Research has shown that sound levels outside of the normal range can be detrimental to the healthy development of premature newborns in the NICU (Bremmer, Byers, & Kiehl, 2003; Johnson, 2003; Philbin & Gray, 2002). Fetuses born after 24 weeks have nearly complete cochlea and sensory organs, and when born prematurely they are thrust out of the quiet atmosphere of the uterus into the noisy environment of the NICU (Abrams & Gerhardt, 2000; Hall, 2000). Elevated sound levels have been shown to interfere with a newborn's sleep and have a negative effect on the infant's vital signs, oxygen saturation, and auditory attention (Bremmer et al., 2003; Gray & Philbin, 2004; Philbin & Gray, 2002). Simple frequent acts, such as closing the incubator door, can reach levels of 100 dB or more, which is equivalent to exposing an infant to the noisy sound levels at a rock concert (American Academy of Pediatrics, 1997). As a result of repeated exposure to loud noise in the NICU, premature newborns are more at risk for sensorineural hearing loss and future developmental delay (Surenthiran et al., 2003). It is important to continually evaluate how to maintain safe sound levels within the NICU (Green et al., 2005; Philbin, 2004; Philbin & Gray, 2002; Philbin, Taber, Hayman, 1996).
Current and Recommended Sound Levels
In an attempt to reduce sound in NICUs, the American Academy of Pediatrics Committee on Environmental Health has recommended safe sound levels (American Academy of Pediatrics, 1997); these 1997 recommendations have been updated by an expert team of practitioners (Graven, 2000). Recommendations specify that sound levels are maintained at (a) an hourly Leq of 50 dB, (b) an hourly L10 of 55 dB, and (c) a 1-second duration Lmax that should not exceed 70 dB. These levels, however, are frequently exceeded because "thousands of nurseries built during the 1970's or with 1970 era-designs and materials continue in use worldwide without modification to improve noise levels" (Philbin & Gray, 2002, p. 455). Researchers have found that nationally, sound levels in NICUs range from an Leq of 50 to 75 dB with peaks (or Lmax) of 105 dB (Abramovich, Gregory, Slemick, & Stewart, 1979; Anagnostakis, Petmezakis, & Messaritakis, 1980; Kellman, 2002; Levy, Woolston, & Browne, 2003; Philbin & Gray, 2002).
Common Sources of Sound in the NICU
Sources of sound within the NICU are typically divided into two categories:
1. Operational (staff or equipment generated sound)
2. Structural (building-generated sound) (Evans & Philbin, 2000; Philbin, 2004; Philbin & Gray, 2002).
Interventions directed at reducing both types of sound have been shown to be effective in reducing the Lmax by >10 dB (Philbin & Gray, 2002). Some of the interventions used to reduce operational sound have included such staff education as (a) moving teaching rounds away from the beside, (b) reducing the volume of neonatal alarms, (c) using plastic instead of metal drawers and waste baskets, and (d) limiting conversations at the bedside (Johnson, 2003) (Table 1). Changing structural sound, however, requires expensive reconstruction of the physical plant to correct (Philbin & Gray, 2002; Robertson, Cooper-Peel & Vos, 1999). Examples of some structural changes for this purpose within an NICU are (a) reducing the number and increasing the space between beds; (b) locating the unit secretary, nurse's station, and consultation places outside the main nursery; (c) using sound-absorbing ceiling and floor materials; and (d) installing quieter ventilation/air conditioning systems (Philbin, 2004).
After our Level III NICU underwent reconstruction, we saw an opportunity to investigate the effectiveness of suggested structural changes. Therefore, our purpose was to compare previously reported sound levels before reconstruction (Krueger, Wall, Parker, & Nealis, 2005) to those obtained after structural reconstruction.
Study Design and Methods
Using established recommendations (Graven, 2000), sound level recordings of the Leq, L10, and Lmax (A-weighted scale) were measured continuously for 8 hours, and the average overall sound levels were compared before and after reconstruction within an NICU located in the Southeast.
Before Reconstruction
Before reconstruction, the NICU had a maximum capacity of 25 bed spaces spread along the arms of a horseshoe configuration. This configuration necessitated taking measurements that followed the shape of the unit. Secretarial and larger workspaces were located outside. High-noise areas identified within the NICU were two entrances to the nursery, a computer desk at which the NICU staff tended to congregate, and metal hand-washing sinks. Based on these higher-noise areas and the horseshoe shape of the NICU, we chose nine locations to set up our sound monitoring equipment. Patient census fluctuated by two patients over the 9 days of measurement. The door of the neonatal unit was found left open to the major work area on two occasions, which resulted in possibly higher sound levels at that time (Krueger et al., 2005).
After Reconstruction
After reconstruction, the nursery maintained the same maximum capacity of 25 bed spaces but was spread along a rectangular, rather than horseshoe configuration. Identified higher noise zones were two larger workspace/computer areas within the NICU in which staff tended to congregate. Based on these potentially higher noise zones and because the NICU was no longer in a horseshoe shape, we chose two locations to set up our sound monitoring equipment. No fluctuations in patient census occurred during the two mornings of measurement.
Several structural changes were made to facilitate sound reduction in the NICU. The ceilings were lowered approximately 4 feet, and the square footage of the facility was increased by 250 square feet; both of these changes increased the noise distribution, thereby decreasing the overall sound levels in the unit. The heating ventilation and air conditioning systems were designed with sound attenuation in all areas of the facility. Ceiling tiles with a high sound absorption rating were placed throughout the entire NICU. In addition, a focus was placed on limiting previously identified sources of sound in the NICU. This focus included placing monitor alarms away from the wall to decrease transmission of sound; widening doorway entrances into the unit to facilitate quiet entry into the NICU; placing computers at each bed space, which reduced staff interaction in these areas; designating private space in close proximity to the NICU for residents and nurse practitioners to discuss patient care, make phone calls, and complete their daily work; and replacing knee-controlled sinks with electronic sinks to decrease the noise level associated with frequent hand washing.
Procedure
A calibrated ANSI type 1 Bruel & Kjaer sound level meter (model #2239) and software (Noise Explorer) were used to record both before and after reconstruction sound level measurements. As recommended by Philbin and colleagues (Gray & Philbin, 2000; Philbin, Robertson & Hall, 1999), the Leq, L10, and Lmax were sampled using an A-weighted, slow response scale of decibel levels. Measurements were performed continuously for eight 1-hour recordings, from approximately 0600 to 1400. Before reconstruction, the sound levels were measured over a total of 9 days; after reconstruction the sound levels were measured over 2 days. The sound level meter was clearly labeled with its microphone placed behind bed spaces on a shelf approximately 2 to 3 feet above the countertop. No attempt was made to conceal data acquisition.
Hourly sound levels were combined from the different locations to calculate the overall average Leq, L10, and Lmax sound levels before and after reconstruction. Analyses performed were descriptive in order to provide a comparison between the known acceptable levels of sound and those before and after reconstruction in the NICU. This analysis is suggested by others in the literature (Philbin & Gray, 2002).
Results
Sound levels were averaged and compared between the before reconstruction (1-hour recordings from nine locations over 8 hours) and after reconstruction (1-hour recordings from two locations over 8 hours) measurements (Table 2). Compared to recommended sound levels (Graven, 2000), the average overall sound levels before reconstruction were Leq M = 60.44 dB, L10 M = 59.26 dB, and Lmax M = 78.39 dB. The average overall sound levels after reconstruction were Leq M = 56.4 dB, L10 M = 60.6 dB, and Lmax M = 90.6 dB.
Clinical Implications
After structural reconstruction was done in the NICU, Leq sound levels decreased as would be predicted (Philbin & Gray, 2002). The Lmin and Lmax did not, however, decrease as predicted, and both remained above recommended sound levels (Leq < 50 dB, L10 < 55 dB, and Lmax < 70dB) (Graven, 2000).
Philbin (2004) acknowledged that every NICU has a particular design and location, and that sound measurement can never be perfect, but stressed that sound can be minimized when reconstructing an NICU. Byers, Waugh, and Lowman (2006) showed a similar decrease in Leq (approximately 5 dB) when measurement occurred at the infant's ear; in the study reported here, measurement occurred approximately 2 to 3 feet behind the infant's bed space. Furthermore, although Philbin and Gray (2002) did not report the Leq, they did find that the average Lmax decreased by >10 dB after a combination of both operational strategies and structural reconstruction. In the present study, the Lmax increased, which may have occurred because no simultaneous operational or staff education programs were offered that targeted reduction in sound levels.
A limitation of the study is that it occurred in only one NICU, but the chance to compare sound levels in a reconstructed NICU offered a unique opportunity. The patient census varied during before reconstruction measurements and not during the after reconstruction measurements, and the shape of the unit changed during the reconstruction. Because of these facts, the comparison of the sound levels is not totally equivalent but rather was design-dependent. Strengths of the study are that the number of beds and the nursing staff remained constant before and after reconstruction of the unit, and the relative acuity of the premature infants in the NICU did not change between measurements. Both consistencies between before and after reconstruction can be used to increase confidence in the findings.
The results of this study demonstrate that structural changes can reduce sound levels in an NICU but not enough by themselves to meet current recommendations (Graven, 2000). The effects of elevated sound levels on the growth and development of the preterm infant warrants continued research to further reduce sound levels and to more specifically investigate what changes (whether operational or structural) are the most effective in maintaining safe sound levels.
Measuring Sound
All sounds are measured using an A-weighted, slow response scale. The Leq is the equivalent steady (dBA) noise level across a 30-second time period. The L10 is a measure of the decibel level exceeded for 10% of the hour; the Lmax is the highest decibel level measured for at least 1/20th of a second duration during the hour. The A-weighted scale is used to measure what an infant's or adult's ear can detect at relatively quiet levels of sound (Philbin, Robertson, & Hall, 1999; Philbin & Gray, 2002).
Acknowledgment
This study was funded by the National Institute of Health (NIH/NINR P20 NR07791, NIH/NCRR MO1 RR00082).
References