Clinical outcomes and risk factors in critically ill children with diaphragmatic dysfunction: an prospective observational study

Background: Diaphragmatic dysfunction (DD) has a great negative impact on clinical outcomes, and it is a well-recognized complication in adult patients with critical illness. However, DD is largely unexplored in the critically ill pediatric population. The aim of this study was to identify risk factors associated with DD, and to investigate the effects of DD on clinical outcomes among critically ill children. Methods: Diaphragmatic function was assessed by diaphragm ultrasound. According to the result of diaphragmatic ultrasound, all enrolled subjects were categorized into the DD group ( n=24 ) and the non-DD group ( n=46 ). Collection of sample characteristics in both groups include age, sex, height, weight, primary diagnosis, complications, laboratory ndings, medications, ventilatory time and clinical outcomes. Results: The incidence of DD in this PICU was 34.3%. The level of CRP at discharge (P=0.003) in the DD group was higher than the non-DD group, and duration of elevated C-reactive protein (CRP) (P<0.001), sedative days (P=0.008) and ventilatory treatment time (P<0.001) in the DD group was signicantly longer than the non-DD group. Ventilatory treatment time and duration of elevated CRP were independently risk factors associated with DD. Patients in the DD group had longer PICU length of stay, higher rate of weaning or extubation failure and higher mortality. Conclusion: DD is associated with poorer clinical outcomes in critically ill childern, which include a longer PICU length of stay, higher rate of weaning or extubation failure and a higher mortality. The ventilatory treatment time and duration of elevated CRP are main risk factors of DD in critically ill children.


Background
The diaphragm is the primary respiratory muscle, generating approximately 60 to 80 percent of the inspiratory capacity [1] . Diaphragmatic dysfunction (DD) has a great negative impact on respiratory function, and it is a well-recognized complication in adult patients with critical illness, especially among patients with mechanical ventilation (MV) [2] . The study performed in adults found that diaphragm thickness decreases rapidly following intubation in nearly 50% of ventilated patients [3] . There is a wide variety of factors which can lead to DD in critically ill adult patients, including MV, sepsis, sedative, steroids, neuromuscular blockers, multiorgan dysfunction syndrome and preadmission injury [2,4,5] . MV is a leading cause of DD, the diaphragm atrophy can be signi cant in 18-69 hours under complete controlled ventilation [6] , And MV is widely used in critical illness patients as a conventional treatment technology of PICU, approximately 30% of children in the pediatric intensive care unit (PICU) receive MV support [7] , and 38% of them have DD [8] . Patients with DD are at higher risk for delayed, di cult weaning and extubation failure, together with an increased risk of longer ICU length of stay, poor functional outcomes, and death [9] . Mariani et al. [10] found that the ICU mortality is higher in patients with DD compared with absence of DD. The study conducted by Lu et al. [11] demonstrated that subjects with DD show longer mechanical ventilation durations and ICU stays.

Page 3/12
A wide variety of adult studies have found the causes and clinical outcomes of DD in critically ill patients.
However, DD is largely unexplored in the critically ill pediatric population. The studies on risk factors of DD and how it impacts clinical outcomes in children are limited [12][13][14] . As a result, there are few clinical experience in the identi cation, appropriate intervention and improvement of prognosis of critically ill children.
The aim of this study was to identify risk factors associated with DD, and to investigate the effects of DD on short-term clinical outcomes in children and to test the hypothesis that DD is associated with poor clinical outcomes among critical illness children in the PICU.

Subjects
This study was conducted in an academic, 57-bed PICU of First Hospital of Jilin University in China.
Study subjects including 70 consecutive patients younger than 18 years old who required invasive MV for more than 24 hours were enrolled between January 2019 and January 2020. The institutional ethics committee of the hospital approved the study protocol (ChiCTR1800020196). The parents or guardians of the eligible children provided written informed consent. An information sheet was provided for the parents or guardians of the participants.
All children met the standard criteria [15] for weaning readiness(improve-ment in the cause of primary disease, PaO 2 /F iO2 > 200, positive end-expiratory pressure (PEEP) ≤ 5-10cm H 2 O, F iO2 ≤ 50%, and hemodynamically stable in the absence of vasopressors) were included in the study. Exclusion criteria include known neuromuscular disease (such as amyotrophic lateral sclerosis, Guillain-Barre, or myasthenia gravis), cervical spinal cord injury, pneumothorax, unwillingness of the parents or guardians to participate in the study.

Study Design
Enrolled subjects underwent a diaphragm assessment by using ultrasonography during the spontaneous breathing trial (SBT), which was performed using pressure support trials with a pressure support (8 cm H 2 O) and 5 cm H 2 O PEEP using a Drager Evita 4 ventilator for 30 min. Ultrasound measurements were taken at the fth minute after the beginning SBT. All enrolled subjects were categorized into the DD group and the non-DD group according to the result of diaphragmatic echo. DD was de ned as a diaphragmatic thickening fraction (DTF) of < 20% during tidal breathing [16] .

Diaphragm ultrasound measurement
The diaphragm ultrasound was performed using a portable ultrasound machine (Mindray, M7 series, China) with a 10HMz linear probe by two experienced sonographers. Only the right hemidiaphragm was measured because the right hemidiaphragm was more feasible and repeatable compared with the left hemidiaphragm [12] . All subjects were placed in a semi-recumbent position with the head of the bed at a 30-degree angle. The probe was placed between the mid-axillary or antero-axillary line, in the 8th to 11th intercostal space, and positioned in a cranio-caudal direction and perpendicular to the skin to achieve the best view of the right hemidiaphragm [17] . At this position, the diaphragmatic ultrasound image was a hypoechoic structure between two echoic lines (the pleural and the peritoneal membrane) (Fig 1). In the Bmode image, diaphragm thickness (Tdi) was measured from the inner edge of the pleural line to the inner edge of the peritoneal line at both end inspiration and end expiration. The calculation formula of DTF was (Tdi-inspiration -Tdi-expiration) / Tdi-expiration [18] . Tdi and body weight (BW) have signi cant positive correlations in children [19] . Therefore, Tdi was standardised by BW (DE/BW).

Patient Characteristics and Clinical Outcomes
We collected the basic demographic of all subjects, beside, primary diagnosis, medications, in ammatory factor levels at discharge and duration of elevated in ammatory factors were also collected, as the previous study demonstrated that systemic in ammation is associated with muscle atrophy in critically ill adult patients [20] . The clinical outcomes we observed include delay or di culty in weaning, extubation failure, total length of time on mechanical ventilation, length of PICU stay and mortality.

Statistical analysis
For comparisons of demographic, clinical characteristics and outcomes between DD and non-DD patient group, continuous variables were compared with Student t-test or Mann-Whitney U test. Categorical variables were compared with Chi-squared test or Fisher's exact test. Data are presented as mean ± standard deviation for continuous variables with a normal distribution and as median with interquartile range for variables with a non-normal distribution. Categorical variables were described as n (%). To determine which factors are signi cantly associated with DD, logistic regression analysis was then performed. All analyses were carried out using IBM SPSS Statistics for Windows, Version 22.0 (IBM Corp, Armonk, NY). and a p-value less than or equal to 0.05 was considered statistically signi cant.

Sample characteristics
One hundred and thirty-three children underwent mechanical ventilation support during the study period. Sixty-three cases were excluded, and seventy patients were included in this study eventually. Eligibles were divided into either the DD group (n=24) or non-DD group (n=46) according to the result of diaphragmatic echo (Fig 2).
All children characteristics were summarized in Table 1 [5.25-15.00] vs 6.00[4.00-8.00], P=0.008) and ventilatory treatment time (13.00[10.00-18.00] vs 5.50[5.00-7.00], P<0.001) in the DD group were signi cantly longer than the non-DD group. However, it should be noted that there were no difference in the proportion of patients with sepsis and methylprednisolone treatment time in the two groups, which are important factors associated with DD in adults.

Risk Factors for the Development of DD
The results of our multivariable logistic regression analysis are shown in Table 2. The independent risk factors associated with the development of DD included ventilatory treatment time (OR, 1.99; 95%CI, 1.18-3.38) and duration of elevated CRP (OR, 1.12; 95%CI, 1.01-1.24).

Comparison of clinical outcomes between the DD group and the non-DD group
It is clearly showd in Table 3

Discussion
To our knowledge, this is the rst study to investigate factors associated with the development of DD and clinical outcomes of patients with DD in the critically ill pediatric population. Our ndings demonstrated that the incidence rate of DD to be 34.3% in children, we also found that ventilatory treatment time and duration of elevated CRP are associated with the development of DD. This shows that MV is not the only cause of DD, and the development of DD in critically ill children is caused by multiple factors. We also report for the rst time that the duration of elevated CRP is associated with DD in critically ill children, which demonstrates that the body's in ammatory response is an important risk factor. In this study, we have also shown that children with DD are associated with poorer outcomes, including a longer PICU length of stay, a higher rate of weaning or extubation failure and a higher mortality.
Our results are similar to studies investigating DD among adult hospitalized patients. We found patients with DD had longer total ventilation time, higher rates of weaning or extubation failure and higher mortality than children without DD [8,21,22] . Limited studies on DD in critically ill children showed that respiratory weakness is independently associated with reintubation and longer ventilation time [23,24] . Our previous study also found that patients with failed weaning have worse diaphragmatic function [25] . This suggests that DD is closely associated with poorer clinical outcomes in both adults and children. The prevalence of diaphragm dysfunction diagnosed with ultrasound was found in 40%-60% of mechanically ventilated adults [10,26,27] . These reports are higher than our result. The main reasons for consideration are as follows: Firstly, the development of auxiliary inspiratory muscles in children are immature, diaphragm playe a greater role in ventilation than adults [28] . Therefore, the baseline function of diaphragm in children is higher than that in adults. Secondly, the majority of patients in adult studies are elderly chronic obstructive pulmonary disease (COPD) patients, whose diaphragmatic muscle bres have a chronic oxidative remodelling process, leading to diaphragm compensatory ability weakness [29] . Thirdly, the average ventilatory treatment time in children with DD is shorter than in adults [ 360 (168-528) vs 576 (374-850) hrs ] [8,25] . Futhermore, our ndings demonstrated that ventilatory time is the independent risk factor of DD in critically ill children, which is similar to adult studies [2,30] . Another important nding in this study is that duration of elevated CRP to be associated with development of DD. In ammation is an important contributor to the pathology of diseases implicated in skeletal muscle dysfunction [31] , some studies among adults found that systemic in ammation is the most signi cant risk factor for ICU-AW [20,32] . And in ammation is prevalent after critical illness and is associated with poor physical recovery [33] . However, the role of in ammation in respiratory muscle is largely unexplored, espically in the critically ill pediatric population. Therefore our ndings demonstrated that duration of in ammatory response may be a main risk factor of DD. Nevertheless, our ndings cannot de nitively con rm causality, which can only be demonstrated in the context of a randomized contral study (RCT).
This study has several limitations. Firstly, we studied a relatively small population, this might restrain the validity. Secondly, ultrasound diagnostic criteria ( DTF < 20% ) for DD comes from adult studies. Whether this reference value applies to children requires further study among pediatric population. Thirdly, our study included only CRP to represent in ammatory response of the patients, the reliability of our nding is limited. Therefore, more in ammatory factors such as interleukin 6 (IL-6), interleukin 8(IL-8), and transforming growth factor (TGF) are need to re ect the in ammatory response more precisely in further study, so that our ndings can be con rmed. Fourthly, due to our small sample size, some important factors in adult research such as sepsis did not produce positive results in our study. This does not demonstrate that sepsis is not associated with DD, instead, we should expand the sample size and conduct the RCT. Finally, we only investigated the clinical outcomes of children with DD during their hospital stay, however, there was a lack of follow-up of the effects of DD on patients after discharge. The study conducted by Dres et al [4] . demonstrated that DD is twice as frequent as limb muscle weakness.
Therefore, it is critical to investigate the impact of DD on long-term prognosis of critically ill children.

Conclusions
DD strongly impacts clinical outcomes in critically ill childern, which includs a longer PICU length of stay, a higher rate of weaning or extubation failure and a higher mortality. The ventilatory time and duration of elevated CRP are main risk factors of development of DD in critically ill children. It is important to identi cate DD earlier among critically ill children in order to protect the diaphragmatic functions. Due to the relatively inadequate research performed in chidren, more studies on DD among pediatric population need to be carried out to better guide clinical and respiratory rehabilitation.  Figure 1 Ultrasound B-mode using a 10 MHz probe in the zone of apposition. Flow chart of this study.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.