Stroke Swallowing Screening Tool Validation

Early detection of aspiration risk in acute stroke is critical as it allows for immediate intervention, thereby reducing mortality, morbidity, length of hospitalization, and healthcare costs. Screening of swallowing prior to the administration of food, liquid or medication, including aspirin, in individuals presenting with stroke symptoms is a guideline American Heart Association/American Stroke Association. In accordance with this guideline, the Veterans Health Administration (VHA) has advocated the screening of swallowing be a quality performance measure in acute stroke. Moreover, the Office of the Inspector General recently issued VHA Directive 2006-032 mandating that the initial nurse assessment must include a screening of swallowing. In response to the directive, many VHA facilities created and implemented some version of a nursing swallowing screening tool (SST), but to our knowledge, none have been validated using an instrumental swallowing examination nor has reliability been established. In developing and establishing a valid and reproducible SST for patients with stroke, clinicians are divided on the need to include trial water swallows as part of the SST. The current notion is that administering trial swallows by disciplines without expertise in dysphagia would compromise patient safety, thus this step is opposed by many speech pathologists and nurses. The prudency of introducing non-validated, non-reproducible SSTs is questionable. The primary objective of this study is to construct a reliable and valid SST to identify risk of aspiration in individuals admitted with suspected stroke. The specific aims of this proposal are to: 1) determine if nurses can make reliable inter-rater judgments of swallowing and non-swallowing features historically used by speech pathologists to make judgments of aspiration, and 2) identify the combination of items that provide the highest level of both sensitivity and specificity in the identification of aspiration risk as measured by a videofluoroscopic swallow study (VFSS) in individuals admitted with suspected stroke. Outcomes of this research will inform as to the execution of a multi-site feasibility study on the implementation of a reliable and valid SST by nurses Methods: Consecutive individuals admitted with suspected stroke (N = 270) will be recruited to participate. Individuals will undergo screening of swallowing and a VFSS. Screening items selected for validation in the identification of risk of aspiration in patients presenting with stroke symptoms were based on extensive literature review using Cochrane and QUADAS guidelines. Eight screening items will be tested for validity and inter-rater reliability: 6 non-swallowing features and 2 swallowing features. Reliability in nursing observations of each screening item will be completed in all participants. Speech pathologists will serve as the reference standard from which to compare reliability with registered nurses who routinely work on the hospital ward with stroke patients.

Inclusion criteria

• Male and female Veterans of all ethnicities admitted to MEDVAMC with a new suspected ischemic or hemorrhagic stroke will be eligible to participate.
• Individuals with a history of prior strokes without dysphagia will be eligible to participate.
• Patients must be medically stable as determined by the attending neurologist.
• Participants with language or cognitive deficits who are judged by the attending neurologist to not have capacity to provide informed consent will be eligible to participate but they must have an authorized representative available within 24 hours of admission to provide consent.

Exclusion criteria

• Individuals with a history of neurological disease other than stroke, head and neck structural surgery, or history of dysphagia unrelated to the current stroke will be excluded from participation.
• Individuals who are obtunded (unable to maintain wakefulness), currently diagnosed with pneumonia, or on ventilator support will be excluded.
• Veterans who are more than 72 hours past MEDVAMC admission will be excluded.