Safety and efficacy of noninvasive ventilation in patients with blunt chest trauma: a systematic review

We searched the MEDLINE (1946 through June 2012), EMBASE (1980 through June 2012) and Cochrane Central Register of Controlled Trials (CENTRAL) databases using the search terms for NIV and blunt chest trauma (see Additional File 1 for the complete search strategy). Reference lists of retrieved articles and personal files were also searched.

We included published studies in any language, regardless of study design, that reported on clinical outcomes (for example, rate of endotracheal intubation, mortality) in patients with blunt chest trauma who were managed with NIV. There were no age restrictions. We also included case series and cohorts with no comparison groups to look at the safety of NIV in this patient setting. We excluded case reports, qualitative studies and economic analyses.

For the purpose of this systematic review, we defined noninvasive ventilation as the use of any degree of positive end-expiratory pressure or pressure support applied by facemask, helmet mask or nasal prongs. Thus, studies using CPAP or NPPV were included. Blunt chest trauma was defined as the presence of pulmonary contusions, rib fractures and flail chest or sternal fractures. The severity of injury was evaluated based on the Injury Severity Score (ISS) or the Thoracolumbar Injury Classification and Severity Score [14, 15]. We ascertained the presence of acute hypoxemia using the partial pressure of arterial oxygen to the fraction of inspired oxygen ratio (PaO₂:FiO₂) of 300 mmHg or less (for ALI) or 200 mmHg or less (for ARDS) (16). Two investigators (AD and PP) independently and in duplicate completed the literature search and located potentially eligible articles.

Two investigators (AD and PP) independently extracted the following data: authors, year and country of publication, ICU type, study design, inclusion and exclusion criteria, number of patients included, severity of hypoxemia and severity of injury (chest and overall trauma). We reviewed clinically relevant outcomes. The primary outcome of interest was the duration of ventilation in patients undergoing NIV compared to mechanical ventilation. Secondary outcomes included in-hospital mortality, ICU and hospital length of stay, development of nosocomial infections and development of any barotrauma. We assessed the use of NIV for ventilatory support in the patients who developed hypoxemic respiratory failure and ARDS and compared it with endotracheal intubation and mechanical ventilation. We also looked at the use of NIV compared to high-flow oxygen through facemask to determine the need for mechanical ventilation. We extracted safety data, including rate of NIV failure, associated mortality, nosocomial infection and barotrauma.

To assess the methodological quality of the included randomized controlled trials (RCTs), we followed the recommendations outlined in the Cochrane Handbook for Systematic Reviews of Interventions (domain-based evaluation of seven components) [17]. To assess the methodological quality of the included observational studies, we used the Newcastle-Ottawa Scale (point-based evaluation of the eight components). We modified the Newcastle-Ottawa Scale to look at the methodological quality of the case series and cohort studies without comparison groups by developing a six-point scale [18].

We could not combine the data from any of the studies because of the clinical heterogeneity that existed during the time period of the intervention as well as the differences in patient selection criteria, severity of injury and comparison groups.

The initial search strategy identified 20 potentially eligible studies (Figure 1). We excluded 11 studies for the following reasons: six studies did not report any clinical outcomes, three were case reports and two were review articles (Additional File 2). Nine studies met our inclusion criteria. Three were RCTs [6, 20, 21], two were retrospective cohort studies [5, 19] and four were observational studies [7‐10].

We report the main characteristics of the nine included studies in Table 1. Three main groups of studies were included. (1) Hernandez et al. evaluated early use of NIV as compared to high-flow oxygen in patients with hypoxemia with or without signs of any respiratory distress within the first 8 h of presentation to the emergency room or ICU [6]. (2) Four studies evaluated patients late in their course of respiratory failure [5, 19‐21]. These studies, in which NIV was introduced following the development of respiratory distress, compared the use of NIV to prevent intubation with early intubation. (3) Another four studies, which did not include a comparator group, assessed the safety of CPAP and NPPV after respiratory distress had developed in the study participants [7‐10]. Accordingly, owing to the differences in the timing of initiation of NIV and the focus on patient safety, we classified these studies into early interventions, late interventions and patient safety assessment categories (Table 1). Analgesia was an integral part of the initial management of these patients. Eight of the studies mentioned the strategies for pain control used in conjunction with NIV (Table 1).

Among the three RCTs, only one [6] was of moderate quality; the other two [20, 21] were of low quality because of concerns regarding blinding, allocation and reporting of outcomes (Table 2). The two retrospective cohort studies were considered to be of low quality because of concerns regarding the assessment of outcomes and follow-up in the cohort studies (Table 3). Because of the high risk of selection bias due to the lack of a comparison group and important shortcomings in their reporting of outcomes, we considered observational studies without a comparison group to be of low quality (Table 4).

Table 5 reports the rates of failure of NIV, nosocomial pneumonia, barotrauma and mortality associated with NIV. Four studies assessed the rate of intubation in patients undergoing treatment with NIV at any time during their hospital admission [6, 7, 10, 20]. The reported rate of NIV failure ranged from 12% to 18%. Hurst et al. found a much lower rate of intubation (6%), but the severity of chest injury in their patients was very low and was not comparable to that reported in the other studies [10]. Gunduz et al. reported no need for intubation in patients undergoing NIV but excluded patients undergoing emergent intubation after randomization from their analysis [20]. If those patients had been included, the rate of intubation would have been 17.1% [20].

Nosocomial pneumonia and pneumothorax were the most commonly reported adverse events associated with NIV use. Nosocomial pneumonia was reported in five studies, and the rate ranged from 8% to 13.8% [5‐7, 20, 21]. The rate of pneumothorax reported in two studies ranged from 5.5% to 24% [6, 21]. Of note, the overall event rate for these two aforementioned events was extremely low. Only two studies reported any deaths associated with the use of NIV, which were also low at 4% and 9% [6, 20].

A single study compared NIV to the use of high-flow oxygen therapy (≥10 L/min) [6]). In this study, patients with ALI and blunt chest trauma were randomized to early NIV or high-flow oxygen therapy. Both groups had similar ISSs and degrees of hypoxemia. The utilization of cointerventions, such as epidurals and patient-controlled analgesia (PCA) for pain control, early chest physiotherapy and need for surgical intervention, was also similar in both groups. Although the study was stopped early, the use of NIV was associated with a lower rate of endotracheal intubation (3 (12%) of 25 patients vs 10 (40%) of 25 patients; P = 0.02). The need for intubation occurred within the first 24 hours in all the patients in the high-flow oxygen group and after 72 hours in all the patients in the NIV group. There were no differences in the rate of pneumothorax, ICU length of stay and mortality.