The design and evaluation of a valid dysphagia screening tool for acute stroke patients

  • Kathryn Head

    Student thesis: Doctoral Thesis


    Screening acute stroke patients for dysphagia (difficulty swallowing) is recommended within 24 hours due to risks of morbidity and mortality. A review of the international literature identified no universal consensus for a valid method of screening. This thesis describes a multi-method Action Research (AR) programme of study focused on the design, development and evaluation of a reliable and valid dysphagia screening tool (the ‘Head Dysphagia Screen for Stroke’ or HeDSS) for use by Registered General Nurses (RGNs). As a component of the assessment phase of the AR programme, a survey of dysphagia screening practices in England and Wales highlighted widely varied screening practices. Many of these practices were based on limited research evidence, reflecting the lack of consensus for valid dysphagia screening criteria reported in the literature. The design phase of the AR programme involved the development of the HeDSS tool, which centred on the use of research-based screening criteria. Focus group activity determined nurses’ perceptions of the design and subsequent refinement of the HeDSS tool. The intervention and evaluation phases of the AR programme followed three empirical stages. Stage one established the inter-rater reliability of the Speech and Language Therapist Researcher’s (SLTR’s) clinical dysphagia assessment, which acted as a reference standard against which the validity of the HeDSS tool was to be measured. Clinical judgements for the presence and absence of dysphagia in the same 30 referred patients were compared between the SLTR and a Speech and Language Therapist (SLT) of equivalent experience. Inter-rater reliability was substantial (k = .71). The second empirical stage established inter-rater reliability of the HeDSS measurement outcomes (indicative signs of dysphagia and appropriateness of referral for SLT clinical dysphagia assessment) when employed by two RGNs compared against the SLTR when screening two samples of 20 acute stroke patients. Rater agreement was substantial (k = .71 and k = .79, for detection of signs of dysphagia and k = .79 and k = .87 for appropriateness of referral). The final empirical stage evaluated the concurrent validity of the HeDSS tool measurement outcomes when employed by a second sample of two RGNs compared with the SLTR’s clinical dysphagia assessment outcomes in a sample of 100 acute stroke patients. The HeDSS tool measurement outcomes correlated highly with the clinical dysphagia assessment outcomes (sensitivity .88 - .96 and specificity .85 - .88 for detection of dysphagia; sensitivity .90 - .96 and specificity .84 - .88 for determining patients appropriate for assessment). Correlation coefficient measures confirmed high concurrent validity for the HeDSS tool (Phi ranged between .76 - .82). This study is the first in the UK to establish a reliable and valid dysphagia screening tool for use with acute stroke patients and has significantly advanced the professional knowledge base within this domain of practice. It is recommended that a multi-centred programme of research be undertaken to replicate this study with a larger nurse and patient sample.
    Date of Award2010
    Original languageEnglish


    • screening
    • strokes
    • patient care
    • dysphagia

    Cite this