Delirium, a form of acute brain dysfunction, occurs in up to 81% of patients receiving mechanical ventilation in the intensive care unit (ICU). Delirium occurring in the ICU is associated with increased functional dependency, cognitive impairment, longer length of hospital stay, and mortality. This study aim to develop a nursing-driven ICU delirium intervention to reduce incidence of delirium, increase the delirium-and coma-free days (DCFDs), and improve ICU patients' function, cognition, and mortality outcomes 3 months following their ICU admission. The modified Hospital Elder life Program at the ICU (mHELP@ICU) will be provided to ensure critically ill patients are cognitively engaged, physically active, and nutritionally well-fed. This three-year study is divided into two phases. The first phase aims to ensure the accuracy of delirium assessment using the Intensive Care Delirium Screening Checklist (ICDSC) by ICU nurses of three participating ICUs. The ICDSC records assessed by ICU nurses will be abstracted from medical records and compared with a gold standard ICDSC evaluation by a well-trained, independent assessor. Cohen's kappa will be reported to represent the consistency of the ICDSC assessment between delirium data from medical records and the independent assessor. When the Cohen's kappa is less than 0.8, a 3-month bedside teaching and real-time feedback education program will be implemented at three ICU units to improve the accuracy of ICDSC assessment by ICU nurses. The second phase will be a clinical trial using a stepped-wedge cluster randomized controlled trial design. Adult (18 years and older) critically ill patients receiving mechanical ventilation will be consecutively enrolled from three mix-medical ICUs at a studied medical center. Estimated 266 participants will be cluster-randomized into the intervention and control groups. Participants in the intervention group will receive a 14-day mHELP@ICU, provided by a trained mHELP nurse, while the participants who received the usual care will serve as controls. Effects of mHELP@ICU will be evaluated using the daily delirium and coma data (max 14 days, or until death or ICU discharge) retrieved from the medical records, along with the participants' mortality, cognitive, and functional outcomes, which a blinded outcome assessor will assess at 48 hours, 14 days, 30 days, and 90 days after ICU admission.