- Email: [email protected]
Finding the red flags: Swallowing difficulties after cardiac surgery in patients with prolonged intubation
Journal of Critical Care 31 (2016) 119–124
Contents lists available at ScienceDirect
Journal of Critical Care journal homepage: www.jccjournal.org
Finding the red flags: Swallowing difficulties after cardiac surgery in patients with prolonged intubation Emma Daly, MSc a, Anna Miles, PhD a,⁎, Samantha Scott, BSc b, Michael Gillham, FCICM, FANZCA b a b
The University of Auckland, Auckland, New Zealand Auckland District Health Board, Auckland, New Zealand
a r t i c l e
i n f o
Keywords: Cardiac surgery Dysphagia Speech-language pathology Silent aspiration Tracheostomy
a b s t r a c t Purpose: This retrospective audit set out to identify referral rates, swallowing characteristics, and risk factors for dysphagia and silent aspiration in at-risk patients after cardiac surgery. Dysphagia and silent aspiration are associated with poorer outcomes post cardiac surgery. Methods: One hundred ninety patients who survived cardiac surgery and received more than 48 hours of intubation were included. Preoperative, perioperative, and postoperative information was collected. Results: Forty-one patients (22%) were referred to speech-language pathology for a swallowing assessment. Twenty-four of these patients (13%) underwent instrumental swallowing assessment, and silent aspiration was observed in 17 (70% of patients diagnosed as having dysphagia via instrumental assessment). Multilogistic analysis revealed previous stroke (P b .05), postoperative stroke (P b .001), and tracheostomy (P b .001) independently associated with dysphagia. The odds ratio for being diagnosed as having pneumonia, if a patient was diagnosed as having dysphagia, was 3.3. Conclusions: Patients identified with dysphagia after cardiac surgery had a high incidence of silent aspiration and increased risk of pneumonia. However, referral rates were low in this at-risk patient group. Early identification and ongoing assessment and appropriate management of dysphagic patients by a speech-language pathologist are strongly recommended. © 2015 Elsevier Inc. All rights reserved.
1. Introduction Dysphagia is a recognized complication of cardiac surgery, with the incidence reported to be between 3% and 67% depending on study inclusion criteria [1–5]. Reported risk factors for dysphagia after cardiac surgery are varied in agreement and include older age [2,4], congestive heart failure [2], sepsis [1,4], perioperative stroke [1], noncoronary bypass surgical procedures [2], transesophageal echocardiography [5], postoperative stroke [5], and intubation time [1,4,5]. Most recently, Skoretz and colleagues [4] conducted a large retrospective study specifically investigating the impact of intubation time on dysphagia prevalence. Prevalence of dysphagia in all patients who underwent cardiac surgery was 6%. However, when intubation time was taken into account, there was a 2-fold increase in the odds of developing dysphagia with increasing lengths of intubation time (analyzed in 12-hour increments). In patients with more than 48 hours of intubation, the prevalence of dysphagia rose to 67% [4].
⁎ Corresponding author at: Speech Science, University of Auckland, Private Bag 92019, Auckland 1142, New Zealand. Tel.: +64 95222620; fax: +64 93737043. E-mail addresses: [email protected] (E. Daly), [email protected] (A. Miles), [email protected] (S. Scott), [email protected] (M. Gillham). http://dx.doi.org/10.1016/j.jcrc.2015.10.008 0883-9441/© 2015 Elsevier Inc. All rights reserved.
Dysphagia after cardiac surgery is associated with poor prognosis [1]. Dysphagia has been observed to delay return to oral intake, triple length of hospital stay, and double hospital costs [1,2]. Patients with dysphagia after cardiac surgery have a higher incidence of pneumonia and postoperative death [6]. Silent aspiration is prevalent in patients in intensive care unit (ICU) settings. In 51 consecutive general ICU patients requiring more than 48 hours of mechanical ventilation, 27 patients aspirated and 12 patients (25%) of these silently aspirated on routine fiberoptic endoscopic evaluation of swallowing (FEES) [7]. Romero and colleagues [8] found a 38% dysphagia rate in nonneurologic patients with a tracheostomy in ICU. Seventy-three percent of those with dysphagia silently aspirated. Aspiration and particularly silent aspiration (aspirating without a protective cough response) have been linked with an increased prevalence of pneumonia in general medical populations [9,10]. In a mixed etiology study, patients with aspiration diagnosed on videofluoroscopic study of swallowing (VFSS) had a 10fold increased risk of pneumonia, whereas those with silent aspiration had a 13-fold increased risk of pneumonia [10]. Newly published data describe the effect of intubation on airway responsiveness in a group of patients after coronary artery bypass graft surgery. Cough reflex testing has been suggested as an indicator of silent aspiration risk [11]. Patients received cough reflex testing prior to surgery, within 2 hours of extubation, and then periodically until their cough reflex returned to
120
E. Daly et al. / Journal of Critical Care 31 (2016) 119–124
baseline. Sixty percent of patients after extubation presented with an absent cough reflex within the first 2 hours, indicating the increased risk of compromised airway protection in this population [12]. Much of the work in dysphagia after cardiac surgery has been conducted in North America, and findings may not equate to different medical systems and patient populations. Inconsistencies within the literature for predictors of dysphagia in this population justify further research. Although sample sizes in these larger North American studies show robust findings, crude measures of dysphagia were used such as modification of diet or bedside swallowing screens [1,4]. Further research using instrumental assessments that are sensitive to the identification of silent aspiration are needed. The aim of this retrospective study was to identify referral rates, swallowing characteristics, and risk factors associated with dysphagia after cardiac surgery in at-risk patients in a single cardiac surgical center in New Zealand. 2. Methods A retrospective audit was performed on consecutive patients who underwent cardiac surgery at Auckland Hospital between 30th June 2012 and 28th June 2014 and who were intubated for more than 48 hours via an endotracheal tube. Patients who died before or after surgery were excluded from the study. Dysphagia was defined as any abnormality of swallowing as confirmed by a speech-language pathologist by bedside assessment and/or instrumental assessment (FEES or VFSS). At the time of this study, no formal screening process existed for referral to speech-language pathology. As standard practice, all patients had a nasogastric tube (NGT) inserted during surgery. When a patient was deemed appropriate to commence oral intake by their medical team, cautious oral intake was trialed with the patient's
nurse. Patients were referred to speech-language pathologist by the multidisciplinary team, most commonly under circumstances of difficulty managing saliva, difficulty swallowing, coughing on intake of oral fluids, or suspected aspiration risk. A standard bedside swallowing evaluation exists at the hospital that includes clinical history, cranial nerve examination, cough reflex test, and oral trials. Frequency of instrumental assessments in individual patients was clinically driven and not controlled. This study received University of Auckland Human Participants Ethics committee and Auckland District Health Board approval (UOAHPEC011721/A+6293). 2.1. Preoperative, perioperative, and postoperative variables The primary author collected demographic variables from the institution's cardiovascular intensive care unit's (CVICU) database. The database is collected prospectively by clinical specialist nurses. Preoperative variables such as the European System for Cardiac Operative Risk Evaluation 2 [13] and New York Heart Association [14] heart failure score were reported on all patients by specialist cardiology staff and were extracted from the database. All perioperative variables, such as surgery type, stroke, and use of transesophageal echocardiogram were also collected from the CVICU database. Postoperative variables were collected from both the CVICU database and scanned clinical records from the patients' inpatient admission. Perioperative or postoperative stroke was confirmed by computed tomographic scan. Presence of pneumonia was established using the National Healthcare Safety Network (2005) “Criteria for Defining Nosocomial Pneumonia” [15] by the primary author. Twentypercent of all files were second marked by the second author with 100% agreement.
Paitent who underwent cardiac surgery with < 48 intubation time (N: 240)
Survived (N:190) 79%
Died (N:50) 21%
Referred to SLP (N:41) 22%
Not referred to SLP (N:149) 78%
Recieved an instrumental assessment (N:24) 59%
Do not recieve an instrumental assessment (N:17) 41%
Silent aspiration observed (N:17) 71%
No silent aspiration observed (N:7) 29% Fig. 1. Inclusion and exclusion.
E. Daly et al. / Journal of Critical Care 31 (2016) 119–124 Table 1 Preoperative characteristics of patients identified and not identified with dysphagia following cardiac surgery Variable
All (n = 190)
Not identified Identified P with dysphagia with dysphagia (n = 157) (n = 33)
Age (y), mean (SD) 61.3 (15.0) 61.2 (14.4) Smoker (%) 27 (14.2) 24 (15.3) Ex-smoker (%) 97 (51.1) 84 (53.5) Diabetes (%) 51 (26.8) 39 (24.8) NYHA score, mean (SD) I 26 (13.7) 23 (14.6) II 72 (37.9) 56 (35.7) III 57 (30) 50 (31.8) IV 35 (18.4) 28 (17.8) Hypertension (%) 135 (71.1) 113 (72.0) Hypercholesterolemia (%) 115 (60.5) 94 (59.9) Previous stroke (%) 13 (6.8) 8 (5.1) Euroscore2 score, mean (SD) 6.9 (8.70) 7.12 (9.3) Ethnicity New Zealand European 86 (45) 70 (45) New Zealand Maori 25 (13) 23 (15) Asian 17 (9) 12 (7) Pasifika 50 (26) 44 (28) Other 16 (7) 10 (5)
63.5 (17.6) 3 (9.1) 13 (39.4) 12 (36.4) 3 (9.1) 16 (48.5) 7 (21.2) 7 (21.2) 22 (66.7) 21 (63.6) 5 (15.2) 5.96 (4.8)
.44 .35 .14 .18 .39 .39 .39 .39 .39 .54 .69 b.05 .49
16 (49) 2 (6) 5 (15) 6 (12) 6 (18)
b.05 b.05 b.05 b.05 b.05
Euroscore2 indicates European System for Cardiac Operative Risk Evaluation 2; NYHA, New York Heart Association.
2.2. Swallowing variables Vocal fold edema and vocal fold paralysis were reported by treating speech-language pathologist and/or otorhinolaryngologist through endoscopic assessment and were extracted from clinical notes. Patients' oral intake was coded using the validated Functional Oral Intake Scale (FOIS) [16] and was extracted from clinical notes. A FOIS score of 1 indicates nil by mouth (NBM) and tube fed, and a score of 7 represents a total oral diet with no restrictions. The validated penetration-aspiration scale [17] was extracted from the original instrumental assessment reports (both VFSS and FEES) that were written by the treating speechlanguage pathologist. A score of 1 indicates no aspiration, and a score of 8 indicates silent aspiration below the vocal cords without any attempt to clear. The secretion rating scale [18] was also extracted from the original FEES reports (0 = no secretions; 3 = severe secretions). 2.3. Statistical analysis Statistical analysis was completed using SPSS version 22 (SPSS, Chicago, Ill). Categorical data are reported in frequencies and percentages with associations analyzed using χ2 test. Where assumptions for χ 2 were violated, the Fisher exact test was used. For continuous and normally distributed data, unpaired, independent t tests are reported
121
with means and SDs. Multiple logistic regressions were applied to the full cohort of 190 patients to evaluate the effect of confounding variables on dysphagia. Binary logistic regressions found that 8 confounding variables reached significance at P b .05 and these variables were included in the model: ethnicity, previous stroke, postoperative stroke, total ventilation time, tracheostomy, duration of tracheostomy, vocal fold paralysis, and vocal fold edema. First, the full model with all confounding factors was fit, and then backward selection was used to select the main effect model. A P b .05 was considered statistically significant for all analyses. 3. Results Two hundred forty patients undergoing cardiac surgery between June 2012 and June 2014 at our institution with more than 48 hours intubation time were included. Of the total 240 patients, 50 were excluded because they died in the hospital (Fig. 1). Two of these patients died with aspiration pneumonia listed as cause of death on their death certificate. Of the remaining 190 cases, there were more male patients (n = 128) than female patients (n = 62). Demographics for the full cohort are displayed in Table 1. Sixty-four patients (34%) had a tracheostomy during their admission (Fig. 2). Thirty-four percent of the cohort has a tracheostomy inserted during their admission. Of the 41 patients referred for a swallowing assessment, 33 were diagnosed as having dysphagia by their treating speech-language pathologist (17.4% of the total 190 audited; Fig. 1). Demographics of the patients with and without dysphagia diagnosis are displayed in Table 1. Perioperative characteristics for patients with and without dysphagia diagnosis are displayed in Table 2 and postoperative characteristics in Table 3. The odds ratio for being diagnosed dysphagia was 7.14 if a patient had a postoperative stroke, 3.34 if a patient had a history of stroke, and 2.50 if they had a tracheostomy (Table 4). Sixty percent of patients identified with dysphagia had had a tracheostomy inserted during their admission. The odds ratio for being diagnosed as having pneumonia during admission if a patient was diagnosed as having dysphagia was 3.3. Return to oral diet was significantly associated with the presence of dysphagia (Table 5). In the group without a dysphagia diagnosis, 7 patients were transferred from our hospital still tracheostomized and tube fed, back to their referral hospitals' ICUs, whereas 4 patients were placed on modified diets but not referred to speech-language pathology for a swallowing assessment. The degree of dysphagia was rated as severe (45%), moderate (30%), or mild (25%) by the treating speech-language pathologist. Nine patients did not receive an instrumental assessment. These patients passed their cough reflex test and were deemed safe to be managed through compensatory strategies at bedside by their treating speechlanguage pathologist. Twenty-four of the 33 patients diagnosed as
Patients N: 190
Tracheostomy (N: 64) 34%
Dysphagia diagnosed (N:20) 31%
No dysphagia diagnosed (N:44) 69%
No tracheostomy (N: 126) 66%
Dysphagia diagnosed (N:13) 10%
No dysphagia diagnosed (N:113) 90%
Fig. 2. Association between tracheostomy and being identified with dysphagia.
122
E. Daly et al. / Journal of Critical Care 31 (2016) 119–124
Table 2 Perioperative characteristics of patients identified and not identified with dysphagia after cardiac surgery Variable
All (n = 190)
Not identified Identified with P with dysphagia dysphagia (n = 157) (n = 33)
Procedure (%) CABG 99 (52.1) 81 (51.6) Valve 58 (30.5) 48 (30.6) Cardiac transplant 9 (4.7) 8 (5.1) Other 24 (12.6) 20 (12.7) Urgency (%) Elective 4.2 (22.1) 39 (24.8) Urgent 81 (42.6) 62 (39.5) Emergency 67 (35.3) 56 (35.7) Bypass time (h), mean (SD) 146.7 (80.4) 148.5 (82.8) Intraoperative TOE (%) 176 (92.6) 145 (92.4) Perioperative complications Stroke (%) 12 (6.3) 4 (2.5) Sepsis (%) 63 (33.2) 53 (33.8)
.96 .96 .96 .96 .96 .07 .07 .07 .07 .52 .75
18 (54.5) 10 (30.3) 1 (3.0) 4 (12.1) 3 (9.1) 19 (57.6) 11 (33.3) 138.5 (68.4) 31 (93.9) 8 (24.2) 10 (30.3)
b.001 .70
CABG indicates coronary artery bypass grafting; TOE, transesophageal echocardiogram.
having dysphagia received an instrumental assessment. This included 23 initial FEES and 1 initial VFSS. During their hospitalization, 14 of these patients received multiple instrumental assessments resulting in a total of 43 FEES and 15 VFSS (Fig. 1). Silent aspiration was identified in 17 (70%) of patients who received an instrumental assessment. Nine patients diagnosed as having dysphagia were also diagnosed as having pneumonia during their admission. All 9 patients silently aspirated (penetration-aspiration scale score of 8) on instrumental assessment and 6 (66%) received a severe secretion rating [3] on FEES. In the group of tracheostomized patients who were diagnosed as having dysphagia, the dysphagia persisted after decannulation (Fig. 3). Aspiration was observed in 11 of the 15 dysphagic patients who received an instrumental assessment after decannulation; all of these patients' aspiration events were silent. Silent aspiration was documented by instrumental assessment to persist after decannulation for a median of 6 days (range, 2-39 days; interquartile range, 3-16). 4. Discussion Dysphagia after cardiac surgery is associated with increased likelihood of further complications and increased hospital stay [1,2]. Only 22% of our cohort was referred to speech-language pathology for a swallowing assessment, and dysphagia diagnosis was associated with prolonged tube feeding, pneumonia, and increased length of hospital stay. Previous stroke, postoperative stroke, tracheostomy, hours of ventilation, and vocal fold paralysis and/or edema were all significantly associated with dysphagia diagnosis. Stroke was strongly associated with postoperative referral to speech-language pathology, with this group 7 times more likely to be diagnosed as having dysphagia than nonstroke patients. Contrary to previous literature, older age [2,4], congestive
Table 4 Multiple logistic regressions of predictors for confirmed dysphagia Variable
P
Odds ratio
Confidence interval (%)
Postoperative stroke History of stroke Tracheostomy
b.001 b.01 b.01
7.14 3.34 2.50
1.79-28.40 1.01-10.91 1.04-5.99
heart failure [2], sepsis [1,4], noncoronary bypass surgical procedures [2], and transesophageal echocardiography [5] were not statistically associated with dysphagia in the current study. In keeping with previous studies, silent aspiration was observed in two thirds of patients who received an instrumental assessment of swallowing [8]. The incidence of silent aspiration is likely far greater because only a fifth of patients at this cardiac unit were referred to speech-language pathology, with only 13% receiving an instrumental assessment. The causal relationship between tracheostomy presence, dysphagia, and aspiration remains controversial [8,19–22]. The most recent evidence suggests that dysphagia, although common in patients with a tracheostomy, is most likely due to concurrent medical factors rather than the tracheostomy itself [19]. Previous studies of at-risk patients after cardiac surgery have excluded patients with tracheostomies due to their likelihood of dysphagia. In this cardiac unit, speech-language pathology referral rates are low, despite high tracheostomy insertion rates (34% in our study compared with 2%) [4]. Most patients referred to speechlanguage pathology had received a tracheostomy during their admission even if it had been removed before their first swallowing assessment. Similarly to previous work in mixed etiology cohorts [19,22], this study demonstrates that dysphagia and silent aspiration can persist long after tracheostomy decannulation in the cardiac population. Pneumonia is a well-recognized risk of dysphagia [6,23]. In this study, pneumonia was significantly associated with dysphagia diagnosis, with a pneumonia rate of 27% in those referred to speechlanguage pathology. All patients who had pneumonia and received an instrumental assessment were silent aspirators, and more than two thirds of this patient group who had a FEES also scored severe on the secretion scale. These findings are concordant with previous studies where secretion accumulation, aspiration, and particularly silent aspiration are significantly associated with pneumonia [10,18]. Upper airway damage is a common complication of endotracheal intubation, with a prevalence of up to 94% in the acute stages and 10% of symptoms persisting at 6 months [24]. In the current study, postsurgical vocal fold paralysis was common in patients referred to speechlanguage pathology. The association between vocal cord paralysis, dysphagia, and aspiration is well established [25]. Referral rates to speech-language pathology were low in our cardiac unit in contrast with the reported incidence of dysphagia in larger comparative studies despite the inclusion of patients who received a tracheostomy [1,4]. This was a retrospective audit, and there was no standard protocol for dysphagia screening or referral to speech-language
Table 3 Postoperative characteristics of patients identified and not identified with dysphagia after cardiac surgery Variable
All (n = 190)
Not identified with dysphagia (n = 157)
Identified with dysphagia (n = 33)
P
Total ventilation hours, mean (SD) Postoperative TOE (%) Tracheostomy (%) Duration of tracheostomy if had one inserted (d), mean (SD) Pneumonia (%) Length of stay in hospital (d), mean (SD) Opiates after oral intake (d), mean (SD) Vocal fold paralysis (%) Vocal fold edema (%) Destination after hospital (%) Home Another hospital/rehabilitation center
159.2 (137.6) 138 (72.6) 64 (33.7) 4.6 (9.1) 25 (13.2) 21.3 (18.9) 3.0 (7.0) 16 (8.4) 11 (5.8)
149.3 (132.7) 116 (73.9) 44 (28) 3.5 (7.2) 16 (10.2) 18.9 (16.3) 0.7 (2.4) 2 (1.3) 1 (0.6)
206.4 (152.1) 22 (66.7) 20 (60.6) 9.9 (14.3) 9 (27.3) 32.6 (25.5) 2.6 (8.6) 14 (42.4) 10 (30.3)
128 (67.4) 62 (32.6)
112 (71.3) 45 (28.7)
b.05 .40 b.001 b.01 b.001 b.001 .21 b.001 b.001 b.01 b.01 b.01
TOE indicates transesophageal echocardiogram.
16 (48.5) 17 (51.5)
E. Daly et al. / Journal of Critical Care 31 (2016) 119–124
123
Table 5 Return to oral intake postsurgery Variable
All (n = 190)
Not identified with dysphagia (n = 157)
Identified with dysphagia (n = 33)
P
NGT duration (d), mean (SD) Duration NBM (d), mean (SD) PEG insertion (%) Duration of modified diet (d), mean (SD) Time until oral fed (d), mean (SD) FOISa[16] on discharge, mean score (SD) 1 2 3 4 5 6 7
10.2 (10.3) 9.9 (9.2) 2 9.0 (12.3) 8.5 (8.2)
8.2 (8.6) 8.3 (8.0) 0 (0.0) 4.5 (2.4) 7.5 (7.5)
19.7 (12.7) 17.2 (10.9) 2 (6.1) 13.7 (16.4) 13.3 (9.8)
b .001 b.001 b.05 b.001 b.001
11 (6) 1 (0.5) 2 (1.1) 6 (3.3) 6 (3.3) – 157 (85.8)
7 (4.6) 0 (0) 0 (0) 2 (1.3) 2 (1.3) – 140 (92.7)
4 (12.5) 1 (3.1) 2 (6.3) 4 (12.5) 4 (12.5) – 17 (53.1)
b.001 b.001 b.001 b.001 b.001 b.001
PEG indicates percutaneous endoscopic gastroscopy. a Recordings of FOIS were only available in 157 of 190 patients. The 33 patients without did not have clear notes as to their discharge diet and therefore were unable to be classified accurately.
pathology in place at the time that patients underwent surgery. With less than a quarter of patients seen by a speech-language pathologist, it is likely that some patients with dysphagia were missed. Patients who died during admission and patients discharged to referring hospitals prior to referral for a swallowing assessment were not included in the dysphagia cohort. A number of the transferred patients were still NBM and tube fed at the time of transfer and would likely receive speech-language pathology input at a later date. It is also worth noting that 8 patients referred for a swallowing assessment were confirmed as having no swallowing difficulties by the speech-language pathologist. These patients give some reassurance that the “threshold” for referral to speech-language pathologist was not set too high. Importantly, in contrast with the less strict criteria of previous studies (initiation of modified diet by staff), those identified as dysphagic in this study all received comprehensive swallowing assessments including cough reflex testing and a high proportion of instrumental assessment [1,4]. Although the total number of dysphagic patients reported is likely to be an underestimate, those within the dysphagia group
have a validated dysphagia diagnosis. In addition, high numbers of patients received tracheostomies and all patients receive NGT feeding postoperatively if they remain intubated for 48 hours or longer. It is possible that this delays commencement of oral intake and referral to speech-language pathology in the unit. In doing so, some patients who had potentially been transiently dysphagic after surgery may have recovered before oral intake was initiated, bypassing the need for a speech-language pathology referral. Prospective observational studies using instrumental swallowing assessment are required to further define dysphagia rates, characteristics, and recovery in patients after cardiac surgery. There are a number of other limitations to the study design. The odds ratios are calculated on the premise that the screening process occurring on the wards was rigorous. It is, therefore, possible that the odds ratios may be an overestimate in view of the possible missed dysphagia cases. Visualization of the vocal cords was not performed on all patients so vocal fold paralysis rates may also be underestimated. Timing and frequency of instrumental assessments were clinically driven and not
Fig. 3. Aspiration in patients who received a tracheostomy and timing of assessment relative to tracheostomy removal (n = 18).
124
E. Daly et al. / Journal of Critical Care 31 (2016) 119–124
controlled. Changes in penetration-aspiration scores on instrumental assessment must therefore be taken as a crude measure of persisting vs resolution of dysphagia. Dysphagia is a serious complication after cardiac surgery and warrants early speech-language pathology input to help mitigate the effects on both the patient and the health system. Improved clinical outcomes for patients with dysphagia, primarily in the stroke population, have been shown through: nurse dysphagia screens [26], formal swallowing clinical pathways [27], and the introduction of a speech-language pathologist-led FEES service [28]. Protocols should be established that screen patients for dysphagia to ensure that those at most risk of aspiration are identified and managed cautiously.
5. Conclusions Patients identified with dysphagia after cardiac surgery had a high incidence of silent aspiration and increased risk of pneumonia. However, referral rates were low in this at-risk patient group. Previous stroke, postoperative stroke, and tracheostomy were independently associated with dysphagia. Early identification of high-risk patients through a standardized speech-language pathology referral protocol may go some way toward optimizing patient outcomes. Future prospective observational studies of patients after cardiac surgery are required to further define dysphagia in this population. References [1] Barker J, Martino R, Reichardt B, Hickey EJ, Ralph-Edwards A. Incidence and impact of dysphagia in patients receiving prolonged endotracheal intubation after cardiac surgery. Can J Surg 2009;52(2):119–24. [2] Ferraris VA, Ferraris SP, Moritz DM, Welch S. Oropharyngeal dysphagia after cardiac operations. Ann Thorac Surg 2001;71(6):1792–6. [3] Hogue Jr CW, Lappas GD, Creswell LL, Ferguson Jr TB, Sample M, Pugh D, et al. Swallowing dysfunction after cardiac operations: associated adverse outcomes and risk factors including intraoperative transesophageal echocardiography. J Thorac Cardiovasc Surg 1995;110(2):517–22. [4] Skoretz SA, Yau TM, Ivanov J, Granton JT, Martino R. Dysphagia and associated risk factors following extubation in cardiovascular surgical patients, 29. Dysphagia Research Society—8th Annual Dysphagia Research Society Meeting; 2014. p. 647–54. [5] Rousou JA, Tighe DA, Garb JL, Krasner H, Engelman RM, Flack III JE, et al. Risk of dysphagia after transesophageal echocardiography during cardiac operations. Ann Thorac Surg 2000;69(2):486–9. [6] Harrington OB, Duckworth JK, Starnes CL, White P, Fleming L, Kritchevsky SB, et al. Silent aspiration after coronary artery bypass grafting. Ann Thorac Surg 1998; 65(6):1599–603.
[7] Ajemian M, Nirmul G, Anderson M, Zirlen D, Kwasnik E. Routine fiberoptic endoscopic evaluation of swallowing following prolonged intubation: implications for management. Arch Surg 2001;136(4):434–7. [8] Romero C, Marambio A, Larrondo J, Walker K, Lira M, Tobar E, et al. Swallowing dysfunction in non-neurologic critically ill patients who require percutaneous dilational tracheostomy. Chest 2010;137:1278–82. [9] Nakagawa T, Sekizawa K, Arai H, Kikuchi R, Katsuhiro M, Hidetada S. High incidence of pneumonia in elderly patients with basal ganglia infarction. Arch Intern Med 1997;157(3):321–4. [10] Pikus L, Levine M, Yang Y, Rubesin S, Katzka D, Laufer I, et al. Videofluoroscopic studies of swallowing dysfunction and the relative risk of pneumonia. Am J Roentgenol 2003;180:1613–6. [11] Miles A, Moore S, McFarlane M, Lee F, Allen J, Huckabee M. Comparison of cough reflex testing against instrumental assessment of aspiration. Physiol Behav 2013;118:25–31. [12] Kallesen M, Psirides A, Huckabee M. Recovery of cough after extubation after coronary artery bypass grafting: a prospective study. J Crit Care 2015;30:758–61. [13] Roques F, Michel P, Goldstone A, Nashef S. The logistic EuroSCORE. Eur Heart J 2003; 24(9):882–3. [14] The Criteria Committee of the New York Heart Association. NYHA functional classification. In: Boston: Little, Brown & Co., editor. Nomenclature and criteria for diagnosis of diseases of the heart and great vessels9th ed. ; 1994. p. 253–6. [15] Centers for Disease Control and Prevention. Pneumonia (Ventilator-associated [VAP] and non-ventilator-associated Pneumonia [PNEU]) Event. Device-associated Module PNEU/VAP; 2015[http://www.cdc.gov.ezproxy.auckland.ac.nz/nhsn/PDFs/ pscManual/6pscVAPcurrent.pdf]. [16] Crary MA, Mann GDC, Groher ME. Initial psychometric assessment of a functional oral intake scale for dysphagia in stroke patients. Arch Phys Med Rehabil 2005; 86(8):1516–20. [17] Rosenbek J, Robbins J, Roecker E, Coyle J, Wood J. A penetration-aspiration scale. 1996;11(2):93–8. [18] Murray J, Langmore S, Ginsberg S, Dostie A. The significance of accumulated oropharyngeal secretions and swallowing frequency in predicting aspiration. Dysphagia 1996;11:99–103. [19] Leder SB, Ross DA. Confirmation of no causal relationship between tracheostomy and aspiration status: a direct replication study. Dysphagia 2010;25:35–9. [20] Elpern EH, Scott MG, Petro MA, Ries MH. Pulmonary aspiration in mechanically ventilated people with tracheostomies. Chest 1994;105(2):563–6. [21] Leder S, Cohn S, Moller B. Fiberoptic endoscopic documentation of the high incidence of aspiration following extubation in critically ill trauma patients. Dysphagia 1998;13:208–12. [22] Kang JY, Choi KH, Yun GJ, Kim MY, Ryu JS. Does removal of tracheostomy affect dysphagia? A kinematic analysis. Dysphagia 2012;27:498–503. [23] Martino R, Foley N, Bhogal S, Diamant N, Speechley M, Teasell R. Dysphagia after stroke: incidence, diagnosis, and pulmonary complications. Stroke 2005;36(12): 2756–63. [24] Colice GL, Stukel TA, Dain B. Laryngeal complications of prolonged intubation. Chest 1989;96:877–84. [25] Morpeth JF, Williams MF. Vocal fold paralysis after anterior cervical discectomy and fusion. Laryngoscope 2000;110(1):43–6. [26] Hinchley J, Shephard T, Furie K, Smith D, Wang D, Tonn S. Formal dysphagia screening protocols prevent pneumonia. Stroke 2005;36:1972–6. [27] Odderson I, Keaton J, McKenna B. Swallow management in patients on an acute stroke pathway: quality is cost effective. Arch Phys Med Rehabil 1995;76:1130–3. [28] Bax L, McFarlane M, Green E, Miles A. SLT-led FEES services: functional outcomes for patients following stroke. J Stroke Cerebrovasc Dis 2014;23(3):195–200.
Contents lists available at ScienceDirect
Journal of Critical Care journal homepage: www.jccjournal.org
Finding the red flags: Swallowing difficulties after cardiac surgery in patients with prolonged intubation Emma Daly, MSc a, Anna Miles, PhD a,⁎, Samantha Scott, BSc b, Michael Gillham, FCICM, FANZCA b a b
The University of Auckland, Auckland, New Zealand Auckland District Health Board, Auckland, New Zealand
a r t i c l e
i n f o
Keywords: Cardiac surgery Dysphagia Speech-language pathology Silent aspiration Tracheostomy
a b s t r a c t Purpose: This retrospective audit set out to identify referral rates, swallowing characteristics, and risk factors for dysphagia and silent aspiration in at-risk patients after cardiac surgery. Dysphagia and silent aspiration are associated with poorer outcomes post cardiac surgery. Methods: One hundred ninety patients who survived cardiac surgery and received more than 48 hours of intubation were included. Preoperative, perioperative, and postoperative information was collected. Results: Forty-one patients (22%) were referred to speech-language pathology for a swallowing assessment. Twenty-four of these patients (13%) underwent instrumental swallowing assessment, and silent aspiration was observed in 17 (70% of patients diagnosed as having dysphagia via instrumental assessment). Multilogistic analysis revealed previous stroke (P b .05), postoperative stroke (P b .001), and tracheostomy (P b .001) independently associated with dysphagia. The odds ratio for being diagnosed as having pneumonia, if a patient was diagnosed as having dysphagia, was 3.3. Conclusions: Patients identified with dysphagia after cardiac surgery had a high incidence of silent aspiration and increased risk of pneumonia. However, referral rates were low in this at-risk patient group. Early identification and ongoing assessment and appropriate management of dysphagic patients by a speech-language pathologist are strongly recommended. © 2015 Elsevier Inc. All rights reserved.
1. Introduction Dysphagia is a recognized complication of cardiac surgery, with the incidence reported to be between 3% and 67% depending on study inclusion criteria [1–5]. Reported risk factors for dysphagia after cardiac surgery are varied in agreement and include older age [2,4], congestive heart failure [2], sepsis [1,4], perioperative stroke [1], noncoronary bypass surgical procedures [2], transesophageal echocardiography [5], postoperative stroke [5], and intubation time [1,4,5]. Most recently, Skoretz and colleagues [4] conducted a large retrospective study specifically investigating the impact of intubation time on dysphagia prevalence. Prevalence of dysphagia in all patients who underwent cardiac surgery was 6%. However, when intubation time was taken into account, there was a 2-fold increase in the odds of developing dysphagia with increasing lengths of intubation time (analyzed in 12-hour increments). In patients with more than 48 hours of intubation, the prevalence of dysphagia rose to 67% [4].
⁎ Corresponding author at: Speech Science, University of Auckland, Private Bag 92019, Auckland 1142, New Zealand. Tel.: +64 95222620; fax: +64 93737043. E-mail addresses: [email protected] (E. Daly), [email protected] (A. Miles), [email protected] (S. Scott), [email protected] (M. Gillham). http://dx.doi.org/10.1016/j.jcrc.2015.10.008 0883-9441/© 2015 Elsevier Inc. All rights reserved.
Dysphagia after cardiac surgery is associated with poor prognosis [1]. Dysphagia has been observed to delay return to oral intake, triple length of hospital stay, and double hospital costs [1,2]. Patients with dysphagia after cardiac surgery have a higher incidence of pneumonia and postoperative death [6]. Silent aspiration is prevalent in patients in intensive care unit (ICU) settings. In 51 consecutive general ICU patients requiring more than 48 hours of mechanical ventilation, 27 patients aspirated and 12 patients (25%) of these silently aspirated on routine fiberoptic endoscopic evaluation of swallowing (FEES) [7]. Romero and colleagues [8] found a 38% dysphagia rate in nonneurologic patients with a tracheostomy in ICU. Seventy-three percent of those with dysphagia silently aspirated. Aspiration and particularly silent aspiration (aspirating without a protective cough response) have been linked with an increased prevalence of pneumonia in general medical populations [9,10]. In a mixed etiology study, patients with aspiration diagnosed on videofluoroscopic study of swallowing (VFSS) had a 10fold increased risk of pneumonia, whereas those with silent aspiration had a 13-fold increased risk of pneumonia [10]. Newly published data describe the effect of intubation on airway responsiveness in a group of patients after coronary artery bypass graft surgery. Cough reflex testing has been suggested as an indicator of silent aspiration risk [11]. Patients received cough reflex testing prior to surgery, within 2 hours of extubation, and then periodically until their cough reflex returned to
120
E. Daly et al. / Journal of Critical Care 31 (2016) 119–124
baseline. Sixty percent of patients after extubation presented with an absent cough reflex within the first 2 hours, indicating the increased risk of compromised airway protection in this population [12]. Much of the work in dysphagia after cardiac surgery has been conducted in North America, and findings may not equate to different medical systems and patient populations. Inconsistencies within the literature for predictors of dysphagia in this population justify further research. Although sample sizes in these larger North American studies show robust findings, crude measures of dysphagia were used such as modification of diet or bedside swallowing screens [1,4]. Further research using instrumental assessments that are sensitive to the identification of silent aspiration are needed. The aim of this retrospective study was to identify referral rates, swallowing characteristics, and risk factors associated with dysphagia after cardiac surgery in at-risk patients in a single cardiac surgical center in New Zealand. 2. Methods A retrospective audit was performed on consecutive patients who underwent cardiac surgery at Auckland Hospital between 30th June 2012 and 28th June 2014 and who were intubated for more than 48 hours via an endotracheal tube. Patients who died before or after surgery were excluded from the study. Dysphagia was defined as any abnormality of swallowing as confirmed by a speech-language pathologist by bedside assessment and/or instrumental assessment (FEES or VFSS). At the time of this study, no formal screening process existed for referral to speech-language pathology. As standard practice, all patients had a nasogastric tube (NGT) inserted during surgery. When a patient was deemed appropriate to commence oral intake by their medical team, cautious oral intake was trialed with the patient's
nurse. Patients were referred to speech-language pathologist by the multidisciplinary team, most commonly under circumstances of difficulty managing saliva, difficulty swallowing, coughing on intake of oral fluids, or suspected aspiration risk. A standard bedside swallowing evaluation exists at the hospital that includes clinical history, cranial nerve examination, cough reflex test, and oral trials. Frequency of instrumental assessments in individual patients was clinically driven and not controlled. This study received University of Auckland Human Participants Ethics committee and Auckland District Health Board approval (UOAHPEC011721/A+6293). 2.1. Preoperative, perioperative, and postoperative variables The primary author collected demographic variables from the institution's cardiovascular intensive care unit's (CVICU) database. The database is collected prospectively by clinical specialist nurses. Preoperative variables such as the European System for Cardiac Operative Risk Evaluation 2 [13] and New York Heart Association [14] heart failure score were reported on all patients by specialist cardiology staff and were extracted from the database. All perioperative variables, such as surgery type, stroke, and use of transesophageal echocardiogram were also collected from the CVICU database. Postoperative variables were collected from both the CVICU database and scanned clinical records from the patients' inpatient admission. Perioperative or postoperative stroke was confirmed by computed tomographic scan. Presence of pneumonia was established using the National Healthcare Safety Network (2005) “Criteria for Defining Nosocomial Pneumonia” [15] by the primary author. Twentypercent of all files were second marked by the second author with 100% agreement.
Paitent who underwent cardiac surgery with < 48 intubation time (N: 240)
Survived (N:190) 79%
Died (N:50) 21%
Referred to SLP (N:41) 22%
Not referred to SLP (N:149) 78%
Recieved an instrumental assessment (N:24) 59%
Do not recieve an instrumental assessment (N:17) 41%
Silent aspiration observed (N:17) 71%
No silent aspiration observed (N:7) 29% Fig. 1. Inclusion and exclusion.
E. Daly et al. / Journal of Critical Care 31 (2016) 119–124 Table 1 Preoperative characteristics of patients identified and not identified with dysphagia following cardiac surgery Variable
All (n = 190)
Not identified Identified P with dysphagia with dysphagia (n = 157) (n = 33)
Age (y), mean (SD) 61.3 (15.0) 61.2 (14.4) Smoker (%) 27 (14.2) 24 (15.3) Ex-smoker (%) 97 (51.1) 84 (53.5) Diabetes (%) 51 (26.8) 39 (24.8) NYHA score, mean (SD) I 26 (13.7) 23 (14.6) II 72 (37.9) 56 (35.7) III 57 (30) 50 (31.8) IV 35 (18.4) 28 (17.8) Hypertension (%) 135 (71.1) 113 (72.0) Hypercholesterolemia (%) 115 (60.5) 94 (59.9) Previous stroke (%) 13 (6.8) 8 (5.1) Euroscore2 score, mean (SD) 6.9 (8.70) 7.12 (9.3) Ethnicity New Zealand European 86 (45) 70 (45) New Zealand Maori 25 (13) 23 (15) Asian 17 (9) 12 (7) Pasifika 50 (26) 44 (28) Other 16 (7) 10 (5)
63.5 (17.6) 3 (9.1) 13 (39.4) 12 (36.4) 3 (9.1) 16 (48.5) 7 (21.2) 7 (21.2) 22 (66.7) 21 (63.6) 5 (15.2) 5.96 (4.8)
.44 .35 .14 .18 .39 .39 .39 .39 .39 .54 .69 b.05 .49
16 (49) 2 (6) 5 (15) 6 (12) 6 (18)
b.05 b.05 b.05 b.05 b.05
Euroscore2 indicates European System for Cardiac Operative Risk Evaluation 2; NYHA, New York Heart Association.
2.2. Swallowing variables Vocal fold edema and vocal fold paralysis were reported by treating speech-language pathologist and/or otorhinolaryngologist through endoscopic assessment and were extracted from clinical notes. Patients' oral intake was coded using the validated Functional Oral Intake Scale (FOIS) [16] and was extracted from clinical notes. A FOIS score of 1 indicates nil by mouth (NBM) and tube fed, and a score of 7 represents a total oral diet with no restrictions. The validated penetration-aspiration scale [17] was extracted from the original instrumental assessment reports (both VFSS and FEES) that were written by the treating speechlanguage pathologist. A score of 1 indicates no aspiration, and a score of 8 indicates silent aspiration below the vocal cords without any attempt to clear. The secretion rating scale [18] was also extracted from the original FEES reports (0 = no secretions; 3 = severe secretions). 2.3. Statistical analysis Statistical analysis was completed using SPSS version 22 (SPSS, Chicago, Ill). Categorical data are reported in frequencies and percentages with associations analyzed using χ2 test. Where assumptions for χ 2 were violated, the Fisher exact test was used. For continuous and normally distributed data, unpaired, independent t tests are reported
121
with means and SDs. Multiple logistic regressions were applied to the full cohort of 190 patients to evaluate the effect of confounding variables on dysphagia. Binary logistic regressions found that 8 confounding variables reached significance at P b .05 and these variables were included in the model: ethnicity, previous stroke, postoperative stroke, total ventilation time, tracheostomy, duration of tracheostomy, vocal fold paralysis, and vocal fold edema. First, the full model with all confounding factors was fit, and then backward selection was used to select the main effect model. A P b .05 was considered statistically significant for all analyses. 3. Results Two hundred forty patients undergoing cardiac surgery between June 2012 and June 2014 at our institution with more than 48 hours intubation time were included. Of the total 240 patients, 50 were excluded because they died in the hospital (Fig. 1). Two of these patients died with aspiration pneumonia listed as cause of death on their death certificate. Of the remaining 190 cases, there were more male patients (n = 128) than female patients (n = 62). Demographics for the full cohort are displayed in Table 1. Sixty-four patients (34%) had a tracheostomy during their admission (Fig. 2). Thirty-four percent of the cohort has a tracheostomy inserted during their admission. Of the 41 patients referred for a swallowing assessment, 33 were diagnosed as having dysphagia by their treating speech-language pathologist (17.4% of the total 190 audited; Fig. 1). Demographics of the patients with and without dysphagia diagnosis are displayed in Table 1. Perioperative characteristics for patients with and without dysphagia diagnosis are displayed in Table 2 and postoperative characteristics in Table 3. The odds ratio for being diagnosed dysphagia was 7.14 if a patient had a postoperative stroke, 3.34 if a patient had a history of stroke, and 2.50 if they had a tracheostomy (Table 4). Sixty percent of patients identified with dysphagia had had a tracheostomy inserted during their admission. The odds ratio for being diagnosed as having pneumonia during admission if a patient was diagnosed as having dysphagia was 3.3. Return to oral diet was significantly associated with the presence of dysphagia (Table 5). In the group without a dysphagia diagnosis, 7 patients were transferred from our hospital still tracheostomized and tube fed, back to their referral hospitals' ICUs, whereas 4 patients were placed on modified diets but not referred to speech-language pathology for a swallowing assessment. The degree of dysphagia was rated as severe (45%), moderate (30%), or mild (25%) by the treating speech-language pathologist. Nine patients did not receive an instrumental assessment. These patients passed their cough reflex test and were deemed safe to be managed through compensatory strategies at bedside by their treating speechlanguage pathologist. Twenty-four of the 33 patients diagnosed as
Patients N: 190
Tracheostomy (N: 64) 34%
Dysphagia diagnosed (N:20) 31%
No dysphagia diagnosed (N:44) 69%
No tracheostomy (N: 126) 66%
Dysphagia diagnosed (N:13) 10%
No dysphagia diagnosed (N:113) 90%
Fig. 2. Association between tracheostomy and being identified with dysphagia.
122
E. Daly et al. / Journal of Critical Care 31 (2016) 119–124
Table 2 Perioperative characteristics of patients identified and not identified with dysphagia after cardiac surgery Variable
All (n = 190)
Not identified Identified with P with dysphagia dysphagia (n = 157) (n = 33)
Procedure (%) CABG 99 (52.1) 81 (51.6) Valve 58 (30.5) 48 (30.6) Cardiac transplant 9 (4.7) 8 (5.1) Other 24 (12.6) 20 (12.7) Urgency (%) Elective 4.2 (22.1) 39 (24.8) Urgent 81 (42.6) 62 (39.5) Emergency 67 (35.3) 56 (35.7) Bypass time (h), mean (SD) 146.7 (80.4) 148.5 (82.8) Intraoperative TOE (%) 176 (92.6) 145 (92.4) Perioperative complications Stroke (%) 12 (6.3) 4 (2.5) Sepsis (%) 63 (33.2) 53 (33.8)
.96 .96 .96 .96 .96 .07 .07 .07 .07 .52 .75
18 (54.5) 10 (30.3) 1 (3.0) 4 (12.1) 3 (9.1) 19 (57.6) 11 (33.3) 138.5 (68.4) 31 (93.9) 8 (24.2) 10 (30.3)
b.001 .70
CABG indicates coronary artery bypass grafting; TOE, transesophageal echocardiogram.
having dysphagia received an instrumental assessment. This included 23 initial FEES and 1 initial VFSS. During their hospitalization, 14 of these patients received multiple instrumental assessments resulting in a total of 43 FEES and 15 VFSS (Fig. 1). Silent aspiration was identified in 17 (70%) of patients who received an instrumental assessment. Nine patients diagnosed as having dysphagia were also diagnosed as having pneumonia during their admission. All 9 patients silently aspirated (penetration-aspiration scale score of 8) on instrumental assessment and 6 (66%) received a severe secretion rating [3] on FEES. In the group of tracheostomized patients who were diagnosed as having dysphagia, the dysphagia persisted after decannulation (Fig. 3). Aspiration was observed in 11 of the 15 dysphagic patients who received an instrumental assessment after decannulation; all of these patients' aspiration events were silent. Silent aspiration was documented by instrumental assessment to persist after decannulation for a median of 6 days (range, 2-39 days; interquartile range, 3-16). 4. Discussion Dysphagia after cardiac surgery is associated with increased likelihood of further complications and increased hospital stay [1,2]. Only 22% of our cohort was referred to speech-language pathology for a swallowing assessment, and dysphagia diagnosis was associated with prolonged tube feeding, pneumonia, and increased length of hospital stay. Previous stroke, postoperative stroke, tracheostomy, hours of ventilation, and vocal fold paralysis and/or edema were all significantly associated with dysphagia diagnosis. Stroke was strongly associated with postoperative referral to speech-language pathology, with this group 7 times more likely to be diagnosed as having dysphagia than nonstroke patients. Contrary to previous literature, older age [2,4], congestive
Table 4 Multiple logistic regressions of predictors for confirmed dysphagia Variable
P
Odds ratio
Confidence interval (%)
Postoperative stroke History of stroke Tracheostomy
b.001 b.01 b.01
7.14 3.34 2.50
1.79-28.40 1.01-10.91 1.04-5.99
heart failure [2], sepsis [1,4], noncoronary bypass surgical procedures [2], and transesophageal echocardiography [5] were not statistically associated with dysphagia in the current study. In keeping with previous studies, silent aspiration was observed in two thirds of patients who received an instrumental assessment of swallowing [8]. The incidence of silent aspiration is likely far greater because only a fifth of patients at this cardiac unit were referred to speech-language pathology, with only 13% receiving an instrumental assessment. The causal relationship between tracheostomy presence, dysphagia, and aspiration remains controversial [8,19–22]. The most recent evidence suggests that dysphagia, although common in patients with a tracheostomy, is most likely due to concurrent medical factors rather than the tracheostomy itself [19]. Previous studies of at-risk patients after cardiac surgery have excluded patients with tracheostomies due to their likelihood of dysphagia. In this cardiac unit, speech-language pathology referral rates are low, despite high tracheostomy insertion rates (34% in our study compared with 2%) [4]. Most patients referred to speechlanguage pathology had received a tracheostomy during their admission even if it had been removed before their first swallowing assessment. Similarly to previous work in mixed etiology cohorts [19,22], this study demonstrates that dysphagia and silent aspiration can persist long after tracheostomy decannulation in the cardiac population. Pneumonia is a well-recognized risk of dysphagia [6,23]. In this study, pneumonia was significantly associated with dysphagia diagnosis, with a pneumonia rate of 27% in those referred to speechlanguage pathology. All patients who had pneumonia and received an instrumental assessment were silent aspirators, and more than two thirds of this patient group who had a FEES also scored severe on the secretion scale. These findings are concordant with previous studies where secretion accumulation, aspiration, and particularly silent aspiration are significantly associated with pneumonia [10,18]. Upper airway damage is a common complication of endotracheal intubation, with a prevalence of up to 94% in the acute stages and 10% of symptoms persisting at 6 months [24]. In the current study, postsurgical vocal fold paralysis was common in patients referred to speechlanguage pathology. The association between vocal cord paralysis, dysphagia, and aspiration is well established [25]. Referral rates to speech-language pathology were low in our cardiac unit in contrast with the reported incidence of dysphagia in larger comparative studies despite the inclusion of patients who received a tracheostomy [1,4]. This was a retrospective audit, and there was no standard protocol for dysphagia screening or referral to speech-language
Table 3 Postoperative characteristics of patients identified and not identified with dysphagia after cardiac surgery Variable
All (n = 190)
Not identified with dysphagia (n = 157)
Identified with dysphagia (n = 33)
P
Total ventilation hours, mean (SD) Postoperative TOE (%) Tracheostomy (%) Duration of tracheostomy if had one inserted (d), mean (SD) Pneumonia (%) Length of stay in hospital (d), mean (SD) Opiates after oral intake (d), mean (SD) Vocal fold paralysis (%) Vocal fold edema (%) Destination after hospital (%) Home Another hospital/rehabilitation center
159.2 (137.6) 138 (72.6) 64 (33.7) 4.6 (9.1) 25 (13.2) 21.3 (18.9) 3.0 (7.0) 16 (8.4) 11 (5.8)
149.3 (132.7) 116 (73.9) 44 (28) 3.5 (7.2) 16 (10.2) 18.9 (16.3) 0.7 (2.4) 2 (1.3) 1 (0.6)
206.4 (152.1) 22 (66.7) 20 (60.6) 9.9 (14.3) 9 (27.3) 32.6 (25.5) 2.6 (8.6) 14 (42.4) 10 (30.3)
128 (67.4) 62 (32.6)
112 (71.3) 45 (28.7)
b.05 .40 b.001 b.01 b.001 b.001 .21 b.001 b.001 b.01 b.01 b.01
TOE indicates transesophageal echocardiogram.
16 (48.5) 17 (51.5)
E. Daly et al. / Journal of Critical Care 31 (2016) 119–124
123
Table 5 Return to oral intake postsurgery Variable
All (n = 190)
Not identified with dysphagia (n = 157)
Identified with dysphagia (n = 33)
P
NGT duration (d), mean (SD) Duration NBM (d), mean (SD) PEG insertion (%) Duration of modified diet (d), mean (SD) Time until oral fed (d), mean (SD) FOISa[16] on discharge, mean score (SD) 1 2 3 4 5 6 7
10.2 (10.3) 9.9 (9.2) 2 9.0 (12.3) 8.5 (8.2)
8.2 (8.6) 8.3 (8.0) 0 (0.0) 4.5 (2.4) 7.5 (7.5)
19.7 (12.7) 17.2 (10.9) 2 (6.1) 13.7 (16.4) 13.3 (9.8)
b .001 b.001 b.05 b.001 b.001
11 (6) 1 (0.5) 2 (1.1) 6 (3.3) 6 (3.3) – 157 (85.8)
7 (4.6) 0 (0) 0 (0) 2 (1.3) 2 (1.3) – 140 (92.7)
4 (12.5) 1 (3.1) 2 (6.3) 4 (12.5) 4 (12.5) – 17 (53.1)
b.001 b.001 b.001 b.001 b.001 b.001
PEG indicates percutaneous endoscopic gastroscopy. a Recordings of FOIS were only available in 157 of 190 patients. The 33 patients without did not have clear notes as to their discharge diet and therefore were unable to be classified accurately.
pathology in place at the time that patients underwent surgery. With less than a quarter of patients seen by a speech-language pathologist, it is likely that some patients with dysphagia were missed. Patients who died during admission and patients discharged to referring hospitals prior to referral for a swallowing assessment were not included in the dysphagia cohort. A number of the transferred patients were still NBM and tube fed at the time of transfer and would likely receive speech-language pathology input at a later date. It is also worth noting that 8 patients referred for a swallowing assessment were confirmed as having no swallowing difficulties by the speech-language pathologist. These patients give some reassurance that the “threshold” for referral to speech-language pathologist was not set too high. Importantly, in contrast with the less strict criteria of previous studies (initiation of modified diet by staff), those identified as dysphagic in this study all received comprehensive swallowing assessments including cough reflex testing and a high proportion of instrumental assessment [1,4]. Although the total number of dysphagic patients reported is likely to be an underestimate, those within the dysphagia group
have a validated dysphagia diagnosis. In addition, high numbers of patients received tracheostomies and all patients receive NGT feeding postoperatively if they remain intubated for 48 hours or longer. It is possible that this delays commencement of oral intake and referral to speech-language pathology in the unit. In doing so, some patients who had potentially been transiently dysphagic after surgery may have recovered before oral intake was initiated, bypassing the need for a speech-language pathology referral. Prospective observational studies using instrumental swallowing assessment are required to further define dysphagia rates, characteristics, and recovery in patients after cardiac surgery. There are a number of other limitations to the study design. The odds ratios are calculated on the premise that the screening process occurring on the wards was rigorous. It is, therefore, possible that the odds ratios may be an overestimate in view of the possible missed dysphagia cases. Visualization of the vocal cords was not performed on all patients so vocal fold paralysis rates may also be underestimated. Timing and frequency of instrumental assessments were clinically driven and not
Fig. 3. Aspiration in patients who received a tracheostomy and timing of assessment relative to tracheostomy removal (n = 18).
124
E. Daly et al. / Journal of Critical Care 31 (2016) 119–124
controlled. Changes in penetration-aspiration scores on instrumental assessment must therefore be taken as a crude measure of persisting vs resolution of dysphagia. Dysphagia is a serious complication after cardiac surgery and warrants early speech-language pathology input to help mitigate the effects on both the patient and the health system. Improved clinical outcomes for patients with dysphagia, primarily in the stroke population, have been shown through: nurse dysphagia screens [26], formal swallowing clinical pathways [27], and the introduction of a speech-language pathologist-led FEES service [28]. Protocols should be established that screen patients for dysphagia to ensure that those at most risk of aspiration are identified and managed cautiously.
5. Conclusions Patients identified with dysphagia after cardiac surgery had a high incidence of silent aspiration and increased risk of pneumonia. However, referral rates were low in this at-risk patient group. Previous stroke, postoperative stroke, and tracheostomy were independently associated with dysphagia. Early identification of high-risk patients through a standardized speech-language pathology referral protocol may go some way toward optimizing patient outcomes. Future prospective observational studies of patients after cardiac surgery are required to further define dysphagia in this population. References [1] Barker J, Martino R, Reichardt B, Hickey EJ, Ralph-Edwards A. Incidence and impact of dysphagia in patients receiving prolonged endotracheal intubation after cardiac surgery. Can J Surg 2009;52(2):119–24. [2] Ferraris VA, Ferraris SP, Moritz DM, Welch S. Oropharyngeal dysphagia after cardiac operations. Ann Thorac Surg 2001;71(6):1792–6. [3] Hogue Jr CW, Lappas GD, Creswell LL, Ferguson Jr TB, Sample M, Pugh D, et al. Swallowing dysfunction after cardiac operations: associated adverse outcomes and risk factors including intraoperative transesophageal echocardiography. J Thorac Cardiovasc Surg 1995;110(2):517–22. [4] Skoretz SA, Yau TM, Ivanov J, Granton JT, Martino R. Dysphagia and associated risk factors following extubation in cardiovascular surgical patients, 29. Dysphagia Research Society—8th Annual Dysphagia Research Society Meeting; 2014. p. 647–54. [5] Rousou JA, Tighe DA, Garb JL, Krasner H, Engelman RM, Flack III JE, et al. Risk of dysphagia after transesophageal echocardiography during cardiac operations. Ann Thorac Surg 2000;69(2):486–9. [6] Harrington OB, Duckworth JK, Starnes CL, White P, Fleming L, Kritchevsky SB, et al. Silent aspiration after coronary artery bypass grafting. Ann Thorac Surg 1998; 65(6):1599–603.
[7] Ajemian M, Nirmul G, Anderson M, Zirlen D, Kwasnik E. Routine fiberoptic endoscopic evaluation of swallowing following prolonged intubation: implications for management. Arch Surg 2001;136(4):434–7. [8] Romero C, Marambio A, Larrondo J, Walker K, Lira M, Tobar E, et al. Swallowing dysfunction in non-neurologic critically ill patients who require percutaneous dilational tracheostomy. Chest 2010;137:1278–82. [9] Nakagawa T, Sekizawa K, Arai H, Kikuchi R, Katsuhiro M, Hidetada S. High incidence of pneumonia in elderly patients with basal ganglia infarction. Arch Intern Med 1997;157(3):321–4. [10] Pikus L, Levine M, Yang Y, Rubesin S, Katzka D, Laufer I, et al. Videofluoroscopic studies of swallowing dysfunction and the relative risk of pneumonia. Am J Roentgenol 2003;180:1613–6. [11] Miles A, Moore S, McFarlane M, Lee F, Allen J, Huckabee M. Comparison of cough reflex testing against instrumental assessment of aspiration. Physiol Behav 2013;118:25–31. [12] Kallesen M, Psirides A, Huckabee M. Recovery of cough after extubation after coronary artery bypass grafting: a prospective study. J Crit Care 2015;30:758–61. [13] Roques F, Michel P, Goldstone A, Nashef S. The logistic EuroSCORE. Eur Heart J 2003; 24(9):882–3. [14] The Criteria Committee of the New York Heart Association. NYHA functional classification. In: Boston: Little, Brown & Co., editor. Nomenclature and criteria for diagnosis of diseases of the heart and great vessels9th ed. ; 1994. p. 253–6. [15] Centers for Disease Control and Prevention. Pneumonia (Ventilator-associated [VAP] and non-ventilator-associated Pneumonia [PNEU]) Event. Device-associated Module PNEU/VAP; 2015[http://www.cdc.gov.ezproxy.auckland.ac.nz/nhsn/PDFs/ pscManual/6pscVAPcurrent.pdf]. [16] Crary MA, Mann GDC, Groher ME. Initial psychometric assessment of a functional oral intake scale for dysphagia in stroke patients. Arch Phys Med Rehabil 2005; 86(8):1516–20. [17] Rosenbek J, Robbins J, Roecker E, Coyle J, Wood J. A penetration-aspiration scale. 1996;11(2):93–8. [18] Murray J, Langmore S, Ginsberg S, Dostie A. The significance of accumulated oropharyngeal secretions and swallowing frequency in predicting aspiration. Dysphagia 1996;11:99–103. [19] Leder SB, Ross DA. Confirmation of no causal relationship between tracheostomy and aspiration status: a direct replication study. Dysphagia 2010;25:35–9. [20] Elpern EH, Scott MG, Petro MA, Ries MH. Pulmonary aspiration in mechanically ventilated people with tracheostomies. Chest 1994;105(2):563–6. [21] Leder S, Cohn S, Moller B. Fiberoptic endoscopic documentation of the high incidence of aspiration following extubation in critically ill trauma patients. Dysphagia 1998;13:208–12. [22] Kang JY, Choi KH, Yun GJ, Kim MY, Ryu JS. Does removal of tracheostomy affect dysphagia? A kinematic analysis. Dysphagia 2012;27:498–503. [23] Martino R, Foley N, Bhogal S, Diamant N, Speechley M, Teasell R. Dysphagia after stroke: incidence, diagnosis, and pulmonary complications. Stroke 2005;36(12): 2756–63. [24] Colice GL, Stukel TA, Dain B. Laryngeal complications of prolonged intubation. Chest 1989;96:877–84. [25] Morpeth JF, Williams MF. Vocal fold paralysis after anterior cervical discectomy and fusion. Laryngoscope 2000;110(1):43–6. [26] Hinchley J, Shephard T, Furie K, Smith D, Wang D, Tonn S. Formal dysphagia screening protocols prevent pneumonia. Stroke 2005;36:1972–6. [27] Odderson I, Keaton J, McKenna B. Swallow management in patients on an acute stroke pathway: quality is cost effective. Arch Phys Med Rehabil 1995;76:1130–3. [28] Bax L, McFarlane M, Green E, Miles A. SLT-led FEES services: functional outcomes for patients following stroke. J Stroke Cerebrovasc Dis 2014;23(3):195–200.