Article Text

Download PDFPDF

Heimlich valve orientation error leading to radiographic tension pneumothorax: analysis of an error and a call for education, device redesign and regulatory action
  1. Joshua S Broder1,
  2. James W Fox2,
  3. Judy Milne3,
  4. Brent Jason Theiling1,
  5. Ann White4
  1. 1Department of Surgery/Division of Emergency Medicine, Duke University School of Medicine, Durham, North Carolina, USA
  2. 2Department of Pediatrics, Duke University School of Medicine, Durham, North Carolina, USA
  3. 3Duke University Hospital, Durham, North Carolina, USA
  4. 4Department of Advanced Clinical Practice, Duke University Hospital, Durham, North Carolina, USA
  1. Correspondence to Dr Joshua S Broder, Duke University School of Medicine, Department of Surgery/Division of Emergency Medicine, DUMC 3096, 2301 Erwin Road, Durham, NC 27710, USA; joshua.broder{at}


Medical errors are commonly multifactorial, with adverse clinical consequences often requiring the simultaneous failure of a series of protective layers, termed the Swiss Cheese model. Remedying and preventing future medical errors requires a series of steps, including detection, mitigation of patient harm, disclosure, reporting, root cause analysis, system modification, regulatory action, and engineering and manufacturing reforms. We describe this process applied to two cases of improper orientation of a Heimlich valve in a thoracostomy tube system, resulting in enlargement of an existing pneumothorax and the development of radiographic features of tension pneumothorax. We analyse elements contributing to the occurrence of the error and depict the implementation of reforms within our healthcare system and with regulatory authorities and the manufacturer. We identify features of the Heimlich valve promoting this error and suggest educational, design, and regulatory reforms for enhanced patient safety.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Key messages

What is already known on this subject?

  • Medical errors have multiple root causes and potential remedies. Emergency physicians may be unfamiliar with systematic analysis of medical errors to enhance patient safety.

What might this study add?

  • We review a case of medical error, perform a root cause analysis, and suggest processes to prevent recurrence. Our description can serve as a model for others to follow in similar circumstances. We specifically highlight a design flaw in the Heimlich valve that can facilitate orientation error leading to a potentially fatal complication of tension pneumothorax. Knowledge of this risk is important to emergency physicians to ensure patient safety.

Case description

A 20-year-old man with autism presented to the emergency department with dyspnoea since awakening that morning. The patient denied any history of trauma or fever. Triage vital signs demonstrated BP 140/95, HR 106, temperature 36.6C, RR 25 and SpO2 100% on room air. The patient's examination was normal with the exception of decreased breath sounds on the patient's right side and evident tachypnoea. His trachea was noted to be midline on exam, and no jugular venous distension was present.

A chest radiograph was obtained (figure 1A), demonstrating a right pneumothorax. The patient underwent percutaneous placement of a 14 French pigtail catheter (Wayne Pneumothorax Tray, Cook Medical, Bloomington, Indiana, USA). Upon insertion of the chest tube, air was heard escaping from the catheter. The catheter was then connected to a chest drainage unit (Express Dry Seal Chest Drain, Atrium Medical Corporation, Hudson, New Hampshire, USA) using a Heimlich valve included in the Wayne Pneumothorax Tray, at which point no air leak was noted. A chest X ray (CXR) performed immediately after the procedure and approximately 2.5 h following the initial CXR (figure 1B) demonstrated appropriate positioning of the pleural catheter, but interval enlargement of the pneumothorax and development of radiographic features of tension pneumothorax, including complete collapse of the lung, lucent hemithorax, mediastinal shift, depressed hemidiaphragm and a deep sulcus sign.1 The patient's clinical condition remained stable. Examination of the chest tube assembly revealed reversal of the intended orientation of the Heimlich valve. The valve was reinserted in the correct orientation, and a repeated X-ray (figure 1C) revealed resolution of the pneumothorax and all radiographic features of tension pneumothorax.

Figure 1

(A) Initial CXR, before placement of thoracostomy tube. A moderate right-sided pneumothorax is visible. (B) 2.5 h later, following insertion of a pigtail catheter with an improperly oriented Heimlich valve. The right lung has completely collapsed, and radiographic features of tension pneumothorax are evident, including a depressed right hemidiaphragm, deep sulcus sign and mediastinal shift towards the patient's left side. (C) 30 min later, after correction of Heimlich valve orientation. The pneumothorax is no longer visible, and all radiographic features of tension pneumothorax have resolved.

The error was disclosed to the patient and family, and the patient was admitted for further management. The error was self-reported by the healthcare provider and independently and anonymously reported by a second party through the institutional safety reporting system. A root cause analysis of the event was conducted (table 1). The device error and risks were reported to the US Food and Drug Administration (FDA), and the device manufacturer (Cook Medical, Bloomington, Indiana, USA) was also notified of a product safety concern.

Table 1

Root cause analysis of events, following a standard rubric2

Following the original case, a second similar event occurred in the institution involving a different provider. In this instance, a patient presented to the emergency department with altered mental status following a motor vehicle collision. The patient was intubated in the emergency department for a GCS score of 5 and underwent CT, revealing intracranial haemorrhage and a pneumothorax. A 14 French pigtail catheter with a Heimlich valve (Wayne Pneumothorax Tray, Cook Medical, Bloomington, Indiana, USA) was placed, and the patient was admitted to an intensive care unit. Over the next several hours, the patient developed radiographically worsened pneumothorax; the patient remained clinically stable and could not verbalise any symptoms because of coma. A delay of over 12 h occurred, with multiple providers from nursing, surgery and critical care assessing the patient before the Heimlich valve was recognised to be oriented incorrectly, at which time the error was corrected and the pneumothorax resolved. While the first case was still being investigated, remediation processes were repeated and escalated.

For publication, exemption from review was obtained from the institutional review board. Informed consent for use of patient data in this manuscript was obtained from the patient and the patient's legally authorised representative.

Dissecting and addressing the error

Medical errors gained broad attention with the 1999 publication of the US Institute of Medicine report, ‘To Err is Human,’ which estimated 44 000–98 000 error-related deaths annually in American hospitals.4 ,5 The relationship between errors and adverse patient outcomes is often difficult to prove, and estimates of preventable harms vary, with poor inter-rater reliability.6 ,7 One recent study suggested 11 859 preventable inhospital deaths in 2009 in England.8 From 2002 to 2007, the rate of preventable patient harms remained high, occurring in between 1 in 10 and 1 in 20 hospital admissions in one study of hospitals in North Carolina (USA).7 Many errors may go unrecognised by the healthcare providers caring for the affected patient, and under-reporting of detected errors is likely widespread.9–13 Often, a medical error may occur but no adverse clinical consequence may ensue, as the error is mitigated by other factors. A ‘Swiss Cheese’ model requiring the simultaneous occurrence and alignment of multiple complimentary failures is widely accepted.14 Multiple contributing factors coincided in the cases described (table 1), including human operator knowledge deficits and error, device design flaws, and limitations of communication and institutional policies.

Without a systematic approach to the detection, mitigation, reporting, analysis, and system-based remediation of medical errors, recurrence is probable, likely resulting in preventable patient harms.

Steps in the remediation and reduction of medical errors

Steps in addressing a medical error are described in table 2. First, an error of omission or commission occurs, with or without clinical adverse consequences or patient harm. Following the occurrence of the error, remedying a medical error begins with detection, requiring healthcare providers to recognise a potential causal relationship between a medical behaviour (omission or commission) and a potential or realised adverse patient outcome. In many cases, a contributing error may go unrecognised, although the adverse patient outcome may be detected. In some cases, providers aware of a medical error may choose not to report it,15 focusing instead on mitigating patient harm. Providers may feel they have fulfilled their ethical duty by addressing the harms immediately resulting from an error, be unaware of error reporting systems, believe that error reporting is not their responsibility, or fear that they may face negative repercussions from error reporting.15 ,16 However, reporting and subsequent steps are essential to prevent future harms to other patients resulting from repetition of similar errors.

Table 2

Stages in addressing a medical error

Challenges in device-related errors

Device-related errors pose a particular challenge because of the burgeoning of medical technology. In the USA, from 1 January 2000 through 31 December 2014, FDA issued greater than 500 new 510(K) decisions (indicating that ‘the device to be marketed is at least as safe and effective, ie, substantially equivalent to a legally marketed device’) annually for medical devices, totally more than 7500 new devices in 15 years.19 By 2014, more than 200 such new devices were approved each month, suggesting that over 24 000 new devices will be approved over the next 10 years in the USA. These figures do not include wholly new devices, dissimilar to existing approved devices. Over the course of a physician's 20–30 years career, device evolution would require continuous re-education, a process that is difficult if not impossible to ensure. In some instances, substantial equivalence to existing devices (the basis for FDA 510(K)) may falsely reassure regulators and healthcare providers using a device that familiarity with related devices and with the general approach to a procedure is sufficient for safe use. If the degree of similarity with existing devices is high, a provider may fail to recognise that a knowledge deficit even exists for a new device, as in the cases described. End users of medical devices may not be involved in institutional decisions to replace one device with another; such decisions may be based on non-medical considerations such as cost or device availability. The first encounter that a healthcare provider may have with a new device may be in an emergency situation with a patient.

Emergency physicians could mitigate the danger of unfamiliar equipment by playing an active role in hospital purchasing decisions for devices. Historically, physicians in many specialties commonly have selected the specific equipment they prefer for their procedures, although this practice may be threatened by cost considerations and concerns over financial conflicts of interest.20 ,21 Whether or not they are involved in device selection, emergency physicians could be made aware of changes in available devices within their institutions and could engage in device-specific training. This process of education may be particularly difficult for physicians who practice in multiple healthcare systems, as is common for some emergency physicians.

Emergency situations pose particular challenges for healthcare providers faced with an unfamiliar device. In an ideal situation, a provider would never proceed with the use of an unfamiliar medical device in a living patient without first undergoing education and training. However, in an emergency situation, the provider may feel forced to act by the medical condition, overriding concerns about lack of familiarity with the device at hand. Medical errors may be more likely to occur in this situation,22 ,23 and policies to prevent use of an unfamiliar device may be unrealistic in a true emergency setting. However, providers should differentiate between true time-sensitive emergencies (such as tension pneumothorax with haemodynamic instability requiring immediate action) and ‘urgent’ medical scenarios (such as the pneumothoraces in the cases described) in which providers have adequate time to halt a procedure and gather more information when presented with an unfamiliar device.

Mandatory and recurring education, documented through an online database, could improve the awareness of device pitfalls and hazards. However, instruction and training cannot alone eliminate errors because the pool of potential users requiring training may be extremely large and difficult to define, even within a single health system or hospital. Moreover, errors can result from a variety of causes, not limited to inadequate medical knowledge.24 Therefore, in addition to education, device designs and labels for emergency scenarios must be engineered to anticipate user cognition and behaviour to reduce the risk of misuse to nearly zero. The ‘usability’ of devices can be measured using a standardised reporting form to predict use errors.25 For reusable devices, errors related to device design may be detected by monitoring of ‘no fault found’ data, in which a device returned for maintenance following an adverse event is found to be functioning properly.26 However, for disposable devices such as the Heimlich valves involved in the described cases, no such process of function review may exist.

Device-specific analysis: the Heimlich valve

Our analysis included device design and labeling features that may have contributed to the medical error. The Heimlich or flutter valve is often used with thoracostomy tubes and other medical devices requiring unidirectional flow.27 The Heimlich valve is a simple mechanical valve, consisting of a flexible and collapsible tube tethered inside a cylindrical chamber. Positive pressure through one end of the valve creates flow through the flexible tube. Negative pressure collapses the tube, preventing flow reversal. The result is a one-way valve (figure 2). The patient's normal respirations create the positive and negative pressure cycles, pumping air or fluid out of the pleural space when the valve is properly oriented and preventing air entry into the thorax. Advantages of the valve include low cost (<$50 per unit)28 and a simple design that does not require the application of a powered suction device. The valve also functions when in multiple different positions (eg, vertically oriented, horizontally oriented).27 Patients with pneumothoraces and thoracostomy tubes can be transported or can be ambulatory while disconnected from wall suction or large pleural drainage units when a Heimlich valve is attached to the thoracostomy tube. In some cases, patients are even discharged from the hospital with thoracostomy tubes and Heimlich valves in place.29

Figure 2

Schematic of a Heimlich valve. A collapsible tube is tethered within a rigid cylindrical chamber. Positive pressure within the tube distends it, allowing air flow. Negative pressure within the tube collapses it, preventing flow reversal.

The design of a Heimlich valve mandates unidirectional flow, the direction of which is determined by the orientation of the valve. Reversal of the valve orientation reverses the allowed direction of flow. In the case of a thoracostomy tube equipped with a Heimlich valve, reversal of the valve results in the entrainment of air through the tube into the pleural space when the patient inspires. That air is then trapped within the pleural space and cannot escape through the Heimlich valve. Within a short period of time, the accumulation of air in the pleural space can lead to an enlarging pneumothorax as in the patient described. Failure of healthcare providers to recognise the error and to correct the valve orientation could result in clinical tension pneumothorax and death. Reversal of the intended valve orientation is possible at the time of initial insertion, during tube manipulation by medical staff (nurses, physicians or other), or even by patients, who may not be under medical supervision.

In 1983, Heimlich asserted that “The construction and function of the valve is easily understood by medical and nursing staffs,”27 yet tension pneumothorax with reversal of a Heimlich valve has been reported previously.30 ,31 Mainini and Johnson31 reported two cases of tension pneumothorax from reversal of Heimlich valves; their report in Chest may not have reached the broad audience of providers who use these devices today. Spouge and Thomas reported a case very similar to that described here and made similar recommendations including device redesign. Their report was published in the Radiology literature. Recent reviews on treatment of pneumothorax describe and advocate for the use of small bore thoracostomy tubes equipped with Heimlich valves and recommend outpatient management of some patients, but do not mention the potential for tension pneumothorax with valve reversal.32 Some case reports include brief mention of the risk of reversal but without adequate emphasis (eg, three sentences in a five-page report).33

Solutions to the described case include improved education and instructions, device redesign, and regulatory action. Zhang et al34 identified 14 heuristics for usability problems in medical devices (table 3), many of which apply to the Heimlich valve, and which could be mandated by regulatory agencies. We employ these in our analysis below.

Table 3

Heuristics for usability problems in medical devices34

Consistency and standards/visibility of system state/use users’ language/minimalist/match between system and world

It goes without saying that medical providers should be expected to have sufficient knowledge of anatomy, physiology, procedures and the specific medical device used to anticipate many complications of a procedure. However, device instructions are important. Labels on the Cook Heimlich valve correctly indicate the direction of flow but are ambiguous and difficult to read, as they are embossed in transparent text upon the transparent plastic of the device in a small font (figure 3A–C). Changing label colours and opacity to enhance contrast would improve legibility. The warning at one end of the device states ‘connect chest drain to this end’—however, the term ‘chest drain’ could be interpreted by users as referring to the thoracostomy tube leading from the patient (the manufacturer's intended definition) or as referring to the chest drainage unit external to the patient. At our institution, the chest drainage unit used in the emergency department (Express Dry Seal Chest Drain, Atrium Medical Corporation, Hudson, New Hampshire, USA) explicitly bears the label ‘chest drain,’ potentially adding to user confusion (figure 4).35 A less error-prone instruction could describe valve orientation relative to the patient, such as ‘orient this end of valve towards patient’, or employ carefully selected symbols to convey the proper orientation. Manufacturers could test labels for clarity of message and legibility before deployment, diminishing the risk of misinterpretation.

Figure 3

(A) One-way Heimlich valve. The intended direction of flow is indicated. However, the instructions are difficult to read because the arrow and text are embossed on a clear background in a small font. The manufacturer's name ‘COOK’ is printed in a larger font than the crucial orientation instructions. The adaptors at each end of the valve are identical in shape, allowing the device to be improperly oriented in the drainage system. The text ‘connect chest drain to this end’ is ambiguous, as a user might interpret ‘chest drain’ to signify either the tube emerging from the patient's chest or the chest drainage unit. The valve mechanism is visible within the valve chamber, but a user may not correctly infer the intended direction of orientation from the valve appearance. (B) Correct insertion of the valve into the catheter leading into the patient's chest. In this correct orientation, the valve allows egress of air or fluid from the patient's chest and prevents entrance of air. (C) Incorrect insertion of the valve into the catheter leading into the patient's chest. In this incorrect orientation, the valve prevents egress of air and fluid from the patient's chest and allows entrance of atmospheric air following the negative intrathoracic pressure generated by patient inspiration. The result can be tension pneumothorax and death.

Figure 4

A common chest drainage unit (Express Dry Seal Chest Drain, Atrium Medical Corporation, Hudson, New Hampshire USA). (A) External product label. (B) The chest drainage unit. Because the term ‘chest drain’ (rectangular outline added for emphasis) can refer to either the chest drainage unit or to the thoracostomy tube, the label ‘connect chest drain to this end’ on the Heimlich valve is not sufficiently explicit to prevent error. Standardising terminology through regulatory reform could reduce the risk of such errors.

Informative feedback/good error messages/reversible actions/clear closure/use users’ language/users in control/help and documentation

Manufacturer instructions for valve insertion state only: “Attach catheter to connecting tube with stopcock, and Cook Chest Drain Valve. Attach Cook Chest Drain Valve in direction indicated by arrow on valve. Note: Chest Drain Valve may be obviated if catheter is to be connected to a water seal suction apparatus or similar mechanical suction device.” We find these instructions ambiguous. These could be improved with a diagram, additional instructions to test valve function after insertion, and explicit warnings. For example, manoeuvres to confirm correct valve orientation should be suggested (eg, ‘After insertion, test Heimlich valve function by having patient cough and take a deep breath. The valve should open with cough and deep breath.’). Such instructions could provide closure to the procedure and allow providers to recognise and recover from improper valve orientation. An explicit warning label (eg, ‘WARNING: IMPROPER ORIENTATION OF VALVE CAN RESULT IN TENSION PNEUMOTHORAX AND DEATH’) could be added to the valve within the kit to draw attention to the correct orientation and risks.

Instructions for patients or first responders about the risks of the valve are also needed, particularly given that the valve may be used in an outpatient setting. Disconnecting an improperly oriented Heimlich valve will convert a tension pneumothorax to an open pneumothorax, relieving high intrathoracic pressures and improving haemodynamics. Instructions to disconnect the valve in case of patient cardiac arrest could be beneficial.

Many medical devices could lead to patient harms if used improperly, and excessive instructions and warnings could lead to alarm fatigue, which itself can endanger patient safety.36 The US FDA provides guidance on the need for device warnings: “Include an appropriate warning if there is reasonable evidence of an association of a serious hazard with the use of the device. A causal relationship need not have been proved.”37 Warnings may be appropriate when a simple error can lead to a device not serving its intended function and to exacerbation of the underlying medical condition, leading to a time-sensitive life threat. In some cases empirical experience such as the recurrence of the same error committed by different providers may provide the imperative for product warning labels. Warnings may be unnecessary when the core function of a product reveals its inherent risks, as for a scalpel (sharp instrument intended to cut; misuse may result in injury) or when complications should be anticipated due to the anatomical properties inherent in the procedure.

Prevent errors/minimise memory load

Fundamental changes in valve design could also provide significant improvement in safety by rendering incorrect orientation difficult or impossible. Multiple vendors produce Heimlich valves; some may have implemented design changes that address the issues identified in our review. However, 32 years have passed without safety improvements to the Heimlich valve described in this case (Cook Medical).27 The Heimlich valve included in the Cook Wayne Pneumothorax Tray and some other vendors is symmetrically designed, with conical ‘Christmas tree’ adaptors at each end (figure 3A). Consequently, the valve can be inserted in the incorrect orientation into the included thoracostomy tube adaptor, which terminates in a funnel shaped opening (figure 3B,C). The valve is similar in appearance to the symmetrical, valve-less adaptor commonly used to connect thoracostomy tubes to chest drainage units, adding to potential confusion. In the cases described, the providers stated that they mistook the Heimlich valve for the valve-less adaptor because of physical similarity. Other vendors manufacture Heimlich valves similar to that encountered in the cases described, which may pose similar safety risks. Becton, Dickinson and Company offers a Heimlich valve ( The website provides clear instructions about correct valve orientation but does not provide any safety warning about the potential adverse clinical outcome of incorrect valve orientation.38 Other vendors have developed more intuitive orientation labels (Vygon, Ecouen, France,, although symmetrical device design could still allow incorrect orientation (figure 5). The Heimlich valve could be redesigned with asymmetrical and unique connections, preventing incorrect orientation even without the need for instructions or labeling. Such one-way valves with asymmetrical connections are available. (eg, Pneumostat Chest Drain Valve, Atrium Medical Corporation, Hudson, New Hampshire, USA, A 5 year review of 98 cases following thoracic surgery found minimal complications of this device, although the study was retrospective, did not clearly describe the chart review technique or complications sought, and may have been confounded by underdocumentation of adverse events in the medical record.39 ,40

Figure 5

A Heimlich valve from another vendor bears more intuitive and legible labels, although the symmetrical device design could still allow incorrect orientation (Vygon, Ecouen, France,

New Joint Commission/International Organization for Standardization standards on connectors for medical devices will soon mandate unique connectors, as tubing misconnections have been recognised as a cause of injury and death.41 However, the new regulations pertain to enteral feeding devices as the first target by early 2016 and do not include the devices described in this report. The roll-out of the broader expectations for delivery systems including neuraxial, limb cuff inflation, and respiratory applications is anticipated at an even later date. More comprehensive and earlier regulatory reform for medical devices is needed.


Medical errors are commonly multifactorial and can be reduced only through a complete cycle of reform, including detection, mitigation of harm to the patient, disclosure, reporting and surveillance, root cause analysis, system modification, regulatory intervention, and engineering and manufacturing reforms. In the cases described, an improperly oriented Heimlich valve in the setting of thoracostomy tube placement resulted in radiographic tension pneumothorax; rapid correction of the error prevented potential patient death. Improved education, user instructions and device designs could mitigate these risks. Common medical references make little mention of the risk of tension pneumothorax. Physicians, physician extenders, nurses, prehospital providers and patients may be unaware of the life-threatening hazard of improper valve orientation and the need for familiarity with this, and any other device, used in the emergency department.



  • Contributors JSB, conceived of the manuscript, wrote the initial draft, performed the associated literature search, and is responsible for the final draft. JWF, JM, BJT and AW participated in case review and analysis, edited the manuscript, and reviewed and approved the final manuscript.

  • Competing interests None declared.

  • Ethics approval Duke University School of Medicine Institutional Review Board.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles