Methods for Measuring and Identifying Sounds in the Intensive Care Unit.

Background Despite many studies in the field examining excessive noise in the intensive care unit, this issue remains an ongoing problem. A limiting factor in the progress of the field is the inability to draw conclusions across studies due to the different and poorly reported approaches used. Therefore, the first goal is to present a method for the general measurement of sound pressure levels and sound sources, with precise details and reasoning, such that future studies can use these procedures as a guideline. The two procedures used in the general method will outline how to record sound pressure levels and sound sources, using sound level meters and observers, respectively. The second goal is to present the data collected using the applied method to show the feasibility of the general method and provide results for future reference. Methods The general method proposes the use of two different procedures for measuring sound pressure levels and sound sources in the intensive care unit. The applied method uses the general method to collect data recorded over 24-h, examining two beds in a four-bed room, via four sound level meters and four observers each working one at a time. Results The interrater reliability of the different categories was found to have an estimate of >0.75 representing good and excellent estimates, for 19 and 16 of the 24 categories, for the two beds examined. The equivalent sound pressure levels (LAeq) for the day, evening, and night shift, as an average of the sound level meters in the patient room, were 54.12, 53.37, and 49.05 dBA. In the 24-h measurement period, talking and human generated sounds occurred for a total of 495 (39.29% of the time) and 470 min (37.30% of the time), at the two beds of interest, respectively. Conclusion A general method was described detailing two independent procedures for measuring sound pressure levels and sound sources in the ICU. In a continuous data recording over 24 h, the feasibility of the proposed general method was confirmed. Moreover, good and excellent interrater reliability was achieved in most categories, making them suitable for future studies