Background/Objectives: Patients with pre-existing chronic obstructive pulmonary disease (COPD) and/or chronic heart failure (CHF) are a particularly high-risk surgical population in whom postoperative pulmonary complications (PPCs) drive a disproportionate share of morbidity, prolonged stay and mortality. Prehabilitation—the structured optimization of functional reserve during the preoperative interval—has emerged as a proactive, largely non-pharmacological strategy to raise that reserve before the surgical insult. Yet the evidence base is organized predominantly by surgical procedure and enrolled in mixed populations, so the patients with the least physiological reserve—those with established COPD or CHF—remain comparatively understudied. This narrative review takes the comorbidity, rather than the incision, as its organizing axis. Methods: We summarize the pathophysiology that links chronic cardio-respiratory disease to perioperative respiratory failure; appraise the principal non-pharmacological interventions, with the respiratory components (inspiratory muscle training, pulmonary rehabilitation, breathing techniques) as the core and exercise, nutritional, psychological, and smoking-cessation elements as the multimodal context; and re-read the surgical evidence through the lens of the underlying disease. Results: A cross-surgical meta-analysis indicates that preoperative exercise training reduces PPCs (relative risk ≈ 0.52) with no significant difference in effect across surgery type or training modality, supporting a comorbidity-centered rather than procedure-centered framework. We address practical determinants of implementation—timing, patient selection, adherence, home-based and telemonitored delivery, and low-resource applicability—and highlight two cross-cutting problems: heterogeneity of intervention prescription and the lack of standardized PPC definitions. Conclusions: The scarcity of disease-specific evidence is itself the central finding, revealing a mismatch between clinical risk and research investment.
Paternò et al. (2026) studied this question.