Impact of Frailty on Postoperative Dysphagia in Patients Undergoing Elective Cardiovascular Surgery

Background Postextubation dysphagia (PED) is a serious postoperative complication following cardiovascular surgery that can lead to a worse prognosis. On the other hand, frailty is a prognostic factor in patients who undergo cardiac surgery. Objectives This study investigated the effect of frailty status on PED and impact of PED on postoperative complications. Methods This single-center retrospective cohort study included 644 consecutive patients who underwent elective cardiovascular surgery between May 1, 2014, and December 31, 2020; they were assigned to the PED or non-PED group based on postoperative swallowing status, and postoperative complications were investigated. Frailty status and physical functions, including walking speed, grip strength, Short Physical Performance Battery, and 6-minute walking distance, were preoperatively assessed; the frailty-status cutoff for predicting PED was determined from the receiver-operating characteristic curve. Results In this study cohort (mean age 67.7 years), the overall PED prevalence was 14.8%; preoperative frailty had a significantly higher prevalence in the PED group (50.0%) than in the non-PED group (20.3%; P < 0.001). PED correlated with a higher incidence of postoperative pneumonia and prolonged intensive care unit or hospital stay (P < 0.05 for all). After adjustment for confounders, multiple regression analysis revealed that preoperative frailty was independently associated with PED (P < 0.001). Conclusions PED occurred commonly after cardiovascular surgery and increased the risk of postoperative complications. Preoperative frailty was independently associated with PED. The 6-minute walking distance was the most powerful predictor of PED. Evaluation of preoperative frailty status is important for risk stratification and prevention of postoperative morbidity in patients undergoing surgery.

D ysfunctional deglutition manifests clinically as dysphagia. Dysphagia on extubation following intubation, called postextubation dysphagia (PED), is a serious complication following cardiovascular surgery. 1 The etiopathogenetic mechanisms underlying PED have not been fully elucidated, although mucosal inflammation, oropharyngeal muscular atrophy, diminished proprioception and sensation, and laryngeal injury associated with prolonged intubation can be attributed to the increase incidence in PED. 2 A recent systematic review reported that the prevalence of PED after cardiac surgery approached 35% although significant heterogeneity exists. 3 PED increases mortality, reintubation, the risk of aspiration, and prolongs hospital and intensive care unit stay. 4 Furthermore, the decreased oral intake associated with PED can worsen the patient's postoperative nutritional status, leading to further functional decline and worse prognosis. 5 Thus, the identification of patients who are at risk for PED is critically important to prevent PED.
Frailty has received considerable attention in recent years because of its impact on postoperative mortality, morbidity, and functional decline. 6 Frailty arises from a multicomponent complex process where the physiological reserve and systemic regulation decline, thereby resulting in a reduced ability to adapt to stressors. 7 Sarcopenia, which is the loss of muscle mass, is the main cause of frailty. 8 Interestingly, sarcopenia might cause dysphagia, which is called "sarcopenic dysphagia," and there is a close relationship between loss of muscle mass and dysphagia. 9 Thus, frailty, sarcopenia, and/or dysphagia are recognized as geriatric syndromes that have a complex inter-relationship. 10  Based on the Food Intake Level Scale score, dysphagia was categorized as no oral intake (score: 1-3), oral intake and alternative nutrition (score: 4-6), and oral intake alone (score: 7-10). The Food Intake Level Scale assessment was undertaken 3 times as follows: 1) within 1 week before surgery; 2) when postoperative oral intake was permitted after the attending cardiologists had confirmed that the  We assessed the preoperative frailty status within the week before the intervention using the Japanese version of the Cardiovascular Health Study frailty index. 16 The frailty phenotype is based on the following 5 components: slowness (gait speed: <1.0 m/s), weakness (grip strength: <28 kg for men and <18 kg for women), weight loss (>2 kg in the past 6 months), exhaustion, and low physical   capacity, the 6-minute walking distance (6MWD) was assessed by a trained physiotherapist who is a registered instructor of cardiac rehabilitation. 20 STATISTICAL ANALYSIS. The sample size to detect 15% difference in prevalence of PED between the 2 groups was determined by using the formula with The statistical significance level was set at P < 0.05.  Figure S1. The mean initial postoperative Food Intake Level Scale score was 8.11 AE 1.55, but there was significant improvement at discharge (8.51 AE 1.05). However, the Food Intake Level Scale score was significant lower in the PED group compared with in the non-PED group even at discharge ( Table 2). The prevalence of frailty was 24.7% in the study cohort and was significantly higher in the PED group than in the non-PED group (50.0% vs 20.3%; P < 0.001). Furthermore, the intergroup difference in the prevalence of frailty between patients with and without dysphagia did not change after propensity score matching (Central Illustration, Supplemental Table S1). Each physical function parameter, such as gait speed, grip strength, SPPB, and 6MWD, was significantly decreased in the PED group compared with that in the non-PED group (P < 0.05 for all).  Table 2.  group than in the non-PED group (P < 0.001 for all).

RESULTS
More patients in the PED group needed longer intensive care unit and hospital stays than patients in the non-PED group did (P < 0.001 for all). The first postoperative evaluation of swallowing was undertaken 1.7 AE 1.1 days after surgery, and oral intake was initiated 1.9 AE 1.1 days after surgery; both durations were prolonged in the PED group (P < 0.05 for all). Table 3

DISCUSSION
The main findings of this study were that 14.8% of the participants developed PED after cardiac surgery, and preoperative frailty was found to be an independent predictor of PED. This is the first study to demonstrate that physical functions, such as gait speed, grip strength, SPPB, and 6MWD, significantly influenced the incidence of PED; the 6MWD, compared with the established swallowing risk score, was the most powerful predictive indicator of PED.
In this study, the prevalence of PED was 14.8%, which was similar to the prevalence reported in previous studies. 15,22 The method of dysphagia assessment differed among previous studies, and the incidence of PED varies by the method of assessment. 1,23 Videofluoroscopic swallow study or a fiberoptic endoscopic evaluation of swallowing are the gold standard for diagnosing swallowing disability, although these assessments are unsuitable for routine use in the clinical setting because of their invasiveness and complexity. 23 Thus, the bedside swallowing evaluation is the primary assessment tool for determining the swallowing function in the clinical setting despite its limitations with regard to its ability to accurately assess the severity of aspiration and to guide prognosis. 23  The relationship between cerebrovascular disease and dysphagia is well documented, and a causative role is attributed to dysphagia. 25 Therefore, we excluded patients with postoperative new-onset stroke. Nevertheless, after excluding patients who had had a stroke, preoperative frailty was independently associated with PED following cardiovascular surgery. Furthermore, frailty was a more powerful predictor of PED than the RODICS score was (Central Illustration). One possible explanation is that vulnerability appears and becomes increasingly problematic after highly invasive surgery, despite the absence of preoperative symptoms of frailty. 6 In particular, dysfunction of deglutition could be one of the commonest problems that is postoperatively exposed. In   28 We previously reported that poor oral status was closely related to physical frailty. 29 The multidimensional components of frailty may manifest as dysphagia. Dysphagia can lead to malnutrition, which in turn can lead to weight loss, decreased healing ability, and increased incidence of other diseases. 30 The definitive evidence presented here suggests that preoperative interventions for frailty may prevent PED, thus breaking the vicious cycle of frailty and dysphagia.
In addition, aortic surgery or prolonged ventilation were independent predictors of PED in this study.   Ogawa et al