NICUS ARE OFTEN LOUD AND NOT CONDUCIVE TO premature infant development. Controlling noise levels in busy technology-dependent nurseries is a continuous challenge. Although new and newly renovated nurseries now have appropriate recommendations to guide their design projects, NICU nurses and hospital leadership are often unaware of how further changes in the environment can alter the sound environment. Almost any environmental change has the potential to affect the sound environment by either directly increasing sound levels or altering the behavior of caregivers (e.g., increased traffic). These environmental changes may include the purchase of new equipment, alterations in communication systems used or a range in the type of flooring material (e.g., carpet vs linoleum).
The purpose of the study reported here was to identify the effect of alterations in an NICU environment on sound levels and to use this information to inform decisions on investing resources in new environmental changes. Research questions addressed included the following: (1) How does the installation of automatic paper towel dispensers affect ambient sound levels in unit pods? (2) How does the installation of a pilot communication system affect ambient sound levels in pods? (3) Are there differences in ambient sound levels between day and night shifts, and, if so, is there a relationship between time of day and environmental change? (4) What are the differences in ambient sound levels inside potential new and existing infant incubators?
Numerous authors have discussed the effects of sound on premature infants in neonatal nurseries, but little research has described how individual changes in the infants' environment affect ambient sound levels.1–8 Ambient sound levels vary widely from unit to unit and within individual NICUs.2,4,5,9 Sources of environmental noise include heating, ventilation, and air-conditioning systems; plumbing lines and fixtures; door mechanisms; surface materials on floors, walls, and ceilings; overhead paging systems; and the locations of desks, storage units, and travel paths.7,8
Much of the sound in the NICU is generated by the equipment used in the care of the infants, including incubators, ventilators, oxygen saturation monitors, and infusion pumps (Table 1).1,6–8,10 An early study revealed that infants cared for in incubators were exposed to peak noise levels created by a variety of “typical” caregiving activities such as setting a plastic feeding bottle on top of an incubator (108 decibels [dB]) and permitting water to bubble in ventilator tubing (87 dB).8 Infants in incubators are exposed to less sound from alarms and equipment than are those in open warmers, and nurseries designed to minimize noise and newer incubators result in a quieter nursery environment.2 Other interventions, such as addition of acoustical foam to the inside of the incubator, can decrease environmental noise measurements inside incubators.11
Studies of ambient noise levels suggest that measurements taken inside an occupied incubator are significantly higher in NICU rooms where staff activity is greatest.4 They also show that mean noise levels are significantly higher in Level III than in Level II NICUs.5 Typical sound levels in the NICU (55–75 A-weighted decibels [dBA]) are well beyond levels that support sleep or stable heart rates in healthy new-borns.7,9,12 Typical sound levels in NICUs vary from 50 to 75 dBA, with peaks of 105 dBA and frequent, prolonged sounds in the 70–80 dBA range.9,13–15
Many studies have explored the physiologic responses of term and premature infants to noise in the neonatal intensive care nursery setting.8,9,12 Most of this research explored the effects of sound on heart rate and respiration.9 Physiologic responses of term newborns to hospital nursery sounds >80 dBA include apnea and bradycardia as well as abrupt fluctuations in heart rate, blood pressure, respiratory rate, perfusion, and oxygen saturation.12 Changes in noise levels have both increased and decreased preterm infants' heart and respiratory rates, decreased transcutaneous oxygen tension, and increased intracranial pressure, with concern that these events may negatively affect short-term pathology and long-term infant development.6,12,16–21 Infants cared for in NICUs may experience many daily sound fluctuations during weeks of hospitalization when they are experiencing rapid growth of their immature brain.12
Based on a review of the literature, recommendations for decreasing sound levels in the NICU to minimize the potential stress for infants include responding quickly to alarms, placing telephones away from patient beds, conducting conversations away from the bedside, and implementing awareness programs for staff.17 Although numerous studies have been conducted, the use of small samples, nonrandomization, and inconsistent use of 24-hour sound measurements have made it difficult to determine the ideal sound environment for infants.9
However, research is beginning to reveal potential long-term effects of the noisy NICU environment. The human auditory system undergoes the majority of its development in utero long before term gestational age, with cochlear function and hearing commencing by about 22–24 weeks.12,22 Throughout fetal development, the uterine environment provides moderately loud acoustic signals across the range of frequencies to which the developing ear is sensitive. The maternal voice provides a distinctive and clear signal against the moderately loud, low-frequency background sounds. This complex intrauterine sensory experience may assist in focusing fetal attention on the mother's voice, which is generalized to a postbirth bias to attend to speech.3
Mechanical aspects of the developing ear and the fluid uterine environment provide the fetus with significant protection from low-, mid-, and high-frequency sound while in utero.2. However, once an infant is born, the sound exposure is different. Unlike intrauterine sounds, traditional nursery sounds are airborne and comprise a wide range of frequencies that are continuous, unpredictable, and strong, even by adult standards.12,23–25 Lickliter has highlighted the importance of normal patterns of perinatal sensory experience, including auditory stimulation, to early perceptual and behavioral capabilities in avian and mammalian species. This animal-based research suggests that alterations of perinatal sensory stimulation may affect development depending on a number of related factors, including (1) the timing of the stimulation relative to the developmental stage of the organism, (2) the amount of stimulation provided or denied the young organism, and (3) the type of sensory stimulation presented (auditory, visual, vestibular).26
Prolonged exposure to the chaotic and strong sensory NICU environment during critical periods of brain development is increasingly implicated as a contributor to attentional difficulties.3 Preterm infants have far more difficulty coming to a quiet alert state and sustaining attention to a signal than do term infants and are further challenged by a sensory environment that is intense and chaotic. Although adults can assign meaning to particular acoustic patterns (e.g., alarms, water, noise, laughing), studies show that infants have a limited ability to make such distinctions, particularly if the background noise is loud. Preterm infants are affected by sounds that do not influence mature listeners, in part because premature infants are not able to “tune out” noise. An acoustical environment similar to that of the third-trimester fetus would be most advantageous for the preterm infant, but the specific acoustic properties of the uterus cannot be duplicated.3
The study was judged by the Duke University Internal Review Board as exempt research. The NICU comprises eight four-bed pods, one three-bed pod, two isolation rooms, and two care-by-family rooms. Each of the pods is open to a common hallway. The unit had been previously renovated to incorporate materials to decrease ambient noise and to contain and absorb transient noise arising within the nursery.12 These steps follow recommendations endorsed by the Committee to Establish Recommended Standards for Newborn ICU Design, chaired by Robert D. White.27
This study utilized a prospective quasi-experimental design and was conducted as part of a larger longitudinal two-group randomized study (awarded to the first author, D.B.) of the effects of cycled light and continuous near-darkness on the health and developmental outcomes of preterm infants. During the 12-month study period, the hospital made a transition to automatic paper towel dispensers in the clinical settings, conducted a trial of a new communication system, and evaluated new incubators for purchase. Sound levels were assessed before and after each environmental change. Because the larger study was evaluating day and night cycling of light, sound level differences between day (7 AM to 7 PM) and night (7 PM to 7 AM) shifts were compared to determine time-of-day differences and potential interactional effects with the environmental changes.
Sound levels were obtained over one 24-hour period each week for a total of 44 weeks between July 1, 2003, and July 31, 2004. Pods that were eligible for the weekly selection each housed at least one infant enrolled in the larger study. The pod that had been monitored least recently was selected for monitoring each week. Some pods were measured more than others, again dependent on a study infant occupying the room. The day of the week for sampling was selected at the discretion of the research study personnel. There were 8 weeks during the year in which sound levels were monitored for the larger study in one of two transitional nurseries, and those weeks were not included in the NICU data analysis.
A single measurement of sound level was conducted inside each of four different incubator models (Drager Caleo [Drager Medical, Lubeck, Germany], Ohmeda Giraffe, Ohmeda Careplus, Ohmeda Omnibed [Ohmeda was bought out by GE Healthcare, United Kingdom, after this research was conducted]) before a decision regarding replacement of the incubator fleet was made.
The Quest model 2900 integrating/logging sound level meter (Quest Electronics, Oconomowoc, Wisconsin) was used to obtain the weekly 24-hour measurements. The measurements were completed using an A-weighting method, a procedure that filters sound according to frequency, which helps sound meter recordings map more closely the sound pressures and frequency ranges to which the human ear is most sensitive.5 Three variables were used to evaluate sound pressure levels (SPLs). The first variable, Leq, measures the steady state dBA (A-weighted, slow response, frequency-weighting filter that approximates the frequency response of the human ear) noise level during a 1-hour period. Leq is a good measure for evaluating average sound levels in the pods. L10 measures the dBA sound level that is exceeded 10 percent of the time over the upper limit of the set range (40–100) during the 24-hour period. The last variable, Lmax, measures the maximum sound level of one-second duration during a one-minute period. Lmax corresponds with an individual's perception of maximum loudness.28 The national recommended levels for NICUs are 45–50 dBA (hourly) for Leq, 55 dBA for L10, and 70 dBA for Lmax.29 Each 24-hour measurement included 1,440 consecutive samples at one-minute intervals.
The meter was calibrated prior to each recording with a sound level calibrator (Quest QC-10) according to the manufacturer's instructions and specifications. The sound level range to calculate the steady state (Leq) for this instrument was set between 40 and 100 dB. Sound levels outside this range are captured and used to calculate L10 and Lmax. The meter was positioned securely on a camera tripod and placed at a bedspace occupied by one of the infants in the larger study. The room number and bedspace number were documented for each recording, as were the date and day of the week. The day of the week was rotated so that each day was sampled over seven weeks.
The single-incubator (Drager Caleo, Ohmeda Giraffe, Ohmeda Careplus, Ohmeda Omnibed) sound measurements were obtained by placing the sound meter in an empty incu-bator with the power on for a one-hour period. The incubator was placed in an unoccupied isolation room to standardize the noise exposure across all incubators. The nursery communication system was active during each of the incubator measurements. Figure 1 depicts the sound sources during the study period. No specific sounds were identified, and all occurred by chance at the bedside.
Weekly recordings were grouped into “pre” - and “post” data sets for both paper towel dispenser and Vocera communication system comparisons based on when during the timeline each change occurred (Figure 2). Weekly recordings before installation of the automatic paper towel dispensers contributed to the predata set, and recordings after installation of the new dispensers but before implementation of the trial communication system contributed to the postdata set. The predata set for the communication system consisted of weekly recordings between installation of the automatic paper towel dispensers and initiation of the communication system trial, and the postdata set comprised the 12 weeks during which the system trial took place. Recordings were also divided into two 12-hour groups to compare day and night sound levels and to evaluate the interaction between time of day and the environmental change. The four 1-hour recordings of the incubators were compared to evaluate noise level differences by type of incubator.
For each of the research questions, regression analyses were completed using PROC MIXED (SAS, Inc., Cary, North Carolina).30 The general linear mixed model is a flexible statistical procedure that is widely used for analyzing continuous longitudinal data and that can account for potential clustering of nested sound observations within one pod. It accommodates for the correlation present across the repeated measures within each measurement. Each of the three sound variables was regressed over paper towel holder (pre vs post), time of day (day vs night), as well as the interaction between time of day and paper towel holders. In addition, each of the three sound variables was regressed over communication system (old vs trial), time of day (day vs night), as well as the interaction between time of day and communication system. For the incubator analysis, each of the three sound variables was regressed over the four different incubator models. The clustering of observations within a pod was treated as random effects in the current study. To reduce the influences of skew and outliers, dependent variables were entered into the regressions in logged form.
Each of the three SPL sound variables (Leq, L10, Lmax] was significantly higher after installation of the automatic paper towel dispensers (Leq—F[1:86, 450] = .24, p <.001; L10—F[l:86, 450] = .24, p < .001; Lmax—F [1:86, 450] = .37, p <.001). This effect remained significant regardless of the time of day (day [7 AM-7 PM] vs night [7 PM – 7 AM]) or the day of the week the measurements took place. L10 and Lmax were significantly higher during the trial of the Vocera communication system (L10—F[1:103, 739] = .05, p<.001; Lmax—F[1:103, 739] = .09, p<.00l). Because the paper towel dispensers were in place for both the pre- and post-communication system data sets, the rise in sound levels with the communication system includes the increase in sound levels following installation of the automatic paper towel dispensers. Like the effect of the paper towel dispensers, the effect of the communication system remained significant regardless of time of day or day of the week. In addition, the mean levels of Leq and L10 for both the towel dispenser and the communication system data sets were higher than recommended national standards.
Levels for each of the three SPL sound variables were higher during the day than during the night (Leq—F[1:86, 459] = .23, p <.001; L10—F[1:86, 459] = .23, p<.00l; Lmax—F[l:86, 459] = .23, p<.001). In analysis for interaction effects between time of day and installation of the paper towel dispensers and time of day and the Vocera communication system trial, the positive effect of the paper towel dispensers and communication system were greater during the daytime (p<.01).
The analysis of the three SPL variables for the four incubator models revealed that the Drager incubator was significantly quieter than the three others tested (Leq—F[1:235] = .23, p<.001; L10—F[1:235] = .23, p<.001; Lmax—F[l:235] = .23, p<.001) (ordered from newest to oldest). Incubator noise levels decreased with the newer model incubators. Only the Drager maintained sound levels on all three SPL variables within the recommended national standards. The Ohmeda Omnibed was within the national standard guidelines for L10 and Lmax.
Previous research has suggested that sound levels are higher when more caregivers are present but are not necessarily higher during the daytime than at night.7 This study did not look at how sound levels changed throughout the day, but on average the daytime sound levels were higher than the nighttime levels. In addition, both the installation of automatic paper towel dispensers and the trial of the communication system resulted in higher sound levels throughout the 24-hour period, but levels were greatest during the daytime. Both of these findings would be expected given the increase in caregivers and unit activity associated with the day shift, with a related increased use of both paper towel dispensers and the communication system.
Although the larger study examined the impact of day-night cycling of light, light levels during the sound recordings were not part of this analysis. In the larger study, the infant intervention is individualized with the use of bed coverings; therefore, the light at each bedspace may differ among infants. It has been suggested that low lighting is related to quieter conversation.7,17 Therefore, the light levels in the pods may have influenced the observed difference between day and night sound levels.
Testing the incubators in the setting in which they would be used permitted evaluation of the impact of the existing unit environment on sound levels inside each of the incubators. Identification of the quietest incubator for this NICU was one criterion for selecting the incubator type for the replacement fleet. Based on the results of this study, the trialed communication system was tabled until additional improvements could be made. Discussions regarding the increased sound levels generated by the automatic paper towel dispensers included the advantages and disadvantages of the product. It was felt that the potential benefits of automatic dispensers to affect nosocomial infection rates and hand-washing compliance (greater capacity, less waste) outweighed the increased noise these devices generate.
This study suggests that all environmental changes must be monitored to ensure that they reduce rather than increase noise levels. It is critical that nurses be aware of and assist in minimizing adverse environmental conditions. Altering care-giving behaviors and modifying the physical environment of the NICU are essential components of controlling the excess sound levels identified in this study. The goal for all NICUs is to provide an environment that promotes sleep, supports neonatal physiologic stability, and reduces potential adverse effects on the auditory development of premature infants.31 Interventions to minimize noise levels should include evaluation of proposed environmental changes whenever possible. When the purchase of new equipment is considered, individuals making the procurement decisions must consider the noise produced by the equipment and whether the noise levels can be modified.
Interventions to decrease sound levels are challenging. It is nearly always more difficult to fix a noise problem than to prevent one, and the costs of the remedy are usually borne by the hospital.32 Hospital administration should consult nurseries when environmental or equipment changes are being considered.
Solving the problems of noise abatement and reduction requires changes in knowledge, attitudes, behaviors, and performance.18 Educational programs should be seen as a first step toward promoting awareness and significantly decreasing sound levels.33 It is recommended that hospitals routinely measure sound levels, including Leq, L10, and Lmax parameters, in nurseries.34–36 Sound measurements should be taken to assess the appropriateness of environmental changes and the selection of new equipment. Levels should be repeated following all environmental changes to evaluate the effectiveness of noise reduction strategies. To minimize infant exposure to unnecessary noise, NICUs must conduct ongoing evaluations of their environment.
Although monitoring of the NICU environment is essential to the maintenance of national standards, future research must focus on the effects of our sound environment on the development of infants and how we provide them care. Changes in an infant's environment have the potential to support and/or impair the neonate's development.
This study suggests that changes in the environment do not always have the intended result. Environmental modifications that are thought to be innocuous may result in small but significant increases in noise levels. Environmental alterations that are expected to decrease noise may not do so. Even small increases in noise levels from individual environmental modifications may have cumulative or exponential effects on the total sound environment. When environmental changes are planned, NICU personnel must include selection criteria that consider the noise they produce.
1. Berens, R. J., & Weigle, C. G. (1997). Noise analysis of three newborn infant Isolettes. Journal of Perinatolojjy, 17,
2. Byers, J. E, Waugli, W. R., & Lowman, L. B. (2006). Sound level exposure of high-risk infants in different environmental conditions. Neonatal Network, 25,
3. Gray, L., & Philbin, M. K. (2004). Effects of the neonatal intensive care unit on auditory attention and distraction. Clinics in Perinatology, 31,
4. Kent, D. T., Tan, A. K., Clarke, M., & Bardell, T. (2002). Excessive noise levels in the neonatal ICU: Potential effects on auditory system development. The Journal of Otolaryngology, 31,
5. Levy, G. D., Woolston, D. J., & Browne, J. V. (2003). Mean noise amounts in Level II vs Level III neonatal intensive care units. Neonatal Network,
6. Lotas, M. J. (1992). Effects of light and sound in the neonatal intensive care unit environment on the low-birth-weight infant. NAACOGS Clinical Issues in Perinatal and Womens Health Nursing, 3(1),
7. Philbin, M. K., & Gray, L. (2002). Changing levels of quiet in an intensive care nursery. Journal of Perinatology, 22,
8. Thomas, K. (1989). How the NICU environment sounds to a preterm infant. MCN: The American Journal of Maternal Child Nursing, 14,
9. Morris, B. H., Philbin, M. K., & Bose, C. (2000). The full term and premature newborn: Physiological effects of sound on the newborn. Journal of Perinatology, 20,
10. Robertson, A., Cooper-Peel, C, & Vos, P. (1999). Sound transmission into incubators in the neonatal intensive care unit. Journal of Perinatology, 19,
11. Johnson, A. (2001). Neonatal response to control of noise inside the incubator. Pediatric Nursing, 27,
12. Philbin, M. K., Robertson, A., & Hall, J. (1999). Recommended permissible noise criteria for occupied, newly constructed, or renovated hospital nurseries. Journal of Perinatology, 19,
13. Abramovich, S. J., Gregore, S., Slemick, M., & Stewart, A. (1979). Hearing loss in very low birth weight infants treated with neonatal intensive care. Archives of Disease in Childhood, 54,
14. Anagnostakis, D., Petmezakis, J., Messaritakis, J., & Matsaniotis, N. (1980). Noise pollution in neonatal units: A potential hazard. Acta Pediatrica Scandinavica, 69, 771–773.
15. Gottfried, A. W., & Gaiter, J. L. (1985). Infant stress under intensive care: Environmental neonatology
(pp. 23–54). Baltimore: University Park Press.
16. Bartoshuk, A. K. (1962). Human neonatal cardiac acceleration to sound: Habituation and dishabituation. Perceptual and Motor Skills, 15,
17. Bremmer, P., Byers, J. F., & Kiehl, E. (2003). Noise and the premature infant: Physiological effects and practice implications. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 32,
18. Graven, S. N. (2000). Sound and the developing infant in the NICU: Conclusions and recommendations for care. Journal of Perinatology, 20(8
19. Schulman, C. A. (1969). Effects of auditory stimulation on heart rate in premature infants as a function of level of arousal, probability of CNS damage, and conceptual age. Developmental Psych obiology, 2,
20. Wharrad, H. J., & Davis, A. C. (1997). Behavioral and autonomic responses to sound in pre-term and full-term babies. British Journal of Audiology, 31,
21. Zahr, L., & Balian, S. (1995). Responses of premature infants to routine nursing interventions and noise in the NICU. Nursing Research, 44,
22. Werner, L. A., & Marean, G. C. (1996). Human auditory development
(pp. 20–27, 41–42). Boulder, CO: Westview Press.
23. Blennow, G., Svenningsen, N. W, & Almqvist, B. (1974). Noise level in infant incubators (adverse effects?). Pediatrics, 53,
24. Ciesielski, S., Kipka, J., & Kidawa, B. (1980). Incubator noise and vibration: Possible iatrogenic influence on the neonate. International Journal of Pediatric Otorhino laryngology, 1,
25. Gray, L., & Philbin, M. K. (1991). Levels of quiet in an intensive care nursery. The Journal of the Acoustical Society of America, 90,
26. Lickliter, R (2000). The role of sensory stimulation in perinatal development: Insights from comparative research for care of the high-risk infant. Journal of Developmental and Behavioral Pediatrics, 21,
27. White, R D. (1997). Recommended standards for newborn ICU design. Committee to establish recommended standards for newborn ICU design. Journal of Perinatology, 19,
28. Gray, L., & Philbin, M. K. (2000). The acoustic environment of hospital nurseries. Journal of Perinatology, 20,
29. Graven, S. N., Bowen, F. W, Brooten, D., Eaton, A., Gramen, M., Hack, M., et al. (1992). The high-risk infant environment: Part 1. The role of the neonatal intensive care unit in the outcome of high-risk infants. Journal of Perinatology, 12,
30. Singer, J. D. (1998). Using SAS PROC MIXED to fit multilevel models, hierarchical models, and individual growth models. Journal of ‘Educational and Behavioral Statistics’, 24,
31. Johnson, A. (2003). Adapting the neonatal intensive care environment to decrease noise. The Journal of Perinatal & Neonatal Nursing, 17,
32. Philbin, M. K. (2004). Planning the acoustic environment of a neonatal intensive care unit. Clinics in Perinatology, 31,
33. Krueger, C, Wall, S., Parker, L., & Nealis, R (2005). Elevated sound levels within a busy NICU. Neonatal Network, 24(6),
34. Kellman, N. (2002). Noise in the intensive care nursery. Neonatal Network, 21(1),
35. Philbin, M. K. (2000). The influence of auditory experience on the behavior of preterm newborns. Journal of Perinatology, 20,
36. Robertson, A., Kohn, J., Vos, P., & Cooper-Peel, C. (1998). Establishing a noise measurement protocol for neonatal intensive care units. Journal of Perinatology, 18,