Study Design

This was a prospective cross-sectional study where the GUSS was translated into Swedish and validated in a Swedish stroke care context. The study was approved by the Swedish Ethical Review Authority (Dnr: 2019–06546, 2021–05043, 2021-06701-02) and conducted following the STARD guidelines [19], see appendix A and the Helsinki Declaration [20]. The study consisted of two parts. In the first part, the GUSS was translated into Swedish to evaluate the content validity of the translated screening tool (see Fig. 1). In the second part (see Fig. 2), the criterion validity of the translated GUSS was explored by comparing the results of screening with FEES results. Further, to obtain convergent validity, the GUSS results were compared with (a) the Functional Oral Intake Scale (FOIS) [21] as per the comprehensive FEES assessment, (b) the Standardized Swallowing Assessment (SSA) [11,12,13], and (c) the National Institutes of Health Stroke Scale (NIHSS) [22, 23]. To evaluate the inter-rater reliability of the translated GUSS, it was used independently by different healthcare professionals in examinations conducted within a period of 2 h. A further aim was to compare the SSA results with the FEES results, since the SSA screening tool has, to the best of the authors’ knowledge, not been validated against the gold standard assessment in an acute stroke population.

Fig. 1

Flow chart of translation procedure. Abbreviations SLP = Speech Language Pathologist, ENT = Ear, Nose and Throat Doctor

Fig. 2

Flow chart of validation procedure. Abbreviations GUSS-S = The Gugging Swallowing Screen (Swedish), SLP = Speech-Language Pathologist, SSA-S = Standardized Swallowing Assessment (Swedish), FEES = Flexible Endoscopic Evaluation of Swallowing

Translation Methodology

The GUSS was translated using a forward-backward multistep approach as recommended by the World Health Organization [24]. The translation process consisted of seven phases. Continuous documentation of all the phases was made during the translation process.

Phase 1: Forward Translation

In the first phase, the forward translation from English into Swedish, two translators actively working in the field of dysphagia who were proficient and fluent in written English, independently translated the GUSS into Swedish.

Phase 2: Expert Panel

Phase two consisted of a review of the two translated versions by an expert panel of nine healthcare professionals working in the field of dysphagia (four SLPs, two nurses, two dietitians, and one otolaryngologist) to identify any discrepancies between, and specific difficulties with, the original scale and the two parallel translations.

Phase 3: Review

In the third step, the two forward translators and another reviewer who had not participated in the expert panel, reviewed the expert panel’s opinions of the two parallel translations in relation to the original version of GUSS. The goal of the third step was to decide on a final version of GUSS-S. The review group and the last author made the final decision regarding the consensus version of the GUSS-S.

Phase 4: Back-translation

A bi-lingual translator with experience of working with dysphagia and blinded to the original version of the GUSS back-translated the final version of GUSS-S.

Phase 5: Review

The back-translation was then compared with the original version of the GUSS by the same review panel as in Phase 3. The goal of the comparison was to identify any discrepancies between the original version and the back translation.

Phase 6: Pretesting

The GUSS-S was pretested in two ways. First, an expert panel of twelve SLPs with experience of working with dysphagia were recruited via convenience sampling. The panel members each rated the translation regarding relevance for screening dysphagia in a Swedish context to enable calculation of the content validity of the GUSS-S [25]. The ratings were collected via a digital survey platform. The SLPs rated seven items of the translated screening tool on a scale from 1 (not relevant) to 4 (very relevant). The ratings were used to measure inter-rater agreement through the calculation of a content validity index (CVI) [25]. Item-CVI (I-CVI) and CVI for the screening (Scale-CVI, S-DVI/Ave) were calculated according to convention. For each item, the number of experts who rated an item as a 3 or 4 was divided by the number of experts in total (n = 12). This gave the proportion in agreement concerning the relevance of each item. Items with an I-CVI of 0.78 or higher are considered to have excellent content validity when the panel consists of 3 or more experts [26]. All items reached excellent content validity since all I-CVI surpassed the limit of 0.78, with values varying between 0.83 and 1.00. Four items were rated with an I-CVI of 1.00, which means that all experts rated the items as relevant for screening dysphagia. S-CVI/Ave was computed by calculating the mean value of all I-CVI. S-CVI/Ave for the screening as a whole was 0.95, which is deemed excellent [25]. These results indicate that the GUSS-S has good-to-excellent CVI per item and an excellent average CVI. There was, therefore, no indication that further revision was needed before using the screening in a Swedish clinical care context.

The GUSS-S was then tested independently by a nurse and a SLP in the screening of five patients admitted to a stroke unit.

Phase 7: Validation

The final phase included validating the GUSS-S in a Swedish stroke care context.

The validation process is described in detail in Fig. 2.

Validation MethodologyParticipants

The validation was carried out on the stroke care unit at Umeå University Hospital. During the study period, a total of 88 patients were eligible to participate. Inclusion criteria were patients diagnosed with a stroke (i.e., ischemic stroke or intracerebral hemorrhagic stroke) or Transient Ischemic Attack (TIA), with no history of dysphagia, and that the assessments could be performed within 96 h of stroke/TIA onset. Eight patients were excluded due to (a) not being diagnosed with a TIA or stroke (n = 4), (b) not giving consent to be examined with FEES (n = 2), and (c) having an unstable medical condition that posed a contraindication for FEES (n = 2). A total of 80 patients were included in the validation process. Before inclusion, the patients signed an informed consent form or agreed to participate verbally if unable to sign by hand. See Fig. 3 for a flow-chart of the subject inclusion and data collection.

Fig. 3

Flow-chart of subject inclusion and data collection, all performed within two hours. Abbreviations GUSS-S = The Gugging Swallowing Screen (Swedish), FEES = Flexible Endoscopic Evaluation of Swallowing, SSA-S = Standardized Swallowing Assessment (Swedish)

GUSS Assessment

The patients were screened using GUSS-S within 96 h of admission to the stroke care unit. The GUSS-S screening was performed in accordance with the original publication by Trapl et al. [15]. The GUSS is divided into two parts: a preliminary assessment and a direct swallowing test consisting of three subtests that are performed sequentially. Each subtest gives a maximum of five points; a score of five points must be achieved to continue to the next subtest. A patient can attain a maximum of 20 points, which denotes normal swallowing ability. Nineteen to 15 points indicates mild dysphagia with a low or minimal risk of aspiration, 14 to10 points indicates moderate dysphagia with risk of aspiration, and 9 to 0 points indicates severe dysphagia with a high risk of aspiration.

Patients were assessed while sitting in an upright position. The preliminary assessment evaluated patients regarding vigilance, voluntary cough, throat clearing, and saliva swallowing. Those who scored five points in this assessment moved on to the second part. Those who scored below five were noted as having a risk of aspiration and testing was terminated. The second part started with a semi-solid swallowing trial where the patient was given water thickened to a consistency of 3 on the IDDSI framework [27]. The patient was first given half a teaspoon followed by an additional 3–5 teaspoons while the investigator observed the patient closely and scored according to the presence or absence of the following signs of aspiration: delayed (> 2 s) or impaired swallowing, voice change, drooling, and cough. If no aspiration signs were detected, the investigator would score accordingly and move on to liquid swallowing. The patient was given increasing amounts of water (IDDSI 0 [27]), starting with 3 ml and, after successful swallowing, continuing to 5, 10, 20 and 50 ml. In the final solid swallowing trial, the patient was instructed to ingest a small piece of dry bread (IDDSI 7 [27]); the time limit for delayed swallowing was ten seconds.

In this study, GUSS-S screening was first performed in 65 of 80 patients by an SLP and in 15 patients by a nurse. In 30 participants of the 65 screened by an SLP, a nurse performed a second screening with GUSS-S within 2 h of the first screening to assess inter-rater reliability. All performers of the GUSS-S were aware of clinical information about the patient from the medical records, but were blinded to all other assessments carried out in the validation procedure.

FEES Assessment

FEES was used as the reference standard for validating the screening tool since it is considered (a) the gold standard for detecting dysphagia and (b) can be performed at bedside [28,29,30]. The FEES assessment was performed at bedside within 2 h of the GUSS-S screening on the stroke unit according to a standardized protocol by Langmore [31]. Two experienced SLPs performed the FEES. The examiners were blinded to the patient’s GUSS-S results and their medical records. Before the FEES assessment, one SLP performed the SSA [11,12,13] for convergent validity testing. The other SLP, blinded to the SSA result, performed the FEES and assessed the patient’s swallowing function. The FEES examination was carried out using a rhino-laryngo video endoscope (ENF-VH, Olympus), a light source (Visera Elite CLV-S190, Olympus), a video processor (OTV-S190, Olympus) and a video monitor (Olympus OEV143). Three bolus consistencies were tested during the FEES, the same as used in the GUSS-S screening: levels 0, 3 and 7 according to the IDDSI framework [27]. The bolus was dyed with green food coloring to enable visibility during the examination. Three teaspoons and three tablespoons of semi-solids (IDDSI 3) and liquid (IDDSI 0) were given in a consecutive order. The patient was then instructed to drink 2–3 mouthfuls of liquid (IDDSI 0) and to masticate and swallow a piece of cracker (IDDSI 7). The protocol was modified or shortened if the SLP determined that the patient’s safety was at risk due to significant aspiration or other severe dysphagia symptoms.

Based on the FEES findings, aspiration risk was classified according to the Penetration-Aspiration Scale (PAS) [32] with a cut-off point of ≥ 5, as per the original publication [15]. The presence of dysphagia was classified according to the Fiberoptic Endoscopic Dysphagia Severity Scale (FEDSS) scoring system [33], as per the revisited publication by Warnecke et al. [17]. The cut-off point of FEDSS > 1 was defined as presence of dysphagia. Based on the comprehensive FEES assessment, the examiner also rated the patient’s ability regarding intake of food and drink according to FOIS.

All assessments (GUSS, SSA, FEES, including FOIS) were performed within a period of 2 h and within 96 h of admission to the stroke unit (see Fig. 2).

SSA-S Assessment

Similarly to the GUSS-S, the SSA-S consists of a preliminary investigation and a secondary direct swallowing test. In the first part, the patient’s vigilance and ability to sit upright and maintain adequate control of the head is assessed. The preliminary investigation also evaluates voluntary cough, control of saliva, mobility of the tongue, breathing, and quality of voice with subsequent termination of screening and SLP referral if one aspect or more is noted to be impaired. In the direct swallowing test, the patient is given teaspoons of water while the investigator observes the following signs of aspiration: absence of swallowing movements, drooling, coughing or shortness of breath, and “gurgly” voice after swallowing. If no signs of aspiration are noted, three teaspoons of water are administered, and the screening ends with the patient drinking half a glass of water.

Statistical Analysis

All statistical analyses were performed using SPSS 28.0 Statistical Package for the Social Sciences [34]. Criterion validity was calculated for sensitivity and specificity in (a) percent and (b) based on the receiver operating characteristic curves (ROC), with the area under the curve being calculated. The analyses were based on the cut-off set of GUSS ≤ 14 points and PAS ≥ 5 for aspiration. For the presence of dysphagia, the analyses were based on the cut-off set of GUSS ≤ 19 points and FEDSS > 1. Convergence validity was calculated with Spearman rank correlation by analyzing the correlation between GUSS scores with PAS, FOIS, SSA, and NIHSS. Inter-rater reliability was analyzed using Cohen’s weighted kappa with quadratic weights. Sensitivity and specificity calculations were also performed in percent for dysphagia (yes/no), identified with SSA, compared with FEES, and assessed with the PAS cut-off ≥ 5.