3.1 Introduction and rationale In 2020 an estimate of 562.700 people had Chronic Obstructive Pulmonary Disease (COPD) in the Netherlands, resulting in 21.335 hospital admissions.\[1\] An acute exacerbation COPD (AECOPD) in combination with hospital admission is associated with high mortality and morbidity. \[2\] Noninvasive ventilation (NIV) has been very effective in the context of AECOPD, since it efficiently offloads respiratory muscles and counteracts dynamic hyperinflation. This method often prevents intubation as a bridge to administering effective therapies (e.g., glucocorticoids, bronchodilators, and antibiotic agents)\[3\], and may reduce mortality.\[4\] Current guidelines recommend NIV for the treatment of acute respiratory failure in patients with respiratory distress, pH \< 7.35, and PaCO2 \> 6 kPA (with exclusion of patients requiring invasive ventilation, see guideline).\[5,6\] National guidelines do not recommend NIV in AECOPD with increased work of breathing without respiratory acidosis due to lack of evidence. \[5\] Ultrasound of the diaphragm can identify atrophy and impaired motion or contractility of the diaphragm\[10\], and has been shown to predict mortality and NIV failure during NIV treatment for respiratory acidosis due to AECOPD.\[11-15\] Whether ultrasound of the diaphragm may predict mortality or need for (non-)invasive ventilation in AECOPD without respiratory acidosis on initial presentation is unknown. During exploratory analysis of a previous study (unpublished; NCT05671198) we might have found a marker that has the potential to predict progression to NIV or death during hospitalization for AECOPD without respiratory acidosis: The difference between diaphragm motion during tidal breathing and maximal breathing, used as a surrogate for inspiratory reserve volume (IRV), was significantly lower in patients requiring NIV or who died in-hospital compared to those who did not (1.59 cm, SD 1.91 vs. 3.14 cm, SD 2.45; p = 0.033). We performed a ROC curve analysis to assess the predictive value of these variables, which yielded an area under the ROC curve (AUROC) of 0.752 (95% CI: 0.535 - 0.968, p = 0.033), indicating a statistically significant discriminatory ability. The best cut-off value, as determined with the Youden's J statistic, was 1.74 cm, with a sensitivity of 86% and specificity of 70% for predicting NIV requirement or inhospital death. However, the number of events was low and the study was primarily powered for another outcome. Therefore, prospective validation is needed before we impose treatment based on this marker. In case the suggested ultrasound measurement (see introduction) proofs to be a discriminatory marker for deterioration during hospitalization or predictive of progression to Non-Invasive Ventilation, we hope to be able to perform a follow-up study in which patient will be randomized to either standard of care or 'elective' NIV (to avoid emergency NIV need) based on yet to determine cut-off values. 3.2 Design (including population, method, confounders and outcomes) A multi-center, prospective observational cohort study conducted at Isala Hospital and UMCG aimed at determining the value of the sonographic motion ratio (tidal/maximum) of the diaphragm in AECOPD after hospital admission. After enrollment, the sonographic diaphragm motion will be assessed as additional measurement during standard-of-care lung ultrasound (POCUS), after which hospital outcomes will be registered. Primary outcome will be the sensitivity of this ultrasound marker for in-hospital deterioration (progression to NIV of death). Secondary analysis will include predictive models. 3.3 Research question What is the sensitivity of the diaphragm motion ratio (tidal/maximum) during ultrasound for in-hospital deterioration (progression to NIV or death) in AECOPD.