The effectiveness and safety of neurally adjusted ventilatory assist iviechanical ventilation compared to pressure support ventilation in optimizing patient venfilator synchrony in critically ill patients: a systematic review and meta-analysis

Abstract

Background: Patient ventilator dyssynchrony is a physical characteristic of suboptimal interaction between patient and ventilator. Some primary clinical studies using neurally adjusted ventilatory assist compared to pressure support suggest it improves patient ventilator synchrony and reduces hospital mortality. With conflicting study outcomes, a systematic review of the effectiveness and safety of neutrally adjusted ventilatory assist is warranted. Objectives: This systematic review aimed to evaluate the effectiveness of neutrally adjusted ventilatory assist (NAVA) compared to pressure support ventilation (PSV) in optimizing patient ventilator synchrony in critically ill adult patients in intensive care unit (ICU). Methods: Seven databases ; the Cochrane Central Register of Controlled Trials, MEDLINE (PubMed), EMBASE, SCOPUS, ClinicalTrials.gov, Web of Science and CINAHL were searched using the following terms: neurally adjusted ventilatory assist, NAVA, neural trigger, interactive ventilatory support, respiration, artificial, mechanical ventilation, patient ventilator asynchrony, synchrony, asynchrony, dyssynchrony. The last search was conducted in April 2018. This review included studies that evaluated the use of NAVA compared with PSV in adult patients who required invasively mechanical ventilation. Outcomes of interest included the frequency of patient ventilator dyssynchrony (PVD) and mortality from all causes. The methodological quality of included studies was assessed, and the data were extracted by using standard forms. Standardized mean differences (SMDs) were calculated for continuous data and risk ratios for dichotomous data, both with 95% CIs. Results: A total of 1,078 articles were identified, for which 210 full text articles were reviewed. In total 17 studies met inclusion criteria. The outcome data were available for approximately 90% of participant (n=398). Neurally adjusted ventilatory assist significantly reduced the AI% by nearly one half of standard deviation; SMD 0.401, 95% CI 0.223 to 0.57, p value 0.000 and I2 0.00% (fixed effect model; two RCTs,128 participants). It was maintained in crossover study group ; SMD 0.304, 95% CI: 0.079 to 0.528, p value 0.008 and I2 75.85% (random effects model, 13 crossover studies, 347 participants). The reduction of the AI% estimated effect size was found to be larger in a sedated group; SMD 0.413, 95% CI: 0.125 to 0.702, p value 0.005 and I2 71.24% than a non-sedated group; SMD 0.225, 95% CI: - 0.208 to 0.659, p value 0.308 and I2 86.76% (random effects model, 10 studies, 248 participants). In addition, a higher reduction of AI% effect size was found in a treatment duration longer than an hour group; SMD 0.413, 95% CI:0.044 to 0.782, p value 0.028 and I2 0.00% than a shorter than an hour group; SMD 0.287, 95% CI:0.069 to 0.505, p value 0.010 and I2 77.62% ( random effects model, 13 studies,301 participants). Similarly, in a 20- minute and longer PVD event-measurement time group found that NAVA reduced AI% more than in a shorter than 20-minute PVD event -measurement time group; SMD 0.389, 95% CI: 0.109 to 0.668, p value 0.006 and I2 0.00% and SMD 0.267, 95% CI: 0.024 to 0.510, p value 0.031 and I2 82.18%, respectively ( random effects model, 13 studies, 301 participants). Neurally adjusted ventilatory assist was associated with a reduction of the risk of AI>10%; OR 0.688,95% CI:0.514 to 0.921, p value 0.012 and I2 21.93%). It significantly reduced the NeuroSync index; SMD 0.745, 95% CI:0.316 to 1.175, p value 0.001 and I2 0.00% (fixed effect model, two studies, 24 participants). In addition, patients in the NAVA group had a lower patient ventilator asynchrony % than in the PSV group in both two levels of assistance; NAVA-low and NAVA-high (Mean ± SD) 7±2% and 7±2%; PSV-low and PSV-high 18±13% and 23±12%, respectively. Patient ventilated with NAVA had a lower ICU mortality compared to the PSV; OR 0.610, 95% CI:0.263 to 1.418, p value 0.251 and I2 0.00% (fixed effect model, two RCTs, 153 participants). Conclusion: Neurally adjusted ventilatory assist is associated with a reduction of PVD frequency compared with PSV. However, effect on lowering the ICU mortality rate is uncertain.