating Improvement through Systems Engineering. Washington D.C : s.n., 2014.

Modelling as an innovation tool Digital methods for better decisions • Only design method which considers the whole system. • Enables a shared understanding of the system by clinicians and managers. • The impact of changes to any part of the system can be evaluated in terms of o capital and operating costs o health services performance can be evaluated.

Modelling as an innovation tool Digital methods for better decisions Healthcare delivery system Staffing Facilities Medical Technology People Patient journey People Communications Logistics Operating protocols Metrics technology

Performance Dashboard Evaluate options using same parameters for full scale operation. PSA enables better decisions by u sing Predictive Simulation and Analytics.

Modelling as an innovation tool Digital methods for better decisions Computer simulation The solution at the The model model level Risk free Experiment in a risk free experimentation space The world of models The real world ? The solution

Modelling as an innovation tool Digital methods for better decisions How does computer simulation relate to digital data ? Prescriptive Search for best outcome / predict Simulation and Business outcome ‘real world’ performance potential Optimisation lue for unseen system configurations Valu > Identify best course of action (local or across a value chain) Predict or classify outcome based on • Advanced Statistics Predictive past performance • Machine Learning > Focus decision making / support • Big Data • Dashboards Structured detailed analysis of past Diagnostic performance • Monitoring > Identify factors and causes • Alerts (typically local) • Data Gathering Aggregated and consumable data Descriptive > Understanding past performance • Curation • Reporting For oresig ight

Unifying Technology • Predictive Analytics • A methodology, – Not technology applications • Compliments and supercharges Lean, 6 Sigma, TOC methodologies • Risk free “Clinical trial” for system design

Reduce costly iteration

Modelling as an innovation tool Digital methods for better decisions Model of innovation Source : Diffusion of Innovation in Service Organisations: Systematic Review and Recommendations. Greenhaigh, Trisha , et al., et al. 4, s.l. : Blackwell Publishing, 2004, The Milbank Quarterly, Vol. 82, pp. 581-629.

Modelling as an innovation tool Digital methods for better decisions Organise Experiment to Develop and stakeholders and create better evaluate design decision -makers system options Process – ideas to implementation Agree brief for Implementation Model existing schematic design planning system

Modelling as an innovation tool Digital methods for better decisions Stakeholder relationships Technical specialists Interaction Clinicians Computer enabling and Simulation exceptional Managers Engineers healthcare Project Managers

Modelling as an innovation tool Digital methods for better decisions Data analytics

Modelling as an innovation tool Digital methods for better decisions Arrivals data Patient data for 41 months – 161,000 attendances

Modelling as an innovation tool Digital methods for better decisions Arrivals data Average arrivals per hour for all attendances peak at 8 – 9 arrivals per hour between 10:00 and 20:00 But what are the peak arrival rates???

Modelling as an innovation tool Digital methods for better decisions Peak arrivals data Peak arrivals per hour for all attendances peak at 11 – 21 in a range between 18 and 22 arrivals per hour

Modelling as an innovation tool Digital methods for better decisions

Modelling as an innovation tool Digital methods for better decisions Attendances per annum Patient Steams Patient groups Urgency category Paediatrics 15 - 34 35-65 Over 65 Total 1 53 99 137 159 449 2 779 1,296 2,410 2,241 6,725 3 2,728 4,198 4,635 4,002 15,562 4 5,857 7,485 6,166 3,537 23,044 5 650 1,103 803 240 2,797 Total 10,067 14,181 14,151 10,179 48,577 Treatment stream Resuscitation 53 99 137 159 449 Paediatrics 10,013 10,013 Emergency 1,296 ,410 2,241 5,946 Acute 4,198 4,635 8,832 Short stay 7,539 7,539 Fast track 8,588 6,969 240 15,798 Total 10,067 14,181 14,151 10,179 48,577

Modelling as an innovation tool Digital methods for better decisions 00:00 to 6:00 Factor Number 7:00 Paediatrics Number of patients 10,013 Ambulance arrivals 6% Private arrival 94% Police arrival 0.1% 3 patients per hour Admitted 14% 16:00 to 21:00 8:00 to 15 Discharged 86% Parameter Parameter Hours of day Distribution 1 2 00 to 6 Negative binomial 4 0.918 7 Negative binomial 8 0.956 22:00 23:00 8 to 15 Negative binomial 13 0.907 16 to 21 Negative binomial 32 0.943 22 Negative binomial 84 0.896 23 Negative binomial 6 0.871

Modelling as an innovation tool Digital methods for better decisions Communication survey

Modelling as an innovation tool Digital methods for better decisions Communications data • Collected by Cisco • 6 days of study (Monday through to Saturday) • 96 hours of observation across all roles • 1721 Communication events recorded • Broad coverage of all the major ED roles

Modelling as an innovation tool Digital methods for better decisions Communications process

Modelling as an innovation tool Digital methods for better decisions Fast track patient - consultation Process mapping for each patient stream

Modelling as an innovation tool Digital methods for better decisions The Model

35

Modelling as an innovation tool Digital methods for better decisions Validation Modelled current NEAT without improved • Validation is checking the model is a communications technology reasonable representation of reality. • Graph compares the actual NEAT performance (brown line) with the result of the model performance. • Comparison of actual 76% compared with the model of 79% Actual NEAT • Difference of 3% is within the ball park

Modelling as an innovation tool Length of Stay Digital methods for better decisions

Modelling as an innovation tool Digital methods for better decisions Le Length of f Stay by Pati tient Type 38