Auditory Exposure in the Neonatal Intensive Care Unit: Room Type and Other Predictors

Abstract

Objective To quantify early auditory exposures in the neonatal intensive care unit (NICU) and evaluate how these are related to medical and environmental factors. We hypothesized that there would be less auditory exposure in the NICU private room, compared with the open ward. Study design Preterm infants born at ≤ 28 weeks gestation (33 in the open ward, 25 in private rooms) had auditory exposure quantified at birth, 30 and 34 weeks postmenstrual age (PMA), and term equivalent age using the Language Environmental Acquisition device. Results Meaningful language (P < .0001), the number of adult words (P < .0001), and electronic noise (P < .0001) increased across PMA. Silence increased (P = .0007) and noise decreased (P < .0001) across PMA. There was more silence in the private room (P = .02) than the open ward, with an average of 1.9 hours more silence in a 16-hour period. There was an interaction between PMA and room type for distant words (P = .01) and average decibels (P = .04), indicating that changes in auditory exposure across PMA were different for infants in private rooms compared with infants in the open ward. Medical interventions were related to more noise in the environment, although parent presence (P = .009) and engagement (P = .002) were related to greater language exposure. Average sound levels in the NICU were 58.9 ± 3.6 decibels, with an average peak level of 86.9 ± 1.4 decibels. Conclusions Understanding the NICU auditory environment paves the way for interventions that reduce high levels of adverse sound and enhancing positive forms of auditory exposure, such as language.

Language Exposure of Preterm Infants in the Neonatal Unit: A Systematic Review

“…The study by Chow and Shellhaas [22] was the only one recruiting and measuring infants of all gestational ages (mean 34 ± 5.1 weeks; the age range was not reported). Accordingly, the mean birthweight in the study by Chow and Shellhaas [22] (2,410 ± 1,070 g) was much higher than that in the studies by Caskey et al [23] (896 ± 195 g) and Pineda et al [37] (855 ± 184 g). The Language Environment Analysis (LENA) digital language processor (LENA Foundation, Boulder, CO, USA) was used in all studies, with a 16-h recording time.…”

“…That by Caskey et al [23] was the only study to be given a strong quality rating, whereas those by Pineda et al [37] and Chow and Shellhaas [22] were rated weak. Statistical analysis and reporting were well documented in Caskey et al [23] and Pineda et al [37], but Chow and Shellhaas [22] reported some outcomes in median values and others in mean values, making overall comparisons difficult.…”

“…A total of 39 articles were excluded and 10 articles reporting language exposure of preterm infants in the neonatal unit were identified as eligible for inclusion. These were categorised by study type and included 3 observational studies in 4 published articles (Table 2) [22][23][24]37] and 6 language intervention studies (Table 3) [26,[38][39][40][41][42]. The reference lists of all included articles were searched and 6 additional intervention studies [43][44][45][46][47][48] were determined to be eligible for inclusion.…”

“…The study by Chow and Shellhaas [22] was conducted in the single-room environment, that by Caskey et al [23] was set in an openbay unit and Pineda et al [37] compared both single rooms and open bays. Pineda et al [37] recruited infants of the earliest gestational age (mean 25.7 ± 1.4 [23][24][25][26][27][28] weeks), followed by Caskey et al [23] (mean 27 ± 2 [23][24][25][26][27][28][29][30] weeks). The study by Chow and Shellhaas [22] was the only one recruiting and measuring infants of all gestational ages (mean 34 ± 5.1 weeks; the age range was not reported).…”

“…The Language Environment Analysis (LENA) digital language processor (LENA Foundation, Boulder, CO, USA) was used in all studies, with a 16-h recording time. Chow and Shellhaas [22] observed the exposure at a single time point, whereas Pineda et al [37] and Caskey et al [23] observed repeated measures at various time points.…”