Prehabilitation for frailty in cardiac surgery: time to move from promise to practice

This invited commentary refers to ‘Prehabilitation for patients with mild to moderate frailty undergoing coronary artery and valve surgery: a pilot randomized controlled trial’ by P.W.C. Li et al. https://doi.org/10.1093/eurjcn/zvag084. Frailty is a clinical state of increased vulnerability to acute stressor events and can be characterized by physical decline and the accumulation of deficits.1,2 One-in-three older adults undergoing cardiac surgery or valve procedures are likely to be frail.3 Consequently, these people with frailty have higher in-hospital mortality, longer hospital stays, higher likelihood of non-home discharges, and worse health outcomes compared to their non-frail counterparts.4 Despite this growing evidence and international consensus,5,6 frailty screening or management isn’t part of routine clinical practice. And with an ageing population, this may well become a major health system issue over the next half-century. But there is hope! Comprehensive evidence demonstrates that multicomponent pre-surgical programmes that include exercise—known as prehabilitation—half the risk of surgical complications, including those outcomes exacerbated by frailty.7 Unfortunately, much of the existing evidence resides in non-frail surgical cohorts,8 providing a clear opportunity to strengthen the cardiac surgery evidence base on route to improved clinical practice and outcomes for patients. Li et al. have sought to bridge this gap with their pilot study of prehabilitation for patients undergoing coronary artery and valve surgery, published in this edition of the European Journal of Cardiovascular Nursing.9 Li et al. designed a two-arm, parallel group, pilot randomized controlled trial of 51 frail patients scheduled for elective coronary artery bypass grafting or valve repair/replacement surgery.9 Eligible patients were mild-moderately frail (4–6 on Clinical Frailty Score), with at least 5 weeks of pre-surgical wait time, living with family, and access to an internet connection (for remote-supervised training). The control group included usual preoperative care such as an educational video and breathing/coughing instruction. In addition to usual care, the >4week frailty-specific prehabilitation programme included weekly sessions on education and stress management, tailored dietary advice, and protein supplementation. In addition, there were thrice weekly exercise sessions, which transitioned from in-person to remote supervised from week 3 onwards, and included low-moderate intensity aerobic, resistance, and balance exercises.9 Outcome assessments were conducted at baseline, one day before surgery, and one/three months after surgery. The primary outcome was the Short Physical Performance battery, a measure combining performance on three functional tasks: walking, balance, and sit-to-stand. Secondary outcomes included intervention feasibility and safety, the frailty status (Fried Phenotype), health-related quality of life, cardiac-related health, blood biomarkers, length of stay, readmissions, and major adverse cardiovascular and cerebral events. The authors should be commended for a rigorous study design, incorporating blinding of outcome assessors and clinical care teams, alongside independent randomization with appropriate allocation concealment. We also commend that older age (>60 years) was not included as an eligibility criterion, particularly in a cardiovascular cohort where a younger age of frailty onset is not unexpected. With the age distribution of the study cohort, this likely enabled the inclusion of younger individuals living with frailty, enhancing the clinical relevance and generalisability of the findings. Future studies may, however, benefit from more explicitly defining cardiovascular contraindications to exercise within the exclusion criteria, such as unrepaired aortic aneurysm, to further strengthen participant safety and reproducibility. The intervention itself represents a novel contribution to the prehabilitation literature. The inclusion of tailored, frailty-specific content distinguishes this programme from more generic prehabilitation approaches, while the use of a multimodal design is well aligned with established evidence for optimal perioperative outcomes. In particular, the delivery model—combining thrice-weekly sessions with a structured transition from in-person to remotely supervised training—is innovative and pragmatically responsive to real-world implementation challenges. Li et al. report several encouraging findings from this pilot study. The multimodal, home-based prehabilitation programme was safe, with no adverse events occurring during exercise, and high programme adherence (89%). These early findings suggest high feasibility and acceptability among frail cardiac surgery patients, and a positive platform for future work.9 Moreover, the intervention produced significant improvements in some clinical outcomes, despite the pilot scale. The most prominent were physical function scores, measured by the Short Physical Performance Battery (SPPB), which improved by around 1 point at both preoperative and 3-month postoperative timepoints. With a moderate-to-large effect size, these improvements are also clinically relevant; associated with >10% reductions in mortality risk and better quality of life in older adults.10,11 The study also showed substantial improvements in frailty (Fried phenotype) and health-related quality of life (MacNew), suggesting that functional improvements translated into measurable changes in frailty status and quality of life. Together, these findings provide early evidence that a tailored, frailty-focused prehabilitation programme is not only safe and feasible but also capable of producing meaningful improvements in function, frailty status, and quality of life before and after cardiac surgery. Although this pilot study offers early and encouraging evidence that frailty-focused prehabilitation is both practical and potentially beneficial, several important areas still need attention to strengthen future research, improve clinical relevance, and support real-world implementation. A key limitation of the Li et al. trial is the absence of a structured process evaluation. Such evaluations are essential for understanding intervention fidelity, feasibility, acceptability, and the contextual factors that influence implementation. Incorporating mixed-methods process evaluations in future studies would strengthen interpretation of outcomes, identify facilitators and barriers to engagement, and guide the optimization of future trials towards scalable and real-world implementation. As expected for a pilot trial, biomarker responses, postoperative complications, readmissions, and length of stay were not meaningfully different between groups—likely due to limited statistical power. Larger, fully powered trials are required to determine whether clinically significant effects emerge when sample sizes and event rates are sufficient. Future investigators may also consider more intensive or higher-dose interventions, particularly those capable of eliciting more robust physiological adaptations within the constrained preoperative window. Future trials should move beyond feasibility to determine the optimal dose, content, and delivery model for frailty-focused prehabilitation. Selecting an appropriate primary endpoint will be fundamental. Debate is warranted between traditional clinical outcomes (e.g. composite postoperative complications) and more patient-centred measures such as functional capacity, frailty status, or quality of life. These decisions will directly influence sample size, data collection burden, and the clinical relevance of findings. Economic evaluation should be incorporated prospectively to assess cost-effectiveness and inform health-system adoption. Optimizing intervention dose represents another high-priority area. Robust physiological adaptation from exercise typically requires 10–12 weeks of training, yet the presurgical window is often significantly shorter. Future research should explore strategies to safely increase training intensity or incorporate more efficient exercise modalities such as circuit-based resistance training or high-intensity interval training. Evidence shows that higher-intensity approaches may yield greater functional gains within limited timeframes alongside a slightly elevated adverse event risk,12 warranting direct comparison with low-to-moderate-intensity programmes. Dose–response analyses and robust reporting of adherence are essential, particularly given that insufficient session attendance has been linked to null effects in recent large trials.13 Finally, prehabilitation should be conceptualized as the first phase of a broader continuum of frailty care rather than a stand-alone intervention. Seamless transition from prehabilitation to cardiac rehabilitation may offer sustained functional benefits and support long-term frailty management, especially in settings where dedicated frailty services remain limited. Therefore, improving access and tailoring cardiac rehabilitation for people with frailty should therefore be a parallel research and service priority. Frailty remains under-managed in cardiac surgery, despite its clear impact on outcomes. The early findings from Li et al. demonstrate that tailored, multimodal prehabilitation can improve function and quality of life in this vulnerable group. The priority now is to build on this promising foundation through fully powered trials that define optimal intervention dose and delivery, incorporate rigorous process and economic evaluations, and develop scalable models that bridge prehabilitation with ongoing cardiac rehabilitation. Doing so will help embed frailty management as a standard component of cardiac surgical pathways—an essential shift as populations age and the demand for equitable, evidence-based perioperative care continues to rise. Matthew A Hollings (Conceptualization, Writing—original draft, Writing—review & editing [lead]), and Julee Mcdonagh (Writing—original draft, Writing—review & editing [supporting]) JM is supported by a National Health and Medical Research Council Investigator Grant (2025-2029), Reference—GNT2034202. This commentary does not contain new data.