Three-step management of pneumothorax: time for a re-think on initial management

Abstract

Pneumothorax is a common disease worldwide, but surprisingly, its initial management remains controversial. There are some published guidelines for the management of spontaneous pneumothorax. However, they differ in some respects, particularly in initial management. In published trials, the objective of treatment has not been clarified and it is not possible to compare the treatment strategies between different trials because of inappropriate evaluations of the air leak. Therefore, there is a need to outline the optimal management strategy for pneumothorax. In this report, we systematically review published randomized controlled trials of the different treatments of primary spontaneous pneumothorax, point out controversial issues and finally propose a three-step strategy for the management of pneumothorax. There are three important characteristics of pneumothorax: potentially lethal respiratory dysfunction; air leak, which is the obvious cause of the disease; frequent recurrence. These three characteristics correspond to the three steps. The central idea of the strategy is that the lung should not be expanded rapidly, unless absolutely necessary. The primary objective of both simple aspiration and chest drainage should be the recovery of acute respiratory dysfunction or the avoidance of respiratory dysfunction and subsequent complications. We believe that this management strategy is simple and clinically relevant and not dependent on the classification of pneumothorax.

INTRODUCTION

Pneumothorax is a common disease worldwide, but surprisingly, its initial management remains controversial. Pneumothorax is generally classified as spontaneous, which occurs without preceding trauma; traumatic, which occurs as a result of direct or indirect trauma and iatrogenic. In addition, spontaneous pneumothorax is subclassified as primary spontaneous, which occurs in young patients without obvious underlying lung disease, or secondary spontaneous, which occurs as a complication of an underlying lung disease. The management of pneumothorax varies depending on whether it is primary or secondary [1, 2]. There are some published guidelines for the management of spontaneous pneumothorax. However, unfortunately, they differ in some respects, particularly in initial management. The consensus process of the American College of Chest Physicians guidelines showed simple aspiration to be rarely appropriate in any clinical circumstance [3]. On the other hand, the British Thoracic Society guideline 2003 recommends simple aspiration as the first-line treatment for all cases of primary spontaneous pneumothorax requiring intervention [4]. Another issue is that, in actual clinical practice, there is considerable deviation from published guidelines, and this is internationally observed [5–7].

There is a need to outline the optimal management strategy for pneumothorax. In this report, we systematically review published randomized controlled trials, point out controversial issues and finally propose a three-step strategy for the management of pneumothorax on the basis of the review and some other reported data. We believe that the management strategy for pneumothorax should be simple and clinically relevant and not dependent on the classification of pneumothorax.

SYSTEMATIC REVIEW AND META-ANALYSIS FOR PRIMARY SPONTANEOUS PNEUMOTHORAX

Some randomized controlled trials and meta-analyses [8–10] of pneumothorax treatment have been conducted. Here, we review these previous reports again. We searched the PubMed database (National Library of Medicine) for reports from January 1999 through August 2011. We used the following search terms: pneumothorax, combined with observation, aspiration or thoracocentesis and chest drain, chest tube or tube drainage. The search was limited to randomized controlled trials. The search and the review finally yielded four randomized controlled trials (Table 1), with a combined total of 331 patients [11–14]. All four trials compared simple aspiration with tube drainage. Meta-analysis was performed by combining the reported success rate, hospital stay and recurrence rate. Relative risk (RR) was used as a summary statistic for dichotomous outcomes and weighted mean difference (WMD) for continuous outcomes. All statistical analyses were conducted using Comprehensive Meta-analysis Software version 2 (Biostat, Englewood, NJ, USA). The software weighs the data from various publications according to their cohort size. The data can be visualized graphically. Each horizontal line represents the 95% confidence interval (95% CI) for each study. The square size represents the cohort size. The diamond at the bottom of the graph represents the overall odds ratio. The width of the diamond is proportional to the overall 95% CI. The data can be analysed using a fixed or a random model. In the fixed model, the effect sizes of each study are conserved. The random model uses a more flexible approach in which the effect sizes are normalized towards an overall mean effect size. We analysed all data using both models. The outcome for success rate could not be precisely combined because of differences in the outcome definitions. A pooled result for success after 1 week or more was considered as the success rate. Integrated analysis revealed that simple aspiration was associated with shorter hospitalization time (WMD, –1.61 days; 95% CI –2.30 to –0.92; Fig. 1). With regard to the success rate, however, simple aspiration tended to be less favourable than tube drainage, but this difference between the two interventions was not statistically significant (RR, 0.91; 95% CI 0.77–1.08; Fig. 2). The recurrence rate did not significantly differ between the two interventions (RR, 0.89; 95% CI 0.58–1.38; Fig. 2). On the basis of our meta-analysis, simple aspiration is recommended for the initial management of pneumothorax because of the shorter hospitalization time, although there is no significant difference in the success and recurrence rates. This conclusion is consistent with that of a previously reported meta-analysis with randomized controlled trials and recent reviews [2, 8].

Table 1:

Summary of the four studies included in the present meta-analysis

Study	Aspiration			Tube drainage
	Total	Immediate success	Success	Total	Immediate success	Success
Harvey and Prescott [11]	35	–	28	38	–	38
Andrivet et al. [12]	33	–	22	28	–	26
Noppen et al. [13]	27	16	25	33	21	28
Ayed et al. [14]	65	40	58	72	49	63

Figure 1:

Results of meta-analysis: hospital stay as weighted mean difference.

Figure 2:

Results of meta-analysis: success rate and recurrence rate.

BACKGROUND BIAS ABOUT PERSISTENT AIR LEAK AND REANALYSIS WITH EXTRACTED DATA

Kjaergard [15] described the cause of spontaneous pneumothorax to be lung perforation and air leak. Many researchers have suggested that one of the principal objectives of pneumothorax treatment is to stop the air leak. If the air leak is stopped, the collapsed lung is observed to gradually expand. From January 2006 to December 2008, our institutional database showed 75 cases of chest tube drainage, of which 30 (40%) did not show air leak until the day after chest tube insertion. We retrospectively assumed that these cases did not have an air leak even before intervention and that we could observe for gradual lung expansion without any intervention.

We think that if there is a bias in the cases in which the air leak stopped before intervention, the treatment cannot be compared. Of the four papers described above, two reported ‘immediate success’ defined as persistent lung expansion after simple aspiration and lung expansion and chest tube removal within 3 days after tube drainage [13, 14]. We presume that these cases of ‘immediate success’ include many that did not need any intervention. We calculated the number of patients showing ‘1-week success’ and excluded those showing ‘immediate success’. The objective of this analysis was to examine the percentage of cases showing a persistent air leak over 3 days that stopped within 1 week after intervention. Ayed's data showed that of 25 cases of persistent air leak, in 18 (72%), the air leak stopped within 1 week in the simple aspiration group; further, of 23 cases of persistent air leak, in 14 (61%), the air leak stopped within 1 week in the tube drainage group. Noppen's data showed that, of 11 cases of persistent air leak, in 9 (82%), the air leak stopped within 1 week in the simple aspiration group and, of 12 cases of persistent air leak, in 7 (58%), the air leak stopped within 1 week in the tube drainage group. Therefore, the data extracted from the above two reports consistently indicated that simple aspiration is more effective than tube drainage in stopping an air leak. This result is not in agreement with the original result of the ‘success’ obtained from the two papers.

Here, it is necessary to address the question why simple aspiration is more effective in stopping an air leak than tube drainage. When air leak is a complication after pulmonary resection, the use of a water seal for tube drainage without expanding the lung too much is more effective than the continuous suction of the drainage [16–18]. Taken together, our clinical experience and the reports indicate that treating the air leak with simple aspiration re-collapses the lung, whereas tube drainage continuously expands the lung. The latter is not desirable to close the perforation of the lung.

COMPARISON OF OBSERVATIONAL AND INTERVENTIONAL TREATMENT

Considering that expanding the lung has adverse effects on stopping the air leak, observation without rapidly expanding the lung would be the best management. Simpson [19] reported that conservative management of much larger pneumothorax is possible if there is no underlying lung disease. Few retrospective studies compared the success and recurrence rates of observational and interventional treatment (Table 2) [20–28]. These reports involved a relatively small number of patients and all the analyses were retrospective. A review of these reports reveals that the success rate of observational treatment is very high and seems to be satisfactory [20, 21, 26–28]. However, the recurrence rate of observational treatment tends to be higher than that after intervention [23–26]. This may be because tube drainage with continuous suction achieves complete lung re-expansion with contact between the visceral and parietal pleura, facilitating closure of the pleural defect. Inducing adhesion formation over the pleural defect to help prevent recurrence has been discussed [24, 29–31]. However, no clear conclusion can be reached because of the potential bias in patient backgrounds.

Table 2:

Success and recurrence rates of observational and interventional treatment

Author and year	Observational treatment	Aspiration	Tube drainage	P-value
Success rate
Hyde [20]	86	–	–
Beumer [21]	91	–	–
Kelly [28]	90	50–83	66–97
Kelly et al. [27]	79 (5)	50 (60)	73 (73)	0.014 (<0.0001)
Chen et al. [26]	95	–	50
Our institutional cases	78 (0)	100 (0)	61 (65)	0.12 (<0.0001)
Integration with four prospective trialsa		83	91
Recurrence rate
Stradling and Poole [22]	6	–	–
Ruckley and McCormack [23]	18	–	15
Seremetis [25]	49	–	38
O'Rourke and Yee [24]	33	0	24
Chen et al. [26]	63		7
Integration with three prospective trialsa		22	25

Data are presented as the percentage of success or recurrence (for a large pneumothorax).

^aIntegration was performed by adding the number of cases of success/recurrence and dividing by the total number of cases.

Our institutional database showed that 40% of cases of chest tube drainage did not show an air leak the day after chest tube insertion. We think that this percentage includes cases without an air leak even before the intervention. As stated above, if there is a bias in the cases in which the air leak was stopped, it is not possible to correctly compare the efficacy of the treatment options. Because of improper evaluation of pneumothorax during the initial stages, we cannot arrive at any conclusions after comparing the results of observational and interventional treatment. It is necessary to develop a clear strategy for the initial management of pneumothorax for implementation in clinical cases and research trials.

RETHINK ABOUT RAPID AIR ELIMINATION

According to Light [1], pneumothorax treatment has two goals: to rid the pleural space of its air and to decrease the likelihood of a recurrence; many other researchers also described that eliminating intrapleural air is the main principle of pneumothorax treatment [32, 33]. Expanding the collapsed lung is not desirable to stop the air leak, and expanding the lung rapidly is necessarily dispensable during initial management because it does not complete treatment. Some investigators also indicated that the presence of intrapleural air in itself is not an indication for intervention [19, 34]. If there is no air leak, the lung gradually expands without any interventional treatment [20, 22]. A review of the data available thus far [19, 22, 27, 35] and our experience during clinical practice show that the lung should not be rapidly expanded unless necessary. It is absolutely necessary to expand the lungs, for example, in cases of tension pneumothorax and cases in which there is a continuous air leak even if tension pneumothorax is not diagnosed during consultation, but is suspected to subsequently occur. We think that the indications for interventional treatment to expand the lung should be much more refined, and fewer patients should need intervention. The purpose of tube drainage and simple aspiration should not be expansion of the lung but recovery of respiratory dysfunction.

Attempts have been made to define the management of pneumothorax according to its size [1, 36, 37]. It is widely accepted that a small primary spontaneous pneumothorax in patients without respiratory symptoms can be conservatively managed. However, the management of a moderately collapsed lung remains controversial [3, 4]. Here, we would like to propose a new strategy that is based on the presence of symptoms of acute respiratory dysfunction and an air leak. Therefore, the size of the pneumothorax at a certain point of time is not very important, but the change in the rate of collapse is important, because the development of lung collapse reveals the existence of the air leak.

INDICATIONS FOR INTERVENTIONAL TREATMENT

Pneumothorax is generally considered to be a benign disease with good prognosis, although sudden death has been reported [24, 38]. Continuous air leak can progress into tension pneumothorax with the risk of sudden death. There are several case reports of diagnostic difficulty or missed diagnoses of tension pneumothorax [39–41]. The incidence is reportedly 1–5% of spontaneous pneumothorax [22, 24, 42–44]. Tension pneumothorax must be treated by immediately evacuating the air. Therefore, the diagnosis of tension pneumothorax is undoubtedly an indication for immediate intervention (Table 3). Although needle decompression is useful for the emergent release of pleural tension, cases of tension pneumothorax that are considered to have a persistent air leak should be eventually treated with tube drainage rather than repeating aspiration.

Table 3:

Recommended initial management depending on clinical condition

Pneumothorax type	Treatment purpose	Recommended treatment
Tension pneumothorax	To recover from acute respiratory dysfunction	Any interventiona First: aspiration (diagnosis) Second: tube drainage
Bilateral pneumothorax	To recover from acute respiratory dysfunction	Drainage of at least one side of the lung
With severe lung disease	To recover from acute respiratory dysfunction	Any interventiona
Suspected air leak	To avoid acute respiratory dysfunction	Any interventiona
Completely collapsed lung	To avoid acute respiratory dysfunction	Any interventiona
Symptomatic	To avoid acute respiratory dysfunction Symptomatic relief	Any interventiona (aspiration > drainage)
Supposed absence of air leak	To maintain stoppage of air leak	Observational treatment

^aTube drainage or repeating aspiration.

Another certain indication of intervention is bilateral pneumothorax that rarely occurs, but also causes potentially life-threatening acute respiratory dysfunction. The incidence of bilateral pneumothorax is reported to be 0.5–1.9% [20, 45]. In these cases, tube drainage for expanding at least one side of the lung is necessary to save the patient's life. Patients with severe underlying lung disease tend to develop severe acute respiratory dysfunction even with a relatively small pneumothorax. Intervention is also necessary in such cases.

A continuous air leak in the absence of tension pneumothorax can be detected with careful examination, such as the patient's symptoms and repeated chest radiographs. We think that the air leak in itself is also an indication for intervention, because lung collapse with continuous air leak finally develops tension pneumothorax. On the basis of the discussion in the second section, we suggest that repeating simple aspiration is more effective for stopping the air leak than tube drainage, although prospective randomized trials with patients having the same background of air leak are necessary to reach a clear conclusion. For preventing the development of tension pneumothorax, tube drainage is more effective.

The initial treatment of a case of totally collapsed lung without severe respiratory dysfunction that is suspected not to be tension pneumothorax is controversial. A recent retrospective study [46] compared complete and partial atelectasis in patients with primary spontaneous pneumothorax and found that 29.4% of the patients with complete atelectasis and that 10% of those with partial atelectasis showed persistent air leak after tube drainage. Thus, when compared with partial atelectasis, complete atelectasis is estimated to more frequently have a persistent air leak, even before tube drainage. If complete atelectasis accompanies a continuous air leak, the patients easily develop tension pneumothorax. Besides, a completely collapsed lung is more likely to subsequently lead to serious complications, such as infection of the lung and pleura and restrictive lung impairment. Therefore, we think that a completely collapsed lung is in itself an indication for tube drainage.

THREE-STEP MANAGEMENT OF PNEUMOTHORAX

We propose that the management of pneumothorax should be divided into three steps depending on the main objective of treatment (Table 4). The objective of the first step is resolving acute respiratory dysfunction. In this first step, we are required to consider whether intervention is necessary to treat acute respiratory dysfunction and other potentially lethal complications. This comes under the ‘initial management’ of the disease, which is mostly performed in the emergency department of a hospital. The most important task in this step is to determine whether there is an air leak, in addition to the patient's symptoms and the size of the pneumothorax in the chest radiograph. If the existence of an air leak is confirmed by the patient's symptoms or physical examination, such as severe chest pain, observation of lung collapse on a chest radiograph and high intrathoracic pressure, any intervention is necessary. Simple aspiration might be permissible for an air leak although insertion of a chest tube is ultimately necessary for continuous drainage of the air. If no air leak is found and the lung has not collapsed completely, it is essential to maintain the stoppage of the air leak, which constitutes the second step. In this step, the objective of treatment is resolving the air leak. Observational treatment to avoid rapid lung expansion can help in sustention of closing the perforation. It must be noted that interventions are never performed for treating the air leak, because such interventions potentially lead to the recurrence of the leak by increasing the size of the perforation. In cases where interventional treatment is given at the first step and a persistent air leak is found, the patient needs affirmative treatment, such as pleurodesis or surgery, to stop the air leak. This procedure is a part of the next aspect of the second step, i.e. targeting the air leak. If an air leak is absent and repeated pneumothorax is observed, aggressive treatment is necessary to prevent the recurrence, which comes under the third step.

Table 4:

Three-step management of pneumothorax

Step	Treatment objective	Recommended treatment
Step 1a	Acute respiratory dysfunction
	To recover from respiratory function  To avoid respiratory dysfunction	Tube drainage (or simple aspiration) Tube drainage (or simple aspiration)
Step 2	Air leak
	To maintain stoppage of air leak  To stop air leak  To stop air leak after drainage	Observational treatment Repeating aspiration (vs tube drainage) Water seal management in tube drainage, surgery (bullectomy, bulla ligation, etc.), pleurodesis, bronchial intervention, others
Step 3	Recurrence
	To prevent recurrence	Pleurodesisb, surgery (bullectomy, pleurectomy, etc.), others

^aStep 1 is defined as initial management.

^bEfficacy of tube drainage in preventing recurrence is controversial.

Traditional pneumothorax management depended on the classification of pneumothorax into primary, spontaneous or secondary; however, classifying pneumothorax is sometimes clinically difficult before computed tomography is performed. Our proposed three-step management strategy can be applied not only to primary spontaneous pneumothorax but also secondary and traumatic pneumothorax. Secondary spontaneous pneumothorax would require more interventions than primary spontaneous pneumothorax in Step 1.

CONCLUSION

We believe that there are three important characteristics of pneumothorax: potentially lethal respiratory dysfunction; air leak, which is the obvious cause of the disease; frequently occurring recurrence. These three characteristics correspond to the three steps of our proposed management in pneumothorax. The central idea is that the lung should not be expanded rapidly, unless necessary. The objective of treatment has not been clarified in published trials, and therefore, it is not possible to compare the treatment strategies between different trails because of inappropriate evaluation of the air leak. The primary objective of both simple aspiration and chest drainage should be the recovery of acute respiratory dysfunction or the avoidance of respiratory dysfunction and subsequent complications. After confirming the air leak, interventions such as simple aspiration or chest drainage and observational treatment may be used for the management of the air leak.

Conflict of interest: none declared.