Rural variations in primary care prostate cancer diagnosis and survival: a cohort study using linked Australian primary care electronic medical record data

Abstract

Background Across the state of Victoria, Australia around one-third of residents live in rural areas. There are persisting disparities and poorer prostate cancer outcomes in rural areas compared with metropolitan areas. Prostate Specific Antigen (PSA) blood tests are commonly used in primary care to investigate patients with suspected prostate cancer. Understanding variations in PSA test use and follow-up of abnormal results between rural and metropolitan areas could identify potential causes of disparities in prostate cancer detection and outcomes.

Methods We conducted a longitudinal cohort study using Australian primary care data linked to cancer registry data. Male patients aged ≥ 18 years, diagnosed with prostate cancer between 30 June 2010 and 31 July 2022 were included if they visited primary care in the year pre-diagnosis and had a primary care PSA test. Logistic regression was used to compare demographic and tumour characteristics between patients in rural and metropolitan areas and the occurrence and timing of abnormal PSA results pre-diagnosis. Poisson regression models were used to estimated trends in PSA test use pre-diagnosis. The Kaplan-Meier method was used to quantify overall and cancer-specific survival rates.

Results The study population consisted of 2,470 patients, (52% rural, 48% metropolitan). Rural patients had higher Gleason scores and were almost twice as likely to have significantly elevated (> 20ng/mL) PSA levels at diagnosis. There were prolonged diagnostic intervals (the time between first abnormal PSA test and cancer diagnosis) of 7 months in metropolitan patients and 8 months in rural patients. In both rural and metropolitan areas, two-thirds of patients with a PSA > 3ng/ml did not have guideline concordant care (a repeat PSA test within 3 months).

Conclusions Rural Victorians with prostate cancer have higher PSA levels and Gleason scores at diagnosis and experience longer diagnostic intervals compared to metropolitan patients. In both rural and metropolitan areas, abnormal PSA tests are frequently detected several months before prostate cancer diagnosis and there is low guideline concordance regarding repeat tests. Targeted interventions to identify and action abnormal PSA blood tests results in primary care could expedite prostate cancer diagnosis and reduce rural-metropolitan disparities in prostate cancer outcomes.