1. How good is CXR for detection of a pneumothorax? Ultrasound? When do you get a chest CT in someone you suspect may have a pneumothorax but has a negative CXR?
The sensitivity of a chest radiograph (CXR) for the detection of a pneumothorax (PTX) depends on how it is taken, with the upright posteroanterior (PA) film being a far better study than the supine x-ray. Interestingly, there are no published studies comparing erect PA chest x-rays with CT as a gold standard. Ball, Kirkpatrick & Feliciano (2009) quote unpublished data that suggest upright PA films have a sensitivity of 92%. Other studies have shown sensitivity closer to 80-85% (Seow, 1996), including one from 1990 in which CT was compared to upright CXR as a non-inferiority trial for CT in detection of pneumothoraces (PTXs) after CT-guided thoracic biopsies (Murphy, 1990). Not surprisingly, they found that CT was as good, and indeed a bit better (though not significantly more so) at detecting PTXs than conventional radiography which caught only about 84% of them.
Supine anteroposterior (AP) chest x-rays–the staple imaging study performed on trauma patients–are a notoriously poor imaging modality for the detection of PTXs. This is due to the tendency of air to track to the least dependent pleural recess, which in the supine patient is the anteromedial space. Though the deep sulcus and double diaphragm signs, among others, can provide clues to the presence of a PTX, they are not prevalent enough to make the supine film a reliable means of detecting a partially collapsed lung. One review quotes sensitivities as low as 36-48% (Wilkerson, 2010), with even lower rates (24%) found when x-rays are interpreted by a trauma team (Ball, 2009) versus by radiologists.
At least one case report has suggested that oblique AP films may be used to aid in the diagnosis of a PTX in the supine trauma patient (Matsumoto, 2010). However, a modality being more rapidly accepted and adopted is ultrasound. While there have been many retrospective studies done evaluating the sensitivity and specificity of ultrasound for this purpose, there have been fewer prospective, blinded studies done. Wilkerson et al. (2010) performed a review of related published literature in 2010 and found four RCTs that met specified criteria. The sensitivities of ultrasound reported in these studies were 86-98%, with specificities of 97-100%. This compared with sensitivities and specificities of 28-75% and 100% respectively for supine AP chest radiographs. Similar results were found in a subsequent prospective single blinded convenience sample study, which found ultrasound to be 81% sensitive, while only 32% of PTXs were identified on the supine CXR (Nagarsheth, 2011).
There is no clean-cut answer as to when/whether to get a CT on a patient who has a negative chest x-ray. In part this is because there is much debate about the significance and management of “occult” PTXs–meaning those that are not visible on X-ray but are seen on CT. This topic will be addressed further in the next question. The British Thoracic Society’s Pleural Disease Guidelines of 2010 states that after a standard erect PA x-ray is taken, “if uncertainy exists, then CT scanning is highly desirable.” (Havelock, 2010).
Bottom line: Upright PA chest x-rays are good at detecting PTXs (~85% sensitive), supine AP films are poor (generally far less than 50% sensitive) and ultrasound is better than either (86-98% sensitive). Whether to get a CT depends on your initial imaging modality combined with your clinical suspicion for a PTX (or other thoracic injury, as in the case of blunt trauma), as well as whether you think it would change your management of the patient.
2. How do you manage occult traumatic pneumothoraces? What if the patient requires mechanical ventilation?
An occult PTX (OPTX) is defined as one that is not suspected clinically and is not evident on plain films but is identified on CT scan. The increased use of CT in trauma has led to a marked increase in the number of PTXs diagnosed (Plurad, 2007). There is much debate about how to manage these. The argument for avoiding tube thoracostomies are the high complication rate (~22%) of the procedure as well as longer associated hospital and ICU stays. Aggressive treatment, on the contrary, may prevent a worsening PTX and the development of a tension PTX.
Advanced Trauma Life Support (ATLS) guidelines state “Any PTX is best treated with a chest tube . . . Observation and aspiration of an asymptomatic PTX may be appropriate, but the choice should be made by a qualified doctor; otherwise, placement of a chest tube should be performed.” An increasing number of studies (both retrospective reviews and prospective RCTs) argue against invasive interventions for OPTX, favoring observation for signs of clinical progression. These studies note a non-significant difference in the progression of OPTX, incidence of pneumonia, and mortality between those with and without thoracostomies.
Supporting this view is a review and analysis of the related RCTs which concluded that observation may be at least as safe and effective as the placement of chest tubes for the management of the occult PTX (Yadav, 2010). Other reviews of current pertinent literature found that it appears safe to observe patients with small to moderate PTXs (Ball, 2009; Mowery, 2011). Unfortunately, most of the available studies and trials have small sample sizes and are typically only Level III evidence. Ideally, the treatment course would be dictated by the ability to predict which OPTX will remain stable and resolve and which will progress; however, no study has been prospectively validated to accomplish this.
While there seems to be increasing agreement in managing small to medium-sized PTXs with observation, there is less consensus on what to do with the occult PTX in a ventilated patient. ALTS guidelines state that all patients with PTXs who are undergoing positive pressure ventilation (PPV) should have chest tubes placed (PPV theoretically can turn a small PTX into a tension PTX). However, the evidence is mixed. Two out of three small randomized control trials reviewed by Yadav, et al. (2010) found no difference in outcomes for patients with OPTX who had PPV. The third (Enderson, 1993) found a higher rate of complications, including tension PTXs, in vented patients. However, recent practice guidelines published in the Journal of Trauma concluded from the same evidence that the studies “would support the notion that the majority of patients with occult pneumothoraces will not have progression regardless of the presence of positive pressure ventilation” (Mowery, 2011). This sentiment seems to be supported by a growing body of retrospective and prospective studies (Barrios, 2008; Mahmood, 2011).
3. In patients with a primary spontaneous pneumothorax, do you automatically place a chest tube? When might you consider a pigtail catheter or even simple needle aspiration?
Historically, most PTXs, including primary spontaneous pneumothoraces (PSPs) were treated with large bore (>24F, as defined by the American College of Chest Physicians) chest tubes. Current recommendations regarding their management are changing based on emerging evidence that more conservative measures such as small-bore catheters (<14F), needle aspiration, or even observation are reasonable and possibly even preferable. When weighing the various approaches, one must consider several factors including: patient stability, whether this is an initial or recurrent PTX, size of the PTX, efficacy of the procedure as well as associated pain/discomfort, complications, likelihood of initial success of the procedure, and likelihood of recurrence of the PTX.
For this discussion, we will take the case of a hemodynamically stable patient with a first occurrence of PTX that is large enough to require intervention. It is generally accepted that in the stable patient with a first episode of a small PTX, observation and outpatient management (assuming no progression after observation) is appropriate (Baumann, 2001; Macduff, 2010). “Small” is defined variably by different groups. The American College of Chest Physicians defines it as less than 3cm apex-to-cupola distance (Baumann, 2001), whereas the British Thoracic Society defines it as >2cm interpleural distance at the level of the hilum (Macduff, 2010).
The American College of Chest Physicians 2001 Delphi Consensus statement on the management of spontaneous PTXs recommends either chest tube or pleural catheter for large PTXs. They state that there is no role for aspiration except possibly in the situation of a small PTX that has progressed during an observation period in an otherwise stable patient (Baumann, 2001). On the other end of the spectrum is the British Thoracic Society (BTS) whose 2003 guidelines recommended simple aspiration as the first line treatment of “all primary pneumothoraces requiring intervention” (Henry, 2003). The 2010 update of the BTS guidelines takes a slightly more nuanced approach stating that while they believe that needle aspiration remains the procedure of first choice in most cases, they also recommend taking into account operator experience and patient choice when deciding on an approach (MacDuff, 2010).
These differences are in part a result of timing of the guidelines and partly due to the paucity of relevant data. There are few retrospective studies and even fewer well-powered and methodologically-sound randomized controlled trials comparing any two–let alone all three–of the modalities being considered here. The most studied comparison is that between needle aspiration and large bore chest tube placement which boasts four RCTs: Noppen, 2002; Ayed, 2006; Harvey, 1994; Andrivet, 1995. Comparatively, there is only one small RCT comparing small bore catheters with aspiration and no RCTs comparing small-bore with large bore chest tubes (although there is an observational study (Inaba, 2012) of chest tubes in trauma that found no difference between small versus large). Additionally, there are four systematic reviews analyzing the four aspiration versus large bore chest tubes RCTs previously mentioned (Zehtabchi, 2008; Wakai, 2007; Aguinalde, 2010; Devanand, 2004).
The available randomized controlled trials comparing aspiration with tube thoracostomy evaluated both immediate and delayed (one week – 1+ years) success rates of each modality for treatment of spontaneous PTXs of all sizes. Immediate success rates were found to be fairly similar between the two, with aspiration succeeding from 59.3%-71% of the time across the four trials, and tube placement being effective 63.6%, 68% and 93% of the time, respectively, in the three studies that reported it (Noppen, Ayed, Andrivet). The high outlier was likely due to the generous definition of success used for tube placement in Andrivet’s trial: a tube only failed if there was a persistent leak after 10 days. One year recurrence rates were not significantly different between needle aspiration and chest tubes (~ 25%) (Noppen, Ayed, Harvey). Reported complication rates for aspiration were generally equal to or lower than that of tube placement, ranging from 0-2%. Patients with aspiration had a significantly lower admission rate (Noppen, Ayed), and either a significantly shorter hospital stay or a trend towards it (Noppen, Ayed, Harvey). They also had lower analgesic requirements and pain scores (Ayed, Harvey).
All these trials suffered from low numbers, with the largest including 137 patients (Ayed). Taken together, they included just 331 randomized patients. Nonetheless, all studies came to the conclusion that simple aspiration was an acceptable, and perhaps even preferred, first line approach to primary PTXs. The systematic reviews that evaluated these studies found there to be insufficient efficacy data to be conclusive, but that the pooled data suggest that there is no significant difference between aspiration and tube thoracostomy in terms of immediate and one-week outcomes or with respect to recurrence rates at one year. They also found that there was a significantly lower admission rate and length of hospital stay among patients who had needle aspiration. They concluded that simple aspiration was a reasonable alternative and possibly preferred strategy in the initial management of primary spontaneous PTXs requiring intervention(Zehtabchi, 2008; Wakai, 2007; Aguinalde, 2010; Devanand, 2004).
But, what of pigtail catheters? The question of whether and when to use small-bore catheters is addressed mostly by retrospective studies. Several such studies found that small-bore catheters (these included pigtail catheters, standard small bore pleural catheters, and single lumen central venous catheters) were no less effective than large bore chest tubes in the treatment of spontaneous PTXs (Vedam, 2003; Liu, 2003; Contou, 2012; Kulvatunyou, 2011). One small RCT out of Singapore that compared small-bore catheters to aspiration found a non-significant trend towards decreased admissions (from ED or from 3-day outpatient re-evaluation) with the catheters (44%) than with aspiration (61%). There was no significant difference in failure rates, complication rates, or pain scores. The study concluded that both methods allowed for safe management of PSPs. Of note, the study included only 48 patients, having failed to reach their target of 100 due to poor recruitment and lack of funding (Ho, 2001).
Bottom Line: Despite the lack of large RCTs, the data from the existing trials and retrospective studies support using either needle aspiration or small bore catheters (especially those with a Heimlich valve which allow the patient to move about or even be discharged with it in place) in lieu of large bore chest tubes in the stable patients with first-time moderate-to-large spontaneous PTXs. There is insufficient data to recommend aspiration versus a small bore tube at this time. Factors associated with failure of aspiration include: patient age > 50, initial aspiration volume > 2.5 L (this suggests a pleural leak), and possibly initial PTX size (some studies have shown higher failure rates with PTX size > 40% as measured on CXR, but this is an actively debated risk factor) (Chan, 2008).
Reflecting these guidelines is a flowchart for managing the spontaneous pneumothorax, taken from the BTS 2010 guidelines (Macduff, 2010).
4. When do you do a needle thoracostomy? Where do you prefer to put the needle?
Needle thoracostomy is indicated when life-threatening tension PTX is suspected. Traditionally, the recommended method for decompressing such a PTX has been to place a standard 5-cm long over-the-needle catheter into the pleural space at the level of the second intercostal space in the mid-clavicular line (ATLS). However, some have expressed concern that this location may not be optimal for reasons both of safety and likelihood for success.
Safety concerns relate to the proximity of major vessels such as the internal mammary artery (IMA) and the subclavian vessels to the traditional needle thoracostomy site. As noted above in Question #3, exceedingly few complications were found in all the RCTs that studied needle aspiration (all of which were performed at the 2nd or 3rd intercostal space (IS) at the mid-clavicular line (MCL)). However, one group reported three cases of life-threatening hemorrhage that occurred in a six-month period in patients who had had a needle thoracostomy performed in the 2nd intercostal space, mid-clavicular line (Rawlins, 2003). Consequently, they propose that the traditional location for tube thoracostomies–the 5th intercostal space just anterior to the mid-axillary line–may be safer for needle decompression as it is farther from large vascular structures (Rawlins, 2003).
Concerns about the efficacy of decompressing a PTX using the traditional method relate to the thickness of the chest wall at the level of the 2nd intercostal space. CT scans demonstrate a chest wall thickness between 4 and 4.5 cm thick at the 2nd intercostal space, mid-clavicular line. The corresponding expected failure rate of penetrating the pleura with a 5cm needle was found be as high as 50%. (Stevens 2009; Givens 2004; Zengerink 2008; Sanchez 2011). All of these studies conclude that a longer needle is likely necessary to increase chances of success at needle thoracostomy, while also acknowledging the associated increased risk of damaging internal structures.
So, is a lateral approach better, given these safety concerns and a potentially high failure rate of the anterior approach? A few studies have addressed this question with conflicting results.
Two imaging-based studies evaluated the chest thickness not only of the anterior wall, but also that of the lateral wall, at the 5th intercostal space along the mid-axillary line (Wax, 2007; Sanchez 2011). Both of these studies found that the chest wall thickness was thinner anteriorly (averaging 3.1cm and 4.6cm in the respective studies) than at the mid-axillary line (3.5cm and ~5.3 cm, respectively), thus making the former a better location for needle thoracostomy. Wax, et al. (2007) also evaluated the distance from these entry sites to key internal structures such as the heart and liver, as well as the distance from the anterior approach to the internal mammary arteries. They concluded that the anterior approach was safest based on its relatively greater distance from key organs and on its sufficient distance from the IMA (>3cm). Both studies recommended using a 7cm needle for increased success rate at penetrating the pleura.
A countering opinion was rendered by Inaba, et al. (2011) based on a cadaveric study in which they performed needle thoracostomies at both the 2nd IS MCL and 5th IS MAL in both right and left chest walls of twenty un-enbalmed cadavers. After having placed the catheters, bilateral thoracotomies were performed to assess penetration into the pleural cavity. They found that 100% of those placed laterally at the 5th intercostal space were successful, versus only 58% of those placed anteriorly at the 2nd intercostal space. They also measured the chest wall thicknesses and found the lateral wall (3.5cm +/- 0.9cm) to be significantly thinner than the anterior wall (4.5cm +/- 1.1cm) (Inaba, 2011).
The reasons for these differing measurements and results are not immediately clear. One possible explanation is that while the exact ages of the cadavers in Inaba, et al.’s study were not known, they were stated to be “higher than the average trauma cohort” and “their muscle mass may have been more atrophic.” Perhaps this changed the relative thicknesses of the various parts of the chest. Another plausible explanation is that the wall thickness along the lateral wall may vary appreciably within relatively small distances. Wax, et al. found that the chest wall thickness along the 5th intercostal space was 3.5 cm at the mid-axillary line but only 2.9cm at the anterior axillary line. This dramatic difference in a short distance may well account for variances in relative chest wall thickness measurements and in success rates in pleural penetration in the cadaver study.
Bottom line: There is ample data demonstrating that either approach to needle thoracostomy may fail when using a standard sized needle. The answer to improving efficacy of needle decompression may lie in using longer needles and catheters, keeping in mind that this increased length will increase risk of damaging other internal structures. Prospective studies are clearly warranted to further study this question, until which time it may well be advisable to use the more traditional anterior approach, especially if using a longer-than-standard needle.