Background: There is increasing recognition of the psychological impacts of cancer. We examine antidepressant therapy post-cancer diagnosis in a cohort of older Australians.

Methods: Our cancer cohort comprised 1,962 Australian Government Department of Veterans' Affairs clients diagnosed with cancer between 2005 and 2007. Outcomes were prevalence of antidepressant therapy, antidepressant initiation in cancer patients post-diagnosis compared with matched non-cancer patients and predictors of antidepressant initiation.  

Results: Point prevalence of antidepressant use in cancer patients was 15% three to six months before cancer diagnosis and 23%, 21 to 24 months post-diagnosis. Antidepressant initiation post-cancer diagnosis was 20% compared with 18% in the comparison cohort (p=0.07 stratified model). Antidepressant initiation was associated with pre-existing comorbidity burden and use of opiate or benziodiazepine anxiolytics during follow-up. However, when each cancer was analysed separately, prostate cancer patients were more likely than their matched comparisons to initiate antidepressants (aOR 1.40, 95% CI 1.15-1.70).  In the model using only cancer patients, overall pre-existing comorbidity burden (aOR 1.10, 95%CI 1.01-1.19) and opiate (aOR 1.72, 95%CI 1.35-2.18) or benzodiazepine anxiolytic use (aOR 2.30, 95%CI 1.68-3.16) not cancer-specific factors, were associated with antidepressant initiation. Median time to antidepressant initiation was 359 days, median number of dispensed antidepressants was 4 (IQR 4-11) and selective-serotonin reuptake inhibitors were the most commonly prescribed antidepressants.

Conclusion: A cancer diagnosis leads to moderate increases in antidepressant therapy. Prescribing is often in the setting of pain and anxiety and pre-existing comorbidity.