Effectiveness of Preoperative Chest Physiotherapy in Patients Undergoing Elective Cardiac Surgery, a Systematic Review and Meta-Analysis

Background and Objectives: Patients undergoing cardiac surgery are particularly vulnerable for developing postoperative pulmonary complications (PPCs). This systematic review and meta-analysis aimed to evaluate the role of preoperative chest physiotherapy in such patients. Materials and Methods: All original articles that assessed patients undergoing elective cardiac surgery, with preoperative chest physiotherapy, and compared them to patients undergoing elective cardiac surgery, without preoperative chest physiotherapy, were included. Animal studies, studies conducted prior to the year 2000, commentaries, or general discussion papers whose authors did not present original data were excluded. Studies assessing physiotherapy regimens other than chest physiotherapy were also excluded. The search was performed using the following electronic resources: the Cochrane Central Register of Controlled Trials, the PubMed central database, and Embase. The included studies were assessed for potential bias using the Cochrane Collaboration’s tool for assessing the risk of bias. Each article was read carefully, and any relevant data were extracted. The extracted data were registered, tabulated, and analyzed using Review Manager software. Results: A total of 10 articles investigating 1458 patients were included in the study. The studies were published from 2006 to 2019. The populations were patients scheduled for elective CABG/cardiac surgery, and they were classified into two groups: the interventional (I) group, involving 651 patients, and the control (C) group, involving 807 patients. The meta-analysis demonstrated no significant differences between the interventional and control groups in surgery time and ICU duration, but a significant difference was found in the time of mechanical ventilation and the length of hospital stay, favoring the interventional group. A significant difference was shown in the forced expiratory volume in 1s (FEV1% predicted), forced vital capacity (FVC% predicted), and maximum inspiratory pressure (Pi-max), favoring the interventional group. Conclusions: This study is limited by the fact that one of the included ten studies was not an RCT. Moreover, due to lack of the assessment of certain variables in some studies, the highest number of studies included in a meta-analysis was the hospital stay length (eight studies), and the other variables were analyzed in a fewer number of studies. The data obtained can be considered as initial results until more inclusive RCTs are conducted involving a larger meta-analysis. However, in the present study, the intervention was proved to be protective against the occurrence of PPCs. The current work concluded that preoperative chest physiotherapy can yield better outcomes in patients undergoing elective cardiac surgery.


Study Design
This is a systematic review and meta-analysis that was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [15]. The review was registered in the Research Registry (reviewregistry1278).

Literature Search Strategy
The included studies were those evaluating the preoperative chest physiotherapy value in adult patients who undergo elective cardiac surgery. The search was performed using the electronic resources; the Cochrane Central Register of Controlled Trials, the PubMed central database, and Embase.

Selection Strategy and Criteria
The search was conducted with the restriction limiting results to original articles published from January 2000 to December 2021. The search was performed using the following keywords: "preoperative care" OR "preoperative" OR "preoperational" OR "pre-habilitation" OR "pre-habilitation" OR "before operation" OR "before surgery" AND "coronary artery disease" OR "CAD" AND "chest" OR "respiratory" OR "lung" OR "pulmonary" AND "physiotherapy" OR "physical therapy" OR "muscular training" OR "muscle training" OR "muscle exercise" OR "muscular exercise" OR "muscle strength" AND "cardiac surgery" OR "open cardiac surgery" OR "open heart surgery" OR "heart surgery" OR "coronary artery bypass graft" OR "coronary artery bypass grafting" OR "CABG" AND "postoperative" OR "following operation" OR "after" OR "after cardiac operation" AND "pulmonary complications" OR "lung complications" OR "lung impairment" OR "respiratory failure" OR "respiratory impairment" OR "impaired respiratory functions" OR "impaired lung functions." The search was performed by two independent reviewers (the first and second authors). Then articles were matched and screened to ensure eligibility. The search strategy in each data base was described in Supplementary File S1.

Inclusion Criteria
Original articles available in English were included. According to PICO, we included the studies meeting the following criteria: study design-all original articles that included randomized controlled trials or observational studies from 2000 until conducting the analysis (mid-2021), participants-patients undergoing elective cardiac surgery, intervention-preoperative chest physiotherapy, control-patients undergoing elective cardiac surgery without preoperative chest physiotherapy, outcome measures-the effect of intervention on PPCs and any other effect.

Exclusion Criteria
Animal studies, studies completed prior to the year 2000, commentaries, or general discussion papers whose authors did not present original data were excluded. Studies assessing physiotherapy regimens other than chest physiotherapy and those applying postoperative physiotherapy programs other than the routine therapies were also excluded.

Data Extraction, Data Collection, and Analysis
Each article was read carefully and any relevant data were extracted (including the study setting, design, research questions, sample size, patients' demographic data, medical history, baseline preoperative data, type and details regarding the intervention, description of the intervention; type, time, duration, rate, and the used device, lung function tests, muscle strength, operative events, length of hospital stay, the occurrence of postoperative pulmonary complications, and the study conclusions). The extracted data were registered, tabulated, and analyzed.

Bias
Methodological quality check lists were used as tools for bias risk assessment. The included studies were assessed for potential bias using the Cochrane Collaboration's tool for assessing the risk of bias.

Summary Measures
The primary outcomes were the incidence of postoperative complications and the changes in the lung function parameters, and the secondary outcomes were the surgery duration, the length of stay in the ICU and hospital, and the time of mechanical ventilation.
Data about the ongoing RCTs related to the study topic was evaluated and described in the Discussion section.

Statistical Analysis
The retrieved data were presented as mean and standard deviation (SD) for numerical data, and frequency and percentage for categorical data.
The meta-analysis and bias assessment were accomplished using the Review Manager software (RevMan version 5.4, the Cochrane Collaboration, London, UK). Dichotomous data were expressed as a risk ratio, with 95% confidence intervals (CIs) to compare intervention and control groups at the study level. For continuous outcomes, the mean differences in effects between the intervention and control groups were computed at the study level and pooled into weighted mean differences (WMDs).

Results
The search of the electronic resources first yielded a total of 24,106 records. After duplication adjustment, the search provided 1123 results. Based on the title, 898 publications were excluded. Then, after checking abstracts, another 199 publications were found not to meet the eligibility criteria, so they were further excluded. After checking the full texts, an additional 16 articles were excluded. Thus, 10 studies were finally eligible for this systematic review ( Figure 1).
The included articles were published from 2006 to 2019. The populations were patients scheduled for elective CABG/cardiac surgery in the Netherlands, Turkey, Taiwan, Iran, Brazil, Pakistan, and China (Table 1).
All patients were adults, with a mean age of 60.36 ± 12.78. Male predominance was noted, as 1019 (69.9%) patients were males, while only 439 patients were females. No significant difference was found among all the included studies regarding either age or male/female ratio (p > 0.05) ( Table 2).
The percentages of smokers ranged from 25% to 70% in the review studies, BMI ranged from 25.66 to more than 30, and the comorbidities were mainly diabetes mellitus, hypertension, and hyperlipidemia. Both groups in all studies were matched according to the prevalence of risk factors and the comorbidities ( Table 3). The included articles were published from 2006 to 2019. The populations were patients scheduled for elective CABG/cardiac surgery in the Netherlands, Turkey, Taiwan, Iran, Brazil, Pakistan, and China (Table 1).     Concerning the type of preoperative intervention in the included studies, some used respiratory training protocols, with an incentive spirometer [16,20,23], one study combined incentive spirometer with a threshold loading device [24], and others used threshold loading devices for chest physiotherapy [16,17,19,20,22,25] (Table 4).
The time frame for preoperative intervention application differed considerably among the included studies, ranging from 5 days to 10weeks. The frequency of performing the interventional program ranged from twice a day [25], to three times every two weeks [18]. The duration of training sessions ranged from 20 [16,21,22,25] to 60 min [18] (Table 4).
In all studies, the basic preoperative pulmonary functions, respiratory muscle test parameters, and ABGs were comparable in the two groups.
The pooled analysis revealed no significant differences between the interventional and control groups in the surgery time ( Figure 2) and the ICU duration ( Figure 3) (p = 0.84 and 0.92, respectively), with no heterogeneity in the results (p = 0.06 and 0.62, respectively). Table 5. Description of the preoperative protocols.

Hulzebos et al., 2006a
Care as usual the day before surgery (i.e., instruction on deep breathing maneuvers, coughing, and early mobilization).
Incentive spirometry, chest physical therapy, and mobilization scheme after operation.

Hulzebos et al., 2006b
Care as usual the day before surgery (i.e., instruction on deep breathing maneuvers, coughing, and early mobilization).
Incentive spirometry, chest physical therapy, and mobilization scheme after operation.

Savci et al., 2011
Mobilization, active exercises of upper and lower limbs, chest physiotherapy.
Chest physical therapy and mobilization scheme after operation.

Tung et al., 2012
Care as usual the day before surgery (i.e., instruction on deep breathing maneuvers, coughing, and early mobilization).
Incentive spirometry, chest physical therapy, and mobilization scheme after operation. Care as usual the day before surgery (i.e., instruction on deep breathing maneuvers, coughing, and early mobilization).
Incentive spirometry, chest physical therapy, and mobilization scheme after operation.

Chen et al., 2019
Both groups received both usual care (i.e., education, coughing and early mobilization) and abdominal breathing training before the surgery.
Chest physical therapy and mobilization scheme after operation.
IMT: inspiratory muscle training. Table 6. The outcomes of the included studies.

Hulzebos et al., 2006a
The primary outcome: the incidence of PPCs. The secondary outcome was duration of postoperative hospitalization. In all studies, the basic preoperative pulmonary functions, respiratory muscle test parameters, and ABGs were comparable in the two groups.
The pooled analysis revealed no significant differences between the interventional and control groups in the surgery time ( Figure 2) and the ICU duration ( Figure 3) (p = 0.84 and 0.92, respectively), with no heterogeneity in the results (p = 0.06 and 0.62, respectively).     The meta-analyses revealed significant differences between the interventional and There were significant differences between the intervention and control groups in the duration of mechanical ventilation ( Figure 4) and the length of hospital stay ( Figure 5) (p < 0.001), favoring the interventional group. The pooled mean differences between groups were 0.76 h and 1.02 days, respectively. The absence of heterogeneity in the meta-analyses of the length of hospital stay (p = 0.1) grants credibility to the results.
The PPCs meta-analysis demonstrated that the intervention had a protective effect on the occurrence of PPCs. The pooled risk ratio was shown to be 47%, with a 95%CI of 36-62%. The overlap between a part of the CI that was shown in the pooled estimates reflected the absence of statistical heterogeneity (I 2 = 0%, p = 0.73) (Figure 9).  There were significant differences between the intervention and control groups in the duration of mechanical ventilation ( Figure 4) and the length of hospital stay ( Figure 5) (p < 0.001), favoring the interventional group. The pooled mean differences between groups were 0.76 h and 1.02 days, respectively. The absence of heterogeneity in the meta-analyses of the length of hospital stay (p = 0.1) grants credibility to the results.       The PPCs meta-analysis demonstrated that the intervention had a protective effect on the occurrence of PPCs. The pooled risk ratio was shown to be 47%, with a 95%CI of 36-62%. The overlap between a part of the CI that was shown in the pooled estimates reflected the absence of statistical heterogeneity (I2 = 0%, p = 0.73) (Figure 9).    The PPCs meta-analysis demonstrated that the intervention had a protective effect on the occurrence of PPCs. The pooled risk ratio was shown to be 47%, with a 95%CI of 36-62%. The overlap between a part of the CI that was shown in the pooled estimates reflected the absence of statistical heterogeneity (I2 = 0%, p = 0.73) (Figure 9).   The PPCs meta-analysis demonstrated that the intervention had a protective effect on the occurrence of PPCs. The pooled risk ratio was shown to be 47%, with a 95%CI of 36-62%. The overlap between a part of the CI that was shown in the pooled estimates reflected the absence of statistical heterogeneity (I2 = 0%, p = 0.73) (Figure 9).  The PPCs meta-analysis demonstrated that the intervention had a protective effect on the occurrence of PPCs. The pooled risk ratio was shown to be 47%, with a 95%CI of 36-62%. The overlap between a part of the CI that was shown in the pooled estimates reflected the absence of statistical heterogeneity (I2 = 0%, p = 0.73) (Figure 9). A funnel plot was constructed (Figure 10), revealing the symmetry in results, as all the involved studies lay within the confidence interval, with a rather symmetric pattern, ensuring the absence of heterogeneity in the results. A funnel plot was constructed (Figure 10), revealing the symmetry in results, as all the involved studies lay within the confidence interval, with a rather symmetric pattern, ensuring the absence of heterogeneity in the results.  [17]. In contrast, Tung et al. (2012) noticed an intervention-related significant improvement in the general QoL(p < 0.001) [18]. Anxiety and depression were evaluated in one study [17], which showed that their expressions were lower in the intervention group. However, this difference was significant only in the anxiety component (p < 0.05).
A summary of the study outcomes is demonstrated in Table 7.   [17]. In contrast, Tung et al. (2012) noticed an intervention-related significant improvement in the general QoL (p < 0.001) [18]. Anxiety and depression were evaluated in one study [17], which showed that their expressions were lower in the intervention group. However, this difference was significant only in the anxiety component (p < 0.05).
A summary of the study outcomes is demonstrated in Table 7. The primary outcome: a statistically significant difference was found between the two groups in the incidence of PPCs (p = 0.02). The secondary outcome: a statistically significant difference was found between the two groups in the LOS (p = 0.02).

Hulzebos et al., 2006b
Primary outcome: the feasibility of the intervention was good. No adverse events were reported. A statistically significant difference was found between the two groups in the satisfaction scores and the muscle strength. Secondary outcome: a statistically significant difference was found between the two groups in the When examining the conclusions reached by the included studies, there was unanimous agreement on the importance and significance of preoperative chest physiotherapy in patients undergoing elective cardiac surgery.
The critical assessment graph and a summary of the risks of bias within each study are shown in Figures 11 and 12. When examining the conclusions reached by the included studies, there was unanimous agreement on the importance and significance of preoperative chest physiotherapy in patients undergoing elective cardiac surgery.
The critical assessment graph and a summary of the risks of bias within each study are shown in Figures 11 and 12.   Figure 11. Risk of bias graph for the included studies. Figure 11. Risk of bias graph for the included studies. Figure 12. Risk of bias summary for the included studies. A square with a green circle means low risk, a square with a red circle means high risk, and an empty square indicates unclear risk.

Discussion
The studies included in this review explored variable methods for preoperative chest physiotherapy. Several strategies are created to intervene in an attempt to prevent PPC development. Interventions may be preoperative, to adjust the physiology of respiration, or intraoperative and postoperative, to minimize the adverse events of surgery and anesthesia [26]. Despite this, there are no established guidelines for preoperative protocols of management. Even if they are present, they are outdated or infrequent [27]. This lack of consensus leads to considerable variation in clinical practice [28].
The current study pooled analysis demonstrated that there was no effect of the intervention on the surgery time or the ICU stay duration, while it favorably affected the mechanical ventilation and the length of hospital stay.
The association of preoperative chest physiotherapy with shorter hospital stay was also documented in previous studies [29,30]. Nardi et al. (2019) observed that the length Figure 12. Risk of bias summary for the included studies. A square with a green circle means low risk, a square with a red circle means high risk, and an empty square indicates unclear risk.

Discussion
The studies included in this review explored variable methods for preoperative chest physiotherapy. Several strategies are created to intervene in an attempt to prevent PPC development. Interventions may be preoperative, to adjust the physiology of respiration, or intraoperative and postoperative, to minimize the adverse events of surgery and anesthesia [26]. Despite this, there are no established guidelines for preoperative protocols of management. Even if they are present, they are outdated or infrequent [27]. This lack of consensus leads to considerable variation in clinical practice [28].
The current study pooled analysis demonstrated that there was no effect of the intervention on the surgery time or the ICU stay duration, while it favorably affected the mechanical ventilation and the length of hospital stay.
The association of preoperative chest physiotherapy with shorter hospital stay was also documented in previous studies [29,30]. Nardi et al. (2019) observed that the length of the postoperative hospital stay in the group that had preoperative training was reduced compared to the control group, but without a statistically significant difference [13].
The short hospital stay affords the patients the chance to continue the recovery in their familiar home environment, saving the hospital resources for new patients to receive health care services [31,32].
In contrast with the findings of our study, the previous meta-analyses did not reveal a significant difference between the interventional and control groups in the mechanical ventilation time [33][34][35]. However, we can state that the significant difference found in our study could be a quasi-significance, due to the heterogeneity found in the results.
The meta-analysis of this study showed that there was a significant difference between the interventional and control groups in the pulmonary functions, including FEV1% predicted , FVC% predicted , and Pi-max, favoring the interventional group. However, only the FEV1% predicted showed homogenous results.
The Pi-max was the most commonly tested lung function parameter in the included studies. Pi-max reflects the inspiratory muscles' functional capacity and has been adopted as a reliable indicator for the weaning from mechanical ventilation in many hospitals [36]. In our meta-analysis, the evidence of the intervention's improving effect was weak. Within the same context, in the meta-analysis conducted by Marmelo et al. (2018), the authors found significant improvement of the Pi-max related to the intervention [33]. On the other hand, Katsura et al. (2015) reported no statistically significant effect in a three-articles meta-analysis, in spite of the fact that a tendency toward a favoring effect was found in all three articles [35].
The current meta-analysis revealed that intervention proved to be protective against the existence of PPCs.
Consistent with our findings, the recent meta-analysis conducted by Odor et al. (2020) disclosed evidence of the prophylactic effect of preoperative physiotherapy against the occurrence of PPCs [28]. Our findings were also congruent with the most recent metaanalysis conducted by Rodrigues et al. (2021), which demonstrated that preoperative chest physiotherapy (breathing interventions) helped to improve postoperative respiratory performance in patients undergoing cardiac surgery. Moreover, the authors concluded that such interventions reduced PPCs and the length of hospital stay [37]. Other previous studies affirmed the effect of preoperative intervention on PPCs. This was investigated in patients who underwent oncologic thoracic surgeries [38], cardiac [33,38] intra-abdominal [39], and cardiac and abdominal surgery [30,40]. In these meta-analyses, a total of 31 studies reported decreased PPCs in the interventional group, while only 8 did not find this relationship. The study of Kamarajah et al. (2019) highlighted that pre-habilitation improved rates of morbidity, including for PPCs, and overall complications after both major abdominal and cardiothoracic surgery [41].
The recently published work of our group showed supporting findings [42]. Our RCT demonstrated overall significant postoperative improvements in lung function and oxygen saturation in the intervention group compared with the control group. An earlier RCT conducted by Sweity et al. (2021) to assess the effect of preoperative incentive spirometry was compatible with our findings, as the study showed a significant difference between the interventional and control groups in the incidence of postoperative atelectasis, mechanical ventilation duration, and hospital LOS. The median of the amount of arterial blood oxygen and oxygen saturation was significantly improved in the intervention group [43].
The QoL outcome exhibited heterogeneity in the measured scale and the obtained results. The variability in the used methods makes it difficult to obtain a consensus about the results [44,45]. The QoL variables' interpretation had to be considered in view of the individual results of the articles assessing this outcome, since we could not conduct a metaanalysis in view of the heterogeneity in the quantification scales. In this regard, Valkenet et al. (2017) reported that the intervention group showed less reduction in QoL values than the control group, but without a statistical significance [24]. Savci et al. (2011) also failed to reveal a statistically significant difference between either groups in the assessed physical component of QoL [17], while Tung et al. (2012) proved a significant improvement in the general QoL [18].
Anxiety and depression was evaluated in the study of Savci et al. (2011) [17], which showed that anxiety and depression tendencies were lower in the intervention group than in the control group. However, this difference was significant only in the anxiety component. In congruence, an earlier study demonstrated that the patients who were preoperatively educated, with guidance on the physiotherapeutic ventilation training, exhibited reduced anxiety levels compared with those did not receive this guidance [46].
It is worth noting that all the included patients were those scheduled for CABG, except for those in the studies of Tung et al. (2012) and Chen et al. (2019), which also included patients with valve surgeries. Both types of surgeries are open heart surgeries, with the indicated general anesthesia, median sternotomy incision, cardiopulmonary bypass, and mechanical ventilation. All of these are factors implicated in the predisposition to PPCs and hence, both types of surgeries were included in the analysis.
Working to prevent or reduce the incidence of pulmonary complications occurring in patients after heart surgery is a major goal among health workers. To achieve this goal, it is recommended to educate patients about how important is to learn the physical therapy techniques. These techniques could help improve the respiratory functions and promote the expansion of the lungs, identifying patients at high-risk for the development of pulmonary complications after surgery. In this vein, physical therapy is highly regarded among the basic treatments and should be offered to the patients in intensive care units. This study confirms the potential performance of a rehabilitation program before cardiac surgery, recommending its availability to all patients, if possible, in order to make the post-operative period less traumatic, and to facilitate a faster functional recovery.

Strengths and Limitations
This study is strengthened by including a meta-analysis in addition to the systematic review of the included studies. Moreover, the pooled analysis included a large number of patients, thus yielding a rather firm conclusion. This study is limited by the fact that one of the included ten studies was not an RCT. Moreover, due to the lack of assessment of certain variables in some studies, the highest number of studies included in a meta-analysis was in the hospital stay length (eight studies), and the other variables were analyzed in a fewer number of studies. The data obtained can be considered as initial results until more inclusive RCTs are conducted involving a larger meta-analysis.

Conclusions
The current work concluded that preoperative chest physiotherapy can yield better outcomes in patients undergoing elective cardiac surgery.
The meta-analysis demonstrated no significant differences between the interventional and control groups in the surgery time and the ICU duration, but a significant difference in the time of mechanical ventilation and the hospital stay length, favoring the interventional group. A significant difference was shown in the FEV1% predicted , FVC% predicted , and Pi-max, favoring of the interventional group. The most notable significance was shown in the analysis of hospital stay length and the FEV1% predicted . The intervention was proved to be protective against the occurrence of PPCs. Funding: Funding for this paper was provided by the Stipendium Hungaricum Scholarship, Doctoral School of Health Science, University of Pecs.