Dr Dan Beckett
Consultant Acute Physician
NHS Forth Valley

Warning signals
◦
◦
◦
◦
◦

Four hour emergency access standard
ED LoS - time profiles
Boarding
(Cancelled elective activity)
(Delayed discharges)
Whole system overview
◦ NHSFV capacity and flow dashboard

Elective vs Emergency imbalance
◦ Optimising patient flow by reducing its variability

Four hour emergency access standard
◦ Useful as an indicator of whole system pressure
◦ Poor compliance indicates with ED overcrowding
 Associated with an increase in mortality both in
patients admitted and patients discharged from the ED
◦ Limited usefulness as an early indicator of pressure
to trigger escalation

ED LoS distribution
◦ Can demonstrate pressure in the system that is not
evident when just looking at compliance with the
four hour emergency access standard
◦ ‘Crisis spike’

ED time curve
◦ Useful for retrospective analysis
◦ Crisis spike correlates with poor performance
◦ Useful for proactive escalation?
 Dynamic monitoring of the proportion of patients
leaving the ED after 210 minutes?
SRI A&E attendances 08:00 - 12:00 hrs
No. of attendances
5
4
3
2
1
BUT STILL 97% COMPLIANT AT
THIS STAGE
0
20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 240 260 280 290 300
Time in ED
SRI A&E attendances 12:00 - 16:00 hrs
No. of attendances
7
27%
6
5
4
3
2
1
0
20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 240 260 280 290 300
Time in ED
SRI A&E attendances 16:00 - 20:00 hrs
No. of attendances
7
6
5
4
91%
3
2
1
0
20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 240 260 280 290 300
Time in ED
SRI A&E attendances 20:00 - 00:00 hrs
No. of attendances
7
6
5
4
3
86%
2
1
0
20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 240 260 280 290 300
Time in ED
SRI A&E attendances 00:00 - 04:00 hrs
No. of attendances
7
6
5
4
3
79%
2
1
0
20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 240 260 280 290 300
Time in ED
SRI A&E attendances 04:00 - 08:00 hrs
No. of attendances
6
5
4
3
2
77%
1
0
20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 240 260 280 290 300
Time in ED

Boarders
◦ Different models of boarding exist
 Exclusively ‘front door’
 Exclusively ‘back door’
 Mixed model
◦ Irrespective of model, increasing numbers of
boarders indicates system pressure and should be
monitored/controlled
◦ Boarded patients have poor outcomes


NHSFV capacity dashboard
Real time information
◦
◦
◦
◦
◦
◦
Pressure vs Capacity
Admissions vs Discharges
Emergency vs Elective
Predicted vs Observed activity
Whole system vs Individual patient
Warning signals across the whole system as a
trigger to escalation



Competition between emergency and elective
flow ‘silos’ can directly lead to ED
overcrowding
Perceived conflict between the 18 week RTT
target and the 4 hour emergency access
standard
Significant variation in numbers of patients
admitted over the week
Hospital admissions, NHSScotland, October 2010
3000
131%
2500
2000
1500
1000
500
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Total
Hospital admissions, NHSScotland, October 2010
3,000
131%
2,500
2,000
1,500
54%
1,000
500
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Emergency
Total
Hospital admissions, NHSScotland October 2010
3,000
2,500
2,000
1,500
1,000
131%
BUT YOU CAN’T
COMPARE WEEKENDS
AND WEEKDAYS! 54%
3288%
500
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Elective
Emergency
Total
Hospital admissions, NHSScotland October 2010
3,000
46%
2,500
2,000
1,500
16%
1,000
237%
500
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Elective
Emergency
Total

Elective admissions display more variability
(artificial variability) than emergency
admissions (natural variability)
◦ Counter-intuitive!


Difficult to plan staffing levels for such high
levels of variation (largely artificial variation)
Invariably staffed for ‘average’ levels of
activity resulting in periods of demand >
capacity (leading to ED overcrowding and
poor outcomes) and capacity > demand
(waste of resources)
Queue
Demand
Capacity
Can’t pass
time
unused capacity
forward to next week
Reducing waiting times in the NHS: is lack of
capacity the problem?
Bevan et al Clinician in Management (2004)

Need to eliminate artificial variation and
manage natural variation
Eliminating artificial variation (Mon-Fri)
NHSScotland October 2010
2500
14%
2000
1500
1000
500
Artificial variability eliminated Monday-Friday
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Elective
Emergency
Total

Reduces overall variation
◦ Reduces ED overcrowding
◦ Less waste


Reduces patient boarding
In 2006 the IOM published a report asking
hospitals to use operational management
tools (queuing theory) to address patient flow
issues that lead to ED overcrowding

Boston Medical Centre
◦ Significant problems with ED overcrowding 2003
◦ Emergency work more predictable and less varied
than elective work
◦ Reprofiled elective cases Monday-Friday
 Subsequently eliminated all block scheduling
◦ Split elective and emergency surgical work
◦ Used queuing theory to guide resources for
emergency work

Boston Medical Centre
◦ Reduced variability in demand for surgical HDU
beds by 55%
◦ Reduced nursing hours – saving $130K per annum
◦ Reduced cancelled/delayed surgery from 334 to 3
(99.5%) for the same time periods April-September
2003/2004 (pre- and post-implementation)
◦ Reduced ED waiting time by 50% and improved ED
throughput by 45 minutes per patient

Now many examples of successful
implementation
◦ Cincinatti Childrens Hospital
 Weekday OR waiting time reduced by 28% (despite an
increase in case volume of 24%)
 Weekend OR waiting time decreased by 34% despite an
increase in volume of 37%)
 Capacity boosted by equivalent of 100 bed expansion
◦ Great Ormond Street Hospital

Assign responsibility for the patient flow
problem
◦ Chief Operations Officer or Vice President





Establish a multidisciplinary team
Collect and analyze data on bottlenecks
Eliminate or smooth artificial variation
Manage natural variation (queuing theory)
www.ihoptimize.org


Managing Capacity and Demand across the
patient journey. Clinical Medicine 2010.
10:1 13-15
Winter Pressures in NHS Scotland 2008-2009.
A report for the Emergency Access Team,
Scottish Government

Professor Derek Bell, Imperial College
Professor Eugene Litvak, Institute for
Healthcare Optimisation
Dr Claire Gordon, NHS Lothian
Bas Gough, Scottish Government
Guy Blackburn, NHSFV

Thanks for listening...



