The following is a summary of “Improving Spontaneous Breathing Trials With a Respiratory Therapist-Driven Protocol,” published in the May 2024 issue of Critical Care by Linke et al.
Protocols led by respiratory therapists (RT) for spontaneous breathing trials (SBTs) have improved patient outcomes.
Researchers conducted a retrospective study investigating the feasibility and long-term effectiveness of implementing an RT-driven SBT protocol to enhance patient outcomes.
They aimed to standardize and re-implement a protocol-driven by RTs for screening MV adult patients in ICU at an academic medical center. RTs conducted daily screening to assess eligibility for SBTs. Patients meeting the criteria received SBTs, with extubation decisions made by the physician team. Data was collected from electronic health records, including demographics, reasons for intubation, SBT eligibility and exclusion factors, ventilator settings during SBTs, start times, durations, and patient outcomes. QI intervention included staff re-education on the protocol, providing documentation guidance, developing process maps, and facilitating open forum discussions across disciplines.
The result showed 128 patients, representing 759 opportunities for safety screen weaning assessments across a baseline sample and three plan-do-study-act (PDSA) cycles. The eligibility of SBTs increased significantly from 25% at baseline to 86% in PDSA cycle 3 (P≤.001). The proportion of eligible patients who received an SBT was 42% at baseline, 35% at PDSA cycle 1, 36% at PDSA cycle 2, and 51% at PDSA cycle 3 (P=.092). Protocolized SBT ventilator settings were more utilized, increasing from 18% to 83% (P≤.001). The percentage of patients starting an SBT before 9 am rose from 41% to 67% (P=.097), and the median duration of SBT decreased from 211 to 64 minutes (P=.008).
Investigators concluded that implementing a standardized RT-driven SBT protocol is feasible despite staffing, communication, and terminology hurdles.
Source: chestcc.org/article/S2949-7884(24)00039-X/fulltext#%20