Understanding Variation and Setting Capacity
Why do we get backlog and queues? Because demand exceeds activity Because we want to be efficient Because we do not take account of variation variation in demand variation in capacity the interaction between the two
3 GP Discharged! Staff skills illness holiday motivation training shifts Patients Kit Process rooms not the same supplies machines not the same age sex race education motivation disease unclear guidelines differ complications anaesthetics 80% is under our control Information transcription transport applications Sources of variation in a clinical system
What variability? GP –Number of patients –Number of problems –Investigations –Length of appointments Outpatients –Number of referrals –Number of staff –Investigations needed –Length of consultation
What variability? Ward –Length of pre-admission stay –Length of post-op stay –Intensity of nursing required –Staffing levels Theatre –Number of cases –Length of cases –Anaesthetic time –Recovery time –Turnaround time
Sources of variation activity Identify examples of variation in your specialty
Bed availability - an example of the problem of variation IN-PATIENT STAYADMISSION DISCHARGE Variation in patient pathways and processes. E.g. in Length of Stay Variation in Admission Patterns - particularly for Elective Care Variation in Discharge - By time of day - By day of week - Seasonal variations
IN-PATIENT STAYADMISSION DISCHARGE Variation in patient pathways and processes. E.g. in Length of Stay Variation in Admission Patterns - particularly for Elective Care Variation in Discharge - By time of day - By day of week - Seasonal variations “We always bring our hips in on Tuesday !”
IN-PATIENT STAYADMISSION DISCHARGE Variation in patient pathways and processes. E.g. in Length of Stay Variation in Admission Patterns - particularly for Elective Care Variation in Discharge - By time of day - By day of week - Seasonal variations “ We’re too busy in the morning and haven’t time to think about discharges. They all get done in the afternoon.
Understanding demand and capacity by hour of the day
time Demand Capacity Queue Can’t pass unused capacity forward to next week Variation mismatch = queue
How should we set capacity? Set capacity to average demand? Set capacity to maximum demand? Something else?
Setting capacity to Average demand Wait Seen Capacity set as average demand 100% utilisation each day = very efficient as long as there is a queue Lots of energy shuffling the queues High risk if we get it wrong (clinical and financial)
Setting capacity to max demand Poor utilisation Large amount of wasted clinic time Very inefficient use of resources Wait Seen Capacity set as maximum demand
Solution: Flex Capacity See today’s work today But how do we cope with the variation in workload from day to day?
Solution: Reduce variation Step 1: Understand why capacity and demand vary. Find the root cause Step 2: Reduce the variation Step 3: Set capacity to the 80th percentile (not the average) Step 4: Flex capacity to match demand Or
80th percentile... 50th percentile (median) 80th percentile Variation Lowest value Highest value Mean
80% of what? 80% of the demand falls below the line: sort the patients/clinics into order draw the line where 80% of the demand will fall below, and 20% above the line Quick equation: minimum demand+ ((Maximum demand – minimum demand) x 0.8) = 80% of variation in demand
A model of variability... “The Variation Model” For model go to analysis.xls
A new definition of capacity... Capacity is not activity Capacity is not “100% utilisation of every resource” Capacity is maximising utilisation of the constraint Capacity should be set at the 80th percentile of the demand