/
How can academic research and modelling add value to NHS de How can academic research and modelling add value to NHS de

How can academic research and modelling add value to NHS de - PowerPoint Presentation

pamella-moone
pamella-moone . @pamella-moone
Follow
385 views
Uploaded On 2017-03-21

How can academic research and modelling add value to NHS de - PPT Presentation

Mr Andrew Fordyce FRCS Dr Mike D Williams Dr Mike Allen How the partnership story began 247 system reliability Built academic clinical partnership Need to save money business question if we make changes aimed to reduce ID: 527794

patients bed hospital ward bed patients ward hospital demand emergency amp hour medical ambulance patient beds los model admissions

Share:

Link:

Embed:

Download Presentation from below link

Download Presentation The PPT/PDF document "How can academic research and modelling ..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.


Presentation Transcript

Slide1

How can academic research and modelling add value to NHS decision makers?

Mr

Andrew Fordyce FRCS, Dr Mike D Williams. Dr Mike AllenSlide2

How the partnership story began

24/7 system reliability

• Built academic – clinical partnership• Need to save money, business question “if we make changes aimed to reduce LoS, can we close some beds?”Slide3

Setting the context

“People working in healthcare increasingly have to do more with less. ...working under conditions they would rather avoid in which the safety margin for those they are caring for has been greatly diminished.”

Runciman

B, Merry A, Walton M., 2007 Safety and Ethics in Healthcare, Ashgate

, Aldershot.Decision makers need assistance in making hard choices in the face of many competing demandsSlide4

Research approach and methods

Taking a systems thinking perspective – complex socio-technical system

• Qualitative – interviews and observations in primary care, ambulance trust and acute hospital – patient pathway

• Quantitative – analysis of hospital PAS data and creation of discrete event simulation model to assess bed occupancySlide5

Key findings for decision makers

When looking at the flow of urgent patients we provided evidence as to some of the reasons why there are daily peaks and variation in demand at the hospital and the problems created

GP working practices• Ambulance prioritising 999• Staffing and productivity of clinical micro systems (clerking)• Impact on wider hospital – discrete event simulation of demand patternsSlide6

GP working practices

Practices facing high demand – prioritise surgery based appointments – no willingness to change

Batch ‘visits’ (create higher number of referral to hospital) as they are not an ‘efficient’ use of doctor time• No standard method of communication to the hospital• Request ambulance – 8mins (999) of 4 hoursSlide7

8 minutes or 8 hours to treatment

GPs visit sickest patients 1 - 3pm – then phone for ambulance (HCP calls)•

Ambulance prioritise 999 response < 8mins therefore GP call as ‘urgent’ <4 hrs• Patient arrives at hospital late afternoon / evening• Patient’s need subordinated to local optimisation of parts of system“Visits are a very inefficient use of GP time.” “Achieving the 999 target is our priority.”Slide8

This area for large pictures/charts/

tables,etc with one line captioning.

Arrival and discharge patterns by hour of day – change demand pattern or design services to copeSlide9

Helping managers understand normal variation around the mean

Panic

– admissions have risen by 7%

no – it is 12%,

some say 15%Acute emergency admissions have been rising at ~1.6% per annumSlide10

A question

How many emergency medical patients does an F1 doctor process (clerk) in A&E on average during an 8 hour, 9 – 5pm shift? Slide11

Clerking Capacity – staffing to meet demand?

This area for pictures/charts/tables,etc

Note: Clerking capacity is estimated based on planned rota of staff assuming an average of 1hr per patient

Weekday

WeekendSlide12

Inefficient clinical micro systems

“...someone will have taken the notes to reception to be photocopied...”

“As an F1, it happens to us all, from nine to five you might see

four

patients. There is a general feeling that if you can see four full patients from scratch and do everything, that’s not bad for an F1 doctor in an eight hour shift. If you actually looked at the amount of time doing medicine it is probably less than a quarter of the time because of the amount of time, you know, you have to spend running around and chasing up on different issues.” “When you take bloods they get left in a pot in A&E, then a porter circulates maybe once every half an hour or forty minutes, so that is half an hour to forty minutes for your blood test sat there not being examined and then they go to the lab to be looked at.” Slide13

Modelling bed occupancy – key themes

Understanding & modelling demand variability at whole hospital and specialty levelDoctors would like bed pools sufficiently large to cope with demand variability for their own specialty

• What are bed requirements given expected changes in systemIncreasing emergency admissions (~2% per annum)Service Improvement Programmes to reduce length of stayCould bed reductions be achieved based on assumptions being made?Slide14

Variability in 2012 emergency admissions

15%CV

4

5% CVSlide15
Slide16

Medical & surgical patients* – midnight count(*Patients categorised by consultant at discharge)

Un-escalated bed stock = 328 (inc EAU & ICU)

Escalated bed stock = 351Slide17

Medical patients – midnight count

Un-escalated bed stock including EAU = 208 beds

Escalated bed stock = 236 bedsSlide18

Model Logic

Placing patient on ward:

Preferred ward(s) for specialtyEscalate preferred ward(s)

Ward of same division (medical/surgical)Escalate ward of same division

Ward of different divisionEscalate ward of different divisionOverflow  Cancel 1 elective procedure for each midnight overflow patientArrivals, routing and lengths of stay are dependent upon specialty & whether elective or emergency admission.# Arrivals adjusted by average for weekday,Outliers are not repatriatedOverflow patients are repatriated once/day Outlier 1 = Non-preferred ward for specialtyOutlier 2 = Ward of different divisionSlide19

This area for large pictures/charts/

tables,etc with one line captioning. Example scenarioSlide20

Model conclusions

Expected LoS reductions (in SIPS) will not allow for closure of beds

In order to close beds LoS reductions significantly greater than anticipated would be required• The model was used to explore a range of scenarios, such asAltering medical/surgical bed balanceVarious bed numbers and LoS reduction combinationsSmoothing elective flow over 6-7 days (in place of 5 days)Differing assumptions on emergency admission growthSlide21

The Impact:

Not ready for bed closures

Speciality to dependency based modelTesting weekend workingBuilding a longer term partnership between NHS in South Devon and the University of Exeter

Contact us:

andrew.fordyce@nhs.netm.allen@exeter.ac.uk; m.d.a.williams@exeter.ac.uk