The impact of a noise reduction quality improvement project upon sound levels in the open-unit-design neonatal intensive care unit

Transcription

1 ORIGINAL ARTICLE The impact of a noise reduction quality improvement project upon sound levels in the open-unit-design neonatal intensive care unit (2010) 30, r 2010 Nature America, Inc. All rights reserved /10 Department of Neonatology, The Children s Hospital of Southwest Florida, Fort Myers, FL, USA, on behalf of the NIC/Q 2005 Physical Environment Exploratory Group Objective: To decrease measured sound levels in the neonatal intensive care unit through implementation of human factor and minor design modification strategies. Study Design: Prospective time series. Two open-unit-design neonatal centers. Result: Implementation of a coordinated program of noise reduction strategies did not result in any measurable improvement in levels of loudness or quiet. Conclusion: Two centers, using primarily human behavior noise reduction strategies, were unable to demonstrate measurable improvements in sound levels within the occupied open-unit-design neonatal intensive care unit. (2010) 30, ; doi: /jp ; published online 10 December 2009 Keywords: sound measurement; incubator; sleep preservation neurodevelopment; newborn; Neonatal Intensive Care Quality Collaborative Introduction There has been growing awareness of the potentially deleterious impact of the environment upon the rapidly developing and vulnerable preterm brain. 1 3 Excessive or inappropriately timed sensory experiences may negatively impact the development of perceptual organization, 4 as well as normal development of sleep cycles and preservation of total sleep. 5 The presence of noise in the neonatal intensive care unit (NICU) has been specifically identified to be of special concern. 6 9 The preterm infant acutely transitions from an acoustically subdued and predominantly lower frequency in utero environment, 10 to the NICU environment, Correspondence: Dr, Department of Neonatology, The Children s Hospital of Southwest Florida, 9981 South HealthPark Drive, Suite 281, Fort Myers, FL 33908, USA. william.liu.md@leememorial.org Received 9 June 2009; revised 30 September 2009; accepted 25 October 2009; published online 10 December 2009 with measured sound levels that may be excessive and potentially hazardous In addition, environmental noise pollution has a negative effect on individual staff attention, inter-staff communication and the risk of medical errors. 18 Reports of measured sound pressure levels (SPLs) in open-unit-design NICUs uniformly are above published recommendations. 26,27 Behavior strategies targeting nursery personnel have been suggested to decrease nursery noise. 16,27 Our goal was to identify and implement human behavior and minor design modification strategies that might diminish the measured SPLs. Our hypothesis was that human behavior and minor unit-design modification strategies for noise reduction would result in a decrease in measured occupied NICU SPLs. Methods From September 2006 to September 2007, The Children s Hospital of Southwest Florida (TCH) and Benefis HealthCare (Benefis) implemented a noise reduction program with a standardized methodology for NICU sound measurement. This was a prospective time series, quasi-experimental design, with ongoing measurement before, during and after intervention cycles. The two participating centers were part of a five-center collaborative ( Senses and Sensibilities: Physical Environment Group) that worked on the development 28 and implementation 29 of potentially better practices to support neurodevelopment in the NICU, under the auspices of the Vermont Oxford Network sponsored Neonatal Intensive Care Quality Collaborative 2005 (NICQ). All five centers participated in center-specific measurement of sources of noise and equipment, minor design modifications and individual noise reduction initiatives targeting human behavior. 29 Noise reduction strategies were developed using insights from this general effort and literature review. The general approach to quality improvement implementation cycles have been described Hospital IRB committees at both centers reviewed and approved this study.

2 490 Noise reduction in the NICU Sound measurement The background concepts of acoustic measurement and terminology have been summarized for the clinician. 20,33 All sound measurements in this paper are expressed in decibels A-weighted (dba). Noise is a subjective assessment that sound levels are excessive or undesired. Loudness is also a subjective measure that is positively influenced not only by sound pressure, but also frequency and duration. With moderate background sound levels, a 3-dB change may be barely perceptible, a 5-dB change would be readily perceived and a 10-dB difference would be perceived as a doubling of loudness by the human ear. This subjective sensitivity increases, as the background sound level decreases. Real-time sound has constant variation over microseconds and this variable sound input is captured by the dosimeter as slow or fast response (1 versus s) over a defined sound measurement period (time history interval). The dosimeter detector measures this dampened sound power and derives an integrated SPL. The root mean-square (r.m.s.) SPL is sampled by the dosimeter (every 1/50th second) and stored in the device for analysis. L max is the highest SPL of at least 1/50th second duration that occurs during the sound measurement period. L min is the lowest SPL of at least 1/50th second duration that occurs during the sound measurement period. L eq is the level of a constant sound, which in a given sound measurement period has the same energy as a time-varying sound over the same time period. L n describes the distribution of sound over a sound measurement period, for example, L 90 is the SPL exceeded for 90%, L 50 is the SPL exceeded for 50% and L 10 is the SPL exceeded for 10% of the defined sound measurement period. As suggested by Philbin and Gray, L 10 reflects the level of loudness, and L 90 reflects the level of quietness. L 50 is a median value for the sample distribution and also reflects more the level of quietness. L min defines the noise floor, a reflection of the intrinsic acoustic qualities of the facility itself. 25 The L eq value reflects more the level of loudness over any defined time period. Both centers used the ANSI compliant-type II Larsen Davis (PCB Piezotronics Div, Provo, Utah) Spark 706 dosimeter, calibrated prior to each usage and set to slow response, A-weighted, 1-s time history interval, 30-dB gain and 3-dB exchange rate; with data reviewed with a proprietary sound analysis software program (Blaze 5.06). TCH had an open-unit-design 42 bed NICU with two main patient areas. The Level-3 unit had a central entry area and two separate patient care areas. The Level-2 unit had a similar design. The Level-3 areas housed all infants requiring some form of positive pressure support, or weighing less than 1250 g, whereas the Level-2 areas were used for larger, more stable, infants not requiring positive pressure support. The staff work area was centrally located within each room, with vinyl tile flooring, painted dry wall and 9-foot suspended ceiling with acoustic panels (NRC 0.55); 40 to 60 sq. ft/level 3 patient area. Benefis was a 20-bed NICU with an open-unit design, with two main areas, designated left and right units, housing both level 3 and level 2 acuity patients. The patient area had similar specifications as that in TCH. The staff work area was separated from patient care areas, with a separate entry area. Our sound measurement protocol was an adaptation of published approaches. 20,23,34 A 24-h measurement period provided a sampling time period to incorporate the usual variations in unit noise due to rounds, staff rotation, admissions and parent visitation. Bi-weekly measurements were obtained, with selection of the day of the week based on convenience. Specifically defined sampling locations were chosen to reflect routine operational areas. At TCH, these defined measurement areas were the following: (1) The Level-3 Bedside: Infant in a radiant warmer or a canopy open Giraffe Omnibed (GE Healthcare, Waukesha, WI, USA) incubator, with the dosimeter placed near the infant head area, either with the microphone suspended about 8 to 12 inches above the infants head, or suspended from an adjacent intravenous fluid stand. Care was taken to avoid direct contact of the dosimeter or microphone with sources of excessive vibration, as might be the case with the use of a high-frequency ventilator. In such cases, the dosimeter was placed on an adjacent non-contiguous shelf, with a foam base and stiff plastic tubing supporting the microphone receiver extended outward and suspended horizontally, in an unobstructed manner, toward the patient care area, about 2 to 3 ft from the patient s head; (2) The Level 3 Incubator: Measurements performed on Level-3 infants who were not initially on positive pressure support, with the microphone suspended vertically within the incubator, with the receiver about half way between the infant s head and the top of the incubator; (3) The Level-3 Staff Work Area: The microphone was suspended from the ceiling, but not higher than 5 ft above desktop level; (4) The Level-3 Central Entry Area: The microphone was suspended above the clerk area, from the ceiling, but no higher than 5 ft above desktop level; and (5) The Level-2 Bedside: The infant was in a bassinette or incubator, with the microphone suspended from the ceiling, about 1 ft above the infant head area if in a bassinette, or in close proximity to the incubator. Measurements at TCH were conducted by four neonatal nurse practitioners trained and demonstrating competency in the above methodology. At Benefis, the bedside measurements were taken in the NICU Left Bedside or NICU Right Bedside, and the Staff Work Area. Microphone placement followed the same guidelines as above. Level-2 or Level-3 areas are not defined in this case with both levels of patients present in both areas. Benefis had one dedicated staff member performing all measurements for the entire study period. Development and implementation of potential strategies for noise reduction TCH and Benefis employed a series of human behavior as well as minor design modification strategies. Both centers used a staff education and minor unit-design-modification approach.

3 Noise reduction in the NICU 491 Staff education included increased awareness of human sources for noise, with peer-directed encouragement to use conversational tones or a library voice, and to move social conversations away from the patient bedside when possible. With incubators, behaviors involving banging the incubator or placing objects on top of the uncovered incubator were discouraged, and care was taken to open and close the portals with care. Pagers were set to vibrate when possible, staff were encouraged to silence alarms when doing patient care and staff and clerks were to respond to alarms and phones promptly; there was increased awareness of noise-generating behaviors, including hand washing, opening disposable equipment (especially within the confines of incubator), open/closing entry doors and doors of storage cabinetry and equipment carts, and disposal of linens and trash, especially glass objects, into the garbage. Minor unit-design modifications included discouraging the use of audio and video equipments within the unit. Cabinet door latches were padded or removed. Metallic garbage cans were replaced with plastic cans. The intercom in the unit was turned off. TCH had three sequential intervention periods defined as the period when coordinated staff education and/or minor unit-design modifications were implemented. Benefis employed two sequential staff education initiatives, with a strategy employing a quiet zone designated as the medication and feeding preparation areas and quiet times, which were 1-h periods in the morning and evening when there was concerted effort to minimize sound-generating activity. Between and during these periods, there was an ongoing attempt to reinforce the above behaviors with modeling and feedback by peers, unit level leaders, nursing management and physicians. The baseline period was defined as all measurements that occurred before the implementation of any major staff educational strategies. The during/after period was defined as measurements following the first implementation cycle to the completion of the study. Sound measurement patterns Using an activity log, TCH documented the details of Level-3 Bedside activity over the 24-h measurement periods, with summarized L eq values viewed at 1-s, 1-min, 5-min and hourly time history intervals. To identify periodicity within the 24-h day, TCH data for all baseline Level-3 Bedside, using hourly L 10, L 50 and L 90 values, were parsed by 1-, 3-, 4- and 6-h time blocks, and these time blocks were compared with each other. Benefis Left NICU Bedside 24-h measurement periods were analyzed similarly. In addition, the weekday (Monday to Thursday) 24-h periods was compared with those collected during the weekend (Friday to Sunday) periods. TCH Level-3 Bedside sound measurements were correlated with total unit census and Level-2 Bedside sound measurements were compared with Level-3 Bedside sound measurements. Statistical analysis Using the dosimeter sound analysis software, the time interval frequency distributions are reported by the defined L 10, L 50 and L 90, values, that is, derived from 4.32 million discrete integrated L eq values obtained within the 24-h measurement periods. The description of a measurement group s central tendency and dispersion are approximated by the L n range and median. In the case of an even total number of measurements, the median is defined as the arithmetic mean of the two central values. The reported decibel values are rounded off to nearest whole number. The Mann Whitney U-test was used for tests of variance between two populations and Kruskal Wallis test was used for multiple populations, with significance level set at P<0.05. Spearman s rank correlation was used to correlate sound data points with the census. SPSS version 16 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. Results TCH and Benefis were unable to demonstrate any significant decrease in sound measurement levels from baseline as compared with during/after implementation periods within any of the NICU sampling locations (Table 1). Sound measurement patterns Activity log documentation of common unit-specific activities could not be associated with specific SPL elevations (reviewed at varying time history intervals), with L eq and L 10 values consistently above 50 dba, as illustrated in a representative TCH Level-3 Bedside 24-h tracing (Figure 1). Shorter time history intervals (down to 1-s intervals) revealed almost continuous SPL elevations at the bedside over a 24-h period (Figure 2). There was periodicity in the baseline TCH Level-3 Bedside measurement periods, with a perceptible disparity (5-dB difference in medians) in periods of loudness (L 10 ) between 1 to 3 AM as compared with 4 to 6 PM (P<0.0001). However, there were no statistical differences in sound measurements within the 24-h measurement periods in the Benefis NICU, nor were there any significant differences in the 24-h L 10, L 50 or L 90 between measurements taken from Monday to Thursday, as compared with the Friday to Sunday period. There was a moderate positive correlation of TCH Level-3 Bedside L 10 (r ¼ 0.495; P ¼ 0.004), L 50 (r ¼ 0.475; P ¼ 0.012) and L 90 (r ¼ 0.404; P ¼ 0.037) to increasing total NICU census. Compared with the TCH Level-3 Bedside areas, the TCH Level-2 Bedside areas were found to be perceptibly quieter, with differences in the median L 90 and L 50 values of 7 and 4 db (P<0.0001). Twenty-four-hour sound measurement periods within the incubator did not improve from baseline as compared with during/after implementation periods (Table 1). In addition, these sound measurements within the incubator had a distinctly

4 492 Noise reduction in the NICU Table 1 Twenty-four-hour sound measurements: before QI compared to during/after QI (dba) Number of 24-h samples L min (range, median) L 90 L 50 L 10 L max TCH NICU location Level-3 Bedside Before Level-3 Bedside During/after P-value NS NS NS NS NS Level-3 Incubator Before Level-3 Incubator During/after P-value NS NS NS NS 0.01 a Level-2 Bedside Before Level-2 Bedside During/after P-value NS NS NS NS a Level-3 staff working area Before Level-3 staff working area During/after P-value NS NS NS NS NS Level-3 central entry area Before Level-3 central entry area During/after P-value NS NS NS NS NS Benefis NICU location NICU Right Bedside Before NICU Right Bedside During/after P-value a a NS NICU Left Bedside Before NICU Left Bedside During/after P-value NS NS NS NS NS Staff work area Before After During/after P-value a NS Abbreviations: NICU, neonatal intensive care unit; NS, non-significant; QI, quality improvement; TCH, The Children s Hospital of Southwest Florida. a A statistically significant db difference that may also be perceptible to the human ear. different pattern characteristic compared with measurements in the bedside area (Figure 2). The TCH Level-3 Bedside had multiple second-by-second SPL elevations that were less frequent within the incubator. The Level-3 Incubator environment was perceptibly less quiet as compared with TCH Level-3 Bedside, with higher L min, L 90 and L 50 levels (median db differences of

5 Noise reduction in the NICU Time History Graph Nursing Rounds Neo Rounds Nursing Rounds Nursing Rounds db X-ray Touch time ABG Blood transfusion started Touch time IV Fluid Touch time Ped Sgy Change X-ray at Bedside ABG ABG Touch time ETT adjusted CBC and Touch time lab work PAL removed :00:00 09:30:00 12:00:00 14:30:00 17:00:00 19:30:00 22:00:00 00:30:00 03:00:00 05:30:00 Leq Max L10 50 db Figure 1 Common NICU activity correlated with 24-h measurement. L 10, L max and L eq with time history interval of 30 min. NICU, neonatal intensive care unit. 7, 6 and 3; P<0.0001). The levels of loudness as measured by L 10 were not significantly different. There was a 7-dB increase in the median L max values within the incubator (P<0.0001). Discussion The AAP has encouraged NICUs to monitor sound levels, and several published studies report measurements in nurseries ,35 All five centers collaborating in the NICQ 2005 Physical Environment Group obtained 24-h sound measurements in their occupied open-unit-design NICUs, guided by the same methodology described. These measurements, as well as published values (all were open unit design NICU s), suggest that the 2006 Recommended Standards for Newborn ICU Design, that recommend hourly L eq of 45 dba, L 10 of 50 dba and L max of 65 dba, 26 remains an elusive goal. All five centers purchased the Larsen Davis dosimeter, but only TCH and Benefis were able to coordinate an ongoing program of consistent 24-h measurements to obtain data that continued through the period of quality improvement initiatives. The other participating centers either had difficulty with consistent and reliable data collection over time, or chose to perform only baseline measurements. Some barriers were the cost in human resources, the learning curve for use of the dosimeter and analytical software as well as compliance with the recommended frequency of measurements performed through the duration of the quality improvement project. A 24-h measurement period is an operationally practical time frame and will include any diurnal periodicity and hour-to-hour temporal variability in sound levels that may occur. Our bi-weekly sampling frequency increased the likelihood of identifying sound patterns. However, in a stable clinical environment employing routine surveillance sound measurements, a less frequent sampling period would be adequate. As suggested by Long and Lucey in 1980, 16 and subsequently by the AAP, 27 ahumanbehavior-targetedstrategytodecreasesound levels made intuitive sense. Why was this strategy unsuccessful? We were unable to demonstrate any improvements, and even had deterioration in some sound measurement sampling locations. We are unable to explain the observed increases in SPL measurements. Our intervention was primarily an educational and organizational initiative seeking to change human behavior. We moved toward culture change by enlisting leadership support, unit level champions and a step-wise process of education to achieve buy-in. Although staff awareness and support were at all time highs during the study, sustained compliance was still anecdotally difficult to maintain. Although human behavior change was encouraged, this change could not be assured, nor were compliance metrics obtained. Figure 3 summarizes the sound measurements among the cited published values and the baseline median values for the five participating NICQ centers, and suggests that these levels of loudness (L 10 or L eq ) and quietness (L 50, L 90 ) for the NICQ centers were in the lower range as compared with historical pre- and postrenovation NICUs. This may reflect a self-selection bias, with the Senses and Sensibilities exploratory group centers highly motivated from the start to decrease excessive sensory stimulation within the NICU environment. An NICU that was perceived as louder from the

6 494 Noise reduction in the NICU Time History Graph INCUBATOR db :00:00 08:46:40 11:33:20 14:20:00 17:06:40 19:53:20 22:40:00 01:26:40 04:13: Time History Graph BEDSIDE db :00:00 18:46:40 21:33:20 00:20:00 03:06:40 05:53:20 08:40:00 11:26:40 14:13:20 50 db Figure 2 Comparison of Level-3 Incubator and Level-3 Bedside SPL patterns. One-second time history interval. SPL, sound pressure level. start due to human behavior factors may have had a greater chance to achieve measurable improvements. Our findings suggest that there may be true limitations to the effectiveness of strategies targeting human behavior within an open-unit-design NICU. Overall loudness and quietness will be dependent not only on human-related sound sources, but also on equipment-derived sound sources, added to the baseline noise floor or NICU unit design characteristics (for example, heating, air conditioning, ventilation system, reverberant surfaces). An openunit-designed configuration may dictate operational patterns that will accentuate the noise impact of human conversation and activity, especially with increasing census or acuity. Although previously published reports have not detected cyclical or periodic changes in sound levels, 19,21,23 25 our results did demonstrate perceptible differences in periods of loudness (L 10 ) in the early morning as compared with early evening at TCH. The negative acoustic impact of higher acuity in the open design unit is suggested by greater noise levels in the Level-3 as compared with Level-2 areas, as well as the positive correlation with census. Other NICUs have also reported this difference. 19,22 This may be related to increased staff activity and increased use of common NICU equipments, within the Level-3 area. The published studies have suggested sound level improvement with major renovation, 21,25,35 and Philbin suggested that human

7 Noise reduction in the NICU Decibels- A weighted TCH Level 3 Bedside (n=25) Benefis Level 3 Bedside (n=19) Miss Bap Level 3 Bedside (n=2) Sunnybrook Level 3 Staff work area (n=2) Wesley Level (n=7) Robertson (1998) Philben: Naïve Nsy (mean) (2002) Philben: After renovation (2002) Walsh-Sukys (before) (2001) Walsh-Sukys (after renovation) (2001) Levy (2003) Gottfried (1981) 30 Lmin L90 L50 L10 Leq Lmax Figure 3 NICQ centers baseline median measurements and comparable published values. Mean value used if median not available. NICQ, Neonatal Intensive Care Quality Collaborative behavior change might be necessary, but not sufficient, to achieve a quieter environment. Stevens 35 demonstrated the acoustic benefits of single-room unit design in the unoccupied NICU. However, their redesigned NICU room, with the addition of running equipment (nasal continuous positive pressure ventilation and high-frequency oscillator) resulted in increased sound levels, with 24-h L eq measurements of 55 to 57 dba within their occupied newly constructed NICU. These preliminary data suggest that even in the single-room design unit, where human activity-related noise might be diminished, the contribution of common NICU equipment may remain substantial. Irrespective of the unit design, the observed sound measurement patterns within the incubator will impact many of our growing preterm infants. There are clear distinctions between the bedside and the incubator sound environment. The incubator noise pattern reveals clusters of elevated SPLs occurring at 1 to 4-h intervals, reflecting routine care times for the infant, whereas the bedside is exposed to SPL elevations throughout the 24-h period (Figure 2). Previous reports have noted excessive sound levels within incubators, 1,13,14 with levels that risk masking of the spoken maternal voice. 36 Incubators provide some attenuation of external sound levels. However, common care-giving practices will generate potentially greater SPL elevations within the incubator, with internal accentuation of sound levels due to reverberation. 15,37 All tested incubators had higher noise floors than the ambient area. It is unclear whether these higher noise floors would have a disruptive effect, or mask lower amplitude SPLs and function as white noise that might be soothing. Strategies to decrease sound levels in the NICU should consider incubator acoustics in addition to the overall unit design. Conclusion Two open-unit-design NICUs were unable to demonstrate measurable improvements in sound levels, despite aggressive noise reduction quality improvement strategies using human behavior and minor design modifications. Conflict of interest The authors declare no conflict of interest. Acknowledgments This work was completed as part of the NIC/Q 2005 Collaborative to Improve Neonatal Care, which was sponsored by the Vermont Oxford Network. Special thanks to Rob Archer, BS, RRT, and Marilyn Farley, NNP, for their individual efforts in coordination of sound measurements at Benefis and TCH. We also thank the representatives of the participating centers that contributed their efforts to the results reported in this paper: Benefis Healthcare: Beckett Perkins, NNP; Vicki Birkeland, RN, BSN; Cheryl Worden, RN; Mississippi Baptist Medical Center: Amanda McGruder; Sanjosa Martin, RN; Jack Owens, MD; Sunnybrook Health Science Center: Daniel Hong, RRT; Michael Dunn, MD; Elizabeth MacMillan-York, The Children s Hospital of Southwest Florida: Pamela Laferriere, NNP; Dee Overbeck, NNP; Lorna Forchin, NNP; Sandy Blackington, RNC, MS; Sandra Eanes-McGugan, RN, BSN; Wesley Medical Center: Paula Delmore, RNC, MSN; Susan Laudert, MD. We thank Stanley Graven, MD, for clinical expertise and James Handyside for expert facilitation.

8 496 Noise reduction in the NICU References 1 Gottfried AW. Environment of newborn infants in special care units. In: Gottfried AW, Gaiter JL, (eds). Infant Stress under Intensive Care: Environmental Neonatology. University Park Press: Baltimore, 1985 pp Avery GB, Glass P. The gentle nursery: developmental intervention in the NICU. J Perinatol 1989; 9: Graven S, Bowen Jr FW, Brooten D, Eaton A, Graven MN, Hack M et al. The high-risk infant environment Part 1. The role of the neonatal intensive care unit in the outcome of high-risk infants. J Perinatol 1992; 12: Lickliter R. Atypical perinatal sensory stimulation and early perceptual development: insights from developmental psychobiology. J Perinatol 2000; 20 (Part 2): S45 S54. 5 Graven S. Sleep and brain development. Clin Perinatol 2006; 33: Miller CI, Byrne JM. Psychophysiologic and behavioral response to auditory stimuli in the newborn. Infant Behav Dev 1983; 6: Gerber SE, Lima CG, Copriviza KI. Auditory arousal in preterm infants. Scand Audiol Suppl 1983; 17: Gadeke R, Doring B, Keller F, Vogel A. Noise levels in a children s hospital and wake-up thresholds in infants. Acta Paediatr Scand 1969; 58: Graven S. Sound and the developing infant in the NICU: conclusions and recommendations for care. J Perinatol 2000; 20: S88 S Richards DS, Frentzen B, Gerhardt KJ, McCann ME, Abrams RM. Sound levels in the human uterus. Obstet Gynecol 1992; 80: Philbin MK. The influence of auditory experience on the behavior of preterm newborns. J Perinatol 2000; 20: S77 S Morris BH, Philbin MK, Bose C. Physiologic effects of sound on the newborn. J Perinatol 2000; 20: S55 S Blennow G, Svenningsen NW, Almquist B. Noise levels in infant incubators (adverse effects?). Pediatrics 1974; 53: Ciesielski S, Kopka J, Kidawa B. Incubator noise and vibration: possible iatrogenic influence on the neonate. Int J Pediatr Otorhinolaryngol 1980; 1: Seleny FL, Streczyn M. Noise characteristics in the baby compartment of incubators. Am J Dis Child 1969; 117: Long JG, Lucey JF, Phillip AGS. Noise and hypoxemia in the intensive care nursery. Pediatrics 1980; 65: Anagnostakis D, Petmezakis J, Messaritakis J, Matsaniotis N. Noise pollution in neonatal units: a potential hazard. Acta Pediatr Scand 1980; 69: Gray L, Philbin K. Effects of the neonatal intensive care unit on auditory attention and distraction. Clin Perinatol 2004; 31: Gottfried AW, Wallace-Lande P, Sherman-Brown S, King J, Coen C, Hodgman JE. Physical and social environment of newborn infants in special care units. Science 1981; 214: Gray L, Philbin MK. Measuring sound in hospital nurseries. J Perinatol 2000; 20: S100 S Walsh-Sukys M, Reitenback A, Hudson-Barr D, DePompei P. Reducing light and sound in the neonatal intensive care unit: an evaluation of patient safety, staff satisfaction and costs. J Perinatol 2001; 21: Byers JF, Lowman LB, Waugh R. Neonatal intensive care unit sound levels, environment, and infant responses. Neonatal Intensive Care 2005; 18: Robertson A, Kohn J, Vos P, Cooper-Peel C. Establishing a noise measurement protocol for neonatal intensive care units. J Perinatol 1998; 18: Levy G, Woolston D, Browne J. Mean noise amounts in level II vs Level III neonatal intensive care units. Neonatal Network 2003; 22: Philbin MK, Gray L. Changing levels of quiet in an intensive care nursery. J Perinatol 2002; 22: Philbin MK, Evans JB. Standards for the acoustic environment of the newborn ICU. J Perinatol 2006; 26: S27 S American Academy of Pediatrics (AAP). Noise: a hazard for the fetus and newborn. Committee on Environmental Health. Pediatrics 1997; 100: Liu W, Laudert S, Perkins B, MacMillan-York E, Martin S, Graven S. The development of potentially better practices to support the neurodevelopment of infants in the NICU. J Perinatol 2007; 27: S48 S Laudert S, Liu W, Blackington S, Perkins B, Martin S, MacMillan-York E et al. Implementing potentially better practices to support the neurodevelopment of infants in the NICU. J Perinatol 2007; 27: S75 S Horbar JD. The Vermont Oxford Network: evidence-based quality improvement for neonatology. Pediatrics 1999; 103: (Available at cgi/content/full/13/1/se1/350). 31 Horbar JD, Rogowski J, Plsek PE, Delmore P, Edwards WH, Hocker J et al. Collaborative quality improvement for neonatal intensive care. NIC/Q Project Investigators of the Vermont Oxford Network. Pediatrics 2001; 107: Horbar JD, Plsek PE, Leahy K. NIC/Q 2000: establishing habits for improvement in neonatal intensive care units. Pediatrics 2003; 111: e397 e410. (Available at: 33 Gray L. Properties of sound. J Perinatol 2000; 20: S6 S Philbin MK, Robertson A, Hall JW. Recommended permissible noise criteria for occupied, newly constructed or renovated hospital nurseries. J Perinatol 1999; 19 (Part 1): Stevens DC, Khan MA, Munson DP, Reid EJ, Helseth CC, Buggy J. The impact of architectural design upon the environmental sound and light exposure of neonates who require intensive care: an evaluation of the Boekelheide neonatal intensive care nursery. J Perinatol 2007; 27: S20 S Robertson A, Stuart A, Walker L. Transmission loss of sound into incubators: implications for voice perception by infants. J Perinatol 2001; 21: Bess FH, Peek BF, Chapman JJ. Further observations on noise levels in infant incubators. Pediatrics 1979; 63: