The evolution of surgical innovation needs to go beyond developments intraoperatively and into a continuum of postoperative recovery and care. While conventional recovery processes include standardization in recovery approaches and have proven effective for populations in which they were studied, they do not account for differences in individual physiology during recovery, behavioral compliance and adherence, and social determinants of health. Similarly to the previous correspondence, this reinforces the importance of integrating patient-centered digital health technologies within the surgical care process, enabling individualized, tailored, responsive, and data-informed recovery pathways1. This manuscript upholds the ethical use and communication of AI-powered technologies in surgical caregiving routines in accordance with the TITAN 2025 Guidelines, focusing on AI reporting in healthcare systems2. Susceptibility to slowed recovery after surgery (or during the recovery process) is unique to the patient based on multiple variables, including individual inflammatory responses, comorbidity scores, preoperative function and functional status, mental or psychological resilience, and overall emotional, mental, and social well-being. While there is great value in Enhanced Recovery After Surgery (ERAS) protocols and procedures for providing evidence-based practices to reduce perioperative morbidity and hospital length of stay, their normative framework will limit the ability of the patient and caregiver to individualize recovery without accommodating real-time, dynamic, and patient-based needs. We see the value of digital health platforms (especially mHealth platforms, wearable biosensors, and AI decision support systems) to continuously monitor postoperative patients and optimize individualized care3. mHealth platforms have grown into sophisticated platforms able to deliver bidirectional communication between patient and provider, real-time analytics, and dynamic educational materials. Algorithms within these applications apply rule-based logic and/or machine learning to classify risk, evaluate symptom deterioration, and generate alerts for certain threshold levels (for example, sustained tachycardia, movement milestones, and delayed wound healing)4. Interventions such as these have improved medication adherence, reduced preventable readmissions, and increased patient-reported satisfaction. With the integration of natural language processing (NLP) into the information exchange between patient and provider, technology could assist in automating triage and analyzing sentiment to detect deteriorating clinical trajectories or psychological distress5. Wearable health devices offer another layer of physiological information by continuously monitoring metrics such as heart variability, respiratory rate, activity levels (steps, gait speed), and circadian disruption. Coupling physiologic data from PPG, electrodermal, and thermal sensors will improve the identification of early inflammatory processes or wound complications6. Together, these data streams, combined with AI algorithms, will allow for predictive modeling of complications. Importantly, these devices can be integrated into existing electronic health records (EHR) via Health Level 7 International (HL7) or Fast Healthcare Interoperability Resources (FHIR) integrations, supporting a seamless clinical workflow and interoperability7. Digital health tools represent a decentralized approach to postoperative care through a systems-level lens. Remote patient monitoring platforms provide real-time triage for complications to allow a clinician to intervene early and possibly prevent emergency department usage. Asynchronous and synchronous telemedicine platforms geographically and financially overcome challenges to care acquisition for underserved and rural populations8. In addition, using adaptive content delivery through artificially intelligent (AI) chatbots or dynamic multi-lingual modules provides a meaningful experience to patients with different literacy and cultural levels to support their recovery across a wide range of variables. This scalable approach is essential for tackling care delivery challenges in low- and middle-income countries (LMICs) or during healthcare disasters, such as the COVID-19 pandemic9. When implementing these technologies, rigorous validation of technology performance and clinical utility will have to occur. A sensor and signal must establish clinical-grade accuracy and fidelity, the data stream must show clinical-grade data latency, and the algorithm must demonstrate clinical-grade transparency if we are to let technology into a surgical care pathway. Further, ethical scaffolding must stay true to cybersecurity, patient and family consent, algorithmic and implementation bias, data sovereignty, and international obligations like the GDPR and HIPAA10. Another important consideration is the incorporation of patient-reported outcome measures (PROMs) into recovery analytics. PROMs, which characterize pain, mobility, fatigue, anxiety, and functional independence, provide subjectively important additional data concerning recovery trajectories. The use of digital methods can allow for PROMs to be captured with a greater frequency, while either dynamic Bayesian models or reinforcement learning could be used to occasionally modify postoperative commands and advice in real-time11. In conclusion, the intersection of surgical science and digitization has tremendous potential to disrupt recovery; recovery as patient-centered, dynamic, and collaborative. Surgical teams would be able to guide patients through postoperative recovery pathways with AI-enabled analytics, biosensing-enabled wearables, and patient-centered interfaces. Recovery could be clinically optimized in contextualized and responsive ways. This communication brings forth a new frontier in postoperative care through the application of AI technologies and digital tools by advocating patient-centered care. It enhances existing literature by offering personalized recovery pathways that transcend the conventional ERAS protocols to improve adaptability, monitoring, engagement, complication detection, and prompt action, especially for marginalized groups and in low-resource healthcare settings.
Thanigaivel et al. (Mon,) studied this question.