Preventing Retained Central Venous Catheter Guidewires: A Randomized Controlled Simulation Study Using a Human Factors Approach

CENTRAL venous catheters are used in health care for monitoring and administering medications. More than 5 million central venous catheters are placed every year in the United States.¹  The standard technique for central venous catheter insertion is the Seldinger method; however, it is associated with guidewire retention, which can cause complications such as arrhythmia, thrombosis, cardiac perforation, and tamponade.²  Retained guidewires have an incidence of 1:3,291 procedures,³  a reported mortality of up to 20%,^2,4  and are considered a never event in the United States and United Kingdom.^5,6  Therefore, they are regarded as preventable errors, which should not occur.^5,6  Case reports of retained guidewire events in the literature describe reasons for the error as inattention,^7,8  distraction,⁹  inexperience,⁷  inadequate supervision,^7,9  high workload,⁹  or staff fatigue.⁸  These authors conclude that guidewire retention is preventable with appropriate retraining of staff by highlighting the importance of guidewire removal,^2,10  by using a two-person approach,^10,11  by introducing a checklist,¹¹  by checking the trolley after the procedure,¹²  by ensuring the clinician grips the proximal end of the guidewire at all times,⁸  by not inserting the guidewire beyond 18 cm,¹³  and by having active senior supervision at all times with trainee doctors.^9,11  Despite these suggestions, the incidence of retained guidewires continues to rise and is reported twice a month in the National Health Service (NHS).¹⁴  These measures do not reliably prevent the error because they are reliant on the human operator remembering to perform the safety check.¹⁵  Humans are fallible and prone to mistakes. Solutions that rely solely on the operator preventing mistakes are unlikely to be completely effective. The transport and energy industries routinely use safety engineering to modify their equipment and design errors out of the system.¹⁶  In health care, for single procedures with specific known errors, it may be possible to use safety engineering to modify the equipment and ensure the operator conducts the procedure by the safest method. We aimed to engineer a safety solution to prevent the error of retained guidewires (see  appendix). The solution was tested in a forced-error simulation study with operators who had no experience of the solution. The design of the simulation study was based on real never-event cases reported to NHS England’s national reporting database.* Forced-error simulation techniques are used in the transport and energy industries to test safety solutions to rare errors, such as air bags in cars or emergency switches in airplane cockpits.^17,18  This technique manipulates the simulated environment to make a rare incident very common to allow preventative solutions to be tested. The participant is initially “forced” into making the error, and then it is determined whether the intervention makes the participant recognize the error and subsequently correct the mistake. For rare errors, occurring in one in several thousand procedures, forced-error simulation is a validated, safe, repeatable, and inexpensive test methodology.^17,18  The primary outcome of this study was incidence of guidewire retention at completion of central venous catheter insertion.

Ethical approval to conduct the research was granted by the University of Cambridge (Cambridge, United Kingdom). The institutional research and development review board approved the simulation study that was conducted at the Queen Elizabeth Hospital (Kings Lynn, United Kingdom) on a single day on September 4, 2015, to ensure confidentiality of study design between participants. Volunteers were requested and self-selected on a single day from the operating room, the intensive care unit, and the general medical and surgical wards. All volunteers were capable of independent central venous catheter placement (n = 20, none excluded; fig. 1A). Signed, informed consent was taken from all participants, who were qualified medical doctors at various degrees of seniority (foundation trainee to consultant level; equivalent to intern to attending physician in U.S. parlance). Participants were randomized to standard practice or locked pack by sealed envelope randomization. Twenty identical envelopes were sealed with control (n = 10) or intervention (n = 10) indicated on paper within and shuffled into a random order. Immediately prior to a participant entering the scenario room, the data collection team (blinded to the participant’s identity) opened an envelope and set up the procedure trolley appropriately.

A scenario was described to the participants using a standardized script prior to entering the room. The scenario outlined was that a colleague had been urgently called away, part way through a routine central venous catheter insertion on a clinically stable patient. Participants were asked to assess the situation, complete the procedure safely, and perform any additional safety checks prior to approving the central venous catheter for use. The simulation utilized a manikin model (Laerdal, USA) adapted for central venous catheter insertion and covered with standard blue surgical drapes (Vygon, United Kingdom) with a clear window for central venous catheter insertion. A central venous catheter (Arrow International, Inc., United Kingdom) was placed in the right internal jugular vein, with the guidewire clearly visible in the transparent portion of the catheter lumen, adjacent to the luer hub (fig. 1B). The guidewire was easily retrievable, if recognized, with fingertips or artery forceps, which were provided upon request. The manikin was connected to an electrocardiogram monitor, which displayed ectopic beats, and an ultrasound machine was available for use if required. A trolley with the equipment required to perform central venous catheter insertion (Rocialle, United Kingdom) was positioned to the right of the patient. Equipment was arranged depending on participant randomization to standard practice or locked pack group. An assistant was available to answer questions and help the participant if required. If participants specifically asked about the location of the guidewire, the assistant stated that they had not seen it. If asked about the ectopic beats, the assistant stated these had commenced during the procedure. In the control group, participants entered the room, assessed the situation, and proceeded to secure the central venous catheter in place and apply the dressings. Upon completing the procedure to their satisfaction and safely disposing of the equipment, participants were asked whether they would perform any additional safety checks prior to using the central venous catheter. Participants randomized to the locked pack group entered the room, assessed the situation, and proceeded to secure the central venous catheter in place. The assistant did not explain how to use the locked pack. If participants specifically asked, the assistant stated that the locked pack was a new safety initiative, but were unsure of its purpose or how to use it. Instructions on the locked pack indicated that the guidewire should be inserted and lifted to open. At the end of the procedure, participants randomized to the locked pack group were given a structured verbal questionnaire that asked their opinion of the locked pack in terms of subjective safety of procedure, convenience, sharps disposal, and guidewire disposal (all categorized as better, same, or worse).

We calculated a power of 0.87 for a statistical significance of 0.05 and for n = 10 to detect a 50% absolute difference in proportions, and we used a two-tailed Fisher exact test to analyze the data (GraphPad Software, Inc., USA). The criterion for statistical significance was 0.05.

The standard group consisted of four women (aged between 23 and 60 yr), six men (aged between 24 and 60 yr), of which four were senior and six were doctors in junior posts. The locked pack group consisted of three women (aged between 25 and 40 yr), seven men (aged between 25 and 60 yr), of which three were senior and seven were doctors in junior posts. Use of the locked pack prevented guidewire retention at completion of central venous catheter insertion. Guidewire retention was prevented in 2/10 (20%) standard versus 10/10 (100%) locked pack, n = 20, P < 0.001. In the standard group, 80% (8/10) of participants failed to recognize the guidewire in the catheter lumen. They secured the central venous catheter, applied the dressings, and were satisfied that they had completed the procedure correctly. In the locked pack group, two participants recognized the guidewire in the lumen. Those who did not (8/10) attempted to complete the procedure. However, inability to access the equipment inside the locked pack triggered a search for the guidewire by the participant. Participants searched the trolley, floor, and sharps bin before looking at the central venous catheter, and finding the guidewire within the catheter lumen. All participants in the locked pack group were able to remove the guidewire, use it to open the locking mechanism (opening procedure took < 10 s in all cases), and finish the procedure. The structured questionnaire of the locked pack group reported that it improved the safety and convenience of central venous catheter insertion and allowed easy disposal of the sharps and guidewire (10/10).

A retained central venous catheter guidewire is a never event that affects patients, clinicians, and hospitals. The reasons for retained guidewires are well documented in the literature and preventative approaches traditionally include appropriate re-education and re-training of staff when the error occurs and highlighting the importance of guidewire removal.^2,10  However, “re-education and training is only as good as the length of time that clinicians remember to do it.”¹⁵  Therefore, this solution necessitates repeated episodes of training and often a cycle of recurrent mistakes. Other suggestions are that loss of the guidewire is preventable with a two-person approach, or with a second person witnessing guidewire removal.^10,11  Requiring two individuals to perform a procedure is both time and cost ineffective, and not always possible in a busy hospital environment without frequent delays to every central line insertion. Furthermore, in terms of rare events, after thousands of procedures with no errors, this leads to creeping complacency by both individuals, with each individual relying on the other to conduct the procedure correctly and leading to unclear accountability.¹²  Several reports suggest having active supervision by senior staff at all times during procedures.^9,11  A number of cases, however, report guidewire retention, despite adequate senior supervision,⁹  and this solution does not address the cases when very experienced senior clinicians make this mistake.⁹ 

Another common suggestion is that guidewire retention is preventable if the clinicians grip the proximal end of the guidewire at all times⁸  or ensure that they do not insert the guidewire beyond 18 cm.¹³  However, it is difficult to always ensure that clinicians act in this way, especially when the lone operator is singlehandedly manipulating equipment or when distractions or clinical emergencies occur.^2,16  Other suggestions have included introducing a checklist for central venous catheter insertion, requiring the documentation of guidewire removal,¹¹  or checking the trolley after the procedure.¹²  Similarly, these measures rely on the clinician remembering to perform the safety step and are prone to fail.

Safety and human-factor principles demonstrate that sole reliance on humans to prevent error inevitably fails.¹⁶  Therefore, rather than introducing complex protocols, which require the clinician to remember to perform the safety action, the best method is to introduce engineered safety systems that allow operators to perform their job safely and prevent the error. Human-factor engineering is used in the energy and transport industries to promote safe practice by anticipating error and redesigning equipment to minimize mistakes.¹⁹  This is possible in health care with an understanding of the procedure, operator, and working environment.²⁰  If guidewire retention is immediately recognized, the guidewire often remains within the catheter lumen (Nationally reported incidents of central line guidewire retention, 2006–2015). At this stage, retrieval is almost always possible by clamping artery forceps at the skin level to the catheter and enclosed guidewire and removing enblock,⁹  reversing the potential of further migration and embolization. In this study, use of the locked pack prevented guidewire retention when compared with standard practice. Use of the locked pack forced participants to recognize guidewire retention because they were unable to complete the procedure without the guidewire ensuring the guidewire was removed each time. The locked pack also was found to be intuitive to use, given that all participants, despite being naive to the device, were able to understand how to use the device. All participants took less than 10 s to utilize the guidewire to unlock the mechanism and access the contents. An interesting observation in the locked pack group was the participant’s behavior when searching for the guidewire. Participants initially searched the trolley, floor, and sharps bin, and finally looked at the central venous catheter lumen. This highlights that the error of a retained guidewire is a low possibility in the clinician’s mind. Commonly proffered solutions in this context include reiteration and emphasis of the importance of guidewire removal. Forcing awareness of a rare error into the mind of the operator increases cognitive load and has the potential to detract awareness from other central venous catheter complications, such as pneumothorax, arterial puncture, dysrhythmias, or air embolism.²¹  The locked pack aided the participant in recognizing the error at an easily correctable point during central venous catheter insertion, without adversely interfering with the procedure. In clinical practice, the locked pack ideally would be incorporated and supplied as part of the central line insertion sterile pack.

The design of the simulation study was based on real never-event cases reported to NHS England’s national reporting database (Nationally reported incidents of central line guidewire retention, 2006–2015) and was performed in such as way as to “force” the error to occur within the scenario. Critics of forced-error simulation may argue that the simulation scenario was unrealistic and the locked pack should be tested replicating routine clinical practice. However, for an error that occurs in 1:3,291 procedures, a power calculation (GraphPad Software, Inc.) shows that a total of 12,000 participants would be required for the study if one were to test the device replicating routine clinical practice. To contextualize this, there are 13,955 anesthetists in the United Kingdom.²²  Replicating routine clinical practice is, therefore, an impractical vehicle to test a safety solution for a rare error in a safe, repeatable, and inexpensive fashion. Therefore, the most cost-effective and practical approach to testing rare errors is to evaluate the intervention in a forced-error simulation; hence, the current scenario was used, which was based on real case events. The energy and transport industries are often called the high reliability industries²³  due to their adoption of a human-factor safety culture and their high safety records. In these industries, similar solutions are tested with forced-error simulation, replicating rare errors to analyze operator behavior and improve or test new safety measures.²⁴  For example, in the car industry, this approach has been used in airbag deployment testing. Industry standards require the replication of the driver being forced into common car crash positions in order to test the effectiveness of the safety technology, in this case air bags.¹⁷  It is impossible to test the ability of equipment designed to improve safety during car crashes, without crashing the car. One could perform routine practice by simply driving the car for thousands of hours and waiting for a crash to occur before determining the safety of the equipment. However, this is unsafe, expensive, and time inefficient. Therefore, the manufacturer forces car crash scenarios, testing the efficacy of the equipment in a repeatable forced-error simulation. Lessons from industry indicate that it is more cost effective to systematically design methods to test safety interventions, where the scenarios are designed to align the latent causes of failure, to make a rare event more common²⁵ —an extrapolation of Reason’s Swiss Cheese Model.²⁶  This engineered sensitization of enhanced risk also removes variability in operator performance and procedure context that may modulate outcome, and thus provides a more tractable context for assessing the intervention.

Despite the majority of reported retained guidewire incidents being single catheter insertions (97.5%; Nationally reported incidents of central line guidewire retention, 2006–2015), another criticism of the locked pack may be that it does not prevent the rarer event in which two or more catheter packs are opened. This happens in two scenarios. The first of these is where there is a planned insertion of multiple central venous catheters within the same procedure, with the use of multiple kits. The second of these is the minority of cases where there has been a difficulty in catheter insertion with the first kit, and a second kit is opened. Although the locked pack would not be able to assuredly prevent guidewire retention in either of these situations, they represent a small minority of reported retained guidewire incidents (2.5%; Nationally reported incidents of central line guidewire retention, 2006–2015), and the locked pack approach could still substantially mitigate the risk of this never event in the substantial majority of cases. To circumvent this problem, one could use an equivalent number of locked packs on the sterile field; however, we are in the process of exploring this area with a view to design improvements to address the problem. Concurrently, we also are exploring the expansion of the locked pack to other Seldinger techniques, which are associated with guidewire retention, such as chest drain insertions.

Importantly, one must remember that although engineered solutions provide greater assurance of safety, no single intervention, on its own, represents a foolproof solution. The introduction of safety initiatives is not designed to remove the need to cognitively engage during the procedure, but to minimize error and aid operators to perform their job safely. If introduced clinically, we believe the locked pack will improve patient safety and protect clinicians from making this error.

A retained central venous catheter guidewire is a never event. Current preventative solutions are dependent on the operator remembering to remove the guidewire. Using human-factor engineering principles, we have designed a safety intervention, a novel locked procedure pack that acts to ensure the operator always removes the guidewire, thereby preventing the never event. We have tested the solution with forced-error simulation testing—a novel clinical application of methodology that is validated in the high-reliability industries. We believe that adoption of this technique can not only improve patient safety, but also protect clinicians from making this error.

Our thanks to Andrew Livesey, M.B.B.S. (Foundation Trainee, Critical Care Department, The Queen Elizabeth Hospital, National Health Service Foundation Trust, Kings Lynn, United Kingdom), Shrestha Sinha, M.B.B.S. (Staff Grade Doctor, Critical Care Department, The Queen Elizabeth Hospital, National Health Service Foundation Trust, Kings Lynn, United Kingdom), Holly Young (Medical Student, University of Cambridge, Cambridge, United Kingdom), and Jay Patel (Medical Student, University of Southampton, Southampton, United Kingdom) for their assistance with data collection. The acknowledged contributors have not been given compensation for their assistance with this work.

This work has been supported by the Eastern Academic Health Science Network, Cambridge, United Kingdom.

Dr. Young has patented and is the inventor of the Venner WireSafe based on the locked pack described in this article. Dr. Mariyaselvam has been selected as a National Health Service Innovation Accelerator Fellow and Clinical Entrepreneur Fellow to support the implementation of the Venner WireSafe. Both have ownership rights to the intellectual property and hope to be involved with supporting and advising on the commercialization of the Venner WireSafe, and they may benefit financially alongside its clinical success, subject to future negotiation and agreement. All other authors declare no competing interests.

Retained guidewires during the central venous catheter placement occur at a “critical point” in the procedure—when the catheter is placed over the guidewire. At this point, if clinicians are inexperienced or distracted, they can forget that the guidewire has not been removed. Without realizing this, they will continue to advance the catheter and finish the procedure by securing the catheter in place. Guidewire retention is then often discovered on or after the initial check radiograph and even at this stage migration may not have occurred and removal is often possible (Nationally reported incidents of central line guidewire retention, 2006–2015).¹⁴ 

A novel procedure pack was designed to ensure removal of the guidewire after the “critical moment.” Normally, after the guidewire is removed from the patient and in order to complete the central venous catheter procedure, the following equipment is needed: suture, suture holder, and antimicrobial dressing. We developed a locked pack with a key-lock mechanism, which contained the suture, suture holder, and antimicrobial dressing (fig. 2). The guidewire is used as a key. It is inserted into the lock, and remaining inside the lock, is used as a handle to open the lid of the locked pack to allow the operator to access the suture, suture holder, and antimicrobial dressing (fig. 3). Therefore, the only way in which the operator can access the contents to complete the procedure is by first removing the guidewire from the patient after the “critical point.” This ensures that clinicians remember to remove the guidewire because they are unable to complete the procedure without doing so. As an additional safety feature, a sharps sticker was placed at the base of the locked pack and instructions for use on the underside. Once the suture, suture holder, and dressings have been removed, the sharps sticker becomes visible (fig. 4). The now empty locked pack becomes a convenient and safe container for collecting the sharps used during the central venous catheter insertion. The used sharps are placed inside the locked pack and the lid is securely closed, with the guidewire safely retained inside the locking mechanism (fig. 4; if required, it is possible to remove the guidewire from the locked pack). The locked pack is then disposed of into the sharps bin (see video, Supplemental Digital Content, https://links.lww.com/ALN/B522, which illustrates the instructions for use of the novel locked pack).

This was a box of dimensions 3 cm height x 10 cm width x 20 cm length (fig. 2). This contained all the equipment required to complete a central venous catheter insertion procedure from the moment immediately after the guidewire should have been removed: the suture, suture holder, and antimicrobial dressings. The lid was designed flush to the box and held shut with a magnet inset into the lid and a metal plate on the box. This made the lid unable to be gripped and opened. The magnet held the lid closed even if inverted and the contents shaken.

A light bulb–shaped channel within the lid accepts a guidewire as large as 10 gauge in size. The guidewire easily passes round the channel and emerges through an identical second hole parallel to the entry hole (fig. 3). Passing a guidewire in this manner leaves two ends, which form a convenient handle. The lid is easy to open with countertraction provided by the angular force of lifting the newly created guidewire handle.

The prototype procedure pack lid firmly closes again so that following the procedure it could be used as a point-of-care sharps receptacle and guidewire retainer to facilitate safe disposal (fig. 4).