Single-Breath Counting Test: a new prognostic tool to evaluate patients with dyspnea and respiratory insufficiency in the Emergency Department

Background Dyspnea represents one of the main causes of E.D. admission; it is important to rapidly stratify the risk of clinical deterioration, the need for hospital admission, respiratory support, and intensive care support. Many validated scores are currently employed for this task (HACOR, REM, MEWS, ROX index, PSI, CURB-65, SOFA, GWTG‐HF and LUS). The Single Breath Counting Test (SBCT) is the measurement of how far an individual can count in a normal speaking voice. SBCT was previously validated as a test of pulmonary function in asthma exacerbations and neuromuscular diseases and estimates the need for non-invasive respiratory strategies in patients with COVID-19. Study objective The purpose of our work is to establish if SBCT could be a useful tool for early risk stratification at E.D. arrival, in case of patients presenting with dyspnea and having an oxygen saturation of < 92% at triage. In this setting, we evaluated SBCT effectiveness to predict the risk of hospitalization, the need for transfer in a higher level of care (HDU/ICU), the need of ventilation support (CPAP, NIV, or invasive mechanical ventilation), and in-hospital mortality risk. Materials and methods We recruited 62 patients during a period of nearly 8 months, and collected clinical data, risk factors and comorbidities. We collected vital parameters, ABG and blood tests results, imaging results, SBCT, and the cited scores and indexes immediately after admission and after 3 hours of treatment. We compared SBCT and the other scores by univariate analysis in groups defined by the four aforementioned outcomes. We evaluated the performance of the scores and SBCT by ROC curves and 95% confidence interval compared by the De Long test. Results HACOR, REM, CURB-65 and SOFA scores were significantly different in admitted vs. discharged patients, whereas SBCT did not reach significance; its effectiveness for this outcome was outperformed by HACOR, REM, MEWS, SOFA, CURB-65 and PSI. Similarly, SBCT was not useful to predict ICU/HDU admission, and was outperformed by MEWS, CURB, PSI, HACOR, and ROX index. SBCT was significantly lower in ventilated patients (3 (2; 6) vs. 5 (3; 8), p = 0,002); SBCT at 3 hours (8 (5; 10) vs. 9 (7; 12), NS); Δ SBCT (-4 (-5; -3) vs. -4 (-5; -2), NS). SBCT demonstrated to be effective in predicting ventilation requirement (AUC 0.6641 (0.5286-0.7995), and its performance was comparable to P/F, CURB-65, REM, ROX, PSI and MEWS at 3 hours. Regarding the risk of in-hospital mortality, patients who survived at the end of hospital stay scored higher at admission SBCT. Overall, REM and MEWS were more predictive of mortality. Discussion According to the results of our pilot study, SBCT could be employed together with the already existing tools to assess patients’ outcome. A low SBCT suggests the possible need of advanced respiratory support, and, result-wise, is comparable with routinely employed scores and indexes, such as P/F.