Abstract
BACKGROUND: Monitor alarms occur frequently but rarely warrant intervention. OBJECTIVE: This study aimed to determine if a safety huddle-based intervention reduces unit-level alarm rates or alarm rates of individual patients whose alarms are discussed, as well as evaluate implementation outcomes. DESIGN: Unit-level, cluster randomized, hybrid effectiveness-implementation trial with a secondary patient-level analysis. SETTING: Children's hospital. PATIENTS: Unit-level: all patients hospitalized on 4 control (n = 4177) and 4 intervention (n = 7131) units between June 15, 2015 and May 8, 2016. Patient-level: 425 patients on randomly selected dates postimplementation. INTERVENTION: Structured safety huddle review of alarm data from the patients on each unit with the most alarms, with a discussion of ways to reduce alarms. MEASUREMENTS: Unit-level: change in unit-level alarm rates between baseline and postimplementation periods in intervention versus control units. Patient-level: change in individual patients' alarm rates between the 24 hours leading up to huddles and the 24 hours after huddles in patients who were discussed versus not discussed in huddles. RESULTS: Alarm data informed 580 huddle discussions. In unit-level analysis, intervention units had 2 fewer alarms/patient-day (95% CI: 7 fewer to 6 more, P = .50) compared with control units. In patient-level analysis, patients discussed in huddles had 97 fewer alarms/patientday (95% CI: 52-138 fewer, P < .001) in the posthuddle period compared with patients not discussed in huddles. Implementation outcome analysis revealed a low intervention dose of 0.85 patients/unit/day. CONCLUSIONS: Safety huddle-based alarm discussions did not influence unit-level alarm rates due to low intervention dose but were effective in reducing alarms for individual children.