ROUTINE RE-SCANNING FOR IMRT HEAD AND NECK …
Download
Report
Transcript ROUTINE RE-SCANNING FOR IMRT HEAD AND NECK …
Incident Reporting and Learning:
THE LIVERPOOL SYSTEM - CLASSIFICATION,
LEARNING & PREVENTION
Anthony Arnold
Director Cancer Services, Illawarra Shoalhaven Local Health District
[email protected]
Context
--- ROSIS Melbourne Australia 2012 ---
2
The Liverpool System
Ref: IJROBP 2010 Volume 78, No 5, Pages 1548-1554
--- ROSIS Melbourne Australia 2012 ---
3
A Problem Worth Solving……
Complexity of radiation oncology
At the time no system of analysis was in place
Lack of clinical governance surrounding
reporting
There was limited openness about reporting
events
The culture was predominantly blame based
Standard reporting systems are ineffective for
radiation oncology
--- ROSIS Melbourne Australia 2012 ---
4
Classification
Prescription
Documentation
Simulation
Computing /
Dosimetry
Shielding
Bolus
Pre-Tmt
Imaging
Treatment
--- ROSIS Melbourne Australia 2012 ---
5
Incorrect Tmt Site Prescribed
Incorrect Energy Prescribed
Incorrect Dose or Dose per Fraction
Bolus not Prescribed when Req’d
Shielding not Prescribed when Req’d
Prescription not signed by Rad Onc
Other: Prescription Related
Incorrect Site Scanned / Simulated
Insufficient Scan Area Applied
Incorrect CT Procedure Applied
Laser Shift / Related Error
Sim Film Marked Incorrectly
Simulation Tattooing Related Error
Inappropriate Pt Positioning Used
Volume / Voluming Related Error
Landmarking Related Error / Omission
Contrast Related Error
Other: Simulation Related
Incorrect CT-Density Conversion Used
Incorrect Weight/Calc/Dose Point Used
Incorrect Normalisation Applied
Other: Computer Planning Error
Incorrect MU Calculation
Incorrect Detail in R+V System
Tray / Wedge Code Missing / Error
Attenuation Factor Missing / Error
QA Check/s Not Completed / Error
Treatment Sheet Annotation Error
TLD Related Error / Omission
Pacemaker Related Error / Omission
Other: Pre-Treatment Related
Geographic Miss: Incorrect Site
Geographic Miss: FSD/SSD Error
Geographic Miss: Incorrect Tattoo
Geographic Miss: Landmarking Error
Geographic Miss: Incorrect Fld Used
Geographic Miss: Field Size Error
Patient Not Treated When Required
Field Not Treated When Required
Incorrect Energy Delivered
Changes / amendments not in R+V
Immobilisation Device Error
History / Chart Check Missed / Error
Other: Treatment Related
Bolus not Applied When Required
Bolus Applied to Incorrect Site
Incorrect Bolus Material Used
Incorrect Bolus Thickness Used
Other: Bolus Related
Shielding Not Applied When Req’d
Incorrect Shielding / Cut-out Used
Shielding Applied to Incorrect Area
Shielding Mounting / Tray Error
MLC Pattern / Related Error
MLC File Missing / Not Attached to Fld
MLC Checks Missed / Not Done
Other: Shielding Related
Image Not taken When Specified
Film / EPI Labelled Incorrectly
Image Not Reviewed When Required
Iso Adjustments Applied Incorrectly
Iso Shifts Not Applied When Req’d
Other: On-line / Off-line Related
Documentation Error: Sim
Documentation Error: Tmt Sheet
Documentation Error: R+V System
Documentation Error: Imaging
Documentation Error: Other
Incorrect Wedge / Wedge Orient
n
Prescription
Simulation
Computing
PreTreatment
Treatment
Bolus
Shielding /
MLC
Imaging
Documentation
6
Classification Advantage
--- ROSIS Melbourne Australia 2012 ---
7
Error / Event Definitions
Event:
“event or circumstance which could have resulted,
or did result in harm to a patient”
Actual Error:
“Error resulting in radiation exposure other than that
intended or prescribed – correctable or otherwise”
Near Miss:
“Error or non-conformance detected before
reaching the patient”
--- ROSIS Melbourne Australia 2012 ---
8
High Level Structure…….PDSA
PLAN
DO
• Classification designed, database constructed, education
• System implementation, clinical leadership and support
• Staff encouraged to report all events irrespective of magnitude
• Supporting governance, openness, process based
• Summary reports analysed monthly across various forums
STUDY • Trend patterns analysed to highlight areas / systems in need
ACT
• The data itself was used to focus QA and improvement activity
• Focussed education, workflow redesign, protocol changes
--- ROSIS Melbourne Australia 2012 ---
9
Reporting and Managing an Event
• Staff detecting initiates report (narratives, tells story)
Detect • Manage patient and situation, immediate actions
• Team review, contributing factors, further actions
Review • Agree on report as a team
Share
• Reverse back through other staff and depts involved
• Learning, prevention, further analysis, additional factors
• Review and classifiy, explore issues, system breaks
Manage • Consider recommendations, initiate change / improvements
--- ROSIS Melbourne Australia 2012 ---
10
Department Analysis
Liverpool Macarthur CTC
(2004-2007)
Illawarra CCC
(2006-2009)
4-5 linear accelerators
Superficial / orthovoltage
Brachytherapy
Widespread conformal 3DCRT
IMRT on horizon
2 linear accelerators
Superficial / orthovoltage
No brachytherapy
Widespread conformal 3DCRT
IMRT widespread clinical use
Large metropolitan centre
Small semi-regional centre
688 reports / 3925 courses
670 reports / 3645 courses
--- ROSIS Melbourne Australia 2012 ---
11
Results - Initial Pilot
688 reports were logged during the study period
155 Actual errors
(23%)
533 Near Miss
(77%)
Attendances
Near Miss
Actual
Error
Analysis of 1st 3 years of operation (May 2004-May 2007)
--- ROSIS Melbourne Australia 2012 ---
12
Results - Subsequent Pilot
670 reports were logged during the study period
67 Actual errors
(10%)
603 Near Miss
(90%)
Attendances
Near Miss
Actual
Error
Analysis of 1st 4 years of operation (2006-2009)
--- ROSIS Melbourne Australia 2012 ---
13
Time Trends Statistics
p-value for
Actual
Error Rate
Difference
p-value for
Near Miss
Rate
Difference
Near
Miss
Total
Errors
No. of
Attendances
%
Actual
Error
Year
1
63
184
247
21788
0.3%
Year
2
58
199
257
38134
0.2%
p<0.01
0.5%
p<0.0001
0.7%
p<0.0001
Year
3
34
150
184
55006
0.1%
p<0.001
0.3%
p<0.0001
0.3%
p<0.0001
% Near
Miss
0.8%
%
Total
Events
p-value for
Total Errors
Rate
Difference
Actual
Error
1.1%
Ref: Simple Interactive Statistical Analysis online statistical calculator. Available at: http://www.quantitativeskills.com/sisa/statistics/t-test.htm. Accessed 29 January 2008
--- ROSIS Melbourne Australia 2012 ---
14
Time Trends Statistics
Actual
Error
Near
Miss
Total
Errors
No. of
Attendances
%
Actual
Error
p-value for
Actual
Error Rate
Difference
% Near
Miss
p-value for
Near Miss
Rate
Difference
%
Total
Events
p-value for
Total Errors
Rate
Difference
Year
1
16
145
161
6221
0.26%
-
2.33%
-
2.59%
-
Year
2
12
173
185
15687
0.08%
0.0016*
1.10%
p<0.0001*
1.18%
p<0.0001*
Year
3
27
128
155
17028
0.16%
0.1695*
0.75%
p<0.0001*
0.91%
p<0.0001*
Year
4
12
157
169
15582
0.08%
0.0017*
1.01%
p<0.0001*
1.08%
p<0.0001*
Ref: Simple Interactive Statistical Analysis online statistical calculator. Available at: http://www.quantitativeskills.com/sisa/statistics/t-test.htm. Accessed 29 January 2008
--- ROSIS Melbourne Australia 2012 ---
15
Time Trends 1st Pilot: Attendances
60000
500
450
400
350
40000
300
30000
250
200
20000
150
100
10000
50
0
04/05
21788
05/06
38134
06/07
55006
Actual Error
63
58
34
Near Miss
184
199
150
No. of Attendances
0
16
Incidents
Attendances
50000
1000
900
800
700
600
500
400
300
200
100
0
Total Courses
Actual Errors
Near Misses
600
500
400
300
200
100
2006
894
47
427
2007
901
12
173
2008
946
27
128
2009
904
12
157
Reported Event Count
Patient Courses of Treatment
Time Trends – 2nd Pilot: Courses
0
17
Study
Our study
Macklis et
al.[17]
Fraass et al.[24]
Actual Error
Rate per
treatment
episode
Total Error Rate
per treatment
episode†
2004-05
0.3%
1.1%
2005-06
0.2%
0.7%
2006-07
0.1%
0.3%
1995
0.2% per
treatment field
NR
Time period
96-97
0.4%
NR
Reporting scope
Comments
Simulation
Prescription
Planning
Treatment
delivery
x
Block errors
most common
Treatment setup and
treatment
accessory errors
most common
x
x
x
97-02
0.3%
NR
x
x
x
Tight
parameters for
error.
Treatment field
errors of >0.5cm
the most
common.
Calandrino et
al.[19]
91-96
0.45% per
treatment
course
3.5% per
treatment
course
x
x
x
MU calculations
only
BarthelemBrichant et
al.[27]
NR
3.5%
NR
x
x
x
Patton et al.[6]
99-00
0.2%
NR
x
x
x
89-90
0.17% per
treatment field
Huang et al.[22]
SwannD’Emilia[25]
NR
x
x
--- ROSIS Melbourne Australia 2012 ---
x
Most errors
were due to
errors in block
18
placement
Outcomes – Key Measures
Reduction
in Errors
Reduction
in Error
Rate
--- ROSIS Melbourne Australia 2012 ---
Improved
Patient
Safety
19
Patient Safety Risk Improvement
REDUCTION IN REPORTED EVENTS as a function of
attendances
Actual Error rate reduced from 0.26% to 0.08%
(p=0.0017)
Near Miss rate reduced from 2.33% to 1.01% (p<0.0001)
IMPROVED RELATIVE PATIENT SAFETY RISK per
treatment course
Actual error rate reduced from 1 in 19 courses to 1 in 75
courses; in other words from 5% down to 1.3% risk of
detectable error (p=0.0003)
Near miss rate reduced from 1 in 2 courses to 1 in 6
courses; in other words from 50% down to 17%
(p<0.0001)
--- ROSIS Melbourne Australia 2012 ---
20
THANK YOU