Journal of Neuroscience Nursing:
RN Dysphagia Screening, a Stepwise Approach
Courtney, Bobby A.; Flier, Lisa A.
Lisa A. Flier, RN MSN, is an acute care nurse practitioner at Associated University Neurosurgeons, Illinois Neurological Institute at OSF Saint Francis Medical Center, Peoria, IL.
Questions or comments about this article may be directed to Bobby A. Courtney, MA, at email@example.com. He is a JD/MPH candidate at the Indiana University School of Law, Indianapolis, IN.
According to the Joint Commission, dysphagia, which occurs in anywhere from 27% to 50% of stroke patients and often leads to aspiration, is a significant quality/safety indicator. As such, it is identified as a standard of performance when awarding disease-specific certification in stroke care. Application for this certification required one institution to assign a multidisciplinary process improvement team to develop and implement a dysphagia screening protocol to replace their current practice, which was being used on an inconsistent basis. This article outlines the steps necessary to identify and tailor a tool for dysphagia screening, develop a protocol to ensure appropriate use of said tool, and use a multiphase implementation process to identify strengths and weaknesses as well as create caregiver support for the practice change.
According to the American Heart Association, stroke accounts for more than 1 of every 16 deaths in the United States, is ranked number three among all causes of death, and is a leading cause of serious, long-term disability (American Heart Association [AHA], 2008). Moreover, approximately 700,000 people experience a stroke each year at an estimated cost of more than $50 billion in direct and indirect costs (Joint Commission on the Accreditation of Healthcare Organizations [JCAH], 2007). As time is a significant factor in reducing the adverse effects of stroke, both pre-and postdiagnosis, a great deal of research has led to clinical practice guidelines and evidence-based protocols, many of which have been incorporated into the performance measure criteria of the Certification Program for Disease-Specific Care (DSC) of the Joint Commission on the Accreditation of Healthcare Organizations (JCAH)(JCAH, 2007).
In 2004, the OSF Saint Francis Medical Center began the process of applying for JCAH's DSC certification in stroke care. This Certificate of Distinction recognizes primary stroke centers that make "exceptional efforts to foster better outcomes for stroke patient care" (JCAH, 2007). In addition, "achievement of certification signifies that the services provided have the critical elements to achieve long-term success in improving outcomes" (JCAH, 2007). The program was created in conjunction with the American Stroke Association and is primarily derived from the Brain Attack Coalition's "Recommendations for the Establishment of Primary Stroke Centers" (JCAH, 2007).
Standards for performance measurement and improvement of DSC certification programs focus on four key areas: (1) creating an organized, comprehensive approach to performance improvement; (2) using comparative data to evaluate program processes and patient outcomes; (3) evaluating participants' perception of care quality; and (4) maintaining data quality and integrity (JCAH, 2007). The performance measurement requirements for DSC certification programs also include two stages: nonstandardized and standardized. The former requires collection and analysis of four or more performance measures, two of which should be evidence-based, clinical measures related to, or identified in, clinical practice guidelines for that program or service. The latter have precisely defined specifications and standardized data collection protocols, meet established evaluation criteria, and can be uniformly adopted for use (JCAH, 2007).
The problem of dysphagia was initially raised at the Medical Center by Dr. David Wang, a faculty neurologist involved in the collection of stroke care data for a study conducted by Hinchey et al. (2005). Because it occurs in approximately 27% to 50% of stroke patients, and of that population, 43% to 54% will experience aspiration (37% leading to pneumonia), dysphagia is a significant quality/safety indicator (JCAH, 2007). The measure itself was derived from an evidence-based clinical practice guideline developed by the Agency for Health Care Policy and Research in 1995 (Agency for Healthcare Research and Quality [AHRQ], 1995). Although no AHRQ recommendations have been made regarding a specific screening modality, current research indicates that formal dysphagia screening protocol, inclusive of a swallow assessment tool, can decrease the risk of pneumonia in patients hospitalized for ischemic stroke threefold (Hinchey et al., 2005). Given the aforementioned 700,000 annual strokes, it has been suggested that a formal dysphagia protocol could save more than 8000 lives and prevent nearly 40,000 cases of pneumonia per year (Hinchey et al., 2005).
Achieving DSC certification at the Medical Center required focused attention on process improvements. Oversight for this endeavor was provided by the Stroke Continuous Process Improvement (CPI) team, which identified several initial concerns. To begin with, team members questioned what qualifications would be necessary for providers conducting a dysphagia assessment. Would a specific skill level or competency be required? In addition, what factors would be included in the tool to accurately identify dysphagia, and how would the tool's efficacy be determined?
When compared with videofluoroscopy, Ramsey, Smithard, and Kalra (2003) noted variable sensitivities, specificities, and predicative values among water-based bedside swallow tests (Ramsey, Smithard, & Kalra, 2003, pp. 1253-1254). As a result, they indicated a need for more reliable screening methods to be used by a range of frontline caregivers, particularly those who are in earlier contact with stroke patients and capable of reducing the number of patients fed while awaiting a more formal assessment (Ramsey et al., 2003). After careful evaluation, the CPI team identified nursing as the common denominator throughout the patient care continuum, and although these caregivers were already feeling pressured for higher productivity and higher patient satisfaction, another protocol meant placing further demands on their current workload. As such, it was determined that a subcommittee would be formed to address this specific issue. Committee members included RN leadership from the CPI team as well as general neurology nurses, neuro critical care nurses, neurology nurse educators, the stroke nurse coordinator, several speech-language pathologists (SPLs), and a stroke neurologist.
Because there was no consistent staff bedside examination/tool used to assess for dysphagia before administration of any oral food or fluids and the assessments being conducted were sporadic with no clearly identified patient criteria, the CPI subcommittee began a literature review and networking through the American Association of Neuroscience Nurses stroke listserv for evidence-based information and tools currently in use. As a result, a meta-analysis was conducted and several studies were reviewed (Tables 1 and 2). Although several published tests existed, such as patient questionnaires/interviews, gag reflex evaluation, observation for overt clinical signs, and decision-making algorithms (American Speech-Language-Hearing Association [ASLHA], 2006), but as a group, and as the Medical Center's individual speech pathologists noted, silent aspiration remained a problem. Although its efficacy had been well documented (DePippo, Holas, Reding, Mandel, & Lesser, 1994), the widely used Burke Dysphagia Screening Test appeared just as problematic because water's lack of color made it difficult to determine whether a patient had drooled, sputtered, or delayed the swallow. Given the lack of an appropriate tool that would specifically fit the Medical Center, a rough plan was developed following a basic model outlined by Sioux Valley Health System in South Dakota, as identified in the study conducted by Hinchey et al. (2005).
The Sioux Valley swallow assessment used a tool called the EATS box (Examine Ability to Swallow) and an accompanying questionnaire to evaluate a patient's swallow; this tool was initially developed by clinicians at Iowa Methodist Medical Center (Wood & Emick-Herring, 1997). The tool was also foodstuff based, making it distinctly different from the traditional water test and therefore particularly appealing to the CPI team. Several key components in the Sioux Valley EATS box were changed as a result of the availability of specific foodstuff as well as the lack of storage space on each unit. New items were selected by the CPI subcommittee's SPL, with special consideration paid to the inclusion of foods with varying consistencies-items that the patient would need to be able to swallow to have any kind of a safe diet. Bold flavoring was also chosen in an attempt to stimulate the desire to swallow, yet remain easy for the nurse to administer. Final items included one container of applesauce, 6 ounces of cranberry juice, one package of graham crackers, and one cup, spoon, and napkin.
The Sioux Valley assessment form was also altered in several ways to better suit the needs of the Medical Center. To begin with, the original protocol allowed for a 24-hour assessment window, which was not appropriate to our institution. Given the acuity level of patients treated, it was determined that assessment should be completed as soon as possible rather than waiting until the patient required oral medication or food. Moreover, the form did not include an initial visual assessment of the patient. Because this permitted the elimination of any unnecessary food intake, it was added to the revision.
The initial protocol developed for use at the Medical Center involved a three-step process that took only minutes to complete and did not require a physician's order. Step 1 required a basic visual assessment of the patient (e.g., is the patient drooling, unable to make a sound, intubated, or recently extubated; indicate difficulty swallowing). If all visual assessment items were negative, a caregiver would obtain the EATS box food items.
Step 2 required elevation of the patient's head of bed to at least 70 degrees (supporting hemiplegic patients with a pillow to ensure an upright position) and serving of the EATS box food items in the following order: (1) one teaspoon of applesauce and watch the swallow, then a second teaspoon if the first is successful; (2) if applesauce is successful, serve a teaspoon-sized sip of cranberry juice and allow the patient to take a second sip independently if the first is successful; (3) if cranberry juice is successful, serve a bite of a graham cracker and serve the rest of the cracker if the swallow is successful.
As the patient is fed the aforementioned foodstuff, it is critical that the caregiver observe the patient's chew and swallow for any of the following signs:
* Food/drink held in the mouth without trying to swallow
* Food/drink is spat out
* Facial droop with bite/drink coming out of mouth
* Food/drink is pocketed in mouth
* Suctioning is required
* Food/drink comes out of the nose
* Choking or persistent coughing starts
* Eyes redden or tear
* Voice is wet
* Breathing is labored/affected
* Patient states that the swallow is painful
If any of these indicators occur at any point while administering the EATS box food items, the assessment is stopped and the patient "fails" the assessment.
Upon completion of the assessment, step 3 required physician notification, speech therapy consult, and nil per os (NPO) for "failed" patients. The protocol for those patients who passed required physician notification and initiation of a mechanical soft diet with thin liquids (advanced as tolerated). The major benefit to this "pass/fail" system was that patients who passed would no longer be delayed feeding and those who were questionable were automatically designated NPO and referred for further swallow evaluation by the SPL. SPLs and nurses work closely together to achieve the same goals through the use of this protocol; however, patient safety and the need for professionals to remain in their scope of practice create limitations for nurses regarding the evaluation of dysphagia in stroke patients. It must be understood that a pass/fail swallow screen by nursing does not replace a complete bedside swallow evaluation by the SPL; rather, it allows nurses to identify abnormalities in a patient's ability to swallow and follow up as they would with any other identified problem.
A pilot project was initiated with a small group of nurses on the Medical Center's Neuroscience Critical Care Unit (NCCU) because this was the typical admission placement for most stroke patients. Over the course of 3 weeks, variations of the tool were used to solicit input on limitations, concerns, and overall effectiveness. Based on feedback received, the tool was revised several times, and the entire NCCU staff, physicians, and SPLs were educated on its purpose and use. For caregivers to become familiar with the new tool, the effort continued to be labeled a pilot program and required screening of every admission to the unit. After 5 months of implementation on the NCCU, staff input was once again requested, the tool was revised for easier readability, and education was repeated. Although most new stroke diagnoses would not be admitted outside of the NCCU, general practitioners often admit their own patients to general units under the diagnosis of "weakness/rule out stroke" or "altered mental status/rule out stroke." As such, the next protocol revision was implemented on the general neurology unit to assess potential barriers in units with mixed patient populations before housewide implementation. This allowed project coordinators to review caregivers' access to the assessment tool and educational information.
After staff working on the general neurology unit was comfortable and skilled with the tool's use, plans to implement the protocol housewide were outlined. Because this was a quality improvement project and did not entail human subject testing, formal institutional review board authorization was not required. However, before a practice change can be initiated at the Medical Center, approval must be received from the organization's Professional Nursing Congress, which serves as the body through which new internal nursing standards are established, upheld, and communicated. Items required for review and approval of a new practice include
* Detailed explanation of the proposed change
* Qualifications of the individual(s) requesting the change
* Background information supporting the need for a change
* Benefits or justification related to the expected outcomes
* Human resources impact
* Cost factors
* Supporting data
* Implementation plan complete with updated policies, procedures, protocols as well as an evaluation and education plan
Upon approval of the practice change, detailed instructions for use of the assessment tool were made available to caregivers (Fig 1), and the hospital policy manual was updated (Fig 2). A Six Sigma/DMAIC (Define, Measure, Analyze, Identify, and Control) model (Fig 3) was used to implement the practice change because it is used at the Medical Center to reduce human variation in particular processes and it significantly reduces the amount of unacceptable outcomes. The dysphagia screen tracking log was also set up in this format.
Given the difficulty in differentiating between the three diagnoses, further education was developed in the form of targeted in-servicing of core nurses, presentations to clinical educators, and a training session available via the Medical Center's intranet that could be submitted for continuing education credit. The online session was only made mandatory for the Emergency Department because they were most likely to care for stroke patients before admission. The next barrier to implementation of the dysphagia screening protocol hospitalwide was to shrink the tool into a single-page document that included all assessment findings and directions.
Initial monitoring on the pilot unit rarely reached 100% on a regular basis, and with the addition of the general hospital units, even more uncertainty remained. A weekly hospitalwide review of all patient charts was reviewed if the admitting diagnoses indicated stroke as a possible source of the patients' "weakness." Frequently, the assessment was not completed. As a result, the CPI subcommittee provided one-on-one education to the nurse caring for that patient as well as other staff on the unit. It was determined that the protocol affected not just the nurses, but the technicians and unit secretaries, and no one should be feeding the patient or ordering food for the patient before completion of the assessment. One month after housewide implementation, it was determined that both formal and informal in-services would be necessary to promote understanding of the tool and its importance and ensure compliance.
Specific Case Examples
Housewide implementation of the protocol had varied levels of compliance; however, numerous case examples confirmed that the CPI subcommittee's work was not without merit.
A male patient was admitted to the NCCU with dense right hemiparesis. The patient was incredibly thirsty, and his speech was thick but intelligible. A resident physician was sent by the attending to assess swallow ability with a cup of water. Finding no difficulty and noting that the patient did not even cough, the resident approved oral administration of medication/food. As nursing staff were aware of the new dysphagia screening protocol, a second test was conducted using the new tool. After identifying the patient's profound facial droop and drooling, he immediately failed the test, and an SPL was consulted. The SLP confirmed the nurse's findings and directed that the patient initially be kept NPO, with advances in food and fluids as his neurologic status changed. This example clearly indicated that additional education was necessary for physician and resident medical staff regarding the protocol, tool, and its effectiveness.
A female patient was admitted to the NCCU with dense right hemiparesis. Her speech was thick but intelligible, and a slight facial droop was recorded; however, no drooling of oral secretions was evident. Oral medication was ordered by the physician, and a bedside swallow assessment was performed by the nursing staff before administration of medication. Findings indicated that the patient was able to tolerate all levels without any symptoms; as a result, she passed the test. A mechanical soft diet was initiated with thin liquids as recommended in the new protocol; however, an SLP was consulted for speech therapy issues. Upon consultation, the SLP reassessed the patient's swallow for quality control and the nurse's assessment was confirmed. This example helped validate the utility and clinical importance of the protocol and tool to caregivers and patients; in addition, it helped to illustrate and support the improved quality and safety standards put in place as a result of applying for JCAH DSC stroke certification.
Education and training regarding safe swallow had originally been provided throughout the Medical Center as a safety issue for stroke patients in any number of settings. However, development of this tool has allowed for improvement in patient assessments and direction for appropriate follow-up. Project administrators are continuing to monitor for compliance regarding completion of the assessment before oral intake in all stroke patients as well as accuracy of administration. There has also been interest on the part of Medical Center administration to initiate use of the assessment for elderly patients regardless of the diagnosis and nurse discretion for any patient who "looks" impaired, including those who were intubated for a length of time that has not yet been determined.
Other organizations have also expressed interest in the Medical Center's dysphagia screening efforts. As such, the protocol and tool have also been distributed via "Get With the Guidelines" (GWTG), a premier hospital-based quality improvement program of the American Heart Association and the American Stroke Association, which empowers providers to consistently treat heart and stroke patients according to the most up-to-date guidelines. In June 2007, project coordinators were invited to a national GWTG teleconference to discuss use of the tool, which led to requests from other institutions across the country. Moreover, the tool has been informally disseminated via conferences, Web sites, and other meetings of professional organizations such as the American Association of Neuroscience Nurses, the American Board of Neuroscience Nursing, the National Stroke Association, and the American Stroke Association/American Heart Association.
Dysphagia assessment in ischemic and hemorrhagic stroke patients and those who have experienced a transient ischemic attack has proven efficacy with regard to quality and safety; however, the varying assortment of tools and protocols available to caregivers can be overwhelming. By using a multidisciplinary team to evaluate the applicability of the various alternatives, it is possible to tailor make a swallow assessment that is designed around the specific institution's practice guidelines and tiered in its evaluative criteria. Perhaps more importantly, accessibility of the tool (e.g., foodstuff kept on the unit), its ease of use (e.g., single page, pass/fail assessment), and increased sensitivity (e.g., food based) make the tool particularly applicable to front-line neuro nursing staff at the bedside-a tool developed and implemented by nurses for nurses. Preliminary postimplementation research and anecdotal evidence from clinicians and Medical Center administrators have shown that this method can also be effective in creating staff participation and commitment to a new or updated process.
The authors thank the staff at OSF Saint Francis Medical Center for their contribution to the development and implementation of this assessment protocol. In addition, the EATS tool, created by Iowa Methodist Medical Center, along with the assistance of clinicians at Sioux Valley Health System, is acknowledged and appreciated.
Agency for Healthcare Research and Quality (formerly Agency for Health Care Policy and Research). (1995). Post-stroke rehabilitation guideline
(AHCPR Publication No. 95-0062: Clinical Guideline Number 16). Washington, DC: U.S. Government Printing Office.
American Heart Association. (2003). Heart disease and stroke statistics-2008 Update
. Dallas, TX: Author.
American Speech-Language-Hearing Association. (2006). FAQ on swallowing screening: Special emphasis on patients with acute stroke
. Retrieved August 6, 2007, from http://www.asha.org
Broadley, S., Croser, J., Creevy, M., Teo, E., Yiu, D., Pathi, R., et al. (2003). Predictors of prolonged dysphagia following acute stroke. Journal of Clinical Neuroscience
Daniels, S., Ballo, L., Mahoney, M.C., & Foundas, A. (2000). Clinical predictors of dysphagia and aspiration risk: Outcome measures in acute stroke patients. Archives of Physical Medicine and Rehabilitation, 81, 1030-1033.
DePippo, K., Holas, M., Reding, M., Mandel, F., & Lesser, M. (1994). Dysphagia therapy following stroke: A controlled trial. Neurology
Garon, B., Engle, M., & Ormiston, C. (1995). Reliability of the 3-oz water swallow test utilizing cough reflex as sole indicator of aspiration. Neurorehabilitation and Neural Repair, 9, 121-126.
Garon, B., Engle, M., & Ormiston, C. (1996). Silent aspiration: Results of 1,000 videofluoroscopic swallow evaluations. Neurorehabilitation and Neural Repair, 10, 121-126.
Hinchey, A., Shephard, T., Furie, K., Smith, D., Wang, D., & Tonn, S. (2005). Formal dysphagia screening protocols prevent pneumonia. Stroke
Joint Commission on the Accreditation of Healthcare Organizations. (2007). Stroke performance measurement implementation Guide
(2nd ed.). Oakbrook Terrace, IL: Author.
Koopman, W., Wiebe, S., Colton-Hudson, A., Moosa, T., Smith, D., Bach, D., et al. (2004). Prediction of aspiration in myasthenia gravis. Muscle and Nerve, 29, 256-260.
Lawrence, E., Coshall, C., Dundas, R., Stewart, J., Rudd, A., Howard, R., et al. (2001). Estimates of the prevalence of acute stroke impairments and disability in a multiethnic population. Stroke
Martineo, R., Foley, N., Bhogal, S., Diamnat, N., Speechley, M., & Teasell, R. (2005). Dysphagia after stroke: Incidence, diagnosis, and pulmonary complications. Stroke, 36, 2756-2763.
Paciaroni, M., Mazzotta, G., Corea, F., Caso, V., Venti, M., Milia, P., et al. (2004). Dysphagia following stroke. European Neurology
Parker, C., Power, M., Hamdy, S., Bowen, A., Tyrrell, P., & Thompson, D. (2004). Awareness of dysphagia by patients following stroke predicts swallowing performance. Dysphagia
Perry, L. (2001). Screening swallowing function of patients with acute stroke. Part one: Identification, implementation and initial evaluation of a screening tool for use by nurses. Journal of Clinical Nursing, 10, 463-473.
Perry, L. (2001). Screening swallowing function of patients with acute stroke. Part two: Detailed evaluation of the tool used by nurses. Journal of Clinical Nursing
Ramsey, D., & Smithard, D. (2004). Assessment and management of dysphagia. Hospital Medicine
Ramsey, D., Smithard, D., & Kalra, L. (2003). Early Assessments of dysphagia and aspiration risk in acute stroke patients. Stroke
Sioux Valley Hospital. (2004). Nutrition and swallowing screen tool. Sioux Valley, SD: Author.
Tohara, H., Saitoh, E., Mays, K., Kuhlmeier, K., & Palmer, J. (2003). Three tests for predicting aspiration without videofluorography. Dysphagia
Wood, P., & Emick-Herring, B. (1997). Dysphagia: A screening tool for stroke patients. Journal of Neuroscience Nursing
© 2009 American Association of Neuroscience Nurses