Transcript Slide 1

National Cancer Intelligence Network
The challenges of coding cancer
of unknown primary
A survey of registration and reporting practices in the UK, Ireland and Australia
Claudia Oehler1, Claire Vajdic2, Nicola Cooper1, John Symons3
1
National Cancer Intelligence Network, Public Health England
2
University of New South Wales, Australia
3
Cancer of Unknown Primary Foundation, UK
CRUK-NCIN Partnership Programme
Content
Introduction to results of the survey on registration and
reporting of Cancer of Unknown Primary (CUP):
2
1.
Overview: what is CUP? why is it problematic?
2.
The survey: what was its purpose?
3.
The results: what did it tell us?
4.
Conclusion: what next?
Coding cancer of unknown primary - a registration and reporting survey in the UK, Ireland and Australia
What is cancer of unknown primary?
Cancer of unknown primary (CUP) is a diagnostic term for
cancer which has spread from its initial location and where
the original cancer cannot be determined with confidence.
Context
 9,762 cases and 10,812 deaths (UK, 2011);
16% 1-year relative survival; 57% emergencies (England, 2006-2010).
 ~8th most common incident cancer (ca.15% cancers present as metastases,
1/3 no obvious origin); ~5th most common cancer cause of death.
 2010 first clinical guideline by National Institute of Clinical Excellence (NICE)
recommending diagnostic pathway and multi-disciplinary teams.
 2013 first Peer Review Measures (internal validation by 133 Trusts).
3
Coding cancer of unknown primary - a registration and reporting survey in the UK, Ireland and Australia
Why is CUP problematic?
Something 'unknown' is difficult to define and treat.
 Clinically heterogeneous and complex presentations; diagnosis often difficult
to determine; morphology and extent of cancer crucial for outcome.
 Clinical uncertainty makes registration/coding of CUP difficult.
 Codes used for reporting CUP vary substantially:
•
•
England: mainly four ICD10 codes C77, C78, C79 (lymphatic, respiratory/abdominal
and 'other' metastases) and C80 (primary cancer site unknown) - regional
variations.
Others: solely C80, or additional ICD10 codes for ill-defined primary cancer sites,
such as C26, C39 or C76 (ill-defined digestive, respiratory/intrathoracic, or generic
sites, e.g. 'head and neck').
 Obscures accurate assessment of CUP burden; nationally & internationally.
Clear definition and consistent recording of CUP crucial for accurate statistics
and clinically meaningful analyses.
4
Coding cancer of unknown primary - a registration and reporting survey in the UK, Ireland and Australia
Purpose of the survey?
The aim was to compare CUP registration and reporting
practices in the UK, Ireland and Australia, with a view to
supporting improved national/international standardisation.
Methodology:
 20 registration services participated: 8 in Australia, the 8 regional NCRS in
England, 1 each in Wales, Scotland, Northern Ireland and Ireland.
 2-part questionnaire:
•
•
5
Section 1 - Registration: coding guidance and system; coding scenarios; death
certificate only; clarification process.
Section 2 - Reporting: codes used for reporting; incidence by source of diagnosis
and morphology.
Coding cancer of unknown primary - a registration and reporting survey in the UK, Ireland and Australia
What did the survey tell us?
Caveat: these are draft results; registration services will be
given the opportunity to comment before publication.
 To preserve the anonymity of individual responses, these are grouped:
•
Registration offices in England and Australia are grouped as that; when individual
answers are shown, these are numbered E1-E8 and A1-A8; these numbers were
randomly assigned.
•
Countries with a single national registry - Scotland, Wales, Northern Ireland and
Ireland - are grouped here to obscure the identity of any individual registry until
permission is obtained to attribute results by nation. When individual responses
are shown, these countries are numbered X1-X4; these numbers were randomly
assigned and do not correspond to the order in which they are listed.
 Limitations: registration systems vary across countries; some of the
differences may be due to a different casemix; a few questions were left
blank / answers were unclear; April 2013, English registries merged to form
National Cancer Registration Service.
6
Coding cancer of unknown primary - a registration and reporting survey in the UK, Ireland and Australia
What did the survey tell us?
Survey confirmed variation in registration and reporting
practices on almost every aspect covered.
For example, coding guidance for CUP:
 Q1_Does your registry have guidelines that specifically cover the
registration of CUP?
England
Yes
No
7
5
3
Scotland, Wales,
Northern Ireland &
Ireland
3
1
Australia
3
5
55%
45%
Coding cancer of unknown primary - a registration and reporting survey in the UK, Ireland and Australia
What did the survey tell us?
Question 4: Does a CUP cause of death get matched
against a prior site-specific cancer registration?
All registries (n = 20)
England (n = 8)
6
6
Number of responses
5
5
4
4
3
3
2
2
1
1
0
0
Yes
Yes
8
No
Depends
length &
type
Depends
length
Depends
type
{Comment
only}
No
Depends Depends Depends
length & length
type
type
Coding cancer of unknown primary - a registration and reporting survey in the UK, Ireland and Australia
What did the survey tell us?
Question 11: If additional information is sought from
notifier(s), who would be contacted for it?
Scotland, Wales, Northern Ireland
England (100% = 8 registries)
& Ireland
England
Q11.1_Notifying institution
Q11.1_Notifying
Q11.1_Notifying
institutioninstitution
Q11.1_Notifying institution
Q11.7_Hospital coding dept
Q11.7_Hospital
coding dept
Q11.7_Hospital
coding dept
Q11.7_Hospital coding dept
Q11.5_Treating doctor
Q11.5_Treating
Q11.5_Treating
doctor doctor
Q11.5_Treating doctor
Q11.2_Referring doctor
Q11.2_Referring
Q11.2_Referring
doctor doctor
Q11.2_Referring doctor
Q11.4_Primary care clinician
Q11.4_Primary
care clinician
Q11.4_Primary
care clinician
Q11.4_Primary care clinician
Q11.3_Pathologist
Q11.3_Pathologist
Q11.3_Pathologist
Q11.3_Pathologist
Q11.6_Radiologist
Q11.6_Radiologist
Q11.6_Radiologist
Q11.6_Radiologist
0
2
4
6
Always
9
8
0
Sometimes
0%
1
2 25%
Rarely
3
50%4
Never
75%0
Australia
100%
2
4
No answer
Coding cancer of unknown primary - a registration and reporting survey in the UK, Ireland and Australia
6
8
What did the survey tell us?
Question 18: Indicate the number of cases broken down by
the basis of diagnosis/registration (source of information).
90%
80%
70%
60%
50%
Pathology
40%
Clinical
DCO
30%
Unknown
20%
10%
0%
E1
E2
E3
E4
E5
England
10
E6
E7
E8
X1
X2
X3
X4
Scotland, Wales,
Northern Ireland &
Ireland
A1
A2
A3
A4
A5
A6
A7
A8
Australia
Coding cancer of unknown primary - a registration and reporting survey in the UK, Ireland and Australia
What did the survey tell us?
Question 18: Indicate the number of cases broken down by
the code used to report CUP [all translated to ICD10].
100%
C26
80%
C39
C76
60%
C77
C78
40%
C79
C80
20%
/6 or /9
0%
E1
E2
E3
E4
E5
England
11
E6
E7
E8
X1
X2
X3
X4
Scotland, Wales,
Northern Ireland &
Ireland
A1
A2
A3
A4
A5
A6
A7
A8
Australia
Coding cancer of unknown primary - a registration and reporting survey in the UK, Ireland and Australia
What next?
Beyond the survey - two key areas: registration processes
and codes/coding systems.
Registration processes
 Guidance for CUP: clarification of notification; source of diagnosis; follow-up
 Specific rules around death certificate notifications and matching to prior
diagnoses (incl. clinical)
 England: impact of National Cancer Registration Service merger; but what
about the wider picture (UK, Ireland, Australia; Europe and beyond)?
Coding systems and codes
 ICDO3 codes ≠ ICD10 C-codes; better depth of data with ICDO3
(particularly metastases)? Australian method?
 Impact of CUP registration definition on mortality statistics?
12
Coding cancer of unknown primary - a registration and reporting survey in the UK, Ireland and Australia
Thanks!
Thanks also go to the Australian registries for their participation.
13
Coding cancer of unknown primary - a registration and reporting survey in the UK, Ireland and Australia