This study has two stages. At the first stage,the investigators will try to use noise detectors to investigate the status of noise exposure in the pediatric intensive care unit(PICU) and to analyze the main noise sources in it. At the second stage,the investigators will try to evaluate the effect of noise management on noise exposure and health outcomes of the pediatric patients in PICU.The investigators hypothesize that noise exposure in PICU is higher than the standard.The investigators also hypothesize that the implementation of the noise management will improve the noise exposure and health outcomes of the pediatric patients in PICU.

This study has two stages. At the first stage,the investigators will try to use noise detectors to investigate the status of noise exposure in the pediatric intensive care unit(PICU) and to analyze the main noise sources in it. At the second stage,the investigators will try to evaluate the effect of noise management on noise exposure and health outcomes of the pediatric patients in PICU.The investigators hypothesize that noise exposure in PICU is higher than the standard.The investigators also hypothesize that the implementation of the noise management will improve the noise exposure and health outcomes of the pediatric patients in PICU.

1.For the investigation of the status of noise exposure:

(1) Preparation of noise monitor equipment: select noise detectors conforming to the national standard. All noise detectors used in the study are of the same brand; (2) Use of noise monitor: set the mode of noise monitor according to the national standard: place the monitor in accordance with the requirements;(3) All the data of noise exposure will be collected and sorted out by the investigator. (4) Spss20.0 software is used for descriptive analysis and statistical inference. Mean ± standard deviation, median and quartile are used to describe measurement data, while frequency and percentage are used to describe counting data. Comparison between groups: when the measurement data obey normal distribution and the variance is homogeneous, two independent sample t-test are used to compare whether there is statistical difference between the two groups of indicators or index difference (measured value baseline value); if it does not obey normal distribution or the variance is uneven, non parametric statistical method is used; counting data: Chi square test is used to compare whether there is statistical difference between the two groups of index values, test level: α = 0.05 2.For the evaluation of the effect of noise management on noise exposure and health outcomes of the pediatric patients in PICU:

(1)General information and clinical data of patients will be collected through hospital electronic information system, ICU nursing records, cases, doctor's advice and consultation. The investigator will evaluate and record the delirium score of ICU children who met the inclusion criteria each shift . At the time of discharge, PICU duration, hospital stay, MV duration and in-hospital mortality were inquired by EMR. (2) Quality control of intervention implementation are as follows:① According to the research results of early promotion strategy, relevant training materials were made to carry out training and education for PICU medical staff;② During the implementation of the intervention, irregular sampling of medical staff was conducted to evaluate the implementation of the intervention and give timely feedback.(3) Collection of information:① The data of patients will be collected by the investigator;② Pediatric delirium score: in order to ensure the quality of the study, the investigator will conduct random sampling and assess the delirium of children in the same patient with the nurses from time to time to investigate the evaluation quality;③ Input the data into EpiData, set reasonable numerical boundary value and logical check items. If there is logical error, check the original data for correction, and import the data into spss20.0 for statistical analysis.

(4) Sample size assessment: the assessment of sample size in experimental research can also be divided into measurement data and counting data. The main outcome measures are delirium rate and noise exposure. Therefore, the incidence of delirium will be used as the rate to calculate the sample size.According to the calculation formula of the rate: n = (U α + U β) 22P (1-p) / (P1-P2) 2, the sample size of the control group and the intervention group are calculated. Uα=1.64，Uβ=0.84。P1 and P2 are the estimated values of delirium incidence before and after the intervention, and P is the total rate. The references showed that P1 = 33%, P2 = 14%.The sample size calculated by the formula is: (1.64 + 0.84) 2 * 2 * 0.47 * 0.53 / 0.192 = 85 cases, plus 20% of the sample loss, that is, 102 cases in the control group and 102 cases in the intervention group (5)Statistical Analysis:①Data set: use EpiData 3.1 software to establish the data set, and clean up and check the data;② The data will be imported into spss20.0 for statistical analysis. Descriptive analysis and statistical inference will be carried out. Mean ± standard deviation, median and quartile are used to describe measurement data, while frequency and percentage are used to describe counting data. Comparison between groups: when the measurement data obeys normal distribution and the variance is homogeneous, two independent sample t-test is used to compare whether there is statistical difference in the indicators or index difference (measured value baseline value) between the two groups; if the measurement data does not obey the normal distribution or the variance is uneven, non parametric statistical method is used; counting data: Chi square test is used to compare whether there is statistical difference between the two groups of patients Level: α = 0.05.

Inclusion Criteria:

• Patients aged ≤ 18 years
• The predicted PICU LOS ≥12 hours
• The informed consent of the child's guardian was obtained

Exclusion Criteria:

• Patients with delirium on admission
• Patients in a coma (Richmond Agitation Sedation Scale[RASS]-4 or -5) all through their stay in PICU
• Patients with mental disorders or mental retardation, who continue to use antipsychotic drugs
• Patients undergoing neurosurgery and neurology surgery
• Hearing or visual impairment
• Non Chinese or English language users
• Missed or incomplete delirium score