Swallowing Function After Stroke
Prognosis and Prognostic Factors at 6 Months
Background and Purpose—Swallowing dysfunction (dysphagia) is common and disabling after acute stroke, but its impact on long-term prognosis for potential complications and the recovery from swallowing dysfunction remain uncertain. We aimed to prospectively study the prognosis of swallowing function over the first 6 months after acute stroke and to identify the important independent clinical and videofluoroscopic prognostic factors at baseline that are associated with an increased risk of swallowing dysfunction and complications.
Methods—We prospectively assembled an inception cohort of 128 hospital-referred patients with acute first stroke. We assessed swallowing function clinically and videofluoroscopically, within a median of 3 and 10 days, respectively, of stroke onset, using standardized methods and diagnostic criteria. All patients were followed up prospectively for 6 months for the occurrence of death, recurrent stroke, chest infection, recovery of swallowing function, and return to normal diet.
Results—At presentation, a swallowing abnormality was detected clinically in 65 patients (51%; 95% CI, 42% to 60%) and videofluoroscopically in 82 patients (64%; 95% CI, 55% to 72%). During the subsequent 6 months, 26 patients (20%; 95% CI, 14% to 28%) suffered a chest infection. At 6 months after stroke, 97 of the 112 survivors (87%; 95% CI, 79% to 92%) had returned to their prestroke diet. Clinical evidence of a swallowing abnormality was present in 56 patients (50%; 95% CI, 40% to 60%). Videofluoroscopy was performed at 6 months in 67 patients who had a swallowing abnormality at baseline; it showed penetration of the false cords in 34 patients and aspiration in another 17. The single independent baseline predictor of chest infection during the 6-month follow-up period was a delayed or absent swallowing reflex (detected by videofluoroscopy). The single independent predictor of failure to return to normal diet was delayed oral transit (detected by videofluoroscopy). Independent predictors of the combined outcome event of swallowing impairment, chest infection, or aspiration at 6 months were videofluoroscopic evidence of delayed oral transit and penetration of contrast into the laryngeal vestibule, age >70 years, and male sex.
Conclusions—Swallowing function should be assessed in all acute stroke patients because swallowing dysfunction is common, it persists in many patients, and complications frequently arise. The assessment of swallowing function should be both clinical and videofluoroscopic. The clinical and videofluoroscopic features at presentation that are important predictors of subsequent swallowing abnormalities and complications are videofluoroscopic evidence of delayed oral transit, a delayed or absent swallow reflex, and penetration. These findings require validation in other studies.
Swallowing abnormalities are a common functional impairment of acute stroke, affecting as many as half of all patients, depending on the timing of the assessment, diagnostic methods and criteria, and case mix.1 2 3 4 5 6 7 8 9 10 11 12 They are not only disabling to patients (compromising diet, nutrition, and hydration) but also have the potential to cause serious complications such as dehydration and aspiration pneumonia.
The few studies that have assessed the natural history of swallowing function after acute stroke suggest that swallowing abnormalities recover quickly, but these studies have relied on bedside clinical examination to diagnose dysphagia and have only assessed swallowing function for short periods, such as 2 weeks, after stroke.1 5 There has been only 1 prospective study of the natural history of dysphagia after acute stroke in which swallowing function was assessed clinically and videofluoroscopically and patients were followed up for persistent swallowing difficulties for up to 6 months.10 11 However, other important outcomes, such as chest infection, were not recorded beyond 7 days after stroke, and videofluoroscopy was not repeated beyond 1 month after stroke.
We have undertaken a slightly larger prospective study of the prognosis over the first 6 months after acute stroke of swallowing function and important complications of swallowing dysfunction, such as chest infection. We have also attempted to identify the important independent clinical and videofluoroscopic prognostic factors at baseline that are associated with an increased risk of swallowing dysfunction and complications. The latter findings may serve to simplify the diagnostic assessment of swallowing function (by distinguishing relevant and irrelevant clinical and videofluoroscopic findings) and thereby may enable patients at risk to be accurately identified at presentation.
Subjects and Methods
Between May 1994 and May 1995, a prospective inception cohort study was undertaken of all patients presenting to our hospital who had an acute first stroke (<7 days since symptom onset), were conscious and medically stable, and had no history of previous swallowing impairment or a medical condition that could affect swallowing function.
All patients were assessed neurologically by the study neurologist to establish the diagnosis of stroke (according to the standardized criteria of the World Health Organization13 ), the clinical syndrome, the pathological and etiologic subtype of stroke, and the functional effects of the stroke, as measured by the Oxford Handicap Scale and Barthel Index.
Swallowing function was assessed clinically, at the bedside, by 2 study speech pathologists and videofluoroscopically by the study radiologist and 1 of the study speech pathologists, using standardized methods and diagnostic criteria (Appendix⇓).
The clinical and videofluoroscopic assessment and diagnostic criteria for the presence and severity of a swallowing disorder and aspiration were developed by us and influenced by the work of Ott and Pikna,14 Linden and Siebens,15 and Logemann.16 Briefly, the clinical assessment of swallowing function was conducted by 2 speech pathologists independent of each other and blinded to the videofluoroscopic and CT brain scan findings. It comprised a history from the patient and/or relatives of swallowing ability and function before and after the stroke; an oral motor-sensory examination assessing the components of swallowing, voice, speech, and language function; and observation of the patient swallowing saliva initially (dry swallow) and subsequently after 5 mL of water, 20 mL of water, and a thickened fluid (if appropriate). The videofluoroscopic assessment was conducted by the study radiologist, who was blinded to the results of the clinical assessment and CT brain scan findings.
Patients were followed up prospectively over the first 6 months after stroke, during which time the occurrence of death, recurrent stroke, and chest infection was recorded. Recurrent stroke was defined as (1) the sudden onset of a new focal neurological deficit with no apparent cause other than that of vascular origin (ie, the deficit could not be ascribed to an intercurrent acute illness, epileptic seizure, or toxic effect) or (2) clinical evidence of the sudden onset of an exacerbation of a previous focal neurological deficit with no apparent cause other than that of vascular origin.13
Chest infection was diagnosed by the attending clinician and based on the presence of ≥3 of the following variables: fever (>38°C), productive cough with purulent sputum, abnormal respiratory examination (tachypnea [>22/min], tachycardia, inspiratory crackles, bronchial breathing), abnormal chest radiograph, arterial hypoxemia (Po2 <70 mm Hg), and isolation of a relevant pathogen (positive gram stain and culture).
At 6 months after stroke, the 112 stroke survivors were assessed clinically by the study speech pathologist. This included a detailed dietary history, with the speech pathologist recording whether the diet was different from the usual prestroke diet. The subgroup of 67 patients who had videofluoroscopic evidence of a swallowing abnormality on the initial assessment underwent a repeated videofluoroscopic assessment of swallowing function by the same assessors who conducted the initial assessment. Any swallowing abnormality (clinical or videofluoroscopic) and aspiration (videofluoroscopic) were defined according to standardized criteria.
Crude associations between the occurrence of categorical outcome events (chest infection, dietary change, swallowing abnormality, and aspiration) and each of the categorical baseline variables were assessed by preliminary cross-tabulations with the χ2 test or Fisher’s exact test. Reverse stepwise multiple logistic regression techniques were used to identify important independent predictors of dietary change, swallowing abnormality, aspiration, and chest infection at 6 months. Time to first chest infection was analyzed with the Kaplan-Meier product limit technique. Stepwise Cox proportional hazards multiple regression analysis was performed to identify significant prognostic factors for the outcome event “chest infection during the first 6 months after stroke” and to measure the relative impact of changes in each factor on the hazard.
One hundred twenty-eight patients presenting within 7 days of acute first-ever stroke prospectively underwent a standardized clinical and videofluoroscopic assessment of swallowing function within a median of 3 days (range, 0 to 16 days) and 10 days (range, 0 to 47 days) of symptom onset, respectively.
Swallowing Problems at Onset
Table 1⇑ shows the baseline demographic, clinical, and videofluoroscopic features of the 128 acute stroke patients. Clinical bedside examination identified a swallowing disorder (dysphagia) in 65 patients (51%; 95% CI, 42% to 60%) and aspiration in 64 patients (50%; 95% CI, 41% to 59%). Videofluoroscopy identified a swallowing disorder in 82 patients (64%; 95% CI, 55% to 72%) and aspiration in 28 patients (22%; 95% CI, 15% to 30%).
During 6 months of follow-up after stroke, 5 patients died (all had initial videofluoroscopic evidence of swallowing dysfunction), 12 (9%) had a recurrent stroke, and 26 (20%) had a chest infection (24 had initial videofluoroscopic evidence of dysphagia). The rate of chest infection was greatest in the first month after stroke (12%; 95% CI, 7% to 18%), but chest infection occurred throughout the follow-up period (Figure⇓) and accounted for 4 of the 5 deaths in the cohort. Recurrent chest infection occurred in 8 patients (6%; 95% CI, 2% to 12%). Cox proportional hazards multiple regression analysis identified delayed or absent swallowing reflex (detected by videofluoroscopy) as the single independent predictor of chest infection during the 6-month period (hazard ratio, 9.8; 95% CI, 2.9 to 33).
At 6 months, the 112 survivors (88%; 95% CI, 80% to 93%) were reassessed clinically, and 11 patients (10%) were lost to follow-up. Ninety-seven of the 112 survivors (87%; 95% CI, 79% to 92%) had returned to their prestroke diet, and 15 (13%; 95% CI, 8% to 21%) had not.
Clinical evidence of a persistent swallowing abnormality was diagnosed in 56 patients (50%; 95% CI, 40% to 60%). Videofluoroscopy was performed in 67 of the 82 patients who had initial videofluoroscopic evidence of a swallowing abnormality. Despite the fact the 97 patients (87%) had returned to their prestroke diet, videofluoroscopy showed persistent swallowing abnormalities in 54 of the 67 patients studied; contrast penetrated the false cords in 34 (63%; 95% CI, 49% to 76%) and the true cords (aspiration) in another 17 (31%; 95% CI, 19% to 46%).
Multiple logistic regression analysis (Table 2⇑) identified the single independent baseline predictor of chest infection at 6 months as videofluoroscopic evidence of a delayed or absent swallowing reflex. The single independent baseline predictor of a failure to achieve the patient’s prestroke diet at 6 months after stroke was videofluoroscopic evidence of delayed oral transit. The independent predictors at baseline of the combined outcome event “swallowing impairment, chest infection, or aspiration at 6 months” were age >70 years, male sex, videofluoroscopic evidence of delayed oral transit, and penetration (Table 2⇑).
The results of this study have been derived from a hospital-referred cohort of stroke patients and therefore may reflect some hospital-referral bias. The results are also likely to be influenced by the diagnostic criteria and the timing of the assessment of swallowing function; clinical assessment was undertaken within a median of 3 days (range, 0 to 16 days) and videofluoroscopic assessment within a median of 10 days (range, 0 to 47 days) of symptom onset.
The only comparable cohort study10 11 was also hospital referred, but the patients were older (median age, 79 years) than those in our cohort (median age, 71 years). Their patients were assessed clinically within a median time of 14 hours (range, 0.5 to 50 hours) and videofluoroscopically within a median time of 1 day (range, 0 to 5 days) of stroke onset. Despite these differences and minor differences in diagnostic criteria, the baseline results of our study are remarkably similar. At presentation (within a median of 1 to 3 days after stroke onset), both studies reported clinical evidence of swallowing dysfunction in 51% of patients and videofluoroscopic evidence of aspiration in 22% of patients. Chest infection was only sought during the first 7 days after stroke by Smithard et al but was nevertheless diagnosed in 25% of patients (33% with clinical evidence of an unsafe swallow and 16% with a safe swallow).10 We followed 112 patients for up to 6 months after stroke and recorded chest infection during that period in 20% (n=26) of patients, of whom 9% (n=10) developed a chest infection in the first 7 days after stroke (Figure⇑). The lower early rate of chest infection in our study may be due to any or all of a number of factors such as chance, diagnostic criteria, and treatment strategies. However, we did find that chest infection occurred most commonly in the first month after stroke.
Follow-up videofluoroscopy, performed at 1 month after stroke by Smithard et al,11 revealed that 12 of the 81 patients (15%) were aspirating, and 40 (49%) demonstrated supraglottic penetration. We performed follow-up videofluoroscopy at 6 months after stroke in the 67 patients who had evidence of a swallowing abnormality on the initial videofluoroscopy and found that 17 of the 67 patients (25%) were aspirating. Both studies reveal a disturbingly high prevalence of unequivocal aspiration at different times after stroke, particularly in view of the selection bias in both studies (for example, patients in our study who did not consent to an additional videofluoroscopy or who had a normal initial videofluoroscopy investigation did not undergo a repeated videofluoroscopy at 6 months).
At 6 months after stroke, we found clinical evidence of a swallowing abnormality in 50% of the 112 survivors, which contrasts with the lower prevalence of 11% among the 73 patients assessed by Smithard et al11 at 6 months. It is unlikelythat the differences reflect differences in diagnostic criteria between the studies, because the assessments at 6 months in each study were made by the same personnel using the same criteria as the baseline assessments, which revealed identical results (a 51% prevalence of swallowing abnormality). Although our estimate of swallowing dysfunction at 6 months appears unexpectedly high, it is noteworthy that 15 patients were still unable to return to their prestroke diet of their own volition, and we found unequivocal videofluoroscopic evidence of contrast penetrating the false cords in 34 patients and the true cords (aspiration) in 17 patients among the 67 patients who underwent videofluoroscopic examination at 6 months. We were unable to determine the relative proportion of swallowing abnormalities that were persistent and new. Recurrent stroke during the 6-month follow-up period may have been a relevant etiologic factor in the development of new swallowing abnormalities in up to 10 patients (it was a significant independent predictor of aspiration at 6 months), but it was not an important independent predictor of residual swallowing abnormality in the multiple regression analysis.
The statistical models derived from our data suggest that older patients (>70 years) with disabling stroke (Barthel index <60/100) may be at increased risk of subsequent swallowing complications and that videofluoroscopic evidence of a delayed or absent swallow reflex, delayed oral transit, and penetration are even stronger markers of risk. Although these models are based on small numbers of patients and are therefore unstable, they are the first attempt to identify important clinical and videofluoroscopic factors at presentation that are independently associated with important clinical outcomes. These and future efforts at modeling should help clinicians not only to identify patients at risk of complications but also to identify which components of our assessment tools are important to elicit and act on.
If these results can be validated externally in other cohorts, they would suggest that swallowing function should be assessed in all acute stroke patients (because swallowing dysfunction and its complications are common) and that the assessment of swallowing function should be both clinical and videofluoroscopic, because videofluoroscopic evidence of a delayed or absent swallow reflex, delayed oral transit, and penetration may be even more predictive of complications of swallowing dysfunction than other important clinical features, such as the age of the patient, the stroke syndrome, and its severity.
This study was supported by grants from the Medical Research Foundation of Western Australia, the Sandoz Gerontological Society, and the Royal Perth Hospital Medical Research Foundation.
- Received November 16, 1998.
- Revision received January 4, 1999.
- Accepted January 4, 1999.
- Copyright © 1999 by American Heart Association
Gordon C, Langton-Hewer R, Wade DT. Dysphagia in acute stroke. BMJ. 1987;295:411–414.
Horner J, Massey EW. Silent aspiration following stroke. Neurology. 1988;38:317–319.
Horner J, Massey EW, Riski JE, Lathrop DL, Chase KN. Aspiration following a stroke: clinical correlates and outcome. Neurology. 1988;38:1359–1362.
Barer DH. The natural history and functional consequences of dysphagia after hemispheric stroke. J Neurol Neurosurg Psychiatry. 1989;52:236–241.
Horner J, Brazer SR, Massey EW. Aspiration in bilateral stroke patients: a validation study. Neurology. 1993;43:430–433.
Kidd D, Lawson J, Nesbitt R, MacMahon J. Aspiration in acute stroke: a clinical study with videofluoroscopy. Q J Med. 1993;86:825–829.
Smithard D, O’Neill P, Park C, Morris J, Wyatt R, England R, Martin D. Complications and outcome after acute stroke: does dysphagia matter? Stroke.. 1996;27:1200–1204.
Logeman JA. Diagnosis and Treatment of Swallowing Disorders. San Diego, Calif: College-Hill Press; 1983.