**Background:** Post-extubation dysphagia (PED) is common in ICU patients, affecting approximately 62% of those who are intubated and mechanically ventilated. PED is associated with higher risk of pneumonia, prolonged parenteral nutrition, reintubation, longer ICU and hospital stays, decreased quality of life, and increased mortality up to one year after ICU admission. Despite established dysphagia screening protocols for stroke patients, no standardized protocols exist for the ICU. The Gugging Swallowing Screen (GUSS) was originally developed and validated for acute stroke patients; this study aimed to modify it for the ICU population and validate the new version (GUSS-ICU) against the gold-standard flexible endoscopic evaluation of swallowing (FEES).
**Methods:** This prospective, monocentric study was conducted at the University Hospital Basel, Switzerland, between September 2020 and February 2021. Forty-five patients who had been intubated for at least 24 hours were consecutively recruited. Inclusion criteria were age ≥18 years, minimal mental status test score of 24, and extubation at least one hour prior. The modified GUSS-ICU was performed twice by two independent speech and language therapists (SLTs). Concurrently, FEES was performed by an otorhinolaryngologist. All three assessments were conducted within a three-hour period in a randomized order determined by computer-generated randomization, and all testers were blinded to each other's results. The GUSS-ICU consists of an indirect swallow test (6 items including Richmond Agitation Sedation Scale 0 to +2, stridor assessment, cough effectiveness, drooling, voice change) and a direct swallow test with four sequential subtests using IDDSI-defined consistencies (moderately thick, thin, solid, mixed solid-liquid). FEES results were graded using the Penetration Aspiration Scale (PAS) and classified into a 4-grade dysphagia severity scale by Warnecke. Dysphagia was defined as FEES grade >0 (PAS ≥1) and GUSS-ICU <10 points.
**Key Results:** The mean age was 63.3 years (±11.7 SD); 47% were female. Mean intubation duration was 9 days, and mean SOFA score at extubation was 2.5. According to FEES, 36 of 45 (80%) participants had dysphagia: 13 severe, 12 moderate, and 11 mild. Using the prespecified cutoffs, the GUSS-ICU showed for the first rater pair: sensitivity 91.7% (95% CI 77.5–98.3%), specificity 88.9% (95% CI 51.8–99.7%), positive predictive value 97.1% (95% CI 83.8–99.5%), negative predictive value 72.7% (95% CI 46.8–89.0%). For the second rater pair: sensitivity 94.4% (95% CI 81.3–99.3%), specificity 66.7% (95% CI 29.9–92.5%), PPV 91.9% (95% CI 81.7–96.6%), NPV 75.0% (95% CI 41.9–92.6%). Positive likelihood ratios were 8.25 (first rater) and 2.83 (second rater); negative likelihood ratios were 0.09 and 0.08, respectively. The area under the ROC curve was 0.923 (95% CI 0.832–1.000) for both rater pairs. Overall test accuracy was 0.89 (rater 1 vs FEES) and 0.91 (rater 2 vs FEES). Dysphagia severity classification showed strong correlation between GUSS-ICU and FEES (Spearman's rho: 0.61 for rater 1, 0.60 for rater 2, both p<0.001). Krippendorff's alpha across all testers was 0.73. Interrater reliability between the two SLTs was good (Cohen's Kappa 0.84, p<0.001). No adverse events occurred.
**Clinical Implications:** The GUSS-ICU is a simple, reliable, and valid multi-consistency bedside swallowing screen for identifying PED in the ICU. Its key advantages include its multi-consistency design (allowing graded dietary recommendations per IDDSI standards), safety (stepwise consistency progression minimizing aspiration risk during testing), and feasibility (can be performed by nurses after validation, addressing the fact that only 4% of dysphagia specialists are dedicated to ICUs and only 66% of ICUs have SLT consultation available). The GUSS-ICU was performed without complications even in tracheostomized patients (with cuff deflated and speaking valve in place). The study's limitations include its monocentric design, relatively small sample size (n=45, though comparable to other validation studies), and heterogeneity of underlying diseases. The GUSS-ICU is not intended to replace instrumental diagnostics but to complement them as part of a systematic dysphagia screening algorithm in the ICU. The authors plan to validate the GUSS-ICU for use by nurses in a future study.