Authors: Naile Akıncı, Esra Eren
Categories: Research, Mitral valve replacement, Breathing exercises, Pulmonary function, Postoperative complications
Source: BMC Surgery
Authors: Naile Akıncı, Esra Eren
Postoperative pulmonary complications remain a major cause of morbidity after cardiac valve surgery. Although incentive spirometry (Triflow) is routinely used in postoperative care, evidence regarding the additional benefits of combining Triflow with deep breathing and coughing exercises remains limited. This study aimed to evaluate the effects of combined Triflow, deep breathing, and coughing exercises on postoperative pulmonary function in adult patients undergoing mitral valve replacement.
This randomized controlled, single-blind trial was conducted between May and August 2025 in a private hospital in Istanbul. A total of 60 adult patients undergoing mitral valve replacement were randomly allocated to an experimental group (n = 30) or a control group (n = 30) using simple randomization. The experimental group performed Triflow combined with deep breathing and coughing exercises, while the control group performed Triflow alone.
Postoperative SpO₂ levels were significantly higher in the experimental group at T1 (p = 0.009; 95% CI: 0.46–3.14), T2 (p < 0.001; 95% CI: 1.57–3.43), and T3 (p < 0.001; 95% CI: 2.72–4.54). The FEV₁/FVC ratio increased significantly in the experimental group compared with the control group at discharge (p < 0.001; 95% CI: 4.46–6.41). Respiratory rate was significantly higher in the experimental group at T1 (p < 0.001; 95% CI: 1.68–4.45), T2 (p < 0.001; 95% CI: 3.34–6.26), and T3 (p < 0.001; 95% CI: 5.23–8.37). Hematocrit levels were significantly lower in the experimental group at T1 (p = 0.039; 95% CI: −8.32 to − 0.24), T2 (p = 0.007; 95% CI: −8.29 to − 1.40), and T3 (p = 0.034; 95% CI: −6.54 to − 0.28). Pain scores were significantly lower in the experimental group at T1 (p < 0.001; 95% CI: −2.82 to − 1.71) and T2 (p < 0.001; 95% CI: −1.98 to − 1.08). Time to first mobilization was significantly shorter in the experimental group (p < 0.001; 95% CI: −2.94 to − 1.26). No postoperative pulmonary complications were observed in either group.
The combined application of Triflow, deep breathing, and coughing exercises was associated with significant improvements in postoperative pulmonary function, oxygen saturation, pain reduction, and earlier mobilization compared with Triflow alone in patients undergoing mitral valve replacement. These findings suggest that a structured, combined respiratory exercise protocol may provide additional clinical benefits in the early postoperative period.
ClinicalTrials.gov Identifier NCT06997224 (First registered on 15 May 2025).
The online version contains supplementary material available at 10.1186/s12893-026-03542-7.
When a healthy lifestyle and pharmacological agents are insufficient for treatment, cardiac surgery is performed [1]. After surgery, various complications may arise, such as respiratory muscle weakness, decreased pulmonary function, and pulmonary infections resulting from sternotomy and extracorporeal circulation [2]. Postoperative pulmonary complications are a significant cause of postoperative morbidity and are associated with restrictive pulmonary disorders resulting from decreased lung volume and impaired gas exchange [3, 4]. With over 40 million people worldwide suffering from mitral or aortic valve disease and more than 180,000 valve replacement surgeries performed annually, the prevalence of postoperative pulmonary complications remains a critical issue among cardiac surgery patients [5].
Mitral valve surgery is a procedure that significantly affects the patient’s cardiac, hemodynamic, respiratory, and multiple organ functions. These physiological impacts are likely to influence both perioperative and long-term outcomes [6]. Borgersen et al. (2023) reported a one-year mortality rate of 12.0% in the overall population after valve surgery. In the same study, age-stratified subgroup analyses showed that the one-year mortality rate was 3.6% in patients younger than 75 years [7]. In a study conducted by Katiyer et al. (2021), significant impairments in FEV₁ and forced vital capacity were observed after cardiac surgery, and these functions had not fully recovered even by the seventh postoperative day [8]. Pulmonary complications following cardiac surgery include atelectasis, pneumonia, pleural effusion, pneumothorax, and pulmonary edema [9]. During surgery, factors such as anesthesia induction and manual compression of the lungs can lead to atelectasis, while ineffective coughing, shallow breathing, interstitial edema, and immobility in the postoperative period further contribute to pulmonary dysfunction [4, 10]. In addition, prolonged apnea and lung collapse during cardiopulmonary bypass may contribute to postoperative pulmonary dysfunction by promoting atelectasis and impaired gas exchange. Postoperative pain and anxiety can also inhibit deep breathing and effective coughing [11]. This ultimately leads to secretion retention and disruption of normal gas exchange mechanisms [12].
To prevent and manage postoperative pulmonary complications, respiratory exercises are routinely prescribed after cardiac surgery. The primary goals of these exercises are to improve ventilation–perfusion matching, increase lung volume, enhance mucociliary clearance, and reduce pain [13]. The most commonly used breathing exercises include deep breathing, effective coughing, pursed-lip breathing, diaphragmatic breathing, and incentive spirometry [14]. Breathing exercises have been shown to have positive effects on postoperative atelectasis, arterial SpO₂ (oxygen saturation), and pulmonary function [15]. They strengthen the respiratory muscles, improve thoracoabdominal movements, and increase lung volume [16]. Studies have shown that postoperative breathing exercises facilitate lung expansion, regulate pulmonary circulation, and affect pulmonary function values [17, 18]. These interventions have also been reported to improve functional capacity, reduce muscle weakness, prevent or minimize postoperative complications, and enhance quality of life after cardiac surgery [19].
Nurses play a critical role in managing postoperative symptoms and ensuring optimal care. Effective management of respiratory complications accelerates the recovery process and enhances the quality of care [20]. Deep-breathing and coughing exercises, which constitute an important component of pulmonary rehabilitation, support lung function by improving secretion clearance, tidal volume, thoracic mobility, and oxygen delivery to tissues [21]. The physiological benefits of these exercises have been extensively described in the literature. In postoperative care, the Triflow device is routinely used as a standard practice in many centers. Triflow is designed to increase inspiratory muscle strength and elevate transpulmonary pressure, serving as a tool that enhances patients’ respiratory motivation [22]. However, Triflow is often applied alone, and complementary techniques such as deep-breathing and coughing exercises frequently remain secondary. The literature suggests that methods used in isolation may be insufficient to fully optimize postoperative respiratory function. This indicates that combining deep-breathing and coughing exercises with the standard use of Triflow may represent a physiologically more rational approach. Deep-breathing exercises promote alveolar expansion, coughing maneuvers facilitate secretion clearance, and Triflow enhances inspiratory volumes, thereby potentiating these mechanisms. Therefore, using all three techniques together may provide multidimensional support for postoperative pulmonary function. Zarneshan et al. (2021) applied stretching exercises targeting the thorax and shoulder girdle in addition to respiratory exercises after coronary artery bypass surgery and reported that this combined approach significantly improved oxygenation [23]. Similarly, Zare et al. (2025) incorporated chest wall stretching and inspiratory muscle mobilization techniques into routine breathing exercises and achieved more pronounced improvements in PCO₂ and oxygen saturation [23]. In the study by Gerçek et al., core stabilization and accessory respiratory muscle strengthening were added to breathing exercises, resulting in an increase in the FEV₁/FVC ratio [24]. The common finding across these studies is that not a single technique but rather the combination of multiple modalities that influence respiration multidimensionally leads to stronger physiological outcomes.
Despite this, research evaluating the combined use of such techniques—particularly in patients undergoing open mitral valve replacement—remains limited [12, 25, 26]. Therefore, the present study aimed to evaluate the effects of systematically adding deep-breathing and coughing exercises to routine Triflow use on postoperative pulmonary function in patients undergoing mitral valve replacement. By demonstrating the effectiveness of combined respiratory exercises that provide multidimensional support to postoperative pulmonary function, this study seeks to offer clinicians more effective rehabilitation strategies. Consequently, a reduction in complications, accelerated recovery, and improved quality of life are anticipated. Rather than proposing a novel therapeutic approach, the present study focuses on systematically evaluating a commonly recommended but inconsistently implemented respiratory rehabilitation strategy within a clearly defined clinical context. In addition, the study is expected to contribute to the literature by providing new scientific data for future research.
This randomized controlled, single-blind trial was conducted between May and August 2025 in a private hospital in Istanbul. ClinicalTrials.gov Identifier: NCT06997224 (first registered on 15 May 2025).
H1: In patients undergoing mitral valve replacement, the combined application of Triflow, deep breathing, and coughing exercises has a significant effect on postoperative respiratory function and complications.H1A (Sub-hypothesis): The combined application of Triflow, deep breathing, and coughing exercises affects postoperative cardiac parameters in patients undergoing mitral valve replacement.H1B (Sub-hypothesis): The combined application of Triflow, deep breathing, and coughing exercises affects the time to first mobilization after surgery.H1C (Sub-hypothesis): The combined application of Triflow, deep breathing, and coughing exercises affects postoperative pain levels in patients undergoing mitral valve replacement.
In determining the sample size for this study, the effect sizes reported in the pilot randomized controlled trial by Alaparthi et al., which compared three different breathing techniques following heart valve replacement surgery, were used as a reference [12]. In that study, a large effect size of Cohen’s d = 1.03 was reported. Based on this value, an a priori power analysis was conducted using G*Power 3.1. Assuming a two-tailed independent samples t-test, a significance level (α) of 0.05, a statistical power (1–β) of 0.95, and an equal group allocation ratio (N₂/N₁ = 1), a minimum of 26 participants per group was required. The analysis yielded a noncentrality parameter (δ) of 3.71, a critical t value of 2.01, degrees of freedom of 50, and an achieved power (1–β) of 0.953. To account for potential dropouts and ensure robust parametric tests, at least 30 participants were included in each group (Fig. 1).Fig. 1CONSORT Flow Diagram
The participants were diagnosed by a multidisciplinary team consisting of an intensive care specialist, a cardiovascular surgeon, and a consulting pulmonologist. All physicians involved have extensive clinical experience in managing cardiac surgery patients and postoperative pulmonary complications.
Voluntary participation.
Literate and proficient in the native language.
Undergoing mitral valve surgery, with the diagnosis of mitral valve disease confirmed by ICD-10 (I34.x) or ICD-11 (BA60.x) codes.
Age ≥ 18 years.
Hemodynamically unstable patients (e.g., those requiring high-dose vasoactive support, experiencing significant arrhythmias, or demonstrating persistent hypotension despite medical management).
Development of postoperative respiratory complications (e.g., Stage 3 pulmonary congestion, pulmonary edema, pneumonia).
Lip or palate anomalies.
Need for reintubation.
Prolonged intubation (> 72 h).
Active respiratory tract infections (e.g., tuberculosis, COVID-19).
Participants were assigned to groups using simple randomization. Random assignment was performed by flipping a those who got heads were assigned to the experimental group, and those who got tails were assigned to the control group. Although all patients were taught how to use the Triflow device, only the intervention group performed additional deep breathing and coughing exercises; Triflow use is part of the clinic’s routine practice. Participants did not know which group they were in, and upon assignment to a group, they signed a consent form specific to that group. In this study, participants were blinded to the patient group they were assigned to, but the researcher was not blinded to the distribution of patients across groups. The study was conducted as a single-blind trial in accordance with the CONSORT guidelines.
The primary outcome of this randomized controlled trial is the effect of a combination of Triflow, deep breathing, and coughing exercises on postoperative respiratory function (SpO₂, FEV₁/FVC) and respiratory complications (atelectasis, pneumonia, hypoxemia, pulmonary congestion, and prolonged oxygen requirement) in patients undergoing mitral valve replacement. “Secondary outcomes include effects on hemodynamic parameters (heart rate, respiratory rate, blood pressure, hematocrit level), time to first ambulation, and pain levels.”
This form, developed by the researchers based on the literature, includes age, gender, marital status, parenthood status, education level, occupation, cohabiting individuals, previous surgical history, presence of chronic pain, BMI (Body Mass Index), tobacco and alcohol use, EF (Ejection Fraction) percentage, surgery duration, cardiopulmonary bypass duration, postoperative intubation duration, intensive care unit stay duration, and reintubation.
It is designed to assess cardiovascular and respiratory parameters, time to first movement, pain, and complications.
The VAS is a one-dimensional measure used to assess subjective parameters. It consists of a 10-cm line drawn vertically or horizontally, with the endpoints representing the extremes of subjective experience (0 = “no pain,” 10 = “worst/unbearable pain”). Patients indicate their condition by marking a point on the line. The VAS is widely accepted as a valid and reliable tool for assessing the severity of acute pain after surgery.
The research was carried out in the cardiovascular unit of a private hospital. All patients stayed in single rooms in the hospital where the procedure was implemented. All procedures were delivered individually to each patient following a standardized protocol, using the same method, duration, and frequency. All participants fully adhered to the intervention protocol, and the treatment compliance/adherence rate was 100%. No participants withdrew from the study, and all patients completed the planned postoperative assessments. During the implementation of the intervention, patients were closely monitored for potential adverse events, including oxygen desaturation, hemodynamic instability, and intolerance to the respiratory exercises. It was predefined that the intervention would be immediately discontinued if any adverse event occurred, and appropriate clinical management would be initiated. However, no adverse events related to Triflow, deep breathing, or coughing exercises were observed throughout the study period. The researcher possesses the necessary theoretical knowledge and practical experience to reliably administer and score these measures.
Before surgery, the experimental group received training on the use of the Triflow device and deep breathing and coughing exercises. Subsequently, the forms included in the appendix were completed. Although Triflow use is standard practice in routine postoperative care at the study hospital, it does not include deep breathing and coughing exercises. If there are no complications, the average length of hospital stay is two days.
Vital signs were measured prior to the intervention, and SpO₂ was assessed using a pulse oximeter Pain was assessed using the VAS, and FEV₁/FVC was measured using spirometry. Blood was drawn for hematocrit assessment. On the first day after surgery, patients performed the Triflow exercise ten times every hour while awake and performed coughing and breathing exercises at alternate hours. Forms were filled out again at the end of the day. Interventions continued until discharge. On the first day, exercises were performed in the patient’s bed. On the second day, patients sat upright in a chair and supported their diaphragm with both hands. They were instructed to take a deep breath through their nose, count to three, and then exhale forcefully through pursed lips. During the exercises, the surgical site was supported with a pillow. Patients were mobilized with active–passive bed exercises. For Triflow use, patients were instructed to place the mouthpiece between their lips, inhale with all their strength to lift the balls, hold at the highest level for three seconds, then remove the device and exhale forcefully while coughing. Fifteen minutes after the intervention, vital signs, SpO₂, and VAS were reassessed. At the end of the day, blood was drawn again for hematocrit measurement. The same assessments, except for spirometry, were repeated in the evening.
Vital signs and SpO₂ were measured. Pain was assessed using the VAS. Triflow and deep breathing and coughing exercises continued. Spirometry (FEV₁/FVC) was repeated, and blood was drawn for hematocrit assessment.
Patients performed only the Triflow exercise. Vital signs and SpO₂ were measured, pain was assessed using the VAS, and FEV₁/FVC was evaluated using spirometry. Blood was drawn for hematocrit assessment. Fifteen minutes after the intervention and in the evening, all tests except spirometry were repeated.
Vital signs and SpO₂ were measured, and pain was assessed using the VAS. Triflow exercises continued. Spirometry and hematocrit measurements were repeated prior to discharge.
The data obtained in the study were analyzed using the SPSS 29.0 statistical software package. To compare demographic and clinical characteristics between the groups, an independent samples t-test was used for normally distributed variables, the Mann–Whitney U test was used for non-normally distributed variables, and Pearson’s chi-square test was applied for categorical variables. To examine the within-group time effect for repeated measurements in the postoperative follow-up form, repeated-measures ANOVA was applied; when the sphericity assumption was violated, the Greenhouse–Geisser correction was used. F values, p values, and partial eta-squared (η²) effect sizes were reported for the time × group interaction and for changes over time. Between-group differences at each time point were evaluated using the independent samples t-test, and the results were presented with 95% confidence intervals and Hedges’ g effect sizes. For analyses of within-group pre-post changes in FEV₁/FVC, the paired samples t-test was used when normality was met, and the Wilcoxon signed-rank test was used when normality was violated; the magnitude of change was reported using Cohen’s d effect size. A significance level of p < 0.05 was accepted for all statistical analyses.
The mean age of the participants was 48.57 ± 17.35 in the experimental group and 58.27 ± 11.64 in the control group. When comparing the experimental group with the control group, significant differences were observed in terms of age, education level, economic status, previous surgical history, EF percentage, surgery duration, cardiopulmonary bypass duration, intubation duration, and intensive care unit stay duration (p < 0.05) (Table 1).Table 1Demographic and clinical characteristics of the study population (N = 60)Experimentaln = 30Controln = 30test valuep valueAge48.57 ± 17.3558.27 ± 11.640.019^m^Sex Female1446.67%1446.67%1.000^x²^ Male1653.33%1653.33%Marital status Married1756.67%2170.00%0.284^x²^ Single1343.33%930.00%Educational level Primary school310.00%1033.33%0.009^x²^ High school826.67%1343.33% Undergraduate1240.00%620.00% Postgraduate723.33%13.33%Economic status Income less than expenses310.00%1446.67%10.1810.006^x²^ Income equal to expenses2273.33%1446.67% Income greater than expenses516.67%26.67%Living arrangement Alone413.33%620.00%4.3310.228^x²^ With spouse only930.00%930.00% With spouse and children1136.67%1446.67% With parents620.00%13.33%Employment status Full-time826.67%1136.67%0.7210.697^x²^ Part-time1136.67%930.00% Unemployed1136.67%1033.33%Previous surgery Yes413.33%1240.00%5.4550.020^x²^ No2686.67%1860.00%Smoking Yes826.67%1136.67%0.6930.405^x²^ No2273.33%1963.33%Alcohol Yes516.67%516.67%0.0001.000^x²^ No2583.33%2583.33%BMI 18.5–24.926.67%516.67%1.4580.482^x²^ 25–29.9.92066.67%1860.00% ≥ 30826.67%723.33%Ejection fraction (%)≤ 55%1136.67%2066.67%5.4060.020^x²^≥ 55%1963.33%1033.33%Duration of surgery (min)3.97±1.255.13±1.530.003^m^Cardiopulmonary bypass time (min)1.87±0.975.40±4.89***< 0.001***^m^Intubation time3.33±3.256.70±3.59***< 0.001***^m^ICU length of stay1.90±0.551.20±0.48***< 0.001***^m^
Table 2Inter-group variations of postoperative patient Follow-Up and evaluation formGroupsExperimental (n = 30)Control (n = 30)t p 95% GAHedges gMeanSdMeanSdPulse Pre90.67015.87585.97013.4741.2360.221–2.91, 12.310.32Pulse T185.43012.53684.60013.1030.2520.802–5.79, 7.460.06Pulse T282.87010.21484.73012.537−0.6320.530–7.78, 4.04−0.16Pulse T386.7309.30785.40011.3000.4990.620–4.02, 6.680.13F, p, η²F(3.87) = 4.048, p = 0.010, η²=0.122****F(3.87) = 0.234, p = 0.873, η²=0.008Bonferroni1 > 2,3; 3 < 4 Mean
Sd
Mean
Sd
t
p
95% GA
Hedges g Respiratory rate Pre11.6002.64710.8003.1781.0600.294–0.71, 2.310.27 Respiratory rate T113.7302.27310.6703.0324.432 0.000 1.68, 4.451.13 Respiratory rate T216.0002.97111.2002.6576.595 0.000 3.34, 6.261.68 Respiratory rate T317.3303.03210.5303.0608.646 0.000 5.23, 8.372.20F, p, η²F(3.87) = 54.654, p < 0.001, η²=0.653****F(3.87) = 0.458, p = 0.712, η²=0.016Bonferroni1 < 2,3,4; 2 < 3,4; 3 < 4 Mean
Sd
Mean
Sd
t
p
95% GA
Hedges g SBP Pre119.3007.949129.03013.430−3.416 0.001 –15.44, − 4.03−0.87 SBP T1119.70019.448130.13015.240−2.313 0.024 –19.46, − 1.40−0.59 SBP T2121.87018.319128.97018.492−1.4940.141–16.61, 2.41−0.38 SBP T3122.07018.984126.87020.774−0.9340.354–15.08, 5.48−0.24F, p, η²F(3.87) = 0.291, p = 0.832, η²=0.010****F(3.87) = 0.386, p = 0.763, η²=0.013 Mean
Sd
Mean
Sd
t
p
95% GA
Hedges g DBP Pre65.40012.89975.20012.734−2.961 0.004 –16.42, − 3.17−0.76 DBP T165.47012.97774.00012.871−2.557 0.013 –15.21, − 1.85−0.65 DBP T258.53019.11171.87010.769−3.329 0.002 –21.35, − 5.31−0.86 DBP T364.23016.05570.27010.693−1.7130.092–13.08, 1.01−0.44F, p, η²F(3,87) = 1.328, p = 0.271, η²=0.044****F(3,87) = 1.905, p = 0.135, η²=0.062 Mean
Sd
Mean
Sd
t
p
95% GA
Hedges g SpO₂ Pre96.1003.80994.5702.8611.7630.083–0.21, 3.270.45 SpO₂ T195.7302.33393.9302.8152.696 0.009 0.46, 3.140.69 SpO₂ T296.9001.56194.4001.9935.408 0.000 1.57, 3.431.38 SpO₂ T397.9301.38894.3002.0707.984 0.000 2.72, 4.542.03F, p, η²F(3.87) = 5.37, p = 0.002, η²=0.156****F(3.87) = 0.50, p = 0.682, η²=0.017Bonferroni1 < 4; 2 < 3,4; 3 < 4 Mean
Sd
Mean
Sd
t
p
95% GA
Hedges g Hct Pre33.1575.65736.1709.276−1.5190.135–6.98, 0.96−0.39 Hct T131.6435.85135.9239.342−2.127 0.039 –8.32, − 0.24−0.55 Hct T230.8305.96035.6807.307−2.817 0.007 –8.29, − 1.40−0.72 Hct T331.7404.50635.1437.260−2.182 0.034 –6.54, − 0.28−0.56F, p, η²F(3,87) = 1.89, p = 0.137, η² = 0.061****F(3,87) = 0.41, p = 0.748, η² = 0.014 Mean
Sd
Mean
Sd
t
p
95% GA
Hedges g Pain Pre5.9002.2347.1001.398−2.494 0.016 –2.16, − 0.24−0.64 Pain T14.5700.9716.8301.177−8.136 0.000 –2.82, − 1.71−2.07 Pain T23.9000.8035.4300.935−6.813 0.000 –1.98, − 1.08−1.73 Pain T33.0001.5543.2301.104−0.6700.505–0.93, 0.46−0.17F, p, η²F(3.87) = 23.52, p < 0.001, η² = 0.448****F(3.87) = 84.39, p < 0.001, η² = 0.744Bonferroni1 > 2,3,4; 2 > 3,4; 3 > 4****1 > 3,4; 2 > 3,4; 3 > 4 Mean
Sd
Mean
Sd
t
p
95% GA
Hedges g FEV1/FVC-Pre-intervention76.2302.62275.3301.4931.6340.109–0.20, 2.000.42 FEV11/FVC-Post-intervention83.1702.21477.7301.48411.165 0.000 4.46, 6.412.85t, p, 95% GA, dt(29) = − 51.34, p < 0.00195% GA: − 7.21,** − 6.65**, Cohen’s d = − 9.37 t(29) = − 26.38, p < 0.001,** 95% GA:–2.586**,** − 2.214**, Cohen’s d = − 4.8
Mean
Sd
Mean
Sd
t
p
95% GA
Hedges g Time to First Mobilization13.7331.33715.8331.877−4.991 0.000 –2.94, − 1.26−1.27Pre: Pre-intervention T2: Post-intervention 2 Hct: Hematocrit DBP: Diastolic blood pressureT1: Post-intervention 1 T3: Post-intervention 3 SBP: Systolic blood pressure
Pre-intervention pain scores were measured as 5.90 ± 2.23 in the experimental group and 7.10 ± 1.39 in the control group, and the difference was statistically significant (t = − 2.494, p = 0.016). The 95% confidence interval for the group difference (–2.16 to − 0.24) indicates that the difference is reliable. In the literature, the minimal clinically important difference (MCID) for pain assessments is reported to be 1.5–2.0 points [27, 28]. In this context, the approximately 1.2-point difference at T0 is close to the MCID threshold and suggests a potentially perceptible clinical difference for patients. At T1 and T2, pain scores were significantly lower in the experimental group (T1: t = − 8.136, p < 0.001; T2: t = − 6.813, p < 0.001), and both differences exceeded the MCID threshold, indicating clinically meaningful improvement. Effect sizes were − 2.07 for T1 and − 1.74 for T2, both of which fall within the “very large” effect category. At T3, no significant difference was detected between groups (p = 0.505), and the difference did not exceed the MCID threshold, indicating a lack of clinical significance (Table 2; Fig. 4).
Pulse values at baseline were 90.67 ± 15.88 in the experimental group and 85.97 ± 13.47 in the control group, and the difference was not statistically significant (t = 1.236, p = 0.221). In the post-intervention measurements at T1, T2, and T3, no statistically significant differences were found between the groups (p > 0.05). All 95% confidence intervals for the group differences included zero, indicating that the differences are consistent with random variation. In this study, pulse changes did not meet the threshold for clinical significance (MCID), suggesting that the intervention did not have a clinically meaningful effect on pulse rate. The low effect sizes (Hedges’ g < 0.32) and small η² values (η² < 0.12) also indicate that the clinical impact is limited (Table 2; Fig. 4).
Clear and progressively increasing statistical differences between groups were observed in respiratory rate. At T1, T2, and T3, respiratory rates were significantly higher in the experimental group compared with the control group (T1: t = 4.432, p < 0.001; T2: t = 6.595, p < 0.001; T3: t = 8.646, p < 0.001). The 95% confidence intervals for the group differences (T1: 1.68–4.45; T2: 3.34–6.26; T3: 5.23–8.37) did not include zero, indicating that the differences are reliable and that the intervention effectively influenced respiratory patterns. In the literature, the MCID threshold for respiratory rate is generally reported as 1–2 breaths per minute [29, 30]. In this study, the differences at T1, T2, and T3 substantially exceeded this MCID range; therefore, the intervention can be considered to have both statistically and clinically significant effects on respiratory rate. Effect sizes were notably high (g = 1.13–2.20), falling within the “very large” clinical effect category (Table 2; Fig. 4).
Systolic blood pressure was lower in the experimental group at baseline (119.30 ± 7.95 vs. 129.03 ± 13.43), and the difference was statistically significant (t = − 3.416, p = 0.001). At T1, the difference also remained significant (p = 0.024), but no significant differences were observed at T2 or T3 (p > 0.05). The 95% confidence interval for systolic blood pressure (–19.46 to − 1.40) largely excluded zero, supporting the reliability of the finding. In the literature, the MCID for systolic blood pressure is generally reported as 5–10 mmHg [31]. Accordingly, the difference observed at T1 is very close to the MCID threshold and may represent a perceptible clinical change for patients. However, the differences at T2 and T3 were below the MCID threshold, indicating a lack of clinical significance (Table 2; Fig. 4).
Diastolic blood pressure was lower in the experimental group at all-time points, with statistically significant differences at T0, T1, and T2 (p < 0.05). At T3, the difference approached but did not reach statistical significance (p = 0.092). The 95% confidence intervals for the group differences (–21.35 to − 1.85) exceeded the MCID range of 5–10 mmHg, demonstrating that the improvement—particularly at T2—is clinically meaningful. Effect sizes ranged from moderate to large (g = 0.65–0.86), indicating that the intervention exerted a moderate-to-large clinical impact on reducing diastolic blood pressure (Table 2; Fig. 4).
Statistically significant differences in SpO₂ were observed in favor of the experimental group. At T1, T2, and T3, p-values were significant (p < 0.01), with particularly pronounced differences at T2 and T3 (t = 5.408 and t = 7.984; p < 0.001). The MCID for SpO₂ is generally accepted as 1–2% points. In this study, the difference at T2 was 2.5 points and at T3 was 3.6 points, both exceeding the MCID threshold and indicating clinically important improvements. Effect sizes were very high (g = 1.38–2.03), demonstrating that the intervention produced a strong and clinically meaningful effect on oxygen saturation (Table 2; Fig. 2).Fig. 2SpO₂ Measurements Comparison by Group
Hematocrit values were significantly lower in the experimental group at T1, T2, and T3 (p < 0.05). The MCID for hematocrit is generally accepted in the literature as 1–2% points. In this study, the group differences at T1, T2, and T3 were approximately 4–5% points, clearly exceeding the MCID threshold. Effect sizes were in the moderate-to-large range (g = 0.55–0.72). These findings indicate that the change in hematocrit is not only statistically significant but also clinically perceptible (Table 2; Fig. 4).
Following the intervention, a very strong increase in the FEV₁/FVC ratio was observed in the experimental group (83.17% vs. 77.73%; t = 11.17, p < 0.001). The 95% CI for the group difference (4.46–6.41) demonstrates high reliability and exceeds the MCID threshold of 3% points. The effect size (Cohen’s d = − 4.80) is considered extremely large in clinical applications. This finding indicates that the intervention produced a clinically meaningful improvement in respiratory function (Table 2; Fig. 3).
Time to first mobilization was significantly shorter in the experimental group (13.73 ± 1.34 vs. 15.83 ± 1.88; t = − 4.991, p < 0.001). The 95% CI for the group difference (–2.94 to − 1.26) did not include zero, indicating high reliability. In the literature, MCID thresholds for mobilization time typically include 1–2 h or a 10–20% improvement. In this study, the approximately 14% difference between groups met the MCID criterion. Given the large effect size (g = − 1.27), the intervention can be considered to provide clinically meaningful benefits for early mobilization (Table 2; Fig. 4).
Fig. 3FEV₁/FVC Ratio Comparison by Group
Fig. 4Clinical Parameter Changes Over Time
Since reductions in respiratory muscle strength and pulmonary capacity may occur following cardiac surgery, early and multicomponent pulmonary rehabilitation programs have become increasingly important to prevent postoperative morbidity and mortality [32, 33, 34]. This study aimed to determine whether postoperative respiratory function in patients undergoing mitral valve replacement improves more effectively through the use of the Triflow device alone or through the combined application of Triflow, deep-breathing exercises, and coughing exercises. The findings demonstrated that the experimental group, which received the combined Triflow–deep breathing–coughing intervention, showed significant improvements in respiratory parameters compared with the control group (H1: Supported).
In our study, the higher SpO₂ and FEV₁/FVC values observed in the experimental group may be attributed to the physiological effects of deep-breathing and coughing exercises, which expand the alveoli, enhance gas exchange, help prevent hypoxemia, support the cough reflex, reduce mucus viscosity, thereby facilitating more effective secretion clearance and maintaining airway patency. The literature similarly indicates that deep-breathing exercises improve oxygenation, reduce dyspnea, and enhance chest wall mobility. This finding is also consistent with previous research reporting that combined respiratory techniques are more effective than single-method interventions [19, 20, 23, 35]. In our study, a statistically significant difference in respiratory rate was observed between groups; however, this finding was not interpreted as an isolated indicator of clinical improvement. Respiratory rate may reflect compensatory mechanisms or physiological stress; therefore, it was evaluated descriptively and in conjunction with objective parameters such as SpO₂, FEV₁/FVC, pain scores, and thoracic expansion rather than as a standalone marker of benefit. Although an increase in respiratory rate alone does not necessarily indicate improved ventilation, when interpreted together with reduced pain levels, enhanced thoracic expansion, and more efficient secretion clearance, it suggests that alveolar ventilation was preserved. In addition, hematocrit values were found to be significantly lower in the experimental group. These changes were reported for descriptive purposes and were not interpreted as clinically meaningful effects of the intervention, given the potential influence of perioperative factors such as fluid balance, bleeding, and transfusion. This may be related to group differences in factors such as duration of surgery, ejection fraction levels, and intubation time, which may have contributed to lower hematocrit values compared with the control group. The study also demonstrated that the intervention had a moderate-to-large clinical effect in reducing diastolic blood pressure. Despite the observed increase in respiratory rate, reduction in hematocrit, and decrease in diastolic blood pressure, the absence of any clinically adverse outcomes suggests that the intervention maintained overall hemodynamic stability (H1A: Supported).
The shorter time to achieve early mobilization in the experimental group indicates that the combined exercises facilitated physical activity and supported overall postoperative recovery (H1B: Supported). A systematic review conducted by Yun et al. reported that respiratory exercises, particularly when applied in combination, improved six-minute walk test performance and inspiratory muscle strength, thereby supporting the findings of the present study [36]. Similarly, the study by Pu et al. demonstrated that deep-breathing exercises significantly reduced postoperative pain and mean arterial pressure, while increasing SpO₂ levels in the intervention group—results that are consistent with the outcomes of the present study [37]. Moreover, in their systematic review, Zhang et al. concluded that early and structured cough training aimed at increasing inhalation volume and expiratory flow may improve lung function and positively influence clinical outcomes, particularly in patients at high risk of ineffective coughing [38].
Furthermore, the lower VAS pain scores observed in the combined-exercise group indicate that these exercises enhanced thoracic expansion, reduced pain perception, and improved patients’ adherence to deep-breathing and coughing practices (H1C: Supported). This suggests that the supportive application of coughing and deep-breathing techniques may reduce postoperative pain, facilitate more comfortable breathing, and promote earlier mobilization. The reduction in pain may also be associated with increased chest expansion. Our findings are consistent with those of Jarah et al., who reported lower pain levels in an intervention group performing deep-breathing exercises following bypass surgery [39]. Similarly, Vahedian et al. demonstrated in their randomized controlled trial that deep-breathing exercises significantly reduced surgical site pain and mean arterial pressure while increasing SpO₂ levels in the experimental group [40].
In the present study, no pulmonary complications were observed in either group. Despite the absence of complications in either group, the improvements observed in respiratory parameters in the experimental group may indicate a protective effect that could reduce the potential risk of postoperative pulmonary complications. The findings of our study confirm the postoperative benefits of using the Triflow device alone; however, they also demonstrate that combining Triflow with deep-breathing and coughing exercises yields superior improvements in respiratory function and pain reduction compared with Triflow use alone.
This study demonstrated that the combined use of Triflow, deep breathing, and coughing exercises positively affected postoperative respiratory function in patients who underwent mitral valve replacement. Significant improvements in SpO₂ and FEV₁/FVC (values were observed in the experimental group, indicating that these exercises support respiratory recovery in the postoperative period. Pulmonary complications (atelectasis, pneumonia, hypoxemia, pulmonary congestion, and prolonged oxygen requirement) were monitored through nursing assessments during the ward stay, while definitive diagnoses and medical decisions were made by the attending physician, and no complications were detected in either group. Despite the increase in respiratory rate, the decrease in hematocrit, and the reduction in diastolic blood pressure, the absence of clinically adverse outcomes suggests that the intervention maintained overall hemodynamic stability. Early mobilization and reduced pain scores further indicate that these exercises contribute to both respiratory function and general well-being. Based on these findings, it is recommended that patients undergoing mitral valve surgery regularly perform Triflow, deep breathing, and coughing exercises together. Further research comparing different combinations of respiratory physiotherapy interventions is warranted, and studies with larger, more diverse samples and longer follow-up periods are needed to strengthen these results.
This study was conducted at a single center, which limits its generalizability to other healthcare settings. The relatively small sample size may have reduced statistical power, and some outcomes showed small effect sizes and wide confidence intervals; further studies with larger sample groups are needed. Although simple randomization was used, allocation concealment could not be fully ensured, which may introduce a risk of sequence predictability. Additionally, because Triflow training was provided to all patients as part of routine practice, the contrast between groups may have been reduced. Excluding patients who developed pulmonary complications in the intensive care unit may have limited the representation of higher-risk patients and potentially influenced effect size estimates. This should be considered when interpreting the results. Therefore, the findings should be interpreted with caution, as the observed effects cannot be attributed solely to the intervention. Only early postoperative outcomes were evaluated; long-term effects on respiratory function, quality of life, and pulmonary complications were not assessed. The effectiveness of respiratory exercises largely depends on patient motivation and compliance, and individual differences may have influenced the results. Participants’ age, BMI, and comorbidities (e.g., diabetes, hypertension) may have affected their response to the exercises. Furthermore, differences in cardiopulmonary bypass time, intubation time, and intensive care unit stay may have affected the results and may not have been fully balanced between groups. Although several patient- and surgery-related factors may influence outcomes, these variables were recorded as baseline characteristics rather than used as exclusion criteria to preserve clinical representativeness. The lack of blinding of the evaluator is another potential methodological limitation that may increase the risk of bias, especially considering the subjective nature of pain as a secondary outcome.
This study also has several strengths. The randomized controlled design enhances the validity of the findings. The limited number of studies comparing respiratory physiotherapy methods in patients undergoing mitral valve replacement makes this research a unique contribution to the literature. The direct comparison of Triflow alone with Triflow combined with deep breathing and coughing exercises provides clinically valuable information for evaluating the effectiveness of the intervention. The assessment of multiple physiological parameters (pulse, respiratory rate, blood pressure, SpO₂, hematocrit) along with pain (VAS), ensured a comprehensive evaluation of the results. Finally, the findings are easily transferable to clinical practice and constitute an important strength of the study by offering recommendations that could improve patient care without additional costs.
Supplementary Material 1.