Effect of a digital biopsychometric risk profile of patients with cancer on the delivery of risk-stratified personalised prehabilitation.

person Laura Halliday Imperial College, London, London, United Kingdom info_outline Laura Halliday, Krishna Moorthy, Alex King, James Dilley, Philippa Lloyd, Nirooshun Rajendran, Jemma O'Reilly, Dipankar Mukherjee, Venetia Wynter-Blyth

Authors person Laura Halliday Imperial College, London, London, United Kingdom info_outline Laura Halliday, Krishna Moorthy, Alex King, James Dilley, Philippa Lloyd, Nirooshun Rajendran, Jemma O'Reilly, Dipankar Mukherjee, Venetia Wynter-Blyth Organizations Imperial College, London, London, United Kingdom, Imperial College Healthcare NHS trust, London, United Kingdom, Barts Health NHS Trust, London, United Kingdom, Barking, Havering and Redbridge NHS Trust, London, United Kingdom, Barking, Havering and Redbridge NHS Trust, Barking, United Kingdom, Onkohealth, London, United Kingdom Abstract Disclosures Research Funding No funding received None. Background: Cancer symptoms and treatment consequences such as fatigue, nausea, distress and breathlessness contribute to inactivity, malnutrition, sarcopenia and ultimately poor clinical outcomes. Multimodal prehabilitation, with a focus on physical activity, nutrition and emotional well-being, is associated with improved treatment outcomes. However, prehabilitation needs to be personalised to maximise adherence and impact. Digital solutions offer an opportunity to collect and consolidate health data to facilitate the personalisation of prehabilitation programmes. Methods: Patients were referred from the NHS to an independent digital prehabilitation provider, Onkohealth Ltd. Baseline health metrics were collected via validated survey tools and wearable technology to develop a biopsychometric profile including the following modifiable factors. Multi-morbidity risk (MMR) was represented as a score from 0-14. Frailty estimation was based on walking speed. Diet quality was based on the World Cancer Research Fund diet score with higher score for better quality food items (range 0-3). Psychological wellbeing was measured via an emotional distress scale (range 0-10) and self-efficacy (range 0-40). Exercise metrics included 1 minute heart rate recovery (1m HRR) and 30 second Sit to Stand test (30sSTS). Risk-stratification was undertaken using the validated thresholds for each of these 9 modifiable risk metrics, including BMI. EuroQoL EQ5D5L was used to report quality of life as health utility (EQ-HU). Results: 303 patients with a range of cancer diagnosis [gastrointestinal (41%), biliopancreatic (10%) gynaecological (16%) and breast (20%) were included. 72% of patients had at least one high-risk metric and 30% had >3 high risk metrics. 42% of patients were obese, 34% had a high MMR and 25% had a slow walking speed. There were strong correlations between age and MMR (rho=0.53; p<0.001) and walking speed (rho=0.33; p=0.004). 16% had a low HRR and 17% had a low STS. STS had significant correlation with walking speed [rho=0.47; p<0.001) and BMI (rho=0.34; p<0.001). 42% of patients reported a low quality diet. High levels of anxiety were recorded in 21% of patients and high levels of depression in 26% of patients. Higher levels of anxiety (rho-0.24; p=0.002) and distress (rho=0.29; p<0.001) and lower levels of self-efficacy (rho= 0.25, p<0.001) were associated with lower EQ HU. Conclusions: This study demonstrated that 7 in 10 cancer patients have at least one risk factor and 1 in 4 patients have more than 3 predictors of poor treatment outcomes. Bio-psychometric profiling highlights modifiable risk factors and provides a focus for prehabilitation to be personalised. The findings also highlighted the importance of psychological wellbeing as high levels of emotional distress were associated with lower reported quality of life.