Serious thoracic trauma accounts for one in four trauma deaths in the UK. This is more than 4000 deaths per year (National Institute for Health and Care Excellence (NICE), 2016). The recognition and successful management of life-threatening complications in this patient group will increase survival.
Tension pneumothorax is a life-threatening condition occurring in 1 in 250 patients experiencing major trauma (Leech et al, 2013), and a failure to recognise and intervene will lead to patient deterioration. A pneumothorax occurs when air leaks into the pleural cavity from the lungs or from outside the body. This becomes a tension pneumothorax when continued inspiration allows more air into the pleural cavity. The negative pressures in the lungs, needed for spontaneous breathing, are unable to overcome the atmospheric pressure in the pleural cavity, and the lung begins to collapse, collapsing further with each breath.
Traumatic cardiac arrest will ultimately occur because cardiac output is impeded by the affected lungs compressing the intrathoracic vessels. This reduces venous return to the heart, causing circulatory instability and eventually collapse.
In these cases prehospitally, needle decompression (NDC) is generally the first-line treatment to reverse the cause or prevent a cardiac arrest. In the UK, NICE guideline NG39 on major trauma recommends the use of a 14 g, 5 cm long cannula as the primary means for NDC (NICE, 2016). The cannula's insertion into the pleural cavity allows trapped air to escape and the lungs to ventilate.
Advice around site selection differs internationally. In the UK, the Faculty for Pre-Hospital Care (FPHC) (Leech et al, 2013) recommends the second intercostal space (ICS) at the mid-clavicular line (MCL) as the primary choice, with the fifth ICS slightly anterior to the mid-axillary line (MAL) as the secondary site. In comparison, the American College of Surgeons' (ACS, 2018) advanced trauma life support (ATLS) guidance and the committee on tactical combat casualty care (TCCC) (Defense Health Agency (DHA, 2020)) recommend the fifth ICS, slightly anteriorly to the MAL, as the first mentioned site for NDC.
NDC can fail for several reasons. The FPHC (Leech et al, 2013) discusses needle obstruction by blood, tissue or kinking, and recommends flushing the cannula with 2 ml of air or water if there is no obvious air release. It can also fail because of incorrect placement; the ATLS (ACS, 2018) guidance reinforces the importance of clinicians being able to find the correct site through landmarks and the need for diligent reassessment because re-tensioning is possible following successful NDC.
Increased success rates are reported by the committee on Tactical Combat Casualty Care (DHA, 2020) than in the other guidance. This may be because it recommends an 8 cm rather than 5 cm needle and catheter. However, catheters of this length are not commonly used in UK prehospital practice, nor are alternative procedures such as a finger thoracostomy.
This literature review will seek to determine the best approach for needle decompression using a standard 14 g, 5 cm cannula.
Methods
This literature review was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidance on identifying and including reviews of research and evaluation of interventions (Moher et al, 2009).
A systematic search of multiple databases from their inception to 2 March 2020 was conducted. Databases included Medline, Cumulative Index to Nursing & Allied Health (CINAHL) Plus and Embase with a hand search of key paramedic texts from the preceding 12 months. The terms of the search (Table 1) were constructed around the population, intervention, comparison and outcome (PICO) principle and implemented using a Boolean search strategy.
| Population | Intervention | Comparison | Outcome |
|---|---|---|---|
| Pneumothorax |
Needle decompression |
Mid-clavicular OR anterior |
Success OR successful decompression |
Concepts were combined using the AND Boolean operator
No date parameters were set around the search to ensure a complete data capture. The search was limited to published research papers in English that focused on needle decompression at the anterior or lateral approach. Articles were excluded if they were focused on a paediatric population or conducted post-mortem. Cadaveric studies were excluded as specimen quality could be questionable; there are likely inconsistencies when compared to live patients, such as those regarding muscle atrophy, and differing cadaveric samples.
The studies included were then critically appraised using a Critical Appraisal Skills Programme (CASP) (2018) checklist for cohort studies. Data were then captured on the mean chest wall thickness (CWT) at the anterior second ICS and the fourth and/or fifth ICS laterally.
Where available, additional data were gathered around age, body mass index (BMI), differences between the sexes and perceived success rates at the anterior and lateral approaches with a 5 cm catheter. The results were tabulated for analysis and descriptive statistics were calculated, then evaluated to identify themes and trends.
The search returned 31 studies; no additional studies were found on hand searching key texts for recent articles that may not have made it into the databases at the time of the searches. Duplicates were removed, leaving 17 studies. Six studies were excluded through title and abstract review and four were removed as the full text was never published. Of the remaining seven, a further three were excluded through the inclusion/exclusion criteria, leaving four articles to be included in the review (Table 2). All articles included were hospital-based studies because computerised tomography (CT) was used for measurements.
| Study | Population, location and cohort | Findings |
|---|---|---|
| Chanthawatthanarak et al (2019) | Srinagarind, Thailand. 155 adults who underwent chest CT between January 2016 and July 2017 | Retrospective review of chest CT, measuring 4 sites. 2nd ICS MCL and 5th ICS AAL bilaterally. Conclusion was average CWT was lower at the 2nd ICS MCL than at the 5th ICS AAL. CWT was greater in women and people with a raised BMI |
| Goh et al (2018) | Singapore. 593 Adults presenting to the emergency department with trauma between 2011 and 2015 | Retrospective review of chest CT, measuring 4 sites. 2nd ICS MCL and 5th ICS MAL bilaterally. Conclusion was mean CWT is significantly shorter in the lateral than in the anterior approach |
| Sanchez et al (2011) | Boston, US. 159 adults presenting to the emergency department with blunt trauma between June 2008 and February 2009 | Retrospective review of chest CT, measuring 6 sites. Anterior 2nd ICS, Lateral 4th and 5th ICS bilaterally. Conclusion was that with commonly available angiocatheters, the lateral approach is less likely to be successful |
| Wax and Leibowitz (2007) | New York, US. 100 randomly selected adults from CT archive in the 12 months before data retrieval | Retrospective review of CT chest, measuring 6 sites. Measurements made at the MHL at the level of the sternal angle, and at the xiphoid process level at AAL and MAL bilaterally. Conclusion states that NDC should be performed at the MHL with a needle ≥7 cm. Study also showed smaller CWT at anterior sites but risk increases with a longer catheter |
AAL: anterior axillary line; CWT: chest wall thickness; ICS: intercostal space; MAL: mid-axillary line; MCL: mid-clavicular line; MH: mid-hemithoracic line
Results
In total, 4546 measurements were taken from 1007 patients. All studies included bilateral measurements at the selected sites (which differed across studies) and all included measurements at the second ICS. Three studies measured the second ICS at the MCL and one at the mid hemithoracic line; this difference was judged to be insignificant and these results were gathered as one data set. At the lateral aspect, two studies gathered data on both the fourth and fifth ICS and two gathered data solely on the fifth ICS. Of the data gathered at the fifth ICS, the majority were collected from the MAL, with one study measuring at the anterior axillary line (AAL).
Wax and Leibowitz's (2007) study collected data at the fourth and fifth ICS at the AAL and MAL, respectively. It is unclear why these two different sites were chosen on the same aspect. Similarly, in Sanchez et al's (2011) study, the MAL was chosen for the measurement of the fourth and fifth ICS. Measurements across the fifth ICS-AAL would have been more in line with recommendations from the FPHC (Leech et al, 2013) and the TCCC (DHA, 2020), although it is uncertain whether local practices or guidelines recommended other approaches at the time the present research was carried out.
The majority of study participants were men (74%); BMI data were available on only 84% of participants and the mean BMI was 24.4 kg/m2. The mean age of participants across the studies was 51.6 years.
Table 3 depicts the mean chest wall thickness at anatomical sites bilaterally as well as by sex. The measurements in all studies were done by CT. The skin-to-pleura distances were shorter at the anterior approach in two studies and at the lateral approach in two studies. There was no apparent difference in chest wall thickness between the left and the right sides in any study. As can be seen in Table 3, Goh et al (2018) did not supply raw data by sex and stated only that there was no difference when the anterior approach was taken. This is unusual as all other studies showed some variation at all sites, some quite substantial, even at the anterior aspect. The lack of appraisable data, with the stated finding unsupported by a published dataset leaves some suspicion of bias. The remainder of the study shows a thorough breakdown of data around ethnicities, leading to the possibility that a matched result is feasible and that the author was not focused on differences between women and men.
| Chanthawatthanarak et al (2019) | Goh et al (2018) | Sanchez et al (2011) | Wax and Leibowitz (2007) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean | Male | Female | Mean | Male | Female | Mean | Male | Female | Mean | Male | Female | |
| Anterior 2nd ICS (left) | 26.8 | 24.9 | 32.3 | 40.4 | No reported difference | 45.2 | 45.9 | 43.6 | 31.0 | 30.0 | 31.0 | |
| Anterior 2nd ICS (right) | 27.1 | 25.2 | 32.6 | 39.2 | No reported difference | 46.3 | 46.9 | 45.0 | 31.0 | 31.0 | 33.0 | |
| 1Lateral 4th ICS (left) | 62.1** | 64.1** | 57.8** | 26.0* | 23.0* | 32.0* | ||||||
| Lateral 4th ICS (right) | 63.7** | 66.6** | 57.6** | 27.0* | 24.0* | 35.0* | ||||||
| Lateral 5th ICS (left) | 35.9* | 32.5* | 54.4* | 35.2** | Mean +6.9 greater in women than men | 52.9** | 53.8** | 51.0** | 34.0** | 33.0** | 37.0** | |
| Lateral 5th ICS (right) | 36.7* | 33.8* | 54.2* | 36.2** | Mean +6.9 greater in women than men | 53.8** | 55.4** | 50.4** | 35.0** | 35.0** | 39.0** | |
ICS: intercostal space
Table 4 shows that, of the studies that measured the lateral fourth and fifth ICS, one shows a mean difference favouring the fifth ICS by 9.5 mm, while the other shows a shorter distance at the fourth ICS by 8 mm. Table 4 also shows that three of the studies show women have a thicker chest wall laterally. One study shows the opposite. At the anterior aspect, two studies show women have a thicker chest wall, one shows no difference and one shows women to have a thinner chest wall than men.
| Chanthawatthanarak et al (2019) | Goh et al (2018) | Sanchez et al (2011) | Wax and Leibowitz (2007) | Overall mean chest wall thickness | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean | Male | Female | Mean | Male | Female | Mean | Male | Female | Mean | Male | Female | ||
| Anterior 2nd ICS | 26.7 | 25.1 | 32.5 | 39.8 | 45.8 | 46.4 | 44.3 | 31.0 | 30.5 | 32.0 | 35.8 | ||
| Lateral 4th ICS | 62.9** | 65.4** | 57.7** | 26.5* | 23.5* | 33.5* | Insufficient data | ||||||
| Lateral 5th ICS | 36.3* | 32.5* | 54.3* | 34.7** | Mean +6.9 greater in women than men | 53.4** | 54.6** | 50.7** | 34.5** | 34.0** | 38.0** | 39.7 | |
ICS: intercostal space
Combined data show the chest wall is marginally thicker at the lateral fifth ICS (39.7 mm) than the anterior second ICS (35.8 mm). There were insufficient data to meaningfully combine the lateral fourth ICS results. The difference between the upper and lower overall mean distances measured at the second ICS was 19.1 mm and at the fifth ICS it was 18.9 mm.
Table 5 shows the perceived success rates at NDC using a 5 cm catheter for each of the studies that contained these data. Perceived success was classified as a 5 cm catheter being able to enter the pleural cavity at the measured site. Combined data from 907 participants and 3628 measurements showed success rates between the anterior second ICS and the lateral fifth ICS to be 79.70% and 80%, respectively.
| Sanchez et al (2011) | Goh et al (2018) | Chanthawatthanarak et al (2019) | Total participants and mean success rates | |
|---|---|---|---|---|
| Study participants | 159 | 593 | 155 | 907 |
| Anterior 2nd ICS | 66.40% | 78.80% | 96.77% | 79.70% |
| Lateral 5th ICS | 44.70% | 88.20% | 84.84% | 80.00% |
ICS: intercostal space
Wax and Leibowitz's (2007) study does not display the data for success with a 5 cm catheter. They do state that in their study, there would be a 50% success at 3.1 cm and 100% with a 7 cm catheter at the anterior approach. The decision to use 3.1 cm as a marker limits the studies' applicability as this does not relate to any method of decompression. They also do not state any expected success rates at the lateral approach on either site other than a catheter >7 cm would be required for some patients. These omissions in data affect the validity of the study to this review, especially when the mean data available show apparent differences between the fourth and fifth ICS and apparent differences between men and women at the fourth ICS-AAL. These all go unexplained and may point toward some bias regarding their selection of the more common anterior site versus the lateral fifth ICS. However, the study may have identified safety issues at the fourth ICS as they were also looking into underlying vital structures and could have ruled this site as inappropriate or too risky.
Discussion
Because a tension pneumothorax is life threatening, it is important that clinicians are aware of the appropriate anatomical sites and the procedure for decompression. It is recognised by NICE (2016) that the prehospital setting creates many additional challenges because of varied and unpredictable environments. However, correct technique and accurate site selection will maximise the efficacy of NDC in these life-threatening cases, regardless of environmental challenges.
The FPHC (2013) favoured the site of the second ICS on the MCL as being likely to be the easiest to access on most patients; landmarks such as the sternal angle at the manubriosternal joint are easy to find, even on obese patients and the likelihood of these prehospital patients experiencing polytrauma and being immobilised often limiting lateral access. The same can be said when transporting a patient to hospital and the limited access around the stretcher.
With varying recommendations on primary site selection from different bodies such as the FPHC, TCCC and ACS, an understanding of the sites and potential differences should help inform clinical decision-making when faced with a patient in severe respiratory compromise or haemodynamic instability.
Chest wall thickness
While there was a range of mean measurements across the studies, differing by 19.1 mm at the anterior second ICS and 18.9 mm at the lateral fifth ICS, the comparison between the left and the right sides showed no apparent difference at the anterior second ICS or the fifth ICS. This was not surprising as, despite individual differences, patients would not be expected to show much disparity regarding chest symmetry.
However, in patients with a tension pneumothorax, an asymmetrical chest could be seen because of subcutaneous emphysema, chest wall trauma or unilateral chest rise and fall caused by the collapse of one lung. While two studies focused on trauma patients, only Sanchez et al (2011) specifically excluded those with subcutaneous emphysema. This factor could have increased the average distances measured on the CT scans in Goh et al's (2018) study. Data were not available on the number of patients scanned with this pathology and it could be argued that this could be more representative of patients requiring NDC. A study specifically focusing on chest wall measurements of patients with chest wall trauma or those with a diagnosed pneumothorax would likely capture a population more representative of those requiring or having received NDC. This kind of study would, by default, capture the correct patient demographics of those studied such as age, sex, BMI and ethnicity through data capture at the hospital.
There is a paucity of data available for the AAL in this study. As this is the recommended site for the lateral approach from both the FPHC (Leech et al, 2013) and TCCC (DHA, 2020), further study is recommended into chest wall thickness at this site, specifically over the more commonly studied MAL, especially when considering the lateral approach as a primary site for NDC.
Patient sex
In the four studies reviewed, only minor differences were found between the chest wall thickness of men and women at the anterior second ICS. At the lateral fifth ICS, Chanthawatthanarak et al (2019) showed women to have a thicker chest wall, with a mean difference of 21.8 mm; the other studies showed differences of <7 mm, with only Sanchez et al (2011) showing men to have thicker chest walls at both sites.
This is in contrast to most studies, which have shown women generally have thicker chest walls. Inaba et al's (2011) cadaveric study showed that chest wall thickness was about 1 cm greater in women than in men. Laan et al's (2016) systematic review of chest wall thickness showed, on data pooling, women had a considerably thicker chest wall than men at both the second ICS-MCL and the fourth/fifth ICS-MAL; no evaluation was made at the fourth/fifth ICS-AAL. However, Inaba et al (2011) showed that chest wall thickness increased with BMI at the AAL.
Sanchez et al (2011) hypothesise that the military population of their study may account for the difference, conceivably through both sexes having a leaner physique, with an increased muscle mass anteriorly for males and decreased subcutaneous tissue laterally in females. Wax and Leibowitz (2007) discuss the possibility that, while women may be expected to have more subcutaneous chest wall (including breast) tissue at all sites than men, men tend to have a higher BMI.
Significance of BMI
A higher BMI was found to correlate directly with a greater chest wall thickness by all the studies that collated this information.
Sanchez et al (2011) found noticeably thicker chest wall averages, but did not collect data on BMI, which may have given the reason for this. World Health Organization (WHO) (2018a) data on BMI shows that 72.7% of Americans are overweight with 37.3% being classed as obese. This is compared to only 21.5% of the population of Southeast Asia being overweight (WHO, 2021).
Unsurprisingly, Powers et al (2014) showed sizeable failure rates in obese patients in the United States using a 5 cm catheter at the second ICS-MCL; failure rates were 25% in underweight patients but 92.9% in obese patients. While a comparison between men and women is of interest, the variable results mean that, when NDC is required, greater consideration should be given to the possibility of failure because of a patient's BMI and less so to their sex.
Needle decompression success rates
Measurements of success at NDC with a 5 cm catheter were available for 90.1% of the study population with measurement of 3628 sites. Success rates between the anterior second ICS and the lateral fifth ICS were shown to be only 0.3% different (favouring the anterior approach). As such, there is no apparent difference between the anterior or lateral approach with a 5 cm catheter.
This is favourable for both the clinician and the patient, as it may be difficult to access a specific site on scene or during transport. Additionally, there may be problems in finding landmarks because of patient injuries or excess subcutaneous tissue. This may require the clinician to choose a site according to the circumstances and this review has shown that success is equally likely using either approach.
Nearly three-quarters (74.3%) of the study population was Asian, limiting the review's applicability to a Western population. While Goh et al (2018) excluded Caucasians from their sample, none of the other studies separated ethnicities for their analysis. The latest available UK census data taken in 2011 from the Office for National Statistics (2018) shows Asian ethnic groups to make up 7.5% of the UK population with >80% being white British.
Regardless of ethnic diversity across the studies, the significance of BMI in successful NDC should not be underestimated. Approximately 30% of the UK population is obese (WHO, 2018b) and, while no specific data were readily available on BMI in UK major trauma, the BMI of the UK's population is more closely aligned to that of the US than Southeast Asia. Therefore, clinicians should expect higher failure rates with a 5 cm catheter than those perceived in this review. Sanchez et al's (2011) study of patients in the United States showed lower success rates with a 5 cm catheter than the other included studies in this review.
Catheter length
Many studies recommend use of 7 cm and even 8 cm catheters for NDC for adults. The TCCC (DHA, 2019) guidelines recommend an 8 cm catheter at the fifth ICS-AAL. The ATLS manual (ACS, 2018) suggests that this length of catheter will be successful in >90% of patients. Both of these publications are American.
Chang et al (2014) raise concerns over the possibility of damage to vital structures such as the heart, liver and major vessels, particularly if an acute entry angle is used. There is the potential to cause major haemorrhage or cardiac tamponade. This may account for the hesitancy in UK practice to recommend a longer needle, given the relatively low incidence of tension pneumothorax and a relatively high complication rate of NDC at ≤11% (Leech et al, 2013).
Commercial products with large-bore, long needles and safety features to reduce the risk of iatrogenic injury are available. Devices such as the ARS Needle (North American Rescue) and the ThoraQuik (BTME Group) chest decompression device offer extended needle lengths of 8.25 cm and 10 cm respectively. However, these are expensive compared to the equipment generally used now.
Limitations
This review is limited by several factors. The lead author is the lone researcher and, as such, has had limited oversight throughout data collection. The retrospective nature of the studies analysed is problematic. There is also a lack of information on patient outcome following decompression and its impact on survival.
The various sites chosen for data collection are not all supported by UK recommended best practice and the heterogeneity of the population is not aligned with UK demographics. This makes the results of the study somewhat difficult to relate to UK practice.
Further studies into UK and European chest wall thickness and success at NDC using a 5 cm catheter are recommended. A prospective designed study would address a limitation of this review as a retrospective analysis of papers of the same kind. Factors such as tension pneumothorax specifically in trauma and the heterogeneity of the population could be addressed, as could prehospital elements.
Additionally, using CT scan measurements presumes perfect placement and does not account for changes in chest wall thickness through respiratory effort or trauma. However, CT gives an accurate measurement of chest wall thickness and gives an otherwise unobtainable view. Its use for all of the included studies has meant that the most accurate measurements are provided. This could not be reliably done through ultrasound because of its compressible nature, nor through cadaveric study because of muscle atrophy in the often elderly cadavers in different conditions.
Recommendations and conclusion
There is no apparent difference between success rates at the anterior or the lateral approach using a 5 cm, 14 g cannula; neither was any meaningful difference found between the left or the right side. NDC can be expected to fail in 20% of patients and deviations from the correct site for insertion or using an oblique angle makes failure more likely. Training to ensure familiarity with the correct landmarks and technique is encouraged to guarantee the highest probability of success.
None of the studies in this review collected data on UK patients. Further study is recommended in a UK trauma population using CT methodology.
The higher the BMI of a patient, the greater the chance of the intervention failing: an increased catheter length should be available for patients with high BMI. This is especially important in the prehospital environment, where additional help is not always nearby and considerations around additional resources, patient extrication and transport times need to be considered.
Successful NDC may not last and the patient may experience tension again. Continued assessment of the patient is vital and the consideration of secondary sites or alternative procedures is an important part of patient care.