Introduction

Tracheal intubation (TI) to secure the airway of severely ill or injured patients is a critical intervention regularly conducted by emergency medical service (EMS) providers throughout the world. This activity is based on the assumption that, in keeping with in-hospital practice, a compromised airway should be secured as early as possible to ensure adequate ventilation and oxygenation. However, because pre-hospital environmental and infrastructural factors can be challenging, intubation success rates are variable [1]. When TI is performed incorrectly, it can provoke adverse events and may worsen outcome in some patient groups [2–4]. Even when performed correctly, suboptimal ventilation following TI may increase the risk of fatal outcomes in certain patient subgroups [5–9].

The use of pre-hospital TI is widely debated [see Additional data file 1], but the majority of TI-related studies are thought to be of limited value [10–12]. The core question therefore remains unanswered: does TI in the pre-hospital setting fail or result in adverse events at rates that exceed the benefits of adequately performed TI?

Rapid sequence induction (RSI) and TI are regarded as the standard of care for airway management during in-hospital emergencies. It seems reasonable that this practice should be applied in the pre-hospital phase to prevent delay in good oxygenation and ventilation. However, because of available expertise and pre-hospital external factors, several alternatives to RSI and TI are practised. Environment, equipment, procedures, provider competence, practical skills, and drug protocols vary between emergency rooms and emergency medical service (EMS) systems [13], among EMS systems [14, 15], and within EMS systems [16, 17]. These variations have been reported to influence the frequency and quality of TI and, in all likelihood, patient outcome [1, 18].

However, the heterogeneity of procedures, providers, patients, systems and monitored outcomes makes the published scientific reports difficult to interpret and compare, and inconsistency in the types of data reported exacerbates the problem. To improve reporting, an international expert panel published a consensus-based, Utstein-style template for the uniform reporting of data on pre-hospital advanced airway management [19]. The template defines inclusion criteria along with 28 core variables and 19 optional variables for documenting and reporting data. The 28 core variables are in three groups: "system variables", "patient variables", and "post-intervention variables" (Table 1). In addition, the template recommends that 12 fixed-system variables be reported (Table 2) to accurately describe the particular EMS system from which the data were collected.

Table 1 The 28 core variables for uniform reporting of data from advanced airway management in the field
Full size table
Table 2 Fixed system variables for uniform reporting of data from advanced airway management in the field, identified by an international expert group
Full size table

The aim of this study was to determine the rate of reported Utstein airway variables (28 core variables and 12 fixed-system variables) found in current scientific publications on pre-hospital TI [19].

Materials and methods

Study eligibility criteria

We included original English language articles pertaining to pre-hospital TI in adult patients. Studies that investigated pediatric cohorts and studies that focused on surgical airways were excluded. Studies that compared TI to other airway devices were also excluded.

Identification and selection of studies: data extraction

A systematic search of Medline and EMBASE databases according to the PRISMA guidelines to identify all relevant studies published prior to 1 September, 2009 was conducted (see Table 3 for search strategy) [20]. All records were converted into an EndNote bibliographic database (EndNote X1© Thompson Reuters, UK). Two reviewers (HML and MR) examined the titles and abstracts of the records for eligibility. The full texts of all potentially relevant studies were obtained, and two reviewers (HML and MR) assessed whether each study met the eligibility criteria. The reference lists of the included studies and a recent relevant Cochrane review were inspected to identify additional relevant studies [11].

Table 3 Search strategy for identification of relevant studies in Medline and EMBASE
Full size table

Study characteristics

One reviewer (HML) used a standardised Excel spreadsheet (© 2007 Microsoft Corporation, USA) and extracted information from the included studies according to the newly published template for uniform reporting of data regarding advanced airway management in the field [19]. Reported variables that matched the Utstein variables were regarded as identical, although definitions sometimes differed or remained unreported.

The data were analysed using the Statistical Package for the Social Sciences, v. 18.0 (SPSS, Inc., Chicago, IL, USA), and the distributions were reported as medians and inter-quartile ranges (IQR). Being a systematic literature review, this study did not need approval from The Regional Committee for Research Ethics or the National Social Science Services.

Results

Literature search

We identified 1,070 records in the initial search. Another six records were identified through other sources. Among these 1,076 records, 75 full-text original papers were assessed. Two of these were excluded from further analysis, one because of qualitative methodology and one being a preliminary report, leaving 73 studies for the final analysis (Figure 1).

Figure 1
figure 1

A search diagram according to the PRISMA statement.

Full size image

Characteristics of the included studies

The majority of the studies (59, 81%) were from North American EMS systems. Of these, 46 (78%) described services in which non-physicians conducted TI. In contrast, 13 (87%) of the 15 non-North American EMS systems, physicians performed the pre-hospital TI. Of the 47 non-physician-manned systems, 25 (53%) performed drug-assisted TI.

Sixty-five studies had applied an observational methodology (89%), of which 29 were conducted prospectively and 36 retrospectively [see Additional data file 1]. We identified two randomised controlled trials (RCT) and six non-RCT interventional studies.

Core variables

None of the included studies presented the complete set of 28 variables recommended in the template [19]. The maximum number of core variables reported in a single study was 21. The minimum number reported was two, whereas the median number of core variables reported from all the studies was 10 (IQR 8 to 12).

The most frequent reported core variable was "patient category", reported in 63 (86%) of the 73 studies (Table 4). The least reported variable was "co-morbidity", reported in only 2 (3%) of 73 studies (Table 4).

Table 4 Number of times (%) each Utstein variable was collected and documented among the 73 studies included
Full size table

Fixed-system variables

Of the 12 fixed-system variables, the maximum number reported in a single study was 11. The median number reported was five (IQR four to eight), and two studies did not report any of the recommended fixed-system variables. The most frequently reported variable was "service mission type", which was reported in 52 (71%) of the 73 studies (Table 4). The least frequently reported fixed system variable was "type of available ventilator", which was only reported in one paper (1%) (Table 4).

All the studies included in the review are listed, and the number of matching core variables and fixed-system variables from each study are presented in Additional file 1.

Discussion

Our systematic literature review of studies pertaining to TI of adults revealed deficient reporting of the Utstein airway core variables as defined by an international expert group. Recommended core variables, such as "post-interventional end-tidal carbon dioxide (ETCO2)", "number of attempts at airway intervention" and "co-morbidity", which are all recognised as being highly associated with efficiency and outcome, were missing in the majority of the papers. Fixed-system variables were incompletely reported or absent in most of the included studies. The low number of reported core variables makes it difficult to compare different scientific reports, assess their validity, and extrapolate to other EMS systems. One could claim that several of the included studies with a low number of documented and reported core variables in fact only report the occurrence and performance of TI within their system and therefore are not reflective of the effects or efficiency of pre-hospital TI.

Several studies have focused on the intricacy of implementing TI in the pre-hospital setting [21–23]. TI represents a complex intervention (Figure 2) that contains several separate but highly interacting components. Scientific studies on this subject are difficult to design and interpret because of tremendous variability in (and insufficient description of) operator experience, technique, and patient case-mix, making it difficult to understand or eliminate confounding factors [24]. Furthermore, neither contemporary interventions nor pre-intervention, per-intervention, or post-intervention factors highly likely to influence outcome are usually documented, analysed, or adjusted for. Key in-hospital factors (likely to be concealed from the investigator) further confound the outcome analysis [25]. This finding may explain why apparently similar studies present conflicting results and reach opposite conclusions.

Figure 2
figure 2

A cause-effect chart and factors influencing the relation between PH-TI and survival.

Full size image

RSI with oral intubation is the standard of care for drug-assisted emergency TI because it is widely recommended to be the safest way of performing this high-risk intervention [26–28]. However, only 19 (31%) of the 73 papers in this study reported the variable "drugs for airway management available on scene". Among papers that reported this variable, the definition and extent of drug assistance varied. Some services had protocols based on administering a muscle relaxant only; some combined this with a small dose of a sedative or analgesic, whereas some administered a traditional RSI. The presence or absence of drug assistance and the availability and dose of specific agents are likely to influence the success rate of TI and the rate and severity of adverse events. This information is essential to correctly assess the reported outcomes.

The majority of the included papers were based on observational studies, commonly referred to as low-quality evidence [29]. In a complex intervention, a true association between a single cause (TI) and an effect (survival) is difficult to prove (Figure 2). The presented results are flawed by multiple confounding factors, and external validity is questionable. Even randomization may fail to exclude the major confounders, a phenomenon demonstrated by Gausche et al. in one of the few randomized trials on pre-hospital TI [30]. The investigators reported no additional effect on survival or neurological outcome when paramedics performed pre-hospital TI compared with traditional bag/valve/mask ventilation in critically ill pediatric patients. The study set out to analyze the effect of the intervention itself, but due to an "intention-to-treat protocol", the intervention group was heavily confounded (abstained intubation, repetitive attempts of intubation, or failed intubation). The study instead demonstrates the effects of suboptimal provider competence and TI complications, and it illustrates the challenges of using traditional analytical techniques when assessing a complex intervention.

Several recent reviews have assessed the evidence of a pre-hospital TI effect [10, 31], including a Cochrane review [11]. They consistently conclude that the available evidence is limited and weak. It has been suggested that the traditional method of systematic review is of limited use in the evaluation of a complex intervention [32]. The lack of a standard definition of pre-hospital TI poses a significant challenge for systematic reviewers and readers of these reviews. With respect to the Cochrane review on pre-hospital TI [11], the number of studies located in our review illustrates that any strict inclusion criteria for a systematic review will exclude the majority of studies published because pre-hospital TI is often performed differently or described inadequately. It also questions the whole evidence base on which current practice is based.

Limitations

We have assessed the included studies assuming that all the recommended Utstein airway core variables are important to document for each study. Some studies focus on particular aspects of pre-hospital TI intervention and may not need to report all the core variables from the template. Nonetheless, understanding the correlations between the intervention and its outcomes presupposes that all the interacting factors are accounted for.

The Utstein airway template still requires validation. Not all the variables relevant to outcome may have been identified. In a systematic review of studies on out-of-hospital cardiac arrest, a large variability in outcome not entirely explained by variability in documented Utstein variables, was found [33].

We also acknowledge that some relevant studies may not have been located during our database search. In the future, more homogenous reporting of studies pertaining to pre-hospital TI may reduce these limitations.

Conclusions

Our systematic literature review of studies investigating TI in adults demonstrated that core data required for proper interpretation of results were frequently not recorded and reported. The inconsistent and imprecise reporting of data may be the explanation for the fact that, despite numerous published studies on this subject, there is an ongoing debate on if, when, how, and by whom pre-hospital advanced airway management should be performed. Pre-hospital TI is a complex intervention, and terminology and study design must be improved to substantiate future evidence-based clinical practice. To support this, there is a significant need for an international standard for documenting and reporting pre-hospital TI in severely ill and injured patients. The newly published template might be a first and important step in this direction [19].

Key messages

  • Studies investigating pre-hospital TI in adults lack the core data required for useful interpretation of results.

  • The published studies investigating pre-hospital TI rarely present high-quality scientific evidence.

  • Pre-hospital TI is a complex intervention, and terminology and study design must be developed to substantiate future evidence-based clinical practice.

  • A recently published template for reporting advanced pre-hospital airway management might be a first and important step in this direction.