405 - Quality Improvement Program to Reduce Alarm Fatigue in the NICU

Poster Number: 405
Publication Number: 405.436

Yanin Fontana Stiglich, Hospital Italiano de Buenos Aires, CABA, Ciudad Autonoma de Buenos Aires, Argentina; PABLO H. BRENER, HOSPITAL ITALIANO BUENOS AIRES, Buenos Aires, Ciudad Autonoma de Buenos Aires, Argentina; María S. Segura, Hospital Italiano de Buenos Aires, Ciudad Autónoma de Buenos Aires, Ciudad Autonoma de Buenos Aires, Argentina; Gonzalo L. Mariani, Hospital Italiano de Buenos Aires, Buenos Aires, Ciudad Autonoma de Buenos Aires, Argentina; Yanina Mazzitelli, Hospital Italiano de Buenos Aires, CABA, Ciudad Autonoma de Buenos Aires, Argentina

Background: Alarm fatigue (AF) in healthcare happens when professionals are exposed to many alarms and they become desensitized to them. It is a multifactorial problem related to proliferation of devices, use of narrow or not standardized alarm limits, high prevalence of “non-actionable alarms”, i.e. false (alarms triggered by equipment issues such as sensor disconnection) or nuisance (actual physiologic change not requiring clinical action). These factors can lead to a longer response time (Tr). After evaluating the situation in our NICU, an alarm management program (AMP) was developed to reduce AF.

Objective: To compare the prevalence of AF in our NICU before and after implementing an AMP, estimating the probability of true critical events (i.e. true alarms /true alarms + non-actionable alarms) and measuring Tr.
To determine variables associated with non-actionable alarms and Tr.

Design/Methods: Quality Improvement study. One hundred baseline observations were collected between 12/2019 and 1/2020. An AMP including webinar sessions, short talks in the NICU and a new alarm management guideline was implemented. Then, 100 new observations were collected between 6/2021 and 8/2021. NICU professionals were blinded to the observations. We estimated true and non-actionable alarm prevalence in both periods. Data was collected throughout every shift. Prespecified demographic and alarm related variables were included on univariate analyses assessing association with non-actionable alarms and Tr (dichotomized on Pc50). Logistic regression was performed to assess independent variables associated with non-actionable alarms.

Results: Population demographics and alarm characteristics are shown in table 1. Probability of critical events reflected as prevalence of true alarms was 31% before AMP and 57% after AMP (p 0.001) while non-actionable alarm prevalence was 69% and 43% respectively (p 0.001). Median Tr was significantly reduced after AMP (35 sec to 12 sec (p 0.001). Table 2 shows univariate analyses for both periods. Before AMP, longer Tr and non-actionable alarms were more frequent in more mature neonates with less intensive care needs. After AMP, Tr was similar for true and non-actionable alarms observations. For both periods, the need of respiratory support was significantly associated with true alarms (p 0.001). In the adjusted analysis, Tr (p=0.001) and respiratory support (p=0.003) remained associated with non-actionable alarms.Conclusion(s): Alarm fatigue was highly prevalent in our NICU and it was significantly reduced after implementing an AMP.
Table 1. Population demographics and alarm characteristics*Mann Whitney ** Chi2 ***Fisher
Table 2. Univariate analyses*Mann Whitney **Chi2 ***Fisher