Professor Jane Young - Beliefs, Behaviours and Systems in Primary Care - NSW Findings from the International Cancer Benchmarking Partnership

2. Beliefs, behaviours and systems in primary care:
NSW findings from the International Cancer
Benchmarking Partnership
Jane Young, Claire McAulay, Ingrid Stacey, Megan Varlow, David Currow
SYDNEY MEDICAL SCHOOL
Jane Young | University of Sydney
Cancer Epidemiology and Cancer Services Research Group
Sydney School of Public Health

3. Background to ICBP
› Evidence of variations in cancer survival between different European
countries
› Example – rectal cancer : 5-year age-standardised relative survival
- Switzerland 61%
- England 52%
- Poland 39%
Sant et al, Eu J Cancer 2009
› ICBP formed to try to answer these questions
› Focus on lung, colorectal, breast and ovarian cancer
3

4. ICBP Module 3 aims
› To explore whether differences in primary care systems might explain
variations in cancer survival between countries
› In NSW, to investigate GPs’:
- beliefs about early diagnosis of cancer and their role in the cancer system
- access to diagnostic tests and specialists in the public and private sectors
- self reported practices for patients presenting with suspicious symptoms
- views of resources that could improve the interface between primary and
specialist care
› To compare responses for urban and rural GPs
4

5. Methods
› Online survey of GPs in
- UK (England, Wales, Northern Ireland)
- Scandinavia (Norway, Sweden, Denmark)
- Canada (British Columbia, Manitoba, Ontario)
- Australia (Victoria and NSW)
› GPs’ behaviours measured using case scenarios
› GPs’ beliefs and systems measured using direct questions
› Additional questions in NSW and Victoria:
- access to services in the public and private sectors
- out-of-pocket expenses for patients
- influences on referral practices
- preferences for resources to improve the interface between primary and specialist care
5

6. NSW sample and recruitment
› Random sample of GPs identified from a commercial list (AMPCo)
› Stratified by urban or regional/rural location based on ARIA+ classification
of practice postcode
› Primer letter, invitation and up to 3 mailed reminders
› GPs were ineligible if:
- not working primarily in clinical general practice
- provided locum services only
- had died, retired, on extended leave
- no longer in NSW
6

7. Response rate
7
2500 names selected
1250 metro, 1250 rural
(10% not eligible)
2246 eligible
Embedded RCT of
response-aiding
strategies
273 responses (12.2%)

8. Characteristics of respondents
8
Characteristic Rural (N = 140) Urban (N= 133)
n (%) n (%)
Female 61 (44) 62 (47)
Age (years) < 35 14 (10) 13 (10)
35–44 28 (20) 22 (17)
45–54 42 (30) 44 (33)
55–64 52 (37) 39 (29)
≥ 65 4 (3) 15 (11)
GP registrar 10 (7) 8 (6)
Sole practitioner 13 (9) 21 (16)
Part time 44 (31) 54 (41)
Years in general practice
< 3 11 (8) 8 (6)
3–5 18 (13) 10 (8)
6–10 12 (9) 21 (16)
11 + 99 (71) 94 (71)

9. Beliefs about timely diagnosis
9
More timely diagnosis of cancer is important to ensure better outcomes
Agree or Strongly agree
Rural (%) Urban (%)
Colorectal 97 98
Melanoma 96 98
Breast 94 96
Ovarian 89 95
Lung 83 89
Prostate 51 56

10. Beliefs about role in cancer system
10
Agree or strongly
agree
Rural (%) Urban (%)
I like to wait until I am sure of a diagnosis before referring
to a specialist
21 17
I am often unclear about when I should refer to a specialist
when I suspect cancer
6 4
Protecting patients from over-investigation is an important
part of my role
44 39
Preventing secondary/specialist care cancer services from
being overloaded is an important part of my role
44 39

11. Influences on management decisions
11
Agree or strongly agree
Rural (%) Urban (%)
Fear of litigation sometimes influences my decisions to
order investigations
38 48
Fear of litigation sometimes influences my decisions to
refer
35 43
I sometimes order cancer investigations that I don’t feel
are indicated due to patient pressure
36 40
I sometimes refer patients due to patient pressure rather
than clinical indication
29 31

12. Access to specialist advice within 48 hours
regarding investigations for suspected cancer
12
Agree or strongly agree
Rural (%) Urban (%)
Public system 51 51
Private system 76 89

13. Specialist referral within 48 hours for patient with
suspected cancer
13
Agree or strongly agree
Rural (%) Urban (%)
Public system 36 50
Private system 69 82

14. Proportion of GPs reporting direct access to GI
diagnostic tests (no specialist referral required)
14
NSW
rural
NSW
urban
VIC
rural
VIC
urban
Upper GI endoscopy 21 30 53 9
Flexible sigmoidoscopy 14 21 29 43
Colonoscopy 21 31 47 78

15. Access to colonoscopy
15
Proportion of GPs reporting average waiting time of 4 weeks or less:
NSW rural NSW urban VIC rural VIC urban
Public
system
14 23 24 18
Private
system
79 88 84 96
Proportion of GPs receiving colonoscopy results within 1 week:
NSW rural NSW urban VIC rural VIC urban
37 40 67 68

16. Proportion of GPs who can arrange colonoscopy
with no out of pocket expenses
16
Rural Urban
Yes, this is easy to organise 25 25
Yes, but with difficulty 51 56
No 19 18
Don’t know 4 2

17. Proportion of GPs reporting time to specialist
appointment within 2 weeks
17
Rural Urban
General surgeon Public
Private
46
85
58
94
Gastroenterologist Public
Private
31
68
52
92
Colorectal surgeon Public
Private
38
69
50
91
Respiratory physician Public
Private
36
66
55
87
Thoracic surgeon Public
Private
29
57
46
84
Gynaecologist Public
Private
46
77
56
90
Gynaecologic oncologist Public
Private
27
52
48
77

18. GPs’ perceived importance of various factors for
selecting a specialist
Factor Rural Urban
Previous experience referring patients to this specialist 84.3 90.2
Length of wait for appointment 70 68.4
Patient preference 49.3 52.6
Colleague recommendation 52.1 48.9
Specialist's hospital has good reputation for cancer care 32.1 57.9
Specialist is member of MDT 33.6 51.1
Specialists’ relevant cancer caseload 29.3 43.6
Know specialist personally 32.9 38.3
Out of pocket costs for patients 32.9 34.6
Distance patient must travel 33.6 26.3
Specialist's hospital has good published outcomes/low complication
17.1 41.4
rates for cancer patients
Specialist is in directory of cancer specialists 6.4 19.5
Specialist is involved in clinical trials 3.6 11.3
18

19. Perceived usefulness of resources for informing
about cancer services and referral pathways
19
Factor Percentage responding “very important”
Rural Urban
Discussion with colleagues 82 76
GP meetings or seminars 60 63
Feedback from patients 41 44
Mailed brochures or info from
36 40
hospitals/specialists
GP publications and newsletters 25 41
Directory of specialists/services 26 39
Internet searches 24 29
Cancer Institute NSW CanRefer
14 19
website

20. Limitations
› Very low response rate
› Participating GPs more positive towards cancer care than non-responders
› Scope of questions limited due to requirements of standardised instrument
20

21. Summary
› GPs expressed strong support for timely diagnosis to improve patient
outcomes for breast, colorectal and melanoma skin cancer, but less for
other cancer types
› Almost half of GPs considered that gatekeeping was an important part of
their role
› There were marked differences in access to diagnostic tests and specialist
services between urban and regional/rural GPs and for patients in the
public and private sectors
› Waiting times were one of the most important factors influencing referral
pathways
› These findings can inform future programs to enhance the interface
between primary and specialist care and provide a baseline to monitor
change
21

22. Acknowledgements
› We thank the GPs who participated in the survey, the ICBP Module 3 team
who developed the core questionnaire and Sigmer UK who developed and
managed the online database
› The study was funded by the Cancer Institute NSW
› JY is supported by Academic Leader in Cancer Epidemiology award
number 08/EPC/1-01 from the Cancer Institute NSW and CM was
employed through this award
› Paper reporting embedded RCT of financial incentives to improve
response rate will be published in Journal of Clinical Epidemiology (in
press)
22