Abstract
Background
In the PORT trial (Perioperative Rehabilitation in Patients Undergoing Elective Cardiac Valve Surgery: A Randomised Controlled Trial), a short-term multidomain perioperative rehabilitation programme was associated with a reduced incidence of postoperative pneumonia in patients undergoing elective cardiac valve surgery. Whether baseline peak oxygen consumption (peak VO₂), a marker of cardiorespiratory reserve, modifies the effect of this intervention remains uncertain.
Objectives
To evaluate whether the association between perioperative rehabilitation and postoperative outcomes differs according to baseline peak VO₂ (≤ 20 vs. > 20 mL·kg⁻¹·min⁻¹).
Methods
This exploratory subgroup analysis included 702 participants from the PORT trial who underwent preoperative cardiopulmonary exercise testing. Patients were stratified by baseline peak VO₂ and randomized to perioperative rehabilitation—comprising education, inspiratory muscle training, active cycle of breathing techniques, and early mobilization—or usual care. The primary endpoint was a composite of in-hospital all-cause mortality, postoperative pulmonary complications (PPCs), and prolonged hospitalization (> 7 days). Logistic and Cox regression models adjusted for prespecified covariates were used to estimate treatment effects within peak VO₂ strata, and treatment-by–peak VO₂ interactions were tested.
Results
Of 702 participants, 202 (29%) had peak VO₂ >20 and 500 (71%) had peak VO₂ ≤20 mL·kg⁻¹·min⁻¹. The composite primary endpoint did not differ significantly between rehabilitation and usual care in either stratum (interaction P = 0.62). Perioperative rehabilitation was associated with a lower incidence of postoperative pneumonia in both peak VO₂ >20 mL·kg⁻¹·min⁻¹ (adjusted OR 0.30; 95% CI 0.13–0.71) and ≤ 20 mL·kg⁻¹·min⁻¹ (adjusted OR 0.65; 95% CI 0.47–0.90), with no significant interaction between peak VO₂ category and treatment effect (P for interaction = 0.12). No significant effect modification was observed for other outcomes, including SICU stay, total hospitalization duration, or 3-month mortality.
Conclusions
In this exploratory analysis of the PORT trial, the reduction in postoperative pneumonia associated with short-term perioperative rehabilitation was consistent across strata of baseline peak VO₂. These findings suggest that the pneumonia benefit of respiratory-focused perioperative rehabilitation may not depend on baseline cardiorespiratory fitness, although adequately powered studies are required to definitively assess effect modification. Although the observed reduction in pneumonia was statistically significant and biologically plausible, it arose from a secondary exploratory analysis without adjustment for multiple comparisons and should therefore be interpreted with caution.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s12871-026-04074-6.
Keywords: Perioperative Rehabilitation, Cardiac Valve Surgery, Postoperative Pulmonary Complications, Peak Oxygen Uptake (Peak VO2)
Introduction
Baseline cardiorespiratory fitness [1], commonly assessed by peak oxygen consumption (peak VO₂) [2], is a key determinant of perioperative risk and postoperative recovery in patients undergoing cardiac valve surgery [3]. Reduced exercise capacity is prevalent in this population [4] and has been consistently associated with higher rates of postoperative complications [5], prolonged hospitalization, and impaired functional recovery [6]. Peak VO₂ reflects the integrated performance of cardiovascular, pulmonary, and peripheral systems and serves as a global marker of physiological reserve [7]. However, the clinical implications of baseline Peak VO₂ in the context of perioperative rehabilitation remain complex and not fully understood.
Despite the established association between peak VO₂ and postoperative risk, it remains uncertain whether baseline peak VO₂ modifies the effectiveness of perioperative rehabilitation. Plausible but competing hypotheses can be proposed. On one hand, patients with lower peak VO₂, reflecting reduced physiological reserve and higher susceptibility to postoperative complications [8], may have greater potential for absolute benefit from structured rehabilitation [9]. On the other hand, severely impaired peak VO₂ may indicate limited tolerance to exercise-based interventions and potentially a blunted response to short-term perioperative rehabilitation [10]. Conversely, patients with higher peak VO₂ may be closer to a functional ceiling, limiting the magnitude of improvement [11], yet they may also be more capable of engaging in and responding to training stimuli. Therefore, the direction and magnitude of rehabilitation benefit across different levels of baseline peak VO₂ remain uncertain. Accordingly, this exploratory subgroup analysis was undertaken to examine whether baseline peak VO₂ is associated with differential effects of a multidomain perioperative rehabilitation intervention on postoperative outcomes, reflecting potential differences in physiological capacity to engage with and respond to the intervention.
Although the PORT intervention primarily comprises respiratory-focused components (education, inspiratory muscle training, active cycle of breathing techniques, and early mobilization) rather than isolated aerobic exercise, baseline peak oxygen consumption (peak VO₂) may nonetheless reflect a patient’s overall physiological reserve and capacity to engage with such interventions. Peak VO₂ integrates cardiovascular performance [12], pulmonary ventilation, respiratory muscle function [13], and peripheral oxygen transport, and thus serves as a global indicator of functional capacity [14–16]. Patients with higher baseline peak VO₂ may have greater ventilatory reserve and tolerance to physical activity [17], potentially facilitating participation in inspiratory muscle training, breathing techniques, and early mobilization, whereas markedly reduced peak VO₂ may indicate limited reserve and impaired functional capacity that could constrain engagement with perioperative rehabilitation [18]. On this basis, baseline peak VO₂ was examined as a potential effect modifier of the association between multidomain perioperative rehabilitation and postoperative outcomes among patients undergoing elective cardiac valve surgery.
Although perioperative rehabilitation has been increasingly evaluated in patients undergoing elective cardiac valve surgery, prior studies have largely focused on average treatment effects at the population level, with limited attention to potential heterogeneity of treatment response according to baseline physiological characteristics [19]. In particular, while peak oxygen consumption (peak VO₂) is widely used for perioperative risk stratification and prognostic assessment in patients undergoing cardiac valve surgery, its role as a potential effect modifier of perioperative rehabilitation—rather than as a predictor of postoperative risk— has not been well characterized. Whether patients with differing levels of baseline peak VO₂ derive comparable benefit from perioperative rehabilitation therefore remains insufficiently evaluated.
The PORT trial demonstrated that a short-term, perioperative, multimodal rehabilitation programme was associated with a reduced incidence of postoperative pneumonia in patients undergoing elective cardiac valve surgery [20]. The trial also incorporated standardized preoperative cardiopulmonary exercise testing with measurement of peak VO₂, and the rehabilitation intervention was not centered on isolated aerobic training but instead targeted multiple functional domains, allowing participation of patients across a wide range of baseline cardiorespiratory fitness [21]. In this context, baseline peak VO₂ may serve as an integrative marker of overall physiological reserve rather than aerobic capacity alone, providing a rationale to explore whether differences in baseline cardiorespiratory fitness are associated with differential intervention effects on postoperative pulmonary outcomes.
Accordingly, this study was conducted as an exploratory subgroup analysis of the PORT trial to assess whether there is evidence that baseline peak oxygen consumption (peak VO₂) is associated with heterogeneity in the effects of perioperative rehabilitation on in-hospital mortality, postoperative pulmonary complications (PPCs), and postoperative length of hospital stay among patients undergoing cardiac surgery for valvular heart disease [20]. We further examined the consistency of intervention-associated effects across strata of baseline peak VO₂ for individual pulmonary complications and related postoperative outcomes, including postoperative pneumonia.
Method
Peak VO₂ Stratification
Peak VO₂ was dichotomised using a threshold of 20 mL·kg⁻¹·min⁻¹, consistent with the Weber functional classification [22], in which values above this level indicate preserved or only mildly impaired cardiac functional capacity.This threshold was selected to provide a clinically interpretable stratification of baseline exercise capacity rather than to define a validated perioperative risk cutoff.
Study design and parent trial overview
This exploratory analysis was conducted within the framework of the previously published PORT trial, a single-center, randomized, attention-controlled trial evaluating a short-term, hospital-based, multidomain perioperative rehabilitation program in patients undergoing elective cardiac valve surgery [20].Following written informed consent and completion of baseline assessments, participants in the parent trial were randomly assigned in a 1:1 ratio to usual care or peri-operative rehabilitation through a secure central web-based randomisation platform (http://crdms.echobelt.org). Treatment allocation was generated through the secure central web-based randomisation platform using a computer-generated stratified permuted block randomisation schedule with variable block sizes (4–6), stratified according to age, left ventricular ejection fraction, and baseline pulmonary function. Allocation concealment was maintained through the centralised web-based randomisation system, which released treatment assignments only after participant enrolment and completion of baseline assessment. Owing to the nature of the intervention, blinding of participants and physiotherapists was not feasible; however, outcome assessors, data collectors, and treating cardiac surgeons remained blinded to group allocation throughout the trial. The trial design, intervention fidelity, and primary outcomes have been published previously; the present study represents a prespecified subgroup analysis examining whether baseline cardiopulmonary fitness modified treatment effects. This trial was conducted in accordance with the Declaration of Helsinki, approved by the Medical Research Ethics Committee of Guangdong Provincial People’s Hospital, and all participants provided written informed consent.
Allocation concealment was maintained through the centralised web-based randomisation system, which released treatment assignments only after participant enrolment and completion of baseline assessment.
Participants
Eligible participants were adults scheduled for elective mitral and/or aortic valve surgery with New York Heart Association (NYHA) functional class I–III symptoms. Major exclusion criteria included preoperative infective endocarditis or pneumonia, urgent or emergency surgery, severe hepatic dysfunction (Child–Turcotte–Pugh class ≥ B), dialysis-dependent renal failure, substantial neurological or musculoskeletal impairment, unstable vital signs, or concurrent enrolment in another clinical trial.
A summary of derivation of the present CPET subgroup cohort from the parent PORT trial population is provided in Supplementary Table 2.
Intervention and comparator
The multidomain perioperative rehabilitation intervention consisted of education, inspiratory muscle training, active cycle of breathing techniques, and early mobilisation, delivered during the perioperative hospitalisation period. Outcomes were compared against a usual-care control group receiving standard perioperative management.
Ancillary CPET cohort and baseline fitness assessment
Baseline cardiopulmonary fitness was assessed using cardiopulmonary exercise testing (CPET) conducted as part of a prespecified ancillary study within the PORT trial. According to the study protocol, participants in both treatment arms were invited to undergo CPET prior to surgery, with approximately 70% of the overall trial cohort anticipated to participate. The CPET protocol has been previously described in detail in cardiovascular populations. Peak oxygen consumption (peak VO₂) obtained from CPET was used as the primary measure of preoperative cardiorespiratory fitness. For participants enrolled in this ancillary study, CPET data were collected before randomisation and initiation of the perioperative rehabilitation intervention. The present subgroup analysis included 702 participants with valid pre-randomisation CPET-derived peak VO₂ measurements, and all analyses were restricted to this cohort.
Of the parent trial analysis population, 702 participants had valid baseline CPET-derived peak VO₂ data and were included in this subgroup analysis. Participants without valid baseline CPET data were not included. Because CPET assessment was performed prior to randomisation and intervention initiation, missing CPET data were unlikely to be related to treatment allocation, although potential selection bias cannot be excluded.
Objective of the present subgroup analysis
The interaction between treatment assignment and peak VO₂ strata was evaluated to assess whether baseline peak VO₂ modified the effect of perioperative rehabilitation on postoperative outcomes. This analysis was not intended to compare outcomes between peak VO₂ strata, but rather to examine the consistency of the intervention effect across different levels of baseline cardiorespiratory fitness.
Reporting standard
This randomised controlled trial is reported in accordance with the CONSORT (Consolidated Standards of Reporting Trials) 2025 guideline.
Statistical methods
Participant characteristics were summarized as mean ± standard deviation (SD), median (interquartile range [IQR]), or number (percentage), as appropriate. Baseline characteristics were compared between treatment groups within strata defined by baseline peak VO₂ (≤ 20 vs. > 20 mL·kg⁻¹·min⁻¹) using Student’s t-tests for continuous variables and chi-square or Fisher’s exact tests for categorical variables.
To evaluate whether baseline peak VO₂ category modified the association between the intervention (perioperative rehabilitation vs. usual care) and early postoperative outcomes, generalized linear models were constructed including terms for treatment group, peak VO₂ category, and their interaction. Models were adjusted for age, sex, left ventricular ejection fraction (LVEF) category (< 50% or ≥ 50%), and Global Initiative for Chronic Obstructive Lung Disease (GOLD) stage, consistent with the primary analysis. Subgroup-specific estimates were derived using least-square means to characterize the pattern of the intervention effect across peak VO₂ strata, and results were reported as adjusted mean differences or odds ratios (ORs) with 95% confidence intervals (CIs). Model assumptions were assessed using residual diagnostics.
For early postoperative outcomes occurring within 7 days (including pneumonia and other pulmonary complications), logistic regression was used as the primary analytical approach, given the fixed follow-up period and the primary interest in whether an event occurred within this time window rather than its exact timing. Results are presented as ORs with 95% CIs.
For time-to-event outcomes (including in-hospital mortality and all-cause death within 3 months), Cox proportional hazards regression models were applied, with results reported as hazard ratios (HRs) and 95% CIs.
For skewed continuous outcomes (SICU length of stay and postoperative hospitalization duration), group comparisons were performed using Wilcoxon rank-sum tests, and median differences with 95% CIs were estimated via bootstrap resampling.
All statistical tests were two-sided, with P < 0.05 considered statistically significant for main effects. The interaction between peak VO₂ category and treatment group was also considered statistically significant at P < 0.05.
Given the exploratory nature of this subgroup analysis, no adjustment for multiple comparisons was performed, and findings for secondary outcomes should be interpreted with caution.
Results
In the primary analysis, no significant difference was observed between the rehabilitation and usual care groups in the primary composite outcome of in-hospital all-cause mortality, postoperative pulmonary complications, and prolonged hospital stay (> 7 days). However, the incidence of postoperative pulmonary complications was lower in the rehabilitation group, driven mainly by a reduction in postoperative pneumonia. No significant differences were observed between groups in other secondary outcomes, including intensive care unit and total hospital length of stay, 3-month mortality, or patient-reported measures of anxiety, depression, and quality of life.
Baseline characteristics
Of the 818 participants enrolled in the PORT study, 702 underwent cardiopulmonary exercise testing and were classified according to baseline peak VO₂. Within each peak VO₂ stratum, baseline characteristics were generally comparable between the perioperative rehabilitation and usual care groups (Table 1). A higher proportion of males was observed in the usual care group among participants with peak VO₂ >20 mL·kg⁻¹·min⁻¹. Key baseline variables, including age, sex, cardiac function (LVEF category), and pulmonary function (GOLD stage), were accounted for in subsequent adjusted analyses.
Table 1.
Baseline characteristics by treatment group within baseline peak VO2 Strata
| Characteristic | Peak VO2>20(mL/min/kg) | p-value | Peak VO2≤20(mL/min/kg) | p-value | |||||
|---|---|---|---|---|---|---|---|---|---|
| All | Usual Care | Perioperative Rehabilitation | All | Usual Care | Perioperative Rehabilitation | ||||
| Demographic characteristics | 202 | 109 | 93 | 500 | 255 | 245 | |||
| Male | 148 ( 73.3) | 87 ( 79.8) | 61 ( 65.6) | 0.034 | 207 (41.4) | 107 ( 42.0) | 100 (40.8) | 0.866 | |
| Age(y) | |||||||||
| < 45 | 81 ( 40.1) | 44 ( 40.4) | 37 ( 39.8) | 0.308 | 112 (22.4) | 65 ( 25.5) | 47 (19.2) | 0.223 | |
| 45–59 | 80 ( 39.6) | 39 ( 35.8) | 41 ( 44.1) | 227 (45.4) | 113 ( 44.3) | 114 (46.5) | |||
| ≥ 60 | 41 ( 20.3) | 26 ( 23.9) | 15 ( 16.1) | 161 (32.2) | 77 ( 30.2) | 84 (34.3) | |||
| BMI(kg/m2) | 22.95 (3.28) | 23.18 (3.22) | 22.68 (3.35) | 0.280 | 23.62 (3.52) | 23.61 (3.60) | 23.64 (3.44) | 0.905 | |
| History of smoking | 30 ( 14.9) | 17 ( 15.6) | 13 ( 14.0) | 0.901 | 69 (13.8) | 32 ( 12.5) | 37 (15.1) | 0.485 | |
| Exercise habit | 89 ( 63.6) | 47 ( 61.0) | 42 ( 66.7) | 0.609 | 255 (57.3) | 127 ( 56.7) | 128 (57.9) | 0.869 | |
| Comorbidities | |||||||||
| Previous valve surgery history | 1 (0.5) | 0 (0.0) | 1 (1.1) | 0.937 | 44 (8.8) | 22 (8.6) | 22 (9.0) | 1 | |
| Ischemic Heart Disease | 4 ( 2.0) | 2 ( 1.8) | 2 ( 2.2) | 1 | 14 ( 2.8) | 5 ( 2.0) | 9 ( 3.7) | 0.374 | |
| Diabetes | 8 ( 4.0) | 7 ( 6.4) | 1 ( 1.1) | 0.114 | 31 ( 6.2) | 13 ( 5.1) | 18 ( 7.3) | 0.391 | |
| Hypertension | 38 ( 18.8) | 21 ( 19.3) | 17 ( 18.3) | 1 | 119 (23.8) | 57 ( 22.4) | 62 (25.3) | 0.503 | |
| Hyperlipidemia | 2 ( 1.0) | 1 ( 0.9) | 1 ( 1.1) | 1 | 7 ( 1.4) | 2 ( 0.8) | 5 ( 2.0) | 0.415 | |
| COPD | 0(0.0) | 0(0.0) | 0(0.0) | NA | 4(0.8) | 1(0.4) | 3(1.2) | 0.364 | |
| Stroke | 5 ( 2.5) | 2 ( 1.8) | 3 ( 3.2) | 0.857 | 24 ( 4.8) | 10 ( 3.9) | 14 ( 5.7) | 0.467 | |
| Peripheral vascular disease | 0(0.0) | 0(0.0) | 0(0.0) | NA | 2(0.4) | 1(0.4) | 1(0.4) | 1 | |
| Chronic Kidney Disease | 0(0.0) | 0(0.0) | 0(0.0) | NA | 1(0.2) | 0(0.0) | 1(0.4) | 0.49 | |
| NYHA Class | |||||||||
| I | 84 ( 41.6) | 47 ( 43.1) | 37 ( 39.8) | 0.517 | 147 (29.4) | 74 ( 29.0) | 73 (29.8) | 0.292 | |
| II | 103 ( 51.0) | 56 ( 51.4) | 47 ( 50.5) | 288 (57.6) | 142 ( 55.7) | 146 (59.6) | |||
| III | 15 ( 7.4) | 6 ( 5.5) | 9 ( 9.7) | 65 (13.0) | 39 ( 15.3) | 26 (10.6) | |||
| Left ventricle function | |||||||||
| Ejection fraction ≥ 50% | 191 ( 94.6) | 103 ( 94.5) | 88 ( 94.6) | 1 | 462 (92.4) | 238 ( 93.3) | 224 (91.4) | 0.526 | |
| Ejection fraction <50% | 11 ( 5.4) | 6 ( 5.5) | 5 ( 5.4) | 38 ( 7.6) | 17 ( 6.7) | 21 ( 8.6) | |||
| Pulmonary function | |||||||||
| FVC(%predicted) | 84.00[76.00, 92.00] | 84.00 [77.00, 92.00] | 83.00 [75.00, 94.75] | 0.693 | 79.00 [69.00, 87.75] | 78.50 [69.00, 85.25] | 80.00 [69.01, 88.00] | 0.263 | |
| FEV1(%predicted) | 87.27 (13.55) | 86.17 (13.08) | 88.52 (14.03) | 0.231 | 80.85 (15.46) | 80.21 (15.51) | 81.48 (15.42) | 0.366 | |
| FEV1/FVC(%predicted) | 108.00 [101.00, 114.00] | 107.00 [101.00, 113.00] | 108.00 [102.50, 114.50] | 0.13 | 109.00 [102.00, 115.75] | 110.00 [102.00, 116.00] | 109.00 [101.00, 115.00] | 0.655 | |
| GOLD | |||||||||
| I(80%≤FEV1%pre) | 155 ( 76.7) | 83 ( 76.1) | 72 ( 77.4) | 0.869 | 287 (57.4) | 142 ( 55.7) | 145 (59.2) | 0.469 | |
| II(50%≤FEV1%pre<80%) | 47 ( 23.3) | 26 ( 23.9) | 21 ( 22.6) | 213 (42.6) | 113 ( 44.3) | 100 (40.8) | |||
| III(30%≤FEV1%pre<50%) | 0 ( 0.0) | 0 ( 0.0) | 0 ( 0.0) | 0 ( 0.0) | 0 ( 0.0) | 0 (0.0) | |||
| Euroscore II | |||||||||
| low-to-medium risk | 172 ( 85.1) | 95 ( 87.2) | 77 ( 82.8) | 0.503 | 366 (73.2) | 182 ( 71.4) | 184 (75.1) | 0.401 | |
| high risk | 30 ( 14.9) | 14 ( 12.8) | 16 ( 17.2) | 134 (26.8) | 73 ( 28.6) | 61 (24.9) | |||
| Exercise tolerance | |||||||||
| Peak VO2(mL/min/kg) | 23.06 [21.47, 25.69] | 23.04 [21.40, 26.00] | 23.07 [21.70, 25.10] | 0.717 | 14.90 [12.40, 16.93] | 14.90 [12.50, 16.81] | 14.90 [12.25, 17.10] | 0.78 | |
| Peak VO2/predicted(%) | 73.00 [64.00, 83.63] | 73.00 [63.00, 82.00] | 74.00 [64.00, 83.84] | 0.926 | 55.00 [45.00, 64.21] | 55.00 [45.00, 64.00] | 54.00 [45.00, 65.00] | 0.6 | |
| Max workload/predicted(%) | 75.00 [65.00, 86.00] | 75.00 [66.00, 86.00] | 75.15 [62.20, 86.48] | 0.464 | 55.00 [44.00, 67.26] | 55.00 [44.87, 67.00] | 56.00 [43.00, 68.00] | 0.983 | |
| Physical frailty (handgrip strength, kg) | 35.90 [27.10, 43.45] | 37.90 [30.10, 43.60] | 34.40 [25.55, 43.30] | 0.149 | 26.45 [20.90, 35.27] | 26.35 [21.58, 34.40] | 26.75 [20.33, 37.40] | 0.99 | |
| Nutrition condition (NRS2002) | 0.00 [0.00, 0.00] | 0.00 [0.00, 0.00] | 0.00 [0.00, 0.00] | 0.296 | 0.00 [0.00, 1.00] | 0.00 [0.00, 1.00] | 0.00 [0.00, 1.00] | 0.723 | |
| Depressive symptoms (PHQ-9) | 2.00 [1.00, 5.00] | 2.00 [1.00, 4.00] | 2.50 [1.00, 6.00] | 0.193 | 3.00 [1.00, 7.00] | 3.00 [1.00, 7.00] | 4.00 [1.00, 7.00] | 0.735 | |
| Anxious symptoms (GAD-7) | 2.00 [0.00, 4.00] | 2.00 [0.00, 4.00] | 2.00 [0.00, 4.00] | 0.457 | 2.00 [0.00, 4.00] | 2.00 [0.00, 4.00] | 2.00 [0.00, 5.00] | 0.36 | |
| Quality of life(SF-36) | |||||||||
| Physical Component Summary(PCS) | 299.00 [244.00, 344.50] | 295.00 [249.00, 346.00] | 300.00 [227.75, 341.50] | 0.609 | 251.00 [191.25, 315.00] | 257.00 [191.50, 315.00] | 242.00 [191.50, 314.50] | 0.567 | |
| Mental Component Summary(MCS) | 320.00 [248.58, 354.38] | 322.00 [262.00, 355.00] | 310.00 [223.00, 352.00] | 0.459 | 296.17 [213.46, 340.00] | 301.17 [217.50, 337.00] | 291.25 [207.13, 342.00] | 0.708 | |
| Surgical procedure | |||||||||
| Aortic valve surgery | 59 ( 29.2) | 30 ( 27.5) | 29 ( 31.2) | 0.678 | 135 (27.0) | 69 ( 27.1) | 66 (26.9) | 1 | |
| Mitral valve surgery | 166 ( 82.2) | 88 ( 80.7) | 78 ( 83.9) | 0.692 | 442 (88.4) | 227 ( 89.0) | 215 (87.8) | 0.763 | |
| Tricuspid valve surgery | 60 ( 29.7) | 32 ( 29.4) | 28 ( 30.1) | 1 | 252 (50.4) | 134 ( 52.5) | 118 (48.2) | 0.373 | |
| With CABG | 6 ( 3.0) | 1 ( 0.9) | 5 ( 5.4) | 0.148 | 19 ( 3.8) | 10 ( 3.9) | 9 ( 3.7) | 1 | |
| Surgical type (%) | |||||||||
| open surgery | 118 ( 58.4) | 58 ( 53.2) | 60 ( 64.5) | 0.138 | 286 (57.2) | 152 ( 59.6) | 134 (54.7) | 0.308 | |
| minimally invasive surgery | 84 ( 41.6) | 51 ( 46.8) | 33 ( 35.5) | 214 (42.8) | 103 ( 40.4) | 111 (45.3) | |||
| Surgery time | |||||||||
| Cardiopulmonary bypass time(minutes) | 144.00 [110.00, 200.00] | 145.00 [112.75, 212.25] | 137.00 [108.00, 179.00] | 0.116 | 150.00 [118.00, 196.50] | 150.00 [118.00, 198.00] | 149.50 [118.25, 195.75] | 0.976 | |
| Aortic cross clamp time(minutes) | 88.00 [64.75, 122.50] | 89.50 [64.75, 125.75] | 84.00 [64.75, 120.25] | 0.449 | 90.00 [67.00, 121.00] | 87.50 [66.75, 124.00] | 91.50 [67.25, 117.75] | 0.847 | |
Data are presented as N (%) or mean ± SD or median[interquartile range, IQR] unless otherwise specified
#: Pvalues calculated by Fisher exact test or chi-square test; P-value is not applicable (NA) due to zero participants in both groups. P value marked in bold indicates statistically significant.
Abbreviations:Peak VO2:Peak oxygen uptake,BMI:body mass index, NYHA:New York Heart Association, FEV1:forced expiratory volume in one second, FVC:forced vital capacity ,GOLD:Global Initiative for Chronic Obstructive Lung Disease, NRS:nutrition risk screening, PHQ-9:Patient Health Questionnaire-9, GAD-7:Generalized Anxiety Disorder-7,SF-36:the 36-item Short FormHealth Survey, MCS:mental component score, PCS:physical component score, CABG: coronary artery bypass graft
Primary and secondary outcomes by treatment group within baseline peak VO₂ strata and test for interaction (Within 7 Days)
Primary and Secondary Outcomes by Treatment Group Within Baseline Peak VO₂ Strata and Test for Interaction are presented in (Table 2). In both peak VO₂ categories, the incidence of the composite primary endpoint—comprising in-hospital all-cause mortality, postoperative pulmonary complications (PPCs), and prolonged hospitalization (> 7 days)—did not differ significantly between the perioperative rehabilitation and usual care groups. Among participants with peak VO₂ >20 mL·kg⁻¹·min⁻¹, the composite endpoint occurred in 46.2% of patients in the rehabilitation group and 52.3% in the usual care group (adjusted OR 0.84; 95% CI 0.47–1.50). Similarly, among those with peak VO₂ ≤20 mL·kg⁻¹·min⁻¹, event rates were 51.4% and 53.3%, respectively (adjusted OR 0.89; 95% CI 0.62–1.27). There was no evidence of effect modification by baseline peak VO₂ (interaction P = 0.62).
Table 2.
Primary and secondary outcomes by treatment group within baseline peak VO2 strata and test for interaction (Within 7 Days)
| Outcomes | Peak VO2>20(mL/min/kg) | Peak VO2≤20(mL/min/kg) | P for Interaction | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Usual Care N(%) |
Perioperative Rehabilitation N(%) |
adjusted Effect size (95%CI) | P-Value | Usual Care N(%) |
Perioperative Rehabilitation N(%) |
adjusted Effect size (OR/HR,95%CI) | P-Value | ||
| Composite outcome | 57 (52.3) | 43 (46.2) | 0.844(0.473,1.503) | 0.563 | 136 (53.3) | 126 (51.4) | 0.886(0.618, 1.268) | 0.509 | 0.62 |
| In-hospital mortality, n(%) | 0(0) | 0(0) | 4 ( 1.6) | 2 ( 0.8) | 0.576(0.105 ,3.173 ) | 0.527 | 1 | ||
| Postoperative hospitalization > 7 days, n(%) | 28 ( 25.7) | 25 ( 26.9) | 1.093(0.570,2.090) | 0.787 | 68 (26.7) | 76 (31.0) | 1.198(0.806, 1.783) | 0.373 | 0.689 |
| any PPCs, n(%) | 49 ( 45.0) | 31 ( 33.3) | 0.645(0.356,1.160) | 0.145 | 120 (47.2) | 99 (40.4) | 0.732(0.509, 1.051) | 0.092 | 0.537 |
| Components of the PPCs | |||||||||
| Pulmonary atelectasis | 6 (5.5) | 1(1.1) | 0.128 ( 0.014, 1.14 ) | 0.065 | 13 ( 5.1) | 9 ( 3.7) | 0.713(0.304 ,1.673 ) | 0.437 | 0.263 |
| Aerothorax | 13 (11.9) | 5 (5.4) | 0.451 ( 0.158, 1.288 ) | 0.137 | 20 ( 7.9) | 14 ( 5.7) | 0.718(0.361,1.426) | 0.344 | 0.459 |
| Pleural effusion | 24 (22.0) | 20 (21.5) | 1.092 ( 0.599, 1.991 ) | 0.775 | 58 (22.8) | 54 (22.0) | 0.954(0.657,1.383) | 0.802 | 0.97 |
| Respiratory failure | 2 (1.8) | 3(3.2) | 2.669(0.410,22.229) | 0.307 | 14 ( 5.5) | 13 ( 5.3) | 0.981(0.439, 2.177) | 0.961 | 0.541 |
| Pneumonia | 24 (22.0) | 6 (6.5) | 0.273 ( 0.11, 0.675 ) | 0.005* | 87 (34.3) | 57 (23.3) | 0.61(0.436,0.853) | 0.004* | 0.096 |
| Secondary Outcome | |||||||||
| SICU length of stay(h), m(P25,P75) | 47.00 [27.00, 71.00] | 42.00 [23.00, 65.00] | 5(2, 26) | 0.069 | 44.00 [26.50, 73.00] | 47.00 [29.00, 73.00] | -3(2,216) | 0.444 | |
| Postoperative hospitalization, m(P25,P75) | 6.00 [5.00, 8.00] | 6.00 [5.00, 8.00] | 0(-1, 13) | 0.748 | 6.00 [5.00, 8.00] | 6.00 [5.00, 8.00] | 00,20) | 0.134 | |
| All-cause death in 3 months, n(%) | 0(0.0) | 2(2.2) | 1.096 ( 0.822, 1.462 ) | 0.532 | 6(2.4) | 2(0.8) | 0.396(0.079,1.984 ) | 0.26 | 1 |
Data are presented as N (%) or median[interquartile range, IQR] unless otherwise specified
P-value is not applicable (NA) due to zero participants in both groups
a:Binary outcomes via logistic regression, skewed continuous outcomes via Wilcoxon rank-sum tests ,survival outcomes via Cox regression. b: Multivariable adjusted effects are adjusted for age, sex, EF category and GOLD category. P value marked with asterisksindicates statistically significant
Abbreviations: CI confidence interval, PPCs postoperative pulmonary complications, SICU surgical intensive care unit
For PPCs, perioperative rehabilitation was associated with lower incidence compared with usual care in both peak VO₂ strata. In participants with peak VO₂ >20 mL·kg⁻¹·min⁻¹, PPCs occurred in 33.3% of the rehabilitation group versus 45.0% of the usual care group (adjusted OR 0.65; 95% CI 0.36–1.16). Corresponding rates among participants with peak VO₂ ≤20 mL·kg⁻¹·min⁻¹ were 40.4% and 47.2% (adjusted OR 0.73; 95% CI 0.51–1.05). No statistically significant interaction between baseline peak VO₂ category and treatment effect was observed (interaction P = 0.54).
Among individual PPCs, a consistent pattern was observed for postoperative pneumonia. In participants with peak VO₂ >20 mL·kg⁻¹·min⁻¹, perioperative rehabilitation was associated with a significantly lower incidence of pneumonia compared with usual care (6.5% vs. 22.0%; adjusted OR = 0.273, 95% CI 0.11–0.675; P = 0.006). Similarly, among participants with peak VO₂ ≤20 mL·kg⁻¹·min⁻¹, rehabilitation was also associated with a reduced incidence of pneumonia (23.3% vs. 34.3%; adjusted OR = 0.61, 95% CI 0.436–0.853; P = 0.004). No statistically significant interaction was observed between baseline peak VO₂ category and the effect of perioperative rehabilitation, suggesting a similar relative reduction in postoperative pneumonia across both peak VO₂ strata (interaction P = 0.10).
For secondary outcomes, including SICU length of stay and postoperative hospitalization duration, no significant differences were observed between treatment groups within either peak VO₂ stratum, and no modifying effect of baseline peak VO₂ was detected (all interaction P > 0.05).
Perioperative rehabilitation versus usual care on postoperative pulmonary complications during hospitalization, stratified by peak VO₂ (Table 3)
Table 3.
Perioperative rehabilitation versus usual care on postoperative pulmonary complications during hospitalization, stratified by peak VO₂
| Outcomes | Peak VO2>20(mL/min/kg) | Peak VO2≤20(mL/min/kg) | P for Interaction | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Usual Care N(%) |
Perioperative Rehabilitation N(%) |
adjusted Effect size (95%CI) | P-Value | Usual Care N(%) |
Perioperative Rehabilitation N(%) |
adjusted Effect size (95%CI) | P-Value | ||
| Composite outcome | 57 (52.3) | 43 (46.2) | 0.844(0.473, 1.503) | 0.563 | 136 (53.3) | 126 (51.4) | 0.886(0.618, 1.268) | 0.509 | 0.62 |
| any PPCs, n(%) | 50 (45.9) | 31 (33.3) | 0.623(0.344, 1.121) | 0.116 | 122 (48.0) | 102 (41.6) | 0.748(0.521, 1.072) | 0.115 | 0.434 |
| Pulmonary atelectasis | 7 (6.4) | 1 (1.1) | 0.112(0.013,0.963) | 0.046 | 15 (5.9) | 10 (4.1) | 0.667(0.299,1.488) | 0.322 | 0.222 |
| Aerothorax | 13 (11.9) | 6 ( 6.5) | 0.553(0.207,1.478) | 0.238 | 22 ( 8.7) | 14 ( 5.7) | 0.638(0.325,1.251) | 0.191 | 0.797 |
| Pleural effusion | 24 (22.0) | 20 (21.5) | 1.084(0.594,1.977) | 0.793 | 61 (24.0) | 56 (22.9) | 0.923(0.641,1.328) | 0.665 | 0.867 |
| Pneumonia | 25 (22.9) | 8 (8.6) | 0.302(0.129,0.706) | 0.006 | 91 (35.8) | 64 (26.1) | 0.65(0.471,0.897) | 0.009 | 0.118 |
Data are presented as N (%)
P value marked in bold indicates statistically significant
Abbreviations: PPCs: postoperative pulmonary complications
In both peak VO₂ subgroups, perioperative rehabilitation was associated with numerically lower incidences of overall PPCs compared with usual care; however, the between-group differences did not reach statistical significance (peak VO₂ >20 mL·kg⁻¹·min⁻¹: 33.3% vs. 45.9%; adjusted OR = 0.62; 95% CI: 0.34–1.12; peak VO₂ ≤20 mL·kg⁻¹·min⁻¹: 41.6% vs. 48.0%; adjusted OR = 0.75; 95% CI: 0.52–1.07; P for interaction = 0.43).
Among individual PPCs, perioperative rehabilitation was associated with a significantly lower incidence of pulmonary atelectasis among patients with higher peak VO₂ (1.1% vs. 6.4%; adjusted OR = 0.11; 95% CI: 0.01–0.96; P = 0.046), whereas no statistically significant difference was observed among those with lower peak VO₂.
For pneumonia, perioperative rehabilitation was associated with a significantly lower incidence in both peak VO₂ strata (peak VO₂ >20 mL·kg⁻¹·min⁻¹: 8.6% vs. 22.9%; adjusted OR = 0.30; 95% CI: 0.13–0.71; P = 0.006; peak VO₂ ≤20 mL·kg⁻¹·min⁻¹: 26.1% vs. 35.8%; adjusted OR = 0.65; 95% CI: 0.47–0.90; P = 0.009). No statistically significant interaction between baseline peak VO₂ category and treatment effect was detected (P for interaction = 0.12), suggesting a consistent association across levels of preoperative exercise capacity.
Figure 1 presents an additional time-to-event analysis, illustrating the cumulative hazard of pneumonia within 7 days after surgery, stratified by baseline peak VO₂ and treatment group. In both peak VO₂ strata, patients receiving perioperative rehabilitation exhibited a lower risk of pneumonia over time compared with those receiving usual care. These findings were consistent with the logistic regression analyses, which showed that perioperative rehabilitation was associated with a significantly lower risk of pneumonia in both patients with peak VO₂ >20 mL·kg⁻¹·min⁻¹ and those with peak VO₂ ≤20 mL·kg⁻¹·min⁻¹ (Table 2). No significant interaction was observed between baseline peak VO₂ category and treatment assignment, indicating that the association between perioperative rehabilitation and reduced pneumonia risk was consistent across levels of preoperative exercise capacity.
Fig. 1.
Cumulative incidence of pneumonia within 7 days, stratified by baseline peak VO₂ and treatment group
Supportive Cox time-to-event analyses yielded hazard ratios that were directionally and quantitatively consistent with the odds ratios observed in the primary logistic regression models.
Discussion
In this exploratory subgroup analysis of the PORT trial, perioperative rehabilitation was associated with a reduced incidence of postoperative pneumonia within 7 days of cardiac valve surgery across strata of baseline peak VO₂. No statistically significant interaction was observed between treatment allocation and peak VO₂ category, suggesting that no clear evidence of effect modification by baseline cardiorespiratory fitness was identified in this exploratory analysis. These findings build on the primary trial results [20] by suggesting broadly similar intervention-associated effects across patients with differing levels of preoperative exercise capacity.
Postoperative pneumonia emerged as the pulmonary complication most responsive to the intervention in a data-driven manner rather than as an a priori hypothesised endpoint. Accordingly, the present analysis aimed to assess the consistency of intervention-associated effects across baseline peak VO₂ strata, rather than to advance pneumonia-specific mechanistic hypotheses.
Although the reduction in postoperative pneumonia is clinically meaningful, it did not translate into significant improvements in the composite primary outcome, length of hospital stay, or short-term mortality. Several considerations may explain this discrepancy. First, the composite endpoint included outcomes with heterogeneous pathophysiological [23, 24] determinants and widely varying event frequencies, such as mortality and prolonged hospitalization, which are strongly influenced by surgical complexity [24], perioperative haemodynamics [25], and non-pulmonary complications. A reduction in a single respiratory component may therefore be insufficient to alter the overall composite outcome. Second, Mechanistically, inspiratory muscle training counteracts anaesthesia-induced diaphragmatic dysfunction [20, 26], improves inspiratory muscle strength, and enhances cough efficacy, thereby reducing mucus retention and the downstream risk of atelectasis and bacterial colonisation [27]. Importantly, these respiratory-specific mechanisms may not exclusively dependent on global aerobic capacity, providing one potential biological explanation for the broadly similar associations observed across strata of baseline peak VO₂. However, length of stay and early mortality are subject to competing risks and broader systemic influences less amenable to short-term respiratory-focused strategies [28, 29]. Third, the programme was intentionally short-term and confined to the peri-hospitalization period, which may be sufficient to influence early respiratory morbidity but insufficient to affect more distal or multifactorial outcomes. Taken together, postoperative pneumonia may represent a comparatively sensitive and biologically plausible outcome through which perioperative rehabilitation effects may be observed in this setting.
The mechanisms potentially underlying the observed association with reduced pneumonia may plausibly operate across a range of baseline peak VO₂ levels. The multimodal programme targeted perioperative respiratory dysfunction through inspiratory muscle training, structured breathing techniques, and early mobilisation—pathways not exclusively dependent on aerobic capacity. Anaesthesia-related diaphragmatic dysfunction, impaired cough efficacy, and postoperative mucus retention are common after cardiac valve surgery regardless of baseline fitness. The consistency of effect across peak VO₂ strata is therefore biologically plausible and suggests that these respiratory-focused components confer benefit independent of preoperative cardiorespiratory reserve.
Consistent with the primary analysis, no differential effects across peak VO₂ strata were observed for other outcomes, including prolonged hospitalisation, ICU stay, or mortality. Duration of hospital stay is influenced by institutional discharge practices, resource availability, and broader postoperative pathways, which are unlikely to be modified by baseline exercise capacity or short-term rehabilitation [30]. Although pneumonia is clinically important, its occurrence in this cohort did not translate into detectable differences in short-term mortality or length of stay.
Previous randomized trials and systematic reviews have shown that inspiratory muscle training reduces postoperative pulmonary complications, particularly pneumonia and atelectasis, in patients undergoing cardiac surgery [23, 31]. While patients with greater cardiopulmonary reserve may theoretically engage more effectively with such interventions, the present findings did not identify clear evidence of effect modification by baseline peak VO₂.
From an implementation perspective, the absence of effect modification supports the potential pragmatic applicability of the PORT rehabilitation programme. The intervention was designed to be short-term, resource-conscious, and adaptable to patient tolerance rather than tailored to predefined fitness thresholds. These findings may suggest that strict stratification by baseline peak VO₂ may not be essential for initial consideration of perioperative rehabilitation eligibility in elective cardiac valve surgery, although prospective confirmation is warranted.
Clinical implications
The absence of clear evidence for effect modification may suggest that the potential utility of perioperative rehabilitation should not necessarily be limited solely on the basis of impaired cardiopulmonary reserve. Patients across a range of baseline peak VO₂ levels demonstrated broadly similar relative associations with reduced postoperative pneumonia risk, which may support consideration of structured perioperative rehabilitation across diverse levels of preoperative functional capacity.
Limitations
This subgroup analysis was exploratory and not powered to detect interaction between treatment allocation and baseline peak VO₂; absence of statistically significant interaction should therefore not be interpreted as evidence of equivalent treatment effects. Peak VO₂ was assessed in a subset of trial participants, introducing potential selection bias. Peak VO₂ was dichotomised according to the Weber functional classification to preserve clinical interpretability; however, categorising a continuous physiological variable may reduce sensitivity to detect more subtle effect modification [32]. Continuous modelling or alternative functional forms may yield additional insights.
Adjusted analyses used the same prespecified covariates as the primary PORT trial to maintain consistency with the original randomized design. Although baseline characteristics were balanced between intervention and control groups within each peak VO₂ stratum, participants with lower versus higher peak VO₂ differed in several clinical features. Such between-stratum differences may introduce residual confounding in interaction analyses and should be considered when interpreting potential effect modification.However, because treatment allocation remained randomized within each peak VO₂ stratum, comparisons between rehabilitation and control groups within strata retain the protection afforded by randomization. Accordingly, within-stratum estimates of relative treatment effect are less susceptible to confounding than cross-stratum comparisons, although formal interaction tests should be interpreted with appropriate caution.
Adherence to the rehabilitation intervention could not be precisely quantified. The programme comprised multiple integrated components delivered throughout the peri-hospitalization period and individualized according to clinical status; a standardized adherence metric was not predefined, and differential adherence across peak VO₂ strata could not be assessed. Future studies incorporating predefined adherence measures or objective monitoring tools are warranted.
Some degree of censoring within the 7-day postoperative period (e.g., due to early discharge) may have occurred; however, given the short and fixed follow-up window, this is unlikely to have materially affected the results.
Because this subgroup analysis was restricted to participants with valid baseline CPET data, selection bias related to CPET availability cannot be entirely excluded.
Conclusion
In conclusion, this exploratory subgroup analysis indicates that the reduction in postoperative pneumonia observed in the PORT trial was consistent across levels of baseline peak VO₂. These findings support the generalisability of the intervention’s pneumonia benefit across a range of preoperative cardiorespiratory fitness levels, while underscoring the need for adequately powered studies specifically designed to evaluate whether baseline physiological reserve modifies response to perioperative rehabilitation.
Supplementary Information
Supplementary Figure S1.Cumulative incidence of any postoperative pulmonary complications within 7 days
Supplementary Table 1. Treatment effects across baseline peak VO₂ strata with corresponding P values for interaction.
Supplementary Table 2. Derivation of the analytic cohort for the present CPET subgroup analysis from the parent PORT trial.
Acknowledgements
The authors gratefully acknowledge the contributions of all co-authors to the design, data collection, analysis, and interpretation of this work. And We are grateful to the Department of Cardiology at Guangdong Provincial People’s Hospital for their assistance in data collection.
Abbreviations
- peak VO₂
Peak oxygen consumption
- PPCs
Postoperative pulmonary complications
- NYHA
New York Heart Association
- LVEF
Left ventricular ejection fraction
- GOLD
Global Initiative for Chronic Obstructive Lung Disease
- CPET
Cardiopulmonary exercise testing
- ICU
Intensive care unit
- SICU
Surgical intensive care unit
Authors’ contributions
All authors contributed to the design of the study. Rou Wang and Zengyan Huang wrote the paper and were involved in study execution. Zhongxing Jiang carried out data analysis and results interpretation. Bingqing Bai, Jinna Chang, Junshuo Zhu, Zengyan Huang participated in the process of data collection. Huan Ma and Kaixuan Lin contributed to performed and supervised the research, provided fund support, analysis tools or data, and revised the paper.
Funding
This study is supported by Su Ke’an Pharmaceutical Research and Development Project (202460), Guangzhou Municipal Science and Technology Program key projects (2023B03J1249), and Special Project for Clinical Collaboration between Traditional Chinese Medicine and Western Medicine of the National Administration of Traditional Chinese Medicine (No. ZDYN-2024-B-010; 2024129).
Data availability
The datasets generated and analysed during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
This study was conducted in accordance with the Declaration of Helsinki and approved by the Medical Research Ethics Committee of Guangdong Provincial People’s Hospital. All participants provided written informed consent.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rou Wang and Zengyan Huang contributed equally to this work.
Contributor Information
Kaixuan Lin, Email: 80379721@qq.com.
Huan Ma, Email: mahuan@gdph.org.cn.
References
- 1.Demchenko I, Prince SA, Merucci K, Cadenas-Sanchez C, Chaput JP, Fraser BJ, Manyanga T, McGrath R, Ortega FB, Singh B, et al. Cardiorespiratory fitness and health in children and adolescents: an overview of systematic reviews with meta-analyses representing over 125 000 observations covering 33 health-related outcomes. Br J Sports Med. 2025;59(12):856–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Mezzani A, Hamm LF, Jones AM, McBride PE, Moholdt T, Stone JA, Urhausen A, Williams MA, European Association for Cardiovascular, Rehabilitation P, et al. Aerobic exercise intensity assessment and prescription in cardiac rehabilitation: a joint position statement of the European association for cardiovascular prevention and rehabilitation, the American association of cardiovascular and pulmonary rehabilitation, and the Canadian association of cardiac rehabilitation. J Cardiopulm Rehabil Prev. 2012;32(6):327–50. [DOI] [PubMed] [Google Scholar]
- 3.Ferguson M, Shulman M. Cardiopulmonary exercise testing and other tests of functional capacity. Curr Anesthesiol Rep. 2022;12(1):26–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Ades PA, Savage PD, Brawner CA, Lyon CE, Ehrman JK, Bunn JY, Keteyian SJ. Aerobic capacity in patients entering cardiac rehabilitation. Circulation. 2006;113(23):2706–12. [DOI] [PubMed] [Google Scholar]
- 5.Steffens D, Ismail H, Denehy L, Beckenkamp PR, Solomon M, Koh C, Bartyn J, Pillinger N. Preoperative cardiopulmonary exercise test associated with postoperative outcomes in patients undergoing cancer surgery: a systematic review and meta-analyses. Ann Surg Oncol. 2021;28(12):7120–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Li C, Meng H, Wei Y, Liang Y, Xu Y, Huang X, Liang W, Quan J, Wu S, Wei X. Postoperative outcomes of preoperative exercise training in patients with operable non-small cell lung cancer: a systematic review and meta-analysis. Front Oncol. 2025;15:1563478. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Datta D, Normandin E, ZuWallack R. Cardiopulmonary exercise testing in the assessment of exertional dyspnea. Ann Thorac Med. 2015;10(2):77–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Yang N, Shi Z, Liu J, Yuan Y. Key cardiopulmonary exercise testing indicators for predicting the risk of postoperative cardiopulmonary complications in patients undergoing thoracoscopic lung resection. Front Surg. 2025;12:1765398. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Smyth E, O’Neill LM, Kearney N, Sheill G, Brennan L, Wade S, Grehan S, Begic S, Egana M, Ryan R, et al. Preoperative exercise to improve fitness in patients undergoing complex surgery for cancer of the lung or esophagus (PRE-HIIT): a randomized controlled trial. Ann Surg. 2025;282(5):690–8. [DOI] [PubMed] [Google Scholar]
- 10.Bojesen RD, Jorgensen LB, Grube C, Skou ST, Johansen C, Dalton SO, Gogenur I. Fit for Surgery-feasibility of short-course multimodal individualized prehabilitation in high-risk frail colon cancer patients prior to surgery. Pilot Feasibility Stud. 2022;8(1):11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Zhang J, Zhang F, Tao J, Yu MTX, Wen FTL. Effects of aerobic combined with resistance exercise on cardiopulmonary function in patients with coronary heart disease: a bayesian meta-analysis based on dose-response, baseline status, and other moderators. BMC Sports Sci Med Rehabil. 2025;17(1):256. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Wang Y, Li H, Wang J, Zhao W, Zeng Z, Hao L, Yuan Y, Lin Y, Wu Y, Wang Z. Normal references of peak oxygen uptake for cardiorespiratory fitness measured with cardiopulmonary exercise testing in Chinese adults. J Clin Med. 2022;11(16):4909. [DOI] [PMC free article] [PubMed]
- 13.Choi H-E, Min E-H, Kim H-K, Kim H-J, Jang H-J. Peak oxygen uptake and respiratory muscle performance in patients with chronic obstructive pulmonary disease: clinical findings and implications. Medicine 2022, 101(42). [DOI] [PMC free article] [PubMed]
- 14.Arena R, Sietsema KE. Cardiopulmonary exercise testing in the clinical evaluation of patients with heart and lung disease. Circulation. 2011;123(6):668–80. [DOI] [PubMed] [Google Scholar]
- 15.Rasch-Halvorsen O, Hassel E, Brumpton BM, Jenssen H, Spruit MA, Langhammer A, Steinshamn S. The association between normal lung function and peak oxygen uptake in patients with exercise intolerance and coronary artery disease. PLoS ONE. 2020;15(5):e0232693. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Prokopidis K, Irlik K, Piasnik J, Michalska Z, Hendel M, Nabrdalik K, Lip GYH. Prognostic impact of peak oxygen consumption in heart failure: a systematic review and meta-analysis. ESC Heart Fail. 2025;12(5):3624–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Murata M, Adachi H, Nakade T, Kobayashi Y, Agostoni P. Relationship between ventilatory pattern and peak VO(2) and area M regulates the respiratory system during exercise. J Cardiol. 2020;76(5):521–8. [DOI] [PubMed] [Google Scholar]
- 18.Puttini F, Pezzuto B, Vignati C. Cardiopulmonary exercise testing in the prognostic assessment of heart failure: from a standardized approach to tailored therapeutic strategies. Med (Kaunas) 2025, 61(10). [DOI] [PMC free article] [PubMed]
- 19.Delimanoli E, Muurlink O, Myrianthefs P, Korompeli A. Cardiac rehabilitation after open heart surgery: a narrative systematic review. J Cardiovasc Dev Dis 2024, 11(11). [DOI] [PMC free article] [PubMed]
- 20.Liu Y, Liu F, Xu M, Zhou H, Wang H, Liu Y, He X, Huang H, Sun T, Jian X et al. Peri-operative rehabilitation in patients undergoing elective cardiac valve surgery: a randomised controlled trial. Anaesthesia 2025. [DOI] [PMC free article] [PubMed]
- 21.Zhou H, Liu F, Liu Y, He X, Ma H, Xu M, Wang H, Zhang G, Cai X, Chen JY, et al. Protocol for the PORT study: short-term perioperative rehabilitation to improve outcomes in cardiac valvular surgery - a randomised control trial. BMJ Open. 2023;13(12):e074837. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Weber KT, Janicki JS, McElroy PA, Reddy HK. Functional classification of chronic heart failure based on peak oxygen consumption and anaerobic threshold. Am J Cardiol. 1985;55(4):A22–31. [Google Scholar]
- 23.Xiang Y, Zhao Q, Luo T, Zeng L. Inspiratory muscle training to reduce risk of pulmonary complications after coronary artery bypass grafting: a systematic review and meta-analysis. Front Cardiovasc Med. 2023;10:1223619. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Head SJ, Osnabrugge RL, Howell NJ, Freemantle N, Bridgewater B, Pagano D, Kappetein AP. A systematic review of risk prediction in adult cardiac surgery: considerations for future model development. Eur J Cardiothorac Surg. 2013;43(5):e121–129. [DOI] [PubMed] [Google Scholar]
- 25.Biccard BM, Rodseth RN. What evidence is there for intraoperative predictors of perioperative cardiac outcomes? A systematic review. Perioperative Med. 2013;2:14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Huai H, Ge M, Zhao Z, Xiong P, Hong W, Jiang Z, Wang J. Early diaphragm dysfunction assessed by ultrasonography after cardiac surgery: a retrospective cohort study. Front Cardiovasc Med. 2024;11:1457412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Dot I, Pérez-Teran P, Samper MA, Masclans JR. Diaphragm dysfunction in mechanically ventilated patients. Arch Bronconeumol. 2017;53(3):150–6. [DOI] [PubMed] [Google Scholar]
- 28.Pato MF, Gelape CL, Cassiano TJM, Carvalho A, Cintra PR, Passaglia LG, Costa HS, Tan TC et al. Santos Junior PCd, Bráulio R : Determinants of prolonged length of hospital stay after cardiac surgery: impact of rheumatic heart disease. Medical Express 2015, 2(3).
- 29.Hadareanu CD, Hadareanu DR, Stoiculescu FM, Berceanu MC, Donoiu I, Istratoaie O, Florescu C, Novac MB, Raicea VC. Predictors of prolonged intensive care unit stay and in-hospital mortality following cardiac surgery: an integrated analysis from the PROCARD-ATI Study. J Clin Med 2025, 14(8). [DOI] [PMC free article] [PubMed]
- 30.Frye RE, Rossignol DA. Identification and treatment of pathophysiological comorbidities of autism spectrum disorder to achieve optimal outcomes. Clin Med Insights Pediatr. 2016;10:43–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Karanfil EO, Møller AM. Preoperative inspiratory muscle training prevents pulmonary complications after cardiac surgery – a systematic review. Dan Med J. 2018;65(3):A5450. [PubMed] [Google Scholar]
- 32.Thoresen M. Spurious interaction as a result of categorization. BMC Med Res Methodol. 2019;19(1):28. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Figure S1.Cumulative incidence of any postoperative pulmonary complications within 7 days
Supplementary Table 1. Treatment effects across baseline peak VO₂ strata with corresponding P values for interaction.
Supplementary Table 2. Derivation of the analytic cohort for the present CPET subgroup analysis from the parent PORT trial.
Data Availability Statement
The datasets generated and analysed during the current study are available from the corresponding author on reasonable request.