Download by Mamta Bajre, Paul McCrone and Mark Pennington

IS RADIOGRAPHER CHEST X-RAY REPORTING
COST-EFFECTIVE?
Mamta Bajre, Paul McCrone, Mark Pennington
King’s Health Economics, King’s College London
Institute of Psychiatry, Psychology and Neuroscience
Background and aims
• Supply of healthcare resources is limited
• Demand for healthcare is high and potentially rising
• Decisions need to be made about how to treat specific
conditions and which conditions to prioritise
• Expanding the role of allied health professionals is one solution
to the challenge of increasing pressures on budgets
• Some concerns over such expanded roles
• Radiographer reporting of CXRs seen as a potential alternative
to radiologist reporting
• What are the cost implications of radiographer reporting of
CXRs for suspected lung cancer?
• How cost-effective is radiographer reporting?
Methods
• Study funded by Society of Radiographers and conducted by
MSc student from City University
• Evaluation options:




Randomised trial
Observational study
Casenote review
Simulation model
• Advantages of models
 Results can be produced quickly
 Models can be adapted to aid generalisability
 Allows a focus on certain key parameters of interest
• Disadvantages of models
 Models are by definition an abstraction from reality
 Data are not always available
Key assumptions
• Time taken to report chest X-rays is 2 minutes for both radiographers
and radiologists
• False negatives present at A&E at a later date at which point disease
has advanced a stage (for patients at stage I to III)
• Sensitivity and specificity of radiographer reporting of chest X-ray and
radiologist reporting of both chest X-ray and CT-scan is independent
of disease stage or other patient characteristics such as age.
• Treatment costs in the year following diagnosis are maintained for
the subsequent four years or until death
• QOL in the year following diagnosis is maintained for the subsequent
four years or until death
• There is no QOL impact arising from false positive reporting
• Findings for non-small cell lung cancer are generalisable to other lung
cancers
Model to assess cost-effectiveness
Stage I
Stage II
CT confirmed lung cancer (TP)
Stage III
Stage IV
True positive
Stage I
Stage II
Lung cancer
CT confirmed no lung cancer (FN)
FN patients are lung cancer positive
they come back to A&E
Stage III
Stage IV
Stage I
Stage II
False negative
Stage III
Lung cancer
patient coming
Stage IV
to A&E
Radiologist reporting
chest X-ray
1
True negative
No lung cancer
Patient sent for CT scan
CT confirmed no lung cancer
Image not clear and radiologist suspects cancer
After initial clinical diagnosis
GP refers for chest X-ray
X-ray imaging is performed
by radiographer
1
False positive
Stage I
Stage II
CT confirmed lung cancer(TP)
Stage III
Model to assess cost-effectiveness
Stage I
Assumed to be 13%
(Field et al, 2013)
Stage II
CT confirmed lung cancer (TP)
Stage III
Stage IV
True positive
Stage I
Stage II
Lung cancer
CT confirmed no lung cancer (FN)
FN patients are lung cancer positive
they come back to A&E
Stage III
Stage IV
Stage I
Stage II
False negative
Stage III
Lung cancer
patient coming
Stage IV
to A&E
Radiologist reporting
chest X-ray
1
True negative
No lung cancer
Patient sent for CT scan
CT confirmed no lung cancer
Image not clear and radiologist suspects cancer
After initial clinical diagnosis
GP refers for chest X-ray
X-ray imaging is performed
by radiographer
1
False positive
Stage I
Stage II
CT confirmed lung cancer(TP)
Stage III
Model to assess cost-effectiveness
Stage I
Assumed to be 13%
(Field et al, 2013)
Stage II
CT confirmed lung cancer (TP)
Stage III
Stage IV
True positive
Stage I
Stage II
Lung cancer
CT confirmed no lung cancer (FN)
FN patients are lung cancer positive
they come back to A&E
Replaced with
radiographer
Stage III
Stage IV
Stage I
Stage II
False negative
Stage III
Lung cancer
patient coming
Stage IV
to A&E
Radiologist reporting
chest X-ray
1
True negative
No lung cancer
Patient sent for CT scan
CT confirmed no lung cancer
Image not clear and radiologist suspects cancer
After initial clinical diagnosis
GP refers for chest X-ray
X-ray imaging is performed
by radiographer
1
False positive
Stage I
Stage II
CT confirmed lung cancer(TP)
Stage III
Estimates of reporting accuracy
Parameter
%
Sensitivity - Radiologist reporting CXR
69.7
Specificity - Radiologist reporting CXR
80.9
Sensitivity - Radiographer reporting CXR
78.1
Specificity - Radiographer reporting CXR
85.2
Sensitivity - Radiologist reporting CT Scan
94.4
Specificity - Radiologist reporting CT Scan
72.6
Sources: Woznitza (2016), Denise et al (2013)
Cost of activities
Activity
£
Chest X-ray
27
Radiologist reporting chest X-ray
32
Radiographer reporting chest X-ray
29
A&E treatment
141
Sources: Oliver et al (2001), RSNA (2016), NHS Reference Costs 2014-15
Cancer prevalence
90
Percentage of cases
80
70
60
50
First presentation
40
Second presentation
30
20
10
0
Stage I
Stage II
Sources: CRUK (2013), assumptions
Stage III
Stage IV
Cancer care costs
45000
40000
35000
Cost (£s)
30000
25000
Year 1
20000
Years 1-5
15000
10000
5000
0
Stage I
Stage II
Sources: CRUK (2014), assumptions
Stage III
Stage IV
Quality-adjusted life years (QALYs) over 5 years
4
3.5
3
QALYs
2.5
2
1.5
1
0.5
0
Stage I
Stage II
0.81
0.77
Sources: Niak et al (2015)
Stage III
0.76
Stage IV
0.76
Results
• At initial presentation there would be:
 95.8 cancer cases identified through radiographer reporting
 85.5 cancer cases identified through radiologist reporting
• Total reporting costs:
 Radiographer £57,302
 Radiologist £65,768
• Total costs including treatment:
 Radiographer £2,576,399
 Radiologist £2,560,795
 Difference = £15,604
• Total QALYs
 Radiographer 196.09
 Radiologist 192.4
 Difference = 3.69
Potential Cost-Effectiveness Results
?
Incremental outcome
Bad
?
Good
Incremental cost
Incremental cost-effectiveness ratio
Cost of treatment A – Cost of treatment B
ICER =
Effect of treatment A – Effect of treatment B
£15,604
ICER =
3.69 QALYs
= £4229 per QALY
Conclusions
• Radiographer reporting of CXRs appears to be a viable
alternative to radiologist reporting
• Costs will rise if accuracy is greater
• Cost per QALY below NICE threshold (£20,000)
• Caveats





Simple model
Data from limited sources
Extra training costs not considered
Earlier diagnosis not assessed
More refined model and robust data required
THANK YOU
[email protected]