Aims: To test the effectiveness of integrating electronic patient-reported outcome measures (ePROMs) within the context of real-world cancer care across Queensland, Australia.
Methods: A pragmatic, multi-site randomised controlled trial conducted across four sites. Between August 2021 and June 2023, adults (≥18 years) with newly diagnosed or recurrent solid cancers were randomised 1:1 to receive either an ePROM intervention or usual care. Intervention participants routinely completed ePROMs assessing cancer symptoms and distress during the first six months of care. Alerts were sent to relevant clinicians to follow-up patients when predefined symptom thresholds were exceeded. All participants completed surveys at baseline and at 3, 6, 12, 18, and 24 months. Co-primary outcomes were unplanned hospital presentations identified through linked administrative health records and physical and functional wellbeing measured using the Functional Assessment of Cancer Therapy-General subscales. Secondary outcomes included symptom burden, psychological distress, overall wellbeing, and healthcare experience. Analyses used mixed-effects repeated-measures models and negative binomial regression.
Results: A total of 538 participants were randomised. The ePROM completion rates across sites ranged from 54% to 91% during the first six months. Compared with usual care, the intervention did not reduce unplanned hospital presentations or improve physical and functional wellbeing at six months. However, intervention participants reported significantly lower symptom burden (adjusted mean difference -0.24, 95% confidence interval -0.40 to -0.08) and a better hospital care experience at six months (88% intervention vs 79% usual care rating it ‘very good’). No significant differences between groups existed at other time points or for the other outcomes.
Conclusions: Integrating ePROMs into routine cancer care resulted in modest short-term improvements in symptom burden and patient experience. While the pragmatic trial enhances real-world relevance, it may explain the smaller effects observed in our study compared with previous trials conducted in settings with a high trial fidelity.