Stella Daniels

Project title: Noise Reduction in an Intensive Care Unit: Assessing the Effects of a Behavior Modification Noise Reduction Program

Degree: MPH | Program: Environmental and Occupational Health (EOH) | Project type: Project
Completed in: 2015 | Faculty advisor: Arnold de Castro


Background: In the intensive care unit (ICU) setting, physiological monitoring and general noise contributes to fragmented sleep in critically ill patients, increasing the risk for sleep deprivation. An environment that increases the risk of sleep deprivation could lead to long-term health consequences especially for the critically ill.   The University of Washington Medical Center (UWMC) 5E and 5SE ICUs implemented noise reduction behavior modification interventions among the nursing staff in three phases. Phase 1 (January 2013) included engagement of key stakeholders, addressing monitor alarms, instituting “Quiet Times” (2-4pm and 10pm-6am), decreasing noise associated with phones, overhead paging, and modifying equipment (e. g., padding the pneumatic tube system). Phase 2 (May 2013) involved installing noise meters throughout ICU locations, and training “Super-Users” to train peers on tailoring alarms for physiological monitoring equipment. And, Phase 3 (November and December 2014) reinforced previous interventions, identified a “Champion” in one ICU, encouraged the nursing staff to be creative, and share successful sleep promoting interventions.   Objective: The objective of this project was to assess the effectiveness of Phase 3 interventions executed in UWMC’s 5E and 5SE ICUs to characterize the sound levels to which patients are likely exposed.   Methods: Noise dosimeters recorded sound levels for four 24-hour periods in 10 designated patient rooms, irrespective of a patient, before and after Phase 3.   The mean and standard deviation describes the sound levels in the two units by before and after Phase 3, specific time frames and the rooms. The fraction exceeding 55 decibels (dB) and the 90th percentile of the estimated Leq, and the fraction of exceedance above the moving average at the 5, 10 and 15 dB levels describes the amount of variability observed. A two-sample t-test tested the pre and post intervention sound levels for significance.   Results: Preliminary results show on 5E a pre-intervention mean of 53.1, post-intervention mean of 51.4, with a 95% CI of 1.68 – 1.76, (p-value <0.001) and on 5SE a pre-intervention mean of 55.7, post-intervention mean of 55.3, with a 95% CI of 0.40 – 0.48, (p-value <0.001).      Conclusion: Phase 3 behavior modification interventions employed in the 5E and 5SE ICUs resulted in a statistically significant decrease in noise levels. The effectiveness of the noise reduction interventions is supported by objective measures and offers promising avenues to creating a more restful and rehabilitative environment in ICU settings.