Alarm fatigue severity and associated factors among intensive care unit nurses: a systematic review and meta-analysis

Alarm fatigue is an important occupational and patient-safety concern in intensive care units (ICUs), where nurses are frequently exposed to false or non-actionable clinical alarms. This systematic review and meta-analysis evaluated the severity of alarm fatigue among ICU nurses and synthesized associated factors and alarm-management interventions. The review followed PRISMA 2020 guidance and was registered in PROSPERO (CRD420261429354). Literature published from January 2001 to May 2026 was searched across major databases and grey-literature sources. ICU-specific studies reporting extractable alarm-fatigue scores were included in the quantitative synthesis. Scores from different instruments were transformed to a common 0–100%-of-maximum-possible (POMP) scale and pooled using a random-effects model. Associated factors were synthesized narratively because of methodological heterogeneity. Twenty-three studies were included, of which 11 contributed to the ICU-specific meta-analysis. The pooled normalized alarm-fatigue score was 51.66 (95% CI: 42.33–60.99), with substantial between-study heterogeneity (I² = 99.6%). Accordingly, the pooled estimate should be interpreted as an exploratory descriptive summary rather than a prevalence estimate or clinical threshold. Frequently reported factors associated with alarm fatigue included false and non-actionable alarms, high alarm burden, inadequate alarm-management knowledge and training, night shifts, overtime, workload, role overload, organizational limitations, and technology-related difficulties. Alarm fatigue was also associated with burnout, compassion fatigue, perceived stress, reduced nursing performance, and medical-error tendency. Randomized studies indicated that structured alarm-management education could reduce alarm fatigue and unnecessary alarm exposure. Alarm fatigue is a substantial and multifactorial concern among ICU nurses. However, confidence in most reported associations remains limited by predominantly cross-sectional, self-reported, and heterogeneous evidence. Further prospective studies incorporating objective alarm data and validated patient-safety outcomes are needed. This systematic review and meta-analysis was registered in PROSPERO under registration number CRD420261429354. Not applicable.