The burden of cancer on primary and secondary health care services before and after cancer diagnosis in New South Wales, Australia

BACKGROUND: Primary and secondary healthcare service usage is assessed in the year before and following a cancer diagnosis, in cancer cases versus matched non-cancer controls in New South Wales (NSW), Australia over 2006–2012, for all invasive cancers collectively and for selected common sites: breast, prostate, colorectal and lung, and melanoma. METHODS: The 45 and Up cohort (n ≈267,000) was linked to NSW Cancer Register (NSWCR), Emergency Department Data Collection (EDDC) and Medical Benefits Schedule (MBS) data using probabilistic record linkage. First-ever malignant cancers diagnosed after enrolment in the 45 and Up study comprised the study cases. Where possible, five controls were randomly selected per case from the 45 and Up cohort, matched by sex and year of birth. Controls comprised those with no cancer recorded on the NSWCR. For each month in the year preceding and following the cancer diagnosis, general practitioner, specialist and specified hospital ED service use was compared between cases and controls using proportions, means, and odds ratios derived from conditional logistic regression. RESULTS: Compared to controls, cases of all cancers combined had a significantly higher likelihood of GP and specialist consultation in the year leading up to diagnosis. This was most pronounced in the 3–4 months leading up diagnosis for all cancers, similarly for lung cancer (GPs and specialists) and melanoma (GPs), and colorectal cancer (specialists). Likelihood of a GP consultation remained significantly higher in cases than controls in the 12 months following diagnosis. During most of the year preceding cancer diagnosis, the likelihood of specified ED presentations was also significantly higher in cases than controls for all cancers, and most pronounced in the 2–3 months before diagnosis. Excepting melanoma, the likelihood of specified ED presentations remained significantly elevated for most of the year following diagnosis for all cancers combined and for the selected cancers. CONCLUSIONS: People with cancer experience a higher use of primary and secondary healthcare services in the year preceding and following diagnosis, with GPs continuing to play a significant role post diagnosis. The higher likelihood of pre-diagnosis GP consultations among cancer cases requires further investigation, including whether signals might be derived to alert GPs to possibilities for earlier cancer detection.