Home > RNJ > 2008 > November/December > Accuracy of a Bedside Dysphagia Screening: A Comparison of Registered Nurses and Speech Therapists

Accuracy of a Bedside Dysphagia Screening: A Comparison of Registered Nurses and Speech Therapists
Janice Weinhardt, MSN BC APRN Susan Hazelett, MS RN Dawn Barrett, MS CCC/SLP • Robert Lada, MD Trish Enos, BSN BS RNC CPHQ Rick Keleman, RN

Evidence-based guidelines suggest that stroke patients should be screened for dysphagia before oral intake. The purpose of this study was to validate a dysphagia screening tool comparing registered nurses (RNs) with speech therapists (STs). All stroke unit patients who received predetermined scores on specific items of the National Institutes of Health Stroke Scale were eligible for screening. The trial consisted of three parts (with swallow, cough, and vocal quality observed during each part): 1 teaspoon lemon ice, 1 teaspoon applesauce, and 1 teaspoon water. RNs performed five screenings that were compared with independent screenings performed on the same patient within 1 hour by a speech therapist (ST). Eighty-three paired screenings were completed, with 94% agreement between the RNs and the STs. This screening identifies patients who are able to swallow and can eat from a safe menu until formally evaluated by an ST while maintaining nothing by mouth (NPO) status for those at risk for aspiration.

Dysphagia affects approximately one-half of stroke patients in the acute phase, although most patients recover within 6 months (Perry & Love, 2001). The reported incidence of dysphagia in studies enrolling acute stroke patients, regardless of lesion location, ranges from 37% to 78%, with a more than three-fold increase for pneumonia risk (Martino et al., 2005). However, studies show that dysphagia is not diagnosed in up to half of cases (Ellul, Barer, & Fall, 1997). Dysphagia has been firmly linked to infections of the lower respiratory tract, presumably because of a link between dysphagia and aspiration, although aspiration alone is insufficient to cause pneumonia (Hinchey et al., 2005; Scottish Intercollegiate Guidelines Network, 2004). Because early diagnosis and implementation of dysphagia protocols during an acute stroke are associated with lower pneumonia rates, it has been recommended that all stroke patients be screened for dysphagia before being given food or drink (Doggett et al., 2001; Martino, Pron, & Diamant, 2000; Rosenvinge & Starke, 2005; Scottish Intercollegiate Guidelines Network). However, withholding food (maintaining nothing by mouth [NPO] status) until a formal swallowing evaluation can be completed by a speech therapist (ST) can cause substantial distress for the patient due to hunger and dry mouth, decreasing patient satisfaction.

Screening for Dysphagia

The accurate identification of patients with dysphagia is the first step in effective dysphagia management. A dysphagia screen usually is performed initially. The American Speech-Language-Hearing Association ([ASHA]; ASHA Division 13, 2006) defines a swallowing screening as a minimally invasive evaluation procedure that quickly documents

  • the likelihood that dysphagia exists
  • the need for further swallowing assessment
  • the safety of patient oral intake
  • the need for alternative nutritional support.

The Joint Commission specifies that poststroke screening methods may include clinical bedside examination, simple water test, Burke water swallow test (DePippo, Holas, & Reding, 1994), simple standardized bedside swallowing assessment (Perry, 2001), barium swallow, videofluoroscopy, double-contrast esophagram, radionuclide studies, and endoscopy.

Those who pass the initial screening may be carefully advanced in their diet, and those who fail are referred for a formal evaluation by an ST (Davies, Taylor, MacDonald, & Barer, 2001; Scottish Intercollegiate Guidelines Network, 2004). Depending on the test used, the sensitivity of bedside screening has been shown to range from 42% to 92% and specificity from 59% to 91% (Ramsey, Smithard, & Kalra, 2003). There is very little consensus among speech-language pathologists regarding specific swallowing screening protocols. Most swallowing screenings require the patient to demonstrate a level of alertness deemed adequate for safe oral intake. Furthermore, the patient must possess basic oral motor and laryngeal integrity, as demonstrated via assessment of the patient’s cough reflex, ability to manage secretions, and communicative effectiveness. If these fundamental requirements are met, the patient can be assessed with small quantities of food or liquid.

The initial screening usually is carried out by a nurse in the acute care setting because bedside testing is safe and easy (Ramsey et al., 2003). However, only one screening tool has been tested for reliability with nurses as screeners (Perry & Love, 2001). It is widely recognized that further validation and reliability testing of bedside screening tools for nurses are needed (Hinchey et al., 2005).


The purpose of this study was to establish the validity of a registered nurse (RN) bedside dysphagia screening by comparing RN ratings with the concurrent results from an ST.


This study was approved by the institutional review board at Summa Health System.

Patient Selection

Before the RN can conduct the swallow evaluation, he or she needs to determine whether it is safe to introduce food or fluids even for screening purposes. A lethargic, unresponsive patient who does not follow commands or sit up independently is at risk for aspiration and should be excluded from screening.

The next step for RNs when determining which patients are eligible for a dysphagia screening involves a neurological assessment using the National Institutes of Health Stroke Scale (NIHSS). The NIHSS was chosen for its usefulness in establishing stroke deficits and severity and providing a standardized, valid assessment tool. The RNs on the stroke unit at this institution have been trained and certified in the administration of the NIHSS, which is routinely performed on admission to establish baseline deficits.

The NIHSS is an 11-item ordinal scale with a numbered score for each item on the scale. A score of 0 on any item means there is no neurological deficit, or a normal exam. Deficits in the following five items have the potential to contribute to dysphagia:

  • wakefulness or level of consciousness
  • level of consciousness commands
  • facial palsy
  • aphasia
  • dysarthria.

Wakefulness or level of consciousness is an assessment of the patient’s degree of alertness. A score of 0 on this item indicates that the patient is keenly responsive and alert. A score of 1 would be given to the person who is drowsy and needs minor stimulation to obey commands. A score of 2 on this item indicates the patient is obtunded or stuporous and needs repeated stimulation to attend. A score of 3 is reserved for patients who are comatose and make only reflexive posturing in response to noxious stimulation. A score of 0 on this item was needed as part of the nursing bedside swallow screening because a fully alert patient is less likely to aspirate than a lethargic patient.

The second item, level of consciousness commands, is evaluated by asking patients to close and open their eyes and to close and open their fist. A score of 0 is achieved if the patient follows both commands correctly. The patient scores 1 if only one command is followed correctly or 2 if neither commands are followed. Patients must follow both commands to be eligible for the dysphagia evaluation. The first stage of swallowing, the oral stage, requires the patient’s voluntary control. A patient unable to follow commands is most likely to allow food or fluids to sit in the oral cavity or fall to the back of the throat, with the bolus not properly prepared for swallowing.

Item 4 on the NIHSS tests for facial palsy by evaluating symmetry of facial features and movement. Weakness of all or part of the face may be the result of a lower motor neuron lesion of the seventh cranial nerve (CN) on the same side of the weakness, a lesion of the cortical spinal tract or motor cortex on the cerebral hemisphere opposite the weakness, or a lesion of one or both sides of the brainstem (The Stroke Group, Inc., 1998). In the oral phase, CN VII mediates salivation to initiate the breakdown of carbohydrates in the mouth and stomach. In addition, CN V and CN VII control the masticatory and mandibular muscles to tear and grind the food. In the pharyngeal stage of swallowing, the motor impulses of CNs V, VII, IX, and X cause the bolus of food to move down the esophagus (Travers, 1999). Thus, facial palsy on the NIHSS was identified as a criterion to determine the safety of performing a bedside swallow. A score of 0 (normal) or 1 (minor paralysis) was required to pass the screening criteria.

Items 9 and 10 on the NIHSS test for language and speech. On item 9, aphasia, the patient can receive a score of 0 (normal), 1 (mild-to-moderate aphasia), or 2 (severe aphasia). A patient with a score of 3 on aphasia would not meet screening criteria because this score correlates with the patient’s inability to follow commands. On item 10, dysarthria, the patient could score 0 (normal) or 1 (mild dysarthria) but not 2. A score of 2 on dysarthria would correlate with facial weakness as described previously.

Dysphagia Screening Protocol

The dysphagia screening protocol used here was developed by a neurologist, an ST, and a clinical nurse specialist from our institution’s stroke unit (Figure 1). Institutional review board approval was obtained before the study was initiated.

Most existing swallowing screening protocols are limited to the presentation of various quantities of water (DePippo, Holas, & Reding, 1992). Although these screenings demonstrate high sensitivity and specificity scores, there is a high potential to introduce a large quantity of water into a patient’s airway. Logemann, Veis, and Colangelo (1999) developed a 28-item checklist (The Northwestern Dysphagia Patient Check Sheet) that includes the presentation of 1 cc thin liquid, 1 cc pudding, and ¼ Lorna Doone cookie (if chewing is possible) as part of a comprehensive swallowing screening. Furthermore, Ramsey and colleagues (2003) outlined a swallowing screening protocol that not only assesses level of consciousness, posture, cognition, cooperation, and gross oral motor function, but also includes presentation of water, yogurt, and foods that would be considered part of a regular hospital diet.

The goal of the RN swallowing screening at our institution was to accurately identify cerebrovascular accident patients exhibiting dysphagia risk factors and to limit the time during which a patient must remain NPO. Because even a normal swallower aspirates liquids more frequently, water is not the first test item presented to the patient. Rather, a 1-teaspoon bolus of lemon ice is presented. Lemon ice provides maximum oral stimulation and heightens a patient’s oral awareness of a bolus while relieving oral cavity dryness. Even when completely melted, lemon ice remains thicker than water, thus reducing aspiration potential. Another reason for starting with lemon ice is that cold and tart items stimulate a better swallow (Daniels, 2005; Selcuk, Uysal, Aydogdu, Akyuz, & Ertekin, 2007). Once the oral cavity has been rehydrated by the lemon ice, bolus formation and oral transit become more efficient and oral residual is lessened. Therefore, a puree texture (applesauce) is subsequently presented if there are no overt signs or symptoms of aspiration with the lemon ice bolus. If the patient achieves full oral clearance after the applesauce bolus, then water can safely be tested (1- teaspoon bolus size). Because the valleculae in a typical adult can accommodate as much as 5 cc, limiting the screening bolus size minimizes aspiration of swallowing screening material into the airway.

All RNs on the stroke unit were trained to perform the dysphagia screen by the ST to ensure patient safety. This training consisted of each nurse demonstrating proficiency in administering the screening on five patients under the direct supervision of the ST. RNs were taught how to assess for prompt swallowing (i.e., feeling the patient’s throat while he or she swallowed), how to assess for cough, and how to assess for vocal quality (i.e., having the patient say “ah”). The stroke unit’s clinical nurse specialist provided the evidence-based guidelines to educate staff and to elicit participation in this study. For patients who met the NIHSS criteria listed above, the RN performed this three-swallow screen (Figure 1):

  1. Give patient 1 teaspoon of lemon ice (observe for swallow, no cough, listen for wet vocal quality).
  2. If patient passes lemon ice swallow, give patient 1 teaspoon of applesauce (observe for swallow, no cough, listen for wet vocal quality).
  3. If patient passes applesauce swallow, give patient 1 teaspoon of water (observe for swallow, no cough, listen for wet vocal quality).

A wet vocal quality after a swallow test indicates that the fluid has entered the airway and made contact with the vocal cords (Lees, Sharpe, & Edwards, 2006; Mann & Hankey, 2001; Ramsey, Smithard, & Kalra, 2003). Vocal quality was assessed by having the patient say “ah.” A cough may occur when weakness of the tongue allows fluid to fall over the base of the tongue into the unprotected airway or when a delay in the swallow reflex allows fluid to enter the airway (Lees et al.; Mann & Hankey; Ramsey et al.). Studies have found that a cough on swallow is the most reliable predictor of aspiration (Lees et al.; Mari et al., 1997; Ramsey, et al.).

After all three screenings, the RN made a recommendation regarding the advancement of the patient’s diet: either remain NPO until a formal swallow evaluation can be done by an ST or advance to a safe diet (Figure 2).

Speech Therapist Evaluation

After the RN had performed the dysphagia screening, an ST was notified to perform an independent dysphagia screening within 1 hour of the RN’s screening to minimize the possibility that differences in screening findings were due to changes in the patient’s condition. STs used the same screening tool as RNs. After this assessment, the ST made a recommendation regarding the advancement of the patient’s diet.

For this study, each nurse performed five dysphagia screenings, which were compared with the ST’s evaluation of the same patient to determine the validity and reliability of the screening.

Safe Menu

If the patient failed the screening, he or she was kept NPO until the ST could perform a formal evaluation.

If the patient passed the screen, then he or she could be offered a safe menu. The safe menu is a one-time meal to hold the patient over from 4:30 pm to 7:30 am the next morning, when an ST is on duty and a formal evaluation can be performed. The RN observes the patient eat approximately 20% of the safe menu before he or she leaves the patient alone to eat. Patients on the safe menu receive routine formal evaluation by the ST the next morning.


Eighty-three paired screenings were completed by RNs and STs. In 78 (94%) cases there was agreement on the dysphagia screening results between the RN and the ST (Figure 3). In three cases the RN had failed the patient on the swallowing screening but the ST had passed the patient, and in two cases the RN had passed a patient whom the ST had failed. A closer examination of the data showed that these two mismatched evaluations were performed by the same RN, who was later retrained.

After the results of the study were shared with hospital administration, it was decided that the screening protocol would become standard practice on the stroke unit. A before-and-after comparison was done to look at aspiration rates before and after implementation of the dysphagia protocol. Table 1 shows the results of that comparison. The rate of aspiration after implementation of the protocol was slightly lower (3.7% vs. 4.2%).


It might be argued that an ST rating should not be the gold standard against which an RN dysphagia screening tool is validated. Indeed, barium swallow, videofluoroscopy, double-contrast esophagram, radionuclide studies, and endoscopy are all unquestionably more reliable methods. However, ST evaluations are the typical means of assessing swallowing capability at this institution, and therefore our comparison most closely reflects clinical practice.

It might also be argued that the results may have been skewed by the fact that the protocol was validated using day-shift RNs only. These RNs typically are the most experienced and may have better assessment skills. However, when the protocol was extended to the night shift, no significant increase in aspiration rates was observed.

It is also possible that patients’ swallowing capability may have changed between the time of the RN and ST screening. Because it was not possible to have the ST present on the stroke unit every time a swallowing screening was performed, we tried to minimize this possibility by having the ST screening performed within 1 hour of the RN screening. Furthermore, changes in patients’ swallowing capability between screenings probably would have only decreased the validity observed here.


The results of this study confirm the validity of the dysphagia screening described in this article and show that it can be accurately performed at the bedside by RNs. Adoption of this bedside screening protocol allows the identification of patients who are able to swallow so that they can eat from a safe menu until they are formally evaluated by an ST while maintaining NPO status for those at risk for aspiration pneumonia. This has decreased the number of patients who are needlessly denied food and water without increasing the rate of aspiration pneumonia.

The generalizability of these results must be established by replicating this study in other institutions. In addition, further testing of the validity and reliability of this dysphagia screening is needed.

Implications for Practice

The dysphagia screening described in this article is most appropriately administered by RNs on a specialized stroke unit for two reasons. First, all RNs on the stroke unit are checked for interrater reliability and certified in the use of the NIHSS, which is necessary to determine eligibility for the dysphagia screen. Second, RNs outside a specialized stroke unit are less likely to encounter enough stroke patients to maintain competency in the use of the dysphagia screen.

The most challenging aspects of implementing the dysphagia protocol at this institution involved educating all hospital staff—including physicians—that NPO means absolutely nothing by mouth until the dysphagia screening is performed. It was a special challenge to change emergency department behavior from giving drugs such as aspirin orally to giving them in suppository form.


We wish to thank the RNs and STs on the stroke unit for their diligence in completing the dysphagia screenings. Partial funding for this project was provided by the Summa Foundation through the Health Services Research and Education Institute.

About the Authors

Janice Weinhardt, MSN GCNS-BC, is stroke coordinator at Summa Health System Hospital in Akron, OH. Address correspondence to her at weinharj@summa-health.org.

Susan Hazelett, MS RN, is a research associate at Summa Health Systems Hospital in Akron, OH.

Dawn Barrett, MS CCC/SLP, is a senior speech pathologist at Summa Health Systems Hospital in Akron, OH.

Robert Lada, MD, is is director of cerebrovascular medicine at Summa Health Systems Hospital in Akron, OH.

Trish Enos, BSN BS RNC CPHQ, is manager of performance improvement and informatics at Summa Health Systems Hospital in Akron, OH.

Rick Keleman, RN, is stroke unit manager at Summa Health Systems Hospital in Akron, OH.


American Speech-Language-Hearing Assocition Division 13 (Swallowing and Swallowing Disorders) Steering Committee. (2006, October 6). Frequently asked questions (FAQ) on swallowing screening: Special emphasis on patients with acute stroke. Retrieved September 16, 2008, from http://www.asha.org/NR/rdonlyres/BBDD4A46-EE96-4821-A710-D19D903E389F/0/DysphagiaFAQ.pdf.

Daniels, S. K. (2005). Treatment of post-stroke dysphagia: Evidence-based approaches. Presented at the International Stroke Conference, New Orleans, LA.

Davies, S., Taylor, H., MacDonald, A., & Barer, D. (2001). An inter-disciplinary approach to swallowing problems in acute stroke. International Journal of Language and Communication Disorders, 36(Suppl.), 357–362.

DePippo, K. L., Holas, M. A., & Reding, M. J. (1992). Validation of the 3 oz. water swallow test for aspiration following stroke. Archives of Neurology, 49, 1259–1261.

DePippo, K. L., Holas, M. A., & Reding, M. J. (1994). The Burke Dysphagia Screening Test: Validation of its use in patients with stroke. Archives of Physical Medicine and Rehabilitation, 75(12), 1284–1286.

Doggett, D. L., Tappe, K. A., Mitchell, M. D., Chapell, R., Coates, V., & Turkelson, C. M. (2001). Prevention of pneumonia in elderly stroke patients by systematic diagnosis and treatment of dysphagia: An evidence-based comprehensive analysis of the literature. ,Dysphagia, 16 279–295.

Ellul, J., Barer, D., & Fall, S. (1997). Improving detection and management of swallowing problems in acute stroke: A multi-centre study. Cerebrovascular Disease, 7(Suppl. 4), 18.

Hinchey, J. A., Shephard, T., Furie, K., Smith, D., Wang, D., Tonn, S., for the Stroke Practice Network Investigators. (2005). Formal dysphagia screening protocols prevent pneumonia. Stroke, 36, 1972–1976.

Lees, L., Sharpe, L., & Edwards, A. (2006). Nurse-led dysphagia screening in acute stroke patients. Nursing Standard, 21(6), 35–42.

Logemann, J. A., Veis, S., & Colangelo, L. (1999). A screening procedure for oropharyngeal dysphagia. Dysphagia, 14, 44–51.

Mann, D., & Hankey, G. J. (2001). Initial clinical and demographic predictors of swallowing impairment following acute stroke. Dysphagia, 16, 208–215.

Mari, F., Matei, M., Ceravolo, M. G., Pisani, A., Montesi, A., & Provinciali, L. (1997). Predictive value of clinical indices in detecting aspiration in patients with neurological disorders. Journal of Neurology, Neurosurgery, & Psychiatry, 63(4), 456–460.

Martino, R., Foley, N., Bhogal, S., Diamant, N., Speechley, M., & Teasell, R. (2005). Dysphagia after stroke: Incidence, diagnosis, and pulmonary complications. Stroke, 36, 2756–2763.

Martino, R., Pron, G., & Diamant, N. (2000). Screening for oropharyngeal dysphagia in stroke: Insufficient evidence for guidelines. Dysphagia, 15, 19–30.

Perry, L. (2001). Screening swallowing function of patients with acute stroke. Part II: Detailed evaluation of the tool used by nurses. Journal of Clinical Nursing, 10, 474–481.

Perry, L., & Love, C. P. (2001). Screening for dysphagia and aspiration in acute stroke: A systematic review. Dysphagia, 16, 7–18.

Ramsey, D. J. C., Smithard, D. G., & Kalra, L. (2003). Early assessment of dysphagia and aspiration risk in acute stroke patients. Stroke, 34, 1252–1257.

Rosenvinge, S. K., & Starke, I. D. (2005). Improving care for patients with dysphagia. Age and Ageing, 34, 587–593.

Scottish Intercollegiate Guidelines Network. (2004). Management of patients with stroke: Identification and management of dysphagia. A national clinical guideline. Edinburgh: Author.

Selcuk, B., Uysal, H., Aydogdu, I., Akyuz, M., & Ertekin, C. (2007). Effect of temperature on electrophysiological parameters of swallowing. Journal of Rehabilitation Research and Development, 44(3), 373–380.

The Stroke Group, Inc. (1998). “Quick & easy,” NIH stroke scale and tutorial program. Littleton, CO: SNR Resources.

Travers, P. (1999). Poststroke dysphagia: Implications for nurses. Rehabilitation Nursing, 24(2), 69–73.