Tension pneumothorax is an uncommon condition with a high mortality rate most frequently reported to occur in prehospital, emergency department, and intensive care unit (ICU) settings.1–4 This condition is frequently lethal in injured and ventilated ICU patients without early diagnosis and treatment.5–7 Although the incidence of tension pneumothorax remains poorly estimated, it may occur in up to 1% to 3% of prehospital, major trauma and ICU patients.1,3,8–12
As many authorities recommend urgent thoracic decompression when the diagnosis is first suspected, health care providers are taught to search for classically described clinical manifestations to recognize patients who may have tension pneumothorax.5,6,13 Although tension pneumothorax is therefore a syndrome diagnosis, available literature sources differ substantially in their descriptions of its clinical presentation.1 Many of these have been generalized from canine studies of the disorder,14,15 and do not account for potential differences in pathophysiology and physical signs based on the ventilatory status of the patient (Fig. 1).1,6,16–21
As misdiagnosis or inappropriate treatment of tension pneumothorax can have devastating consequences,6,22–26 a comprehensive description of its clinical presentation may improve patient care.1 Thus, we conducted a systematic review to determine whether the reported clinical presentation (and resultant management and outcomes) of tension pneumothorax differs between patients who are breathing unassisted (ie, breathing spontaneously and not receiving positive pressure ventilation) versus receiving assisted (ie, positive pressure) ventilation. Our primary objective was to determine whether available clinical data support animal study observations of potentially important differences between subjects of varying ventilatory status in time to, severity, and frequency of presenting hemodynamic complications.16,17,19 As systematic reviews of case reports and series of other uncommon/emergent conditions have guided clinical practice and future research,27–31 we synthesized and analyzed data reported by these types of studies alongside a systematic review of observational studies.1,32–35
A published protocol details our study methods.1 This protocol was registered in the PROSPERO Register of Systematic Reviews (registration number: CRD42013005826) and developed according to recommendations for conducting systematic reviews and meta-analyses.1,35–37
With assistance from a medical librarian (H.L.R.), we searched Ovid MEDLINE and EMBASE, PubMed, and the Cochrane Library from their first available dates to October 15, 2013 without restrictions (see our protocol1 for details regarding database search strategies). To identify additional/ongoing studies, we searched personal files, wrote to colleagues and content experts, and investigated 2 clinical trials registries (ClinicalTrials.gov and www.Controlled-Trials.com). We also used the PubMed “related articles” and Google Scholar “cited by” features and manually searched reference lists of included articles and relevant review papers identified during the search.
Two physicians (D.J.R. and C.B.) independently reviewed titles and abstracts of citations identified by the search and selected articles for full-text review. Potentially relevant non-English language articles were translated into English. We included observational (cohort, case-control, and cross-sectional) studies and case reports and series38,39 that reported original data on clinical manifestations of tension pneumothorax. We defined clinical manifestations as patient-level findings gathered during medical interview, physical examination, or through invasive monitoring or treatment equipment or diagnostic studies.1,40 Studies and reports of fatal cases were included if the condition causing death was attributed by authors to be tension pneumothorax and associated with expulsion of air on thoracic decompression or determined by a pathologist to be present on autopsy.1 We excluded studies and reports not describing patient ventilatory status and those involving children (defined as age <12 years41), as their clinical presentation likely differs from older patients because their mediastinum and thoracic wall are more compliant.1,5,16 We also excluded studies and reports involving patients with conditions that could misrepresent the more common clinical manifestations of tension pneumothorax.1
Disagreements regarding study eligibility were resolved by discussion after the entire article had been reread in full. Inter-investigator agreement was quantified using kappa (κ) statistics,42 and the κ-statistic interpretation guidelines suggested by Altman.43
Two physicians (D.J.R. and C.B.) independently extracted data from included studies and case reports using a data extraction spreadsheet.1 Data extracted included study and case characteristics and reported clinical manifestations, initial management, and outcomes of tension pneumothorax.1 For case reports, the 2 physicians independently categorized times from onset of symptoms, a deterioration in clinical status, or iatrogenic production of a pneumothorax to respiratory decompensation/arrest or hypotension/cardiac arrest according to whether they were described to occur suddenly (approximately 0 to 5 minutes), acutely (approximately >5 to 60 minutes), subacutely (approximately >60 to 180 minutes), or in a more delayed fashion (approximately >180 minutes). Clinical manifestations data were abstracted as proximal as possible to author's descriptions of pretreatment diagnoses of tension pneumothorax. When partial pressure of arterial oxygen (PaO2) and fraction of inspired oxygen (FiO2) values were not provided, these were estimated from reported arterial oxygen saturation (SpO2) values and oxygen delivery device flow rates using conversion tables.44
Risk of Bias Assessment
For observational studies, 2 physicians (D.J.R. and C.B.) independently evaluated whether tension pneumothorax diagnoses were supported by radiographic findings/response to thoracic decompression and whether overlap existed between diagnostic criteria and reported clinical manifestations.1 They also evaluated settings from which patients were recruited to determine whether they were likely representative of the population of tension pneumothorax patients.1 Finally, they assessed whether reported frequencies of clinical manifestations were precise [(by assessing widths of associated confidence intervals (CIs)] and whether clinical manifestations were sought thoroughly and consistently (by determining methods by which they were gathered and whether this was done similarly across all study patients).1
For case reports, the 2 physicians independently determined whether patient presentations satisfied a published tension pneumothorax definition.1 According to this definition, a tension pneumothorax was defined as one “that results in significant respiratory or hemodynamic compromise that reverses (or at least significantly improves) on thoracic decompression alone.”1,6
Observational Study Data Synthesis
As included observational studies were limited by clinical heterogeneity, planned observational study meta-analyses1 were not conducted. Results of these studies were instead described narratively. Exact, 95% CIs surrounding dichotomous clinical manifestations variables reported by observational studies were determined using the Clopper-Pearson method.45
Case Reports and Series Data Synthesis and Analysis
We summarized characteristics of reported cases and their described clinical manifestations, management, and outcomes as proportions, medians (with interquartile ranges), and means (with standard deviations). Dichotomous and continuous variables were compared using Fisher exact and Wilcoxon rank sum or matched-pairs signed-ranks tests, respectively.
We estimated unadjusted and adjusted mean differences and odds ratios (ORs) comparing hemodynamic events at tension pneumothorax diagnosis between cases who were breathing unassisted and receiving assisted ventilation. To accommodate for clustering of clinical manifestation variables in case series, we conducted these comparisons using generalized estimating equations with independent correlational data structures.46 Model covariates for adjusted analyses included age; antihypertensive or vasopressor administration before diagnosis; preexisting shock; history of hypertension, heart failure, or pulmonary disease; and concomitant diagnosis of hemothorax, other pleural effusion, or new pulmonary disease.1 A separate clustered logistic regression model was used to determine whether subcutaneous emphysema, tracheal deviation, jugular venous distention, ipsilateral percussion hyperresonance, hypoxia, hypotension, respiratory arrest, or cardiac arrest independently predicted ventilatory status across included case reports.
To test the robustness of our findings, we conducted sensitivity analyses in which we recalculated the aforementioned comparisons using only those cases that satisfied the published tension pneumothorax definition.6 We also explored whether adjusted ORs varied in magnitude or direction among subgroups of cases.1 We considered 2-sided P values of less than 0.05 significant. Stata MP version 13.1 (Stata Corp, LP, College Station, TX) was used for statistical analyses.
Among 4160 citations identified by the search, we included 5 cohort studies (n = 310 total patients),12,47–50 29 case series (median, 2 cases per series; range, 1–5), and 127 case reports in the systematic review (Fig. 2). Inter-investigator agreement on full-text article inclusion was good (κ-statistic, 0.75; 95% CI, 0.68–0.82). We requested supplementary information on study procedures or reported cases from 11 authors, and 10 responded.3,11,12,48,51–56 After excluding 25 cases that failed inclusion criteria from within included case series, 183 cases were included in the synthesis and analysis of case reports and series data.
Description of Included Cohort Studies and Case Reports
Characteristics of included cohort studies are presented in Table 1. Studies were published between 2005 and 2014. Three exclusively enrolled prehospital trauma patients treated with needle thoracostomy for suspected tension pneumothoraces,12,48,50 1 included only injured patients who received prehospital tube thoracostomy,49 and 1 enrolled ICU patients with both ventilator-associated simple and tension pneumothoraces.47 Mean patient ages ranged from 31.5 to 67 years. Three studies included patients receiving assisted ventilation,47–49 whereas only 1 reported separately on patients who were breathing unassisted versus receiving assisted ventilation.50 The fifth study included 2 groups of patients of which 61% and 87% were receiving assisted ventilation.12
Among the 183 included cases, 86 (47.0%) were breathing unassisted and 97 (53.0%) receiving assisted ventilation (see Table in Supplemental Digital Content 1, available at https://links.lww.com/SLA/A690, for details regarding case ventilatory statuses). Most (75.4%) cases were reported after the year 1990. The proportion of reported cases who were breathing unassisted increased across the study search period from a minority of the total reports before 1994 to the majority of them thereafter (see Figure in Supplemental Digital Content 2, available at https://links.lww.com/SLA/A691).
The Table in Supplemental Digital Content 3, available at https://links.lww.com/SLA/A692, provides a bibliography of included case reports/series and characteristics of individual cases. The demographics and medical history of cases were similar between ventilatory status groups (Table 2). Mean age of all cases was 45.5 years (standard deviation, 20.2 years). A total of 3.5% of cases who were breathing unassisted received general anesthesia before diagnosis versus 55.7% receiving assisted ventilation. Bilateral tension pneumothoraces were less frequent among cases breathing unassisted (2.3%) versus receiving assisted ventilation (24.4%).
Trauma was a relatively commonly reported cause of tension pneumothorax among all cases (Table 2). However, spontaneous pneumothoraces and gastrointestinal perforation were more frequently described etiologies among cases breathing unassisted whereas barotrauma and attempted central venous catheter insertions were more commonly reported causes among cases receiving assisted ventilation.
Risk of Bias Assessment
An overview of the risk of bias of included cohort studies is shown in Table 3. In all studies, tension pneumothorax diagnoses were partly supported by diagnostic imaging findings and/or the cardiorespiratory response of the patient to thoracic decompression. Described patient presentations or case definitions partly satisfied the published tension pneumothorax definition in 3 cohort studies.48–50 One study combined some data on clinical manifestations of simple and tension pneumothoraces together.49
The 2 physicians independently agreed that clinical conditions of 103 (57.2%) included cases satisfied the published tension pneumothorax definition (κ-statistic, 0.89; 95% CI, 0.82–0.95). Characteristics of these cases were similar when compared with all cases (see Table in Supplemental Digital Content 4, available at https://links.lww.com/SLA/A693).
Reported Clinical Presentation of Tension Pneumothorax
Signs and Symptoms
Tables 1 and 4 summarize signs and symptoms of tension pneumothorax reported by included cohort studies and case reports, respectively. Signs and symptoms reported by all included case reports were similar to those that satisfied the published tension pneumothorax definition (see Table in Supplemental Digital Content 5, available at https://links.lww.com/SLA/A694).
Symptoms and Respiratory Vital Signs
Symptoms reported among case reports of patients breathing unassisted included chest pain (52.3%), dyspnea (38.4%), and shortness of breath (31.4%). Respiratory distress was described in 41.9% of cases breathing unassisted versus 8.3% receiving assisted ventilation.
Many (46.5%) case reports of patients breathing unassisted described tachypnea. Hypoxia or requirement for supplemental oxygen was reported among 43 (50.0%) cases who were breathing unassisted versus 89 (91.8%) receiving assisted ventilation (P < 0.001). Hypoxia was also reported among 25.0% of patients breathing unassisted versus 11.1% to 50.9% receiving assisted ventilation across 2 included cohort studies.49,50 The median PaO2/FIO2 ratio among all included case reports was 73.0 (interquartile range, 48.4–152.6). One included cohort study of mechanically ventilated patients with tension pneumothoraces reported a mean PaO2/FIO2 ratio of 150.47 Respiratory arrest occurred in 9.3% of case reports of patients breathing unassisted.
Head and Chest Examination
Jugular venous distention (7.1%) and contralateral tracheal deviation (9.3%) were uncommonly reported by included case reports and were not described by any of the included cohort studies. As compared with cases who were breathing unassisted, subcutaneous emphysema was noted more often (10.5% vs 30.9%; P = 0.001) and contralateral tracheal deviation was noted less often (17.9% vs 2.9%; P = 0.004) among cases receiving assisted ventilation. In one included cohort study, subcutaneous emphysema was reported among 27.3% of patients receiving assisted ventilation.49
Ipsilateral decreased air entry, percussion hyperresonance, and decreased thoracic excursions/mobility were the most commonly reported chest examination findings among case reports of unilateral tension pneumothoraces. Ipsilateral decreased air entry was also reported among 50.0% to 54.5% of patients receiving assisted ventilation across 2 included cohort studies.48,49 Hyperresonance to percussion was more commonly described among case reports of patients breathing unassisted versus receiving assisted ventilation (26.7% vs 8.3%; P = 0.001).
Cardiovascular Vital Signs
Unadjusted systolic, diastolic, and mean arterial blood pressures were substantially higher among cases who were breathing unassisted versus receiving assisted ventilation (Fig. 3). After adjustment, cases who were breathing unassisted had higher reported systolic (126 mm Hg vs 94 mm Hg; difference = 32 mm Hg; 95% CI, 19.8–45.0 mm Hg; P < 0.001) and mean arterial blood pressures (95.0 mm Hg vs 62.8 mm Hg; difference = 32.8 mm Hg; 95% CI, 22.0–43.7 mm Hg; P < 0.001) than those receiving assisted ventilation. Moreover, when compared with cases who were breathing unassisted, the adjusted odds of hypotension (defined a priori as a mean arterial pressure ≤60 mm Hg1) and cardiac arrest were 12.6 (95% CI, 5.8–27.5) and 17.7 (95% CI, 4.0–78.4) times higher among those receiving assisted ventilation, with the most commonly reported initial arrest rhythm being pulseless electrical activity (75.0%). These increased odds were robust to a number of sensitivity analyses (see Table in Supplemental Digital Content 6, available at https://links.lww.com/SLA/A695). One included cohort study also reported that none of the included patients with a tension pneumothorax who were breathing unassisted versus 39.6% of those receiving assisted ventilation presented without an arterial pulse.50
Clustered logistic regression suggested that contralateral tracheal deviation was independently associated with an increased odds of breathing unassisted (OR, 33.3; 95% CI, 3.0–364.5; P = 0.004) whereas hypotension (OR, 8.6; 95% CI, 3.5–31.5; P < 0.001) and subcutaneous emphysema (OR, 5.9; 95% CI, 1.9–18.4; P = 0.002) were independently associated with an increased odds of having received assisted ventilation.
Approximate times to development of hypotension/cardiac arrest could be determined for 20 (90.9%) case reports of patients breathing unassisted versus 54 (72.0%) receiving assisted ventilation. In contrast to cases who were breathing unassisted, the majority (70.4%) of those receiving assisted ventilation who experienced hypotension or cardiac arrest developed these signs within minutes of clinical presentation (Fig. 4).
A chest radiograph was obtained before treatment in 55 (64.0%) case reports of patients who were breathing unassisted versus 44 (45.4%) receiving assisted ventilation. A pneumothorax occupying greater than 50% of hemithorax volume (55.8%), contralateral mediastinal deviation (57.1%), and ipsilateral hemidiaphragm flattening (22.2%) were the most commonly reported chest radiography findings (see Table in Supplemental Digital Content 7, available at https://links.lww.com/SLA/A696).
Presenting electrocardiographic findings were described by 17 (9.3%) included case reports (see Table in Supplemental Digital Content 7, available at https://links.lww.com/SLA/A696). Commonly reported findings included sinus tachycardia (41.2%), ST segment/T wave changes (41.2%), and decreased voltage (29.4%).
Invasive Cardiopulmonary Measurements
Several case reports of patients receiving assisted ventilation described changes in cardiopulmonary variables between baseline and diagnosis of tension pneumothorax (see Table in Supplemental Digital Content 8, available at https://links.lww.com/SLA/A697). These changes included significant increases in peak inspiratory and central venous pressures, rises in pulmonary vascular resistance, and decreases in cardiac index.
Management and Outcomes
Needle (46.3% vs 36.2%) and tube (46.3% vs 50%) thoracostomy were the most frequently reported initial interventions among cases who were breathing unassisted versus receiving assisted ventilation, respectively. The reported mortality of tension pneumothorax in cases breathing unassisted (6.7%) was substantially lower than that for patients receiving assisted ventilation (22.7%) (P = 0.003). One cohort study reported that development of tension pneumothorax among mechanically ventilated patients was associated with a 7.4 (95% CI, 2.2–24.6) times increase in the adjusted hazard of mortality.47
In this systematic review, after considering findings of animal studies of tension pneumothorax,14–17,19–21 we synthesized information on the clinical presentation of patients with tension pneumothorax reported by 5 cohort studies and 183 case reports (n = 86 breathing unassisted, n = 97 receiving assisted ventilation). When summarized, these studies highlight a number of reported differences in clinical presentation depending on the ventilatory status of the patient (see Table 5 for a summary). They also highlight how clinicians reportedly diagnose and manage these patients in practice.
Cases who were breathing unassisted were frequently reported to present with shortness of breath, dyspnea, tachypnea, respiratory distress, hypoxemia, and ipsilateral decreased air entry and percussion hyperresonance. Pulmonary dysfunction progressed to respiratory arrest in 9% of cases breathing unassisted. Hypotension and cardiac arrest were reported among only 16% and 2% of included cases who were breathing unassisted, respectively, and among none of the 12 breathing unassisted patients in one included cohort study.50 When these outcomes did occur, more than half of the cases seemed to develop them in a relatively delayed fashion, and nearly two-thirds of clinicians obtained a chest radiograph before performing thoracic decompression. Despite this, half of the cases were managed first with needle thoracostomy.
Hypoxemia, subcutaneous emphysema, and ipsilateral decreased air entry were also commonly described among case reports and cohort studies of patients receiving assisted ventilation. However, the clinical presentation of these patients differed substantially from those who were breathing unassisted, potentially as a result of their requirement for ventilatory support and/or greater illness severity. Similar to animal study findings (Fig. 1), when compared with case reports of patients who were breathing unassisted, hypotension and cardiac arrest were significantly more commonly reported to be present at the time of tension pneumothorax diagnosis. These outcomes were also frequently described to occur within minutes of a sudden clinical deterioration (eg, a decrease in SpO2) or an iatrogenic creation of a pneumothorax. Interestingly, however, half of the cases receiving assisted ventilation underwent chest radiography before thoracic decompression and half were initially managed with tube thoracostomy.
Our findings may have implications for improving the diagnosis and treatment of tension pneumothorax. In contrast to classical medical teaching, contralateral tracheal deviation and jugular venous distention are uncommonly reported clinical manifestations of tension pneumothorax. Tension pneumothorax may have to be considered in patients who are breathing unassisted who present with predominantly respiratory signs and symptoms. As those who are breathing unassisted have seldom been reported to present with sudden hemodynamic compromise, it may be appropriate to obtain a chest radiograph in a monitored setting to confirm the diagnosis and lateralize the disease instead of performing urgent thoracic decompression for patients who are not in extremis.5,6,13 Thoracic ultrasonography may be superior to chest radiography for this purpose, as it has a sensitivity of approximately 80% to 90% for detection of pneumothoraces (versus approximately 50% for supine chest radiography) and can be performed rapidly at the bedside.57,58 Conversely, clinicians should be prepared to perform urgent thoracic decompression without chest radiographic confirmation in patients suspected of a tension pneumothorax who are receiving assisted ventilation, as these patients have frequently been reported to present with sudden hemodynamic compromise and/or cardiac arrest.
Our synthesis and analysis of case reports/series data has several potential limitations. Our estimates of the frequency of clinical manifestations of tension pneumothorax may have been influenced by underreporting of relatively common presentations of tension pneumothorax or overreporting of presentations that manifested more unusual or interesting clinical features.1,59 However, as we can think of no reason why under- or overreporting would depend on case ventilatory status, it seems unlikely that selection bias would have influenced our between-group comparisons. Furthermore, although some of the included case reports could be argued not to represent tension pneumothorax, our findings were robust to sensitivity analyses that included only cases satisfying a published definition. Similarly, as we included case reports of patients with less common etiologies of tension pneumothorax (eg, gastrointestinal perforation), the validity of combining all cases together may be questioned.1 Despite this, we are unsure why patients with a less common etiology would present with different clinical manifestations when compared to those with more common etiologies.1,16,17,19–21 Finally, although some may argue that our findings may be due to unmeasured confounding,59 this seems unlikely given that any unmeasured confounder that could account for the observed magnitude of the association between ventilatory status and hypotension/cardiac arrest would have to be very strongly associated with patient ventilatory status and highly predictive of hypotension and cardiac arrest. Thus, as our findings are consistent with results from animal studies,16,17,19–21 we believe them to be clinically important.
The reported clinical presentation of tension pneumothorax depends on the ventilatory status of the patient. This may have implications for improving the diagnosis and treatment of this uncommon yet catastrophic clinical condition.
Dr Roberts had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. The authors thank the staff of the University of Calgary Health Sciences Library for obtaining copies of articles identified throughout the conduct of the systematic review, Sandy Cochrane at Multimedia Services at the University of Calgary for assisting with creation of Figure 1, and Kelly Mrklas, MSc, for assisting with translation of non-English language articles. The authors also thank Eddy S. Lang, MDCM in the Department of Emergency Medicine at the University of Calgary for reviewing the article and providing critical input on its findings before submission for peer review. Dr Roberts is supported by an Alberta Innovates—Health Solutions Clinician Fellowship Award, a Knowledge Translation Canada Strategic Training in Health Research Fellowship, and funding from the Canadian Institutes of Health Research. These funders had no role in the design or conduct of the study; collection, management, analysis, or interpretation of the data; or preparation, review, or approval of the article.
1. Roberts DJ, Leigh-Smith S, Faris PD, et al. Clinical manifestations of tension pneumothorax
: protocol for a systematic review and meta-analysis. Syst Rev. 2014;3:3.
2. Cameron PA, Flett K, Kaan E, et al. Helicopter retrieval of primary trauma patients by a paramedic helicopter service. Aust N Z J Surg. 1993;63:790–797.
3. Coats TJ, Wilson AW, Xeropotamous N. Pre-hospital management of patients with severe thoracic injury. Injury. 1995;26:581–585.
4. Eckstein M, Suyehara D. Needle thoracostomy in the prehospital setting. Prehosp Emerg Care. 1998;2:132–135.
5. American College of Surgeons Committee on Trauma. Advanced Trauma Life Support (ATLS): Ninth Edition. Chicago, IL: American College of Surgeons; 2012.
6. Leigh-Smith S, Harris T. Tension pneumothorax
—time for a re-think? Emerg Med J. 2005;22:8–16.
7. Chen KY, Jerng JS, Liao WY, et al. Pneumothorax in the ICU: patient outcomes and prognostic factors. Chest. 2002;122:678–683.
8. Fleming WH, Bowen JC. Early complications of long-term respiratory support. J Thorac Cardiovasc Surg. 1972;64:729–738.
9. Kumar A, Pontoppidan H, Falke KJ, et al. Pulmonary barotrauma during mechanical ventilation. Crit Care Med. 1973;1:181–186.
10. Ludwig J, Kienzle GD. Pneumothorax in a large autopsy population. A study of 77 cases. Am J Clin Pathol. 1978;70:24–26.
11. Warner KJ, Copass MK, Bulger EM. Paramedic use of needle thoracostomy in the prehospital environment. Prehosp Emerg Care. 2008;12:162–168.
12. Ball CG, Wyrzykowski AD, Kirkpatrick AW, et al. Thoracic needle decompression for tension pneumothorax
: clinical correlation with catheter length. Can J Surg. 2010;53:184–188.
13. Waydhas C, Sauerland S. Pre-hospital pleural decompression and chest tube placement after blunt trauma: a systematic review. Resuscitation. 2007;72:11–25.
14. Hilton R. Some effects of artificial pneumothorax on the circulation. J Pathol Bacteriol. 1925;37:1–8.
15. Simmons DH, Hemingway A, Ricchiuti N. Acute circulatory effects of pneumothorax in dogs. J Appl Physiol. 1958;12:255–261.
16. Rutherford RB, Hurt HH Jr, Brickman RD, et al. The pathophysiology of progressive, tension pneumothorax
. J Trauma. 1968;8:212–227.
17. Gustman P, Yerger L, Wanner A. Immediate cardiovascular effects of tension pneumothorax
. Am Rev Respir Dis. 1983;127:171–174.
18. Subotich D, Mandarich D. Accidentally created tension pneumothorax
in patient with primary spontaneous pneumothorax—confirmation of the experimental studies, putting into question the classical explanation. Med Hypotheses. 2005;64:170–173.
19. Barton ED. Tension pneumothorax
. Curr Opin Pulm Med. 1999;5:269–274.
20. Martin M, Satterly S, Inaba K, et al. Does needle thoracostomy provide adequate and effective decompression of tension pneumothorax
? J Trauma Acute Care Surg. 2012;73:1412–1417.
21. Nelson D, Porta C, Satterly S, et al. Physiology and cardiovascular effect of severe tension pneumothorax
in a porcine model. J Surg Res. 2013;184:450–457.
22. Mines D, Abbuhl S. Needle thoracostomy fails to detect a fatal tension pneumothorax
. Ann Emerg Med. 1993;22:863–866.
23. Bailey RC, Esberger D. Development of tension pneumothorax
after chest drain insertion. J Accid Emerg Med. 1998;15:128.
24. Friend KD. Prehospital recognition of tension pneumothorax
. Prehosp Emerg Care. 2000;4:75–77.
25. Rawlins R, Brown KM, Carr CS, et al. Life threatening haemorrhage after anterior needle aspiration of pneumothoraces. A role for lateral needle aspiration in emergency decompression of spontaneous pneumothorax. Emerg Med J. 2003;20:383–384.
26. Riwoe D, Poncia HD. Subclavian artery laceration: a serious complication of needle decompression. Emerg Med Australas. 2011;23:651–653.
27. Bybee KA, Kara T, Prasad A, et al. Systematic review: transient left ventricular apical ballooning: a syndrome that mimics ST-segment elevation myocardial infarction. Ann Intern Med. 2004;141:858–865.
28. Lyrer P, Engelter S. Antithrombotic drugs for carotid artery dissection. Cochrane Database Syst Rev. 2010;(10):CD000255. doi: 10.1002/14651858.CD000255.pub2.
29. Holty JE, Bravata DM, Liu H, et al. Systematic review: a century of inhalational anthrax cases from 1900 to 2005. Ann Intern Med. 2006;144:270–280.
30. Andersohn F, Konzen C, Garbe E. Systematic review: agranulocytosis induced by nonchemotherapy drugs. Ann Intern Med. 2007;146:657–665.
31. Aronson JK, Hauben M. Anecdotes that provide definitive evidence. BMJ. 2006;333:1267–1269.
32. Jenicek M. Clinical Case Reporting in Evidence-Based Medicine. New York, NY: Oxford University Press; 2001.
33. Gagnier JJ, Kienle G, Altman DG, et al. The CARE guidelines: consensus-based clinical case report guideline development. J Clin Epidemiol. 2014;67:46–51.
34. Selvaraj SA, Chairez E, Wilson LM, et al. Use of case reports and the Adverse Event Reporting System in systematic reviews: overcoming barriers to assess the link between Crohn's disease medications and hepatosplenic T-cell lymphoma. Syst Rev. 2013;2:53.
35. Higgins JPT, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions Version 5.0.2. The Cochrane Collaboration 2009. Available at: www.cochrane-handbook.org
. Accessed November 25, 2014.
36. Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis of Observational Studies in Epidemiology (MOOSE) group. JAMA. 2000;283:2008–2012.
37. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. Ann Intern Med. 2009;151:W65–W94.
38. Oleckno WA. An overview of epidemiologic study designs. In: Oleckno WA, ed. Epidemiology: Concepts and Methods. Long Grove, IL: Waveland Press, Inc.; 2008:55–84.
39. Dekkers OM, Egger M, Altman DG, et al. Distinguishing case series from cohort studies. Ann Intern Med. 2012;156:37–40.
40. Richardson WS, Wilson MC, Williams JW Jr, et al. Users' guides to the medical literature: XXIV. How to use an article on the clinical manifestations of disease. Evidence-Based Medicine Working Group. JAMA. 2000;284:869–875.
41. Stedman's Medical Dictionary. 28th ed. Baltimore, MD: Lippincott Williams & Wilkins; 2006.
42. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977;33:159–174.
43. Altman DG. Practical Statistics for Medical Research. London, United Kingdom: Chapman & Hall; 1991.
44. EPIC II Group of Investigators. Conversion tables. Available at: http://www.intensive.org/epic2/Documents/Estimation%20of%20PO2%20and%20FiO2.pdf
. Accessed January 1, 2014.
45. Clopper CJ, Pearson ES. The use of confidence or fiducial limits illustrated in the case of the binomial. Biometrika. 1934;26:404–413.
46. Repeated measures and longitudinal data analysis. In: Vittinghoff E, Glidden DV, Shiboski SC, et al. eds. Regression Methods in Biostatistics: Linear, Logistic, Survival, and Repeated Measures Models Second Edition. New York, NY: Springer; 2012:261–308.
47. Hsu CW, Sun SF, Lin HS, et al. Clinical characteristics, hospital outcome and prognostic factors of patients with ventilator-related pneumothorax. Minerva Anestesiol. 2014;80:29–38.
48. Mistry N, Bleetman A, Roberts KJ. Chest decompression during the resuscitation of patients in prehospital traumatic cardiac arrest. Emerg Med J. 2009;26:738–740.
49. Massarutti D, Trillo G, Berlot G, et al. Simple thoracostomy in prehospital trauma management is safe and effective: a 2-year experience by helicopter emergency medical crews. Eur J Emerg Med. 2006;13:276–280.
50. Davis DP, Pettit K, Rom CD, et al. The safety and efficacy of prehospital needle and tube thoracostomy by aeromedical personnel. Prehosp Emerg Care. 2005;9:191–197.
51. Tocino IM, Miller MH, Fairfax WR. Distribution of pneumothorax in the supine and semirecumbent critically ill adult. AJR Am J Roentgenol. 1985;144:901–905.
52. Barton ED, Epperson M, Hoyt DB, et al. Prehospital needle aspiration and tube thoracostomy in trauma victims: a six-year experience with aeromedical crews. J Emerg Med. 1995;13:155–163.
53. Clark S, Ragg M, Stella J. Is mediastinal shift on chest x-ray of pneumothorax always an emergency? Emerg Med (Fremantle). 2003;15:429–433.
54. Lockey D, Crewdson K, Davies G. Traumatic cardiac arrest: who are the survivors? Ann Emerg Med. 2006;48:240–244.
55. Allison K, Porter KM, Mason AM. Use of the Asherman chest seal as a stabilisation device for needle thoracostomy. Emerg Med J. 2002;19:590–591.
56. Wildgruber M, Rummeny EJ. Bilateral tension pneumothorax
. Emerg Med J. 2012;29:752.
57. Roberts DJ, Niven DJ, James MT, et al. Thoracic ultrasonography versus chest radiography for detection of pneumothoraces: challenges in deriving and interpreting summary diagnostic accuracy estimates. Crit Care. 2014;18:416.
58. Lichtenstein DA. Lung ultrasound in the critically ill. Ann Intensive Care. 2014;4:1.
59. Richason TP, Paulson SM, Lowenstein SR, et al. Case reports describing treatments in the emergency medicine literature: missing and misleading information. BMC Emerg Med. 2009;9:10.