Rapid Fire Session 2014 World Cancer Congress

Reducing time to cancer diagnosis in rural Western Australia (WA): Results of a cluster randomised controlled trial of community and primary care interventions (#391)

Jon Emery 1 2 , Christobel Saunders 3 , Victoria Gray 1 4 5 , Terry Slevin 4 , Emma Croager 4 , Shelley Cheetham 1 , Fiona Walter 6 , Max Bulsara 7 , Kirsten Auret 8 , D'Arcy Holman 5 , Jennifer Walker
  1. School of Primary, Aboriginal and Rural Health Care, University of Western Australia, Perth, WA, Australia
  2. University of Melbourne, Melbourne, VIC, Australia
  3. School of Surgery, University of Western Australia, Perth, WA, Australia
  4. Cancer Council WA, Perth, WA, Australia
  5. School of Population Health, University of Western Australia, Perth, WA, Australia
  6. University of Cambridge, Cambridge, UK
  7. University of Notre Dame, Perth, WA, Australia
  8. Rural Clinical School Western Australia, Albany, WA, Australia

Background:

Rural cancer patients in Australia have significantly poorer outcomes than metropolitan patients and this may be partly due to later presentation and diagnosis.  Interventions aimed at raising community awareness of cancer symptoms and improving GPs assessment of symptoms could reduce time to cancer diagnosis but there have been few RCTs previously reported.

Aim:

To test a complex intervention targeted at the community and at primary healthcare providers to reduce the time to diagnosis for common cancers in rural WA.

Methods:

A 2x2 factorial cluster randomised trial of: 1. A community symptom awareness campaign to promote earlier help-seeking for specific cancer-related symptoms; 2. A general practice based academic-detailing intervention aimed at implementing cancer symptom risk models (Hamilton CAPER charts) and best practice diagnostic routes.  Two matched geographical areas of rural WA were randomised to the community intervention or control region.  General practices within both geographical regions were randomised to receive multiple academic detailing visits or no intervention.  The primary outcome measure is the total diagnostic interval (i.e. the time from first symptom/attendance for cancer screening to date of cancer diagnosis) of newly diagnosed cancer patients from both trial arms, calculated from the Symptom questionnaire and medical records.   Additional measures include TNM stage, measures of campaign coverage/impact, and health economic outcomes. 

Results:

Over 1,100 people with breast, lung, bowel or prostate cancer had been recruited to the trial after 20 months and we estimate a final accrual of 1,250 by March 2014. We will report the combined and separate effects of the community and provider-based interventions on total diagnostic interval.

Conclusions:

This is the largest trial known to date to test this type of intervention using such rigorous methods.  Results will be of international relevance on the role of symptom awareness campaigns and GP interventions to achieve earlier cancer diagnosis.