Advertisement

Defibrillator Design and Usability May Be Impeding Timely Defibrillation

Open AccessPublished:July 03, 2018DOI:https://doi.org/10.1016/j.jcjq.2018.01.005

      Background

      Timely defibrillation is the only rhythm-specific therapy proven to increase survival to hospital discharge following cardiac arrest secondary to ventricular tachyarrhythmia. Delayed defibrillation occurs in more than 30% of this population. A study was conducted to test the hypothesis that unintuitive defibrillator design and lack of usability are barriers to timely defibrillation, as measured by time to defibrillation and the proportion of defibrillations delivered within 2 minutes.

      Methods

      A mixed-methods (qualitative and quantitative) prospective usability study was performed to evaluate the use of a defibrillator in a simulated hospital environment. Participants were asked to perform two simulated tasks typical of in-hospital cardiac arrest care: defibrillation and synchronized cardioversion.

      Results

      The average time to defibrillation was 4 minutes 21 seconds. Only 9.1% of participants (2/22) performed a defibrillation within 2 minutes. Participants had difficulty with several aspects of defibrillator use, including attaching the hands-free defibrillator electrode pads and selecting an appropriate display. Participants rated defibrillator design 4.2 ± 1.8 (mean, standard deviation) on a perceived usability scale (1 = “poorly designed”; 9 = “perfectly designed”).

      Conclusion

      Most participants were unable to perform a simulated defibrillation within 2 minutes. This delay in defibrillation was likely at least partially the result of poor defibrillator design and lack of usability. Expert observation and participant feedback were largely congruent in terms of which aspects of defibrillator design do not suit the end user. Modification of future defibrillator design, along with improved education and training, may result in more timely defibrillation.
      Each year, between 370,000 and 750,000 American patients suffer in-hospital cardiac arrest with attempted cardiopulmonary resuscitation. Approximately 20% of these patients will have suffered a ventricular tachyarrhythmia. In this population, the only rhythm-specific therapy proven to increase survival to hospital discharge is timely defibrillation.
      • Eisenberg MS
      • Mengert TJ
      Cardiac resuscitation.
      • Girotra S
      • et al.
      Trends in survival after in-hospital cardiac arrest.
      • Chan PS
      • et al.
      Delayed time to defibrillation after in-hospital cardiac arrest.
      Guidelines published by the American Heart Association recommend the defibrillation of in-hospital cardiac arrest secondary to ventricular tachyarrhythmia occur within 2 minutes of recognition of the arrest.
      • Chan PS
      • et al.
      Delayed time to defibrillation after in-hospital cardiac arrest.
      • Neumar RW
      • et al.
      Part 1: executive summary: 2015 American Heart Association guidelines update for cardiopulmonary resuscitation and emergency cardiovascular care.
      Patients who experience delayed defibrillation beyond 2 minutes are less likely to survive to hospital discharge, and among survivors are more likely to have disabilities in neurologic and functional status.
      • Chan PS
      • et al.
      Delayed time to defibrillation after in-hospital cardiac arrest.
      • Herlitz J
      • et al.
      Very high survival among patients defibrillated at an early stage after in-hospital ventricular fibrillation on wards with and without monitoring facilities.
      ,
      • Spearpoint KG
      • McLean CP
      • Zideman DA
      Early defibrillation and the chain of survival in ‘in-hospital’ adult cardiac arrest; minutes count.
      A retrospective analysis of the National Registry of Cardiopulmonary Resuscitation—a large American registry of more than 14,000 in-hospital cardiac arrests—demonstrated that delayed defibrillation occurs in more than 30% of this patient population.
      • Chan PS
      • et al.
      Delayed time to defibrillation after in-hospital cardiac arrest.
      • Peberdy MA
      • et al.
      Cardiopulmonary resuscitation of adults in the hospital: a report of 14720 cardiac arrests from the National Registry of Cardiopulmonary Resuscitation.
      At our institution, the average time to defibrillation is 8 minutes in a critical care setting and 11 minutes on a medical ward.
      • Godbout J
      • et al.
      Innovative use of AED by RNs and RTs during in-hospital cardiac arrest: phase II.
      Identifying barriers to timely defibrillation—and implementing solutions to circumvent these problems—is an opportunity to improve quality of care. However, limited information is available regarding the system-related and patient-related factors that are associated with delayed defibrillation.
      • Chan PS
      • et al.
      Delayed time to defibrillation after in-hospital cardiac arrest.
      • Chan PS
      • et al.
      Hospital variation in time to defibrillation after in-hospital cardiac arrest.
      ,
      • Peberdy MA
      • et al.
      Survival from in-hospital cardiac arrest during nights and weekends.
      Usability testing is a technique in which users interact with a product under controlled conditions, and behavioral data are collected.
      • Wichansky AM
      Usability testing in 2000 and beyond.
      This information is then applied to better understand human performance capabilities and limitations, and how product design can be modified to meet these needs. Although recognized as an essential component of safety engineering in other fields such as nuclear power or aerospace, usability testing has been reported to be underutilized in health care.
      • Karsh BT
      • Scanlon M
      When is a defibrillator not a defibrillator? When it's like a clock radio . . . the challenge of usability and patient safety in the real world.
      • Rousek JB
      • Hallbeck MS
      The ergonomics of “Code Blue” medical emergencies: a literature review.
      • Gosbee J
      Who left the defibrillator on?.
      We hypothesized that flaws in defibrillator design contribute to a delay in timely defibrillation, and we aimed to identify such flaws via usability testing in a simulated hospital environment.

      Methods

      Design

      This mixed-methods (qualitative and quantative) prospective usability study evaluated the use of a manual-mode defibrillator in a simulated hospital environment. The study protocol was approved by the Ottawa Health Science Network Research Ethics Board (protocol 20150868). Informed consent was obtained for all participants. Participants were assigned a unique identification number to facilitate de-identification of data, and results were made available only to the investigators.

      Setting

      The study was conducted in a high-fidelity clinical exam room at the University of Ottawa Skills and Simulation Centre (uOSSC). Operated in conjunction with the University of Ottawa and The Ottawa Hospital (TOH), the uOSSC is the largest medical simulation center in Canada and one of the largest in North America.
      University of Ottawa Skills and Simulation Centre
      Facts and Figures.
      TOH is an academic tertiary care regional referral center with more than 1,100 inpatient beds. Simulation was performed using the Laerdal full-size ALS [Advanced Life Support] Simulator and operated using the SimPad remote (Laerdal Medical, Stavanger, Norway) (Figure 1). The SimPad software can generate more than 1,400 cardiac rhythms. Simulated rhythms are then transmitted and displayed using a clinical monitor or manual-mode defibrillator. The Laerdal ALS Simulator allows for the palpation of synchronized pulses. Other functions of the Laerdal ALS Simulator, such as endotracheal intubation and spontaneous breathing, were not used.
      Laerdal Medical
      ALS Simulator (Discontinued).
      The Philips HeartStart XL (Koninklijke Philips N.V., Amsterdam) was chosen as the defibrillator of interest, as it is the sole device used at our institution (Figure 2). The defibrillator was equipped with both hands-free defibrillator electrode pads and handheld defibrillator paddles in accordance with local practice.
      Fig 1
      Figure 1Panel A of the figure shows the Laerdal full-size ALS [Advanced Life Support] Simulator, and the SimPad remote is shown in panel B. Images reproduced with permission. Source: Laerdal. ALS Simulator (Discontinued). 2018. Accessed Jun. 22, 2018. https://www.laerdal.com/us/ALS#/Images.
      Fig 2
      Figure 2The Philips HeartStart XL manual-mode defibrillator is shown. Image reproduced with permission. Source: Southwest Medical. Philips HeartStart XL Defibrillator – Refurbished. 2018. Accessed Jun 22, 2018. https://swmedical.com/products/philips-heartstart-xl-defibrillator-refurbished.

      Participants

      Twenty-two internal medicine residents postgraduate year (PGY) 1–3 were recruited to participate. Potential participants (all University of Ottawa internal medicine residents enrolled at the time of simulation; N = 72) were contacted via e-mail, and participation was voluntary. All participants were certified in advanced cardiovascular life support (ACLS) and had completed ACLS certification training within the preceding 12 months. Senior residents (PGY-2–PGY-3) are cardiac arrest team leaders at our institution, while junior residents (PGY-1) are members of the cardiac arrest team. Both senior and junior residents are often tasked with defibrillator use during cardiac arrest.
      Staff physicians and resident physicians in other disciplines were not recruited. At our institution staff physicians are not in-house 24 hours per day and thus not formal members of the cardiac arrest team. Resident physicians in other disciplines—such as critical care or cardiology—are occasionally unable to leave critically ill patients in the ICU or coronary care unit (CCU) to attend to cardiac arrest on the ward. As such, they are not members of the cardiac arrest team.

      Procedure

      Prior to the simulation, participants were briefed regarding the objective of the study (to assess the usability of a manual-mode defibrillator) and instructed to follow standard ACLS protocol.
      • Neumar RW
      • et al.
      Part 1: executive summary: 2015 American Heart Association guidelines update for cardiopulmonary resuscitation and emergency cardiovascular care.
      All materials required to use the defibrillator with either the hands-free defibrillator electrode pads or with the handheld defibrillator paddles were provided. Participants were encouraged to express their thoughts throughout the simulation, and information gathered via this means was included in the qualitative analysis.
      Participants were provided with a written scenario (Sidebar 1): “You are on call for the internal medicine service at your local hospital. A code blue (cardiac arrest) is called overhead. You happen to be standing immediately outside the room for which the code blue was activated. Please enter the room and follow the direction provided to you by the cardiac arrest team leader.” The scenario noted that, “unless explicitly requested by the cardiac arrest team leader,” participants were not required to perform other tasks typical of cardiac arrest care (for example, chest compressions, airway management, IV access, administration of medication). When participants arrived in the simulation room, further instruction was provided by a confederate playing the role of cardiac arrest team leader: “Please attach the defibrillator to the patient, perform a rhythm check, and deliver either an immediate defibrillation or immediate synchronized cardioversion dependent on the rhythm identified. Please assume all rhythms are unstable and require immediate management.”
      Scenario
      You are on call for the internal medicine service at your local hospital.
      A code blue (cardiac arrest) is called overhead. You happen to be standing immediately outside the room for which the code blue was activated.
      Please enter the room and follow the direction provided to you by the cardiac arrest team leader.
      You are NOT required to perform any other tasks typical of cardiac arrest care (Eg: chest compressions, airway management, IV access, administration of medication) unless explicitly requested by the cardiac arrest team leader.
      Upon arrival into the simulation room, participants were instructed by a confederate playing the role of cardiac arrest team leader:
      “Please attach the defibrillator to the patient, perform a rhythm check, and deliver either an immediate defibrillation or immediate synchronized cardioversion dependent on the rhythm identified. Please assume all rhythms are unstable and require immediate management.”
      No other confederates were present during the simulation. Participants were instructed to assume that other tasks typical of cardiac arrest care were ongoing throughout the simulation. One participant was evaluated per simulation, and no other participants observed.
      Following the simulation, participants completed a post-exposure questionnaire regarding their perceived usability of the manual-mode defibrillator (Appendix 1, available in online article).

      Outcomes

      The primary outcome was time to defibrillation (including the proportion of participants able to deliver a defibrillation within 2 minutes).
      Secondary outcomes included objective observer evaluations reported by a critical care physician with specialized training in human factors engineering [G.D.] and a general internal medicine fellow with specialized training in defibrillator use [M.R.]. Participants were evaluated on their ability to complete six core functions of defibrillator use: ability to turn on the defibrillator, ability to attach the defibrillator electrode pads, ability to use the defibrillator paddles, ability to select an appropriate display, ability to deliver a defibrillation, and ability to deliver a synchronized cardioversion. Participants were deemed to have “difficulty” completing a specific function if a delay of greater than 1 minute was encountered in completing the task. The ability to attach the hands-free defibrillator electrode pads was further subdivided into four subfunctions for which participants were evaluated: ability to identify the electrode pads, ability to identify the adaptor cable, ability to toggle the cable locking mechanism, and ability to attach the electrode pads to the mannequin.
      Perceived usability of the manual-mode defibrillator was assessed via a post-exposure questionnaire. Participants were asked to rate the ease or difficulty of completing task 1 (defibrillation) and task 2 (synchronized cardioversion) on a scale of 1–9 (1 = “very difficult”; 9 = “very easy”). Participants further rated their confidence in effectively using this defibrillator to complete defibrillation or synchronized cardioversion in a real-world environment. Finally, participants graded the overall usability of the defibrillator.
      Qualitative data were collected via direct observation, comments provided by participants during the simulation, and from open-ended questions included in the post-exposure questionnaire.

      Statistics

      Mean values, along with standard deviation of the mean, were calculated for the time to defibrillation, objective observer evaluations, and perceived usability scores. Qualitative information was transcribed and thematically coded by the study investigators.

      Results

      Demographics

      Of the 22 participants, most were senior trainees (7/22 PGY-2 and 11/22 PGY-3). Prior to completing our study, participants had participated in an average 13.1 ± 7.4 real-world cardiac arrest resuscitations and acted as cardiac arrest team leader on 4.7 ± 2.8 occasions. All participants had completed ACLS certification within the preceding 12 months. ACLS training at our institution integrates use of the Philips HeartStart XL manual-mode defibrillator in a simulated environment. All participants had previously discharged the Philips HeartStart XL manual-mode defibrillator in either a real-world or simulated environment. Unfortunately, we were unable to determine the number of times participants had previously discharged a defibrillator, as this information is not systematically recorded at our institution.

      Time to Defibrillation

      The average time to defibrillation was 4 minutes 21 seconds ± 138 seconds. This includes two participants who were unable to deliver a successful defibrillation within 10 minutes. When 10 minutes had elapsed, guidance was provided by the investigators to facilitate completion of the task. Excluding these participants, the average time to defibrillation was 3 minutes 52 seconds ± 98 seconds. The average time to defibrillation for senior trainees was similar to that of the group as a whole at 3 minutes 56 seconds ± 138 seconds. Ultimately, only 9.1% of participants (2/22) could perform a simulated defibrillation within 2 minutes as recommended by the American Heart Association.
      The average time to synchronized cardioversion was 1 minute 55 seconds ± 59 seconds. The second synchronized cardioversion was delivered on average 18 seconds ± 6 seconds after the first.

      Objective Observer Evaluations

      Delay in defibrillation resulting from the inability to perform one or more core functions of defibrillator use in a timely manner was common (Table 1). Participants had particular difficulty with specific functions such as selecting an appropriate display, using the handheld defibrillator paddles, and attaching the hands-free defibrillator electrode pads. All nine participants who attempted to use the handheld defibrillator paddles had difficulty. No participants had difficulty with other cognitive functions such as rhythm identification or selection of an appropriate voltage for defibrillation. No participants unintentionally delivered a defibrillation when attempting to deliver a synchronized cardioversion.
      Table 1Objective Observer Evaluations of the Core Functions
      Core FunctionProportion of Participants with Difficulty
      Difficulty completing a specific function was defined as a delay of greater than 1 minute.
      Completing Task (%)
      Ability to turn on the defibrillator27.3
      Ability to attach the defibrillator electrode pads77.3
      Ability to use the defibrillator paddles100.0
      Ability to select an appropriate display54.5
      Ability to deliver a defibrillation4.5
      Ability to deliver a synchronized cardioversion13.6
      low asterisk Difficulty completing a specific function was defined as a delay of greater than 1 minute.
      Participants found the process of attaching the hands-free defibrillator electrode pads to be particularly troublesome. In fact, of those participants who were unable to deliver a defibrillation within 2 minutes, 85.0% had difficulty with this task. To evaluate this further we subdivided the task into four subfunctions (Table 2).
      Table 2Objective Observer Evaluations of Those Who Had Difficulty Attaching the Hands-Free Defibrillator Electrode Pads
      SubfunctionProportion of Participants with Difficulty
      Difficulty completing a specific subfunction was defined as a delay of greater than 1 minute.
      Completing Task (%)
      Ability to identify the electrode pads17.6
      Ability to identify the adaptor cable88.2
      Ability to toggle the cable locking mechanism82.4
      Ability to attach the electrode pads to the skin0.0
      low asterisk Difficulty completing a specific subfunction was defined as a delay of greater than 1 minute.

      Perceived Usability

      Perceived usability scores are summarized in Table 3. Participants were more comfortable performing a synchronized cardioversion than a defibrillation. Participants also felt more confident in their ability to effectively use this defibrillator in a real-world environment than the simulated environment used for this study.
      Table 3Post-Exposure Questionnaire
      Difficulty“Very Difficult” … “Very Easy”
      How would you rate the ease or difficulty in performing a defibrillation?3.9 ± 2.1
      How would you rate the ease or difficulty of performing a synchronized cardioversion?4.8 ± 2.6
      Confidence“Not Confident” … “Very Confident”
      How confident are you in your ability to effectively utilize this defibrillator to perform a defibrillation in a real-world environment?4.8 ± 2.2
      How confident are you in your ability to effectively utilize this defibrillator to perform a synchronized cardioversion in a real-world environment?5.3 ± 2.4
      Usability“Poorly Designed” … “Perfectly Designed”
      How would you rate the overall usability of this defibrillator?4.2 ± 1.8
      Qualitative participant feedback was analyzed and grouped according to theme. Many participants noted the user interface when commenting on elements of defibrillator design with a high degree of usability. Screen size, resolution, and brightness were cited on several occasions, with 22.7% of participants perceiving the physical display as having very strong usability, and 63.6% noting the intuitive sequencing and distinctive labeling of the buttons. The latter observation was summarized nicely by a participant who stated, “The synchronization, charge, and shock buttons are obvious and located beside each other in a logical order.”
      Although not specifically examined in this study, several participants also commented on the usability of the automated external defibrillator (AED) mode.
      Negative participant feedback focused almost exclusively on the process of attaching the hands-free defibrillator electrode pads to the defibrillator. All 22 participants commented on this function in the post-exposure questionnaire. Common themes included the requirement for an adaptor cable, inaccessible location of the cable locking mechanism, and difficulty toggling the locking mechanism for cable detachment. Descriptors of the process included “unintuitive,” “inconvenient,” and “awkward.” Several participants questioned why the defibrillator does not come attached to the hands-free defibrillation electrode pads as the default configuration. Others commented as to why the hands-free defibrillator electrode pads require an adaptor cable and are not manufactured as a single component.

      Discussion

      More than 90% of participants in our study were unable to perform a simulated defibrillation within 2 minutes. This is significant because recent literature indicates that patients with delayed defibrillation are less likely to survive to hospital discharge (22.2% when defibrillation was delayed vs. 39.3% when defibrillation was not delayed, in the National Registry of Cardiopulmonary Resuscitation
      • Chan PS
      • et al.
      Delayed time to defibrillation after in-hospital cardiac arrest.
      ). It has been further noted that the likelihood of survival to hospital discharge decreases incrementally with delayed defibrillation beyond 2 minutes, with research suggesting an approximate 7%–10% decline in survival for each additional minute of ventricular fibrillation.
      • Adams BD
      • Anderson PI
      • Stuffel E
      ‘‘Code Blue’’ in the hospital lobby: cardiac arrest teams vs. public access defibrillation.
      Even among those who survive, delayed defibrillation is associated with a lower likelihood of having no major disabilities in neurologic (adjusted odds ratio, 0.74) or functional status (adjusted odds ratio, 0.74).3
      The response to an adverse event or near miss involving a product or device is often to blame the user. This may be particularly true in health care where the principles of human factors engineering and usability have not been adopted universally.
      • Karsh BT
      • Scanlon M
      When is a defibrillator not a defibrillator? When it's like a clock radio . . . the challenge of usability and patient safety in the real world.
      • Rousek JB
      • Hallbeck MS
      The ergonomics of “Code Blue” medical emergencies: a literature review.
      • Gosbee J
      Who left the defibrillator on?.
      Strategies such as root cause analysis, increased user training, and policy modification are often employed to reduce the likelihood of similar events recurring in the future. Nonetheless, even the most expert of users may continue to make mistakes when confronted with a device that is illogical or poorly designed.
      • Rousek JB
      • Hallbeck MS
      The ergonomics of “Code Blue” medical emergencies: a literature review.
      • Gosbee J
      Human factors engineering and patient safety.
      ,
      • Sawyer D.
      Do It by Design: An Introduction to Human Factors in Medical Devices.
      Participants in our study were predominantly senior trainees with a high degree of familiarity with the Philips HeartStart XL defibrillator and ACLS protocol. Despite this, the average time to defibrillation was more than twice the time recommended by the American Heart Association. Based on these results, it appears likely that delay in defibrillation at our institution is at least partially the result of poor defibrillator design and lack of usability. Expert observer evaluation and qualitative participant feedback were largely congruent in regard to which aspects of defibrillator design do not suit the end user. Delay in defibrillation was largely the result of an inability to attach the hands-free defibrillator electrode pads to the defibrillator and select an appropriate display. Design factors influencing the usability of these components is discussed in detail below.
      More than three quarters (77.3%) of participants had trouble connecting the hands-free defibrillator electrode pads to the defibrillator. This represents the single most common cause for delayed defibrillation encountered in our study. When not in use, the Philips HeartStart XL manual-mode defibrillator is often attached to an external battery pack. Modifying device configuration to allow for attachment of the hands-free defibrillator electrode pads is a multistep process involving three separate components: the hands-free defibrillator electrode pads, an adaptor cable, and the defibrillator (Figure 3). The requirement for an adaptor cable is not immediately intuitive nor clearly labeled on the defibrillator or hands-free defibrillator electrode pads. Furthermore, the packaging for hands-free defibrillator electrode pads used at our institution is labeled “Defibrillation Electrodes,” while the packaging for a separate component designed to prevent electrical burns when using the handheld defibrillator paddles is labeled “Defib-Pads” (Figure 4). This somewhat ambiguous packaging is confusing for users and is likely the result of sourcing components from multiple suppliers.
      Fig 3
      Figure 3Hands-free defibrillator electrode pads are shown in panel A, while panel B shows the simulation hands-free defibrillator electrode pads connected to an adaptor cable, which is subsequently attached to the defibrillator. The requirement for an adaptor cable is not intuitive nor clearly labeled on the defibrillator or hands-free defibrillator electrode pads.
      Fig 4
      Figure 4The hands-free defibrillator electrode pads in panel A are labeled as “Multi-Function Defibrillation Electrodes,” and the handheld defibrillator pads in panel B are labeled as “Defib-Pads.” This somewhat ambiguous packaging is confusing for users and is likely the result of sourcing components from multiple different suppliers. Panel 4B reproduced with permission. Source: 3M. 3M™ Defibrillator Pads, 2346N, 4-1/2 in × 6 in (11.4 cm × 11.4 cm). 2018. Accessed Jun 22, 2018. https://www.3mcanada.ca/3M/en_CA/company-ca/all-3m-products/~/3M-Defibrillator-Pads-2346N-4-1-2-in-x-6-in-11-4-cm-x-11-4-cm-/?N=5002385+8711096+3293805730+3294529206&rt=rud.
      After the adaptor cable is connected to the hands-free defibrillator electrode pads, it must be attached to the defibrillator. This connection point includes a cable locking mechanism to prevent inadvertent detachment during use (Figure 5). Although relatively simple to operate, this mechanism is located immediately adjacent to the surface on which the defibrillator sits and is thus somewhat difficult to access. Sixty-four percent of participants had difficulty connecting the hands-free defibrillator electrode pads to the defibrillator. In several instances, we observed participants manually lift the defibrillator to change its orientation, allowing better access to the cable locking mechanism. This was deemed a safety hazard, as the defibrillator almost fell off the cart on several occasions because of these actions.
      Fig 5
      Figure 5The cable locking mechanism (yellow circle) for hands-free defibrillator electrode pad attachment is relatively simple to operate but is located immediately adjacent to the surface on which the defibrillator sits and is thus somewhat difficult to access.
      Poor usability was not limited to hands-free defibrillator electrode pads. All nine participants who attempted to use the handheld defibrillator paddles had difficulty. This was largely the result of an inability to remove the paddles from their casing on the side of the defibrillator. Plastic notches designed to prevent inadvertent removal seem to make deliberate removal more difficult, as they necessitate an equal force to be direct upward and outward simultaneously, similar to the design of child-resistant packaging (Figure 6). This seems to be particularly difficult to operate during periods of high stress such as cardiac arrest resuscitation.
      Fig 6
      Figure 6Plastic notches (yellow circles) designed to prevent inadvertent removal of the handheld defibrillator paddles seem to make deliberate removal more difficult, as they necessitate an equal force to be directed upward and outward simultaneously, similar to the design of child-resistant packaging.
      The Philips HeartStart XL defibrillator has the capability to display cardiac rhythms from multiple inputs, including electrocardiogram (ECG) leads, hands-free defibrillator electrode pads, and handheld defibrillator paddles. The default display is ECG lead II. When using either the hands-free defibrillator electrode pads or handheld paddles, the input must be changed manually to that of the desired component. The majority (54.5%) of participants had difficulty displaying a cardiac rhythm because of an improperly inputted display. Neither of the two participants who were unable to deliver an unassisted defibrillation within 10 minutes recognized the need to modify the default display from that of ECG lead II.
      The validity of our study is dependent on obtaining behavioral data from actual users of the device in the environment in which it is intended to be used. Environmental factors such as ambient noise, lighting, and temperature affect device usability. Similarly, intrinsic characteristics of the user such as confidence, stress, or fatigue will influence performance.
      • Gosbee J
      Human factors engineering and patient safety.
      • Sawyer D.
      Do It by Design: An Introduction to Human Factors in Medical Devices.
      Medical Device and Diagnostic Industry
      Eleven Keys to Designing Error-Resistant Medical Devices.
      We have attempted to account for many of these variables. At our institution, a resident physician is usually tasked with attaching and using the defibrillator (there are four resident physicians per cardiac arrest team), thus allowing other team members such as respiratory therapists or nursing to focus on other tasks typical of cardiac arrest care. We therefore elected to evaluate resident physicians’ use of this device, while excluding other physicians who are not routinely involved in cardiac arrest care. We replicated many of the environmental factors typical of in-hospital cardiac arrest through the use of a high-fidelity simulation room. We did not simulate other tasks typical of in-hospital cardiac arrest care (for example, chest compressions, airway management, IV access, administration of medications), allowing participants to focus solely on defibrillator use free of potential distraction.
      There are few published studies with which to compare ours. Fairbanks et al. evaluated the use of two manual-mode defibrillators (the Lifepak 10 and Lifepak 12) by paramedics in a simulated environment.
      • Fairbanks RJ
      • et al.
      Usability study of two common defibrillators reveals hazards.
      For both devices, the difficulty of performing a defibrillation was rated 6.9 on a scale of 1–9 (1 = “very difficult”; 9 = “very easy”).
      • Fairbanks RJ
      • et al.
      Usability study of two common defibrillators reveals hazards.
      This contrasts with our results in which difficulty of performing a defibrillation with the Philips HeartStart XL was rated 3.9. Difficulty of performing a synchronized cardioversion was rated 5.3 for the Lifepak 10 and 6.7 for the Lifepak 12,
      • Fairbanks RJ
      • et al.
      Usability study of two common defibrillators reveals hazards.
      while our study showed a rating of 4.8 for the Phillips HeartStart XL. Interestingly, half of all participants in the Fairbanks et al. study (7/14) performed at least one unsynchronized defibrillation when attempting to perform a synchronized cardioversion using the Lifepak 10 or Lifepak 12,
      • Fairbanks RJ
      • et al.
      Usability study of two common defibrillators reveals hazards.
      while none of the 22 participants in our study did so using the Philips HeartStart XL. We hypothesize that these results may reflect intuitive design of the Philips HeartStart XL or prior education and training of our participants. Experiential learning is an unlikely cause for this observation, as participants had no interaction with the synchronization function during the initial defibrillation simulation.
      Hunt et al. evaluated the performance of pediatric residents during simulated cardiopulmonary resuscitation, which included defibrillation using the ZOLL M Series semiautomatic defibrillator.
      • Hunt EA
      • et al.
      Delays and errors in cardiopulmonary resuscitation and defibrillation by pediatric residents during simulated cardiopulmonary arrests.
      The median time from recognition of cardiac arrest to defibrillation was 1 minute and 50 seconds. Approximately 83% of participants (58/70) made at least one error while operating the defibrillator. The only variable independently associated with time to defibrillation was having previously discharged a defibrillator in either a real-world or simulated setting.
      • Hunt EA
      • et al.
      Delays and errors in cardiopulmonary resuscitation and defibrillation by pediatric residents during simulated cardiopulmonary arrests.
      All participants in our study had previously discharged the Phillips HeartStart XL defibrillator.
      Fidler and Johnson evaluated the use of three manual-mode defibrillators (ZOLL R Series Plus Monitor, Lifepack 15, and Phillips HeartStart MRx) in a simulated environment.
      • Fidler R
      • Johnson M
      Human factors approach to comparative usability of hospital manual defibrillators.
      The average time from recognition of cardiac arrest to defibrillation was 39.3 seconds for the Zoll R Series Plus, 33.5 seconds for the Lifepack 15, and 19.0 seconds for the Phillips HeartStart MRx. Participants were not required to attach the defibrillator to the mannequin.
      • Fidler R
      • Johnson M
      Human factors approach to comparative usability of hospital manual defibrillators.
      This differs from our study in which delay in defibrillation secondary to difficulty attaching the hands-free defibrillator electrode pads was common.
      It remains unclear as to why so many of our participants were unable to compete a simulated defibrillation within 2 minutes. The National Registry of Cardiopulmonary Resuscitation revealed that delayed defibrillation occurred in 30.1% of cases. The median time to defibrillation was approximately 1 minute.
      • Chan PS
      • et al.
      Delayed time to defibrillation after in-hospital cardiac arrest.
      This contrasts with our institution in which the average time to defibrillation is 8 minutes in a critical care setting and 11 minutes on a medical ward.
      • Godbout J
      • et al.
      Innovative use of AED by RNs and RTs during in-hospital cardiac arrest: phase II.
      In our study, average time to defibrillation was more than 4 minutes, while the average time to synchronized cardioversion was nearly 2 minutes.
      Indisputably, there is a role for improved training and education. The time to synchronized cardioversion was significantly less than that of defibrillation—likely the result of experiential learning and recency (by protocol, defibrillation occurred first and was followed by synchronized cardioversion). This observation may also explain why participants were more confident performing a synchronized cardioversion than a defibrillation and performing defibrillation in the real-world as opposed to our simulated environment. It remains uncertain whether users can be trained frequently enough to maintain such proficiency. Our results also call into question the accuracy of timekeeping during real-world cardiopulmonary resuscitation.
      • Kaye W
      • Mancini ME
      • Truitt TL
      When minutes count—the fallacy of accurate time documentation during in-hospital resuscitation.
      • Cordell WH
      • et al.
      Does anybody really know what time it is? Does anybody really care?.

      Limitations

      Our study has several important limitations. Although novel in terms of participant and device selection, this is a single-center trial with a relatively modest number of participants. Further work will be required to confirm the reproducibility of our results. For example, the generalizability of our results to other defibrillator models remains uncertain, particularly given the discrepancy in time to defibrillation at our institution relative to that of the National Registry of Cardiopulmonary Resuscitation. Despite this, we suspect unintuitive design and poor usability are not limited to this model of defibrillator, as demonstrated previously by Fairbanks et al., Hunt et al., and Fidler and Johnson.
      • Fairbanks RJ
      • et al.
      Usability study of two common defibrillators reveals hazards.
      • Hunt EA
      • et al.
      Delays and errors in cardiopulmonary resuscitation and defibrillation by pediatric residents during simulated cardiopulmonary arrests.
      • Fidler R
      • Johnson M
      Human factors approach to comparative usability of hospital manual defibrillators.
      Second, participants were recruited on a voluntary basis, raising the possibility of selection bias. Logically one would expect such a bias to reduce the average time to defibrillation, as participants more familiar with defibrillator use would be thought more likely to participate. This is contrary to our results. Participants were also relatively homogenous in terms of background and training. Further research is required to assess the perceived usability of this manual-mode defibrillator from the perspective of other allied health professionals.
      Finally, study observers were not blinded to the study objectives, raising another possibility of bias. We attempted to account for this by defining “difficulty” in completing a specific task by an objective measure (time greater than 1 minute). This measure has not been externally validated.

      Conclusion

      Most participants in our study were unable to perform a simulated defibrillation within 2 minutes. This delay in defibrillation is likely at least partially the result of poor defibrillator design and lack of usability. Expert observation and qualitative participant feedback were largely congruent in terms of which aspects of defibrillator design do not suit the end user. Modification of future defibrillator design, along with improved education and training, may result in more timely defibrillation.
      The intent of our study was not to critique one specific model of defibrillator, population of user, or health care institution. Rather, we set out to reinforce the importance of intuitive design and usability in the provision of health care. Defibrillators are used infrequently relative to many other products and devices encountered in the hospital setting, but they are of the utmost importance given their proven efficacy in reducing morbidity and mortality during cardiac arrest. This emphasizes the importance of intuitive design. Gathering behavioral data regarding user capabilities, limitations, and preferences may help device manufacturers improve the design and usability of future defibrillator models. In the interim, knowledge of defibrillator usability and associated limitations should help inform training strategies and institutional policy. We recommend that health care institutions involve experts with training in human factors engineering along with real-world users of the device when procuring the next generation of defibrillators.

      Funding

      All funding was provided by a University of Ottawa Department of Medicine Patient Quality and Safety research grant and by The Ottawa Hospital Division of Critical Care.

      Acknowledgments

      The authors would like to thank the University of Ottawa Department of Medicine (Patient Quality and Safety research grant) and The Ottawa Hospital Division of Critical Care for providing funding for this project. The authors also thank the staff of the University of Ottawa Skills and Simulation Centre for their expertise and assistance.

      Conflicts of Interest

      All authors report no conflicts of interest.

      Appendix. Supplementary materials

      References

        • Eisenberg MS
        • Mengert TJ
        Cardiac resuscitation.
        N Engl J Med. 2001 Apr 26; 344: 1304-1313
        • Girotra S
        • et al.
        Trends in survival after in-hospital cardiac arrest.
        N Engl J Med. 2012 Nov 15; 367: 1912-1920
        • Chan PS
        • et al.
        Delayed time to defibrillation after in-hospital cardiac arrest.
        N Engl J Med. 2008 Jan 3; 358: 9-17
        • Neumar RW
        • et al.
        Part 1: executive summary: 2015 American Heart Association guidelines update for cardiopulmonary resuscitation and emergency cardiovascular care.
        Circulation. 2015 Nov 3; 132: S325-S367
        • Herlitz J
        • et al.
        Very high survival among patients defibrillated at an early stage after in-hospital ventricular fibrillation on wards with and without monitoring facilities.
        Resuscitation. 2005; 66: 159-166
        • Spearpoint KG
        • McLean CP
        • Zideman DA
        Early defibrillation and the chain of survival in ‘in-hospital’ adult cardiac arrest; minutes count.
        Resuscitation. 2000; 44: 165-169
        • Peberdy MA
        • et al.
        Cardiopulmonary resuscitation of adults in the hospital: a report of 14720 cardiac arrests from the National Registry of Cardiopulmonary Resuscitation.
        Resuscitation. 2003; 58: 297-308
        • Godbout J
        • et al.
        Innovative use of AED by RNs and RTs during in-hospital cardiac arrest: phase II.
        CJEM. 2015; 17: S5
        • Chan PS
        • et al.
        Hospital variation in time to defibrillation after in-hospital cardiac arrest.
        Arch Intern Med. 2009 Jul 27; 169: 1265-1273
        • Peberdy MA
        • et al.
        Survival from in-hospital cardiac arrest during nights and weekends.
        JAMA. 2008 Feb 20; 299: 785-792
        • Wichansky AM
        Usability testing in 2000 and beyond.
        Ergonomics. 2000; 43: 998-1006
        • Karsh BT
        • Scanlon M
        When is a defibrillator not a defibrillator? When it's like a clock radio . . . the challenge of usability and patient safety in the real world.
        Ann Emerg Med. 2007; 50: 433-435
        • Rousek JB
        • Hallbeck MS
        The ergonomics of “Code Blue” medical emergencies: a literature review.
        IIE Trans Healthc Syst Eng. 2011; 1: 197-212
        • Gosbee J
        Who left the defibrillator on?.
        Jt Comm J Qual Saf. 2004; 30: 282-285
        • University of Ottawa Skills and Simulation Centre
        Facts and Figures.
        (Accessed Jun 7)
        • Laerdal Medical
        ALS Simulator (Discontinued).
        (Accessed Jun 7)
        • Adams BD
        • Anderson PI
        • Stuffel E
        ‘‘Code Blue’’ in the hospital lobby: cardiac arrest teams vs. public access defibrillation.
        Int J Cardiol. 2006 Jun 28; 110: 401-402
        • Gosbee J
        Human factors engineering and patient safety.
        Qual Saf Health Care. 2002; 11: 352-354
        • Sawyer D.
        Do It by Design: An Introduction to Human Factors in Medical Devices.
        US Department of Health and Human Services, Public Health Service, Food and Drug Administration, Center for Devices and Radiological Health, Washington, DC1996
        • Medical Device and Diagnostic Industry
        Eleven Keys to Designing Error-Resistant Medical Devices.
        Wiklund ME, 2002 (May 1 Accessed Jun 7, 2018)
        • Fairbanks RJ
        • et al.
        Usability study of two common defibrillators reveals hazards.
        Ann Emerg Med. 2007; 50: 424-432
        • Hunt EA
        • et al.
        Delays and errors in cardiopulmonary resuscitation and defibrillation by pediatric residents during simulated cardiopulmonary arrests.
        Resuscitation. 2009; 80: 819-825
        • Fidler R
        • Johnson M
        Human factors approach to comparative usability of hospital manual defibrillators.
        Resuscitation. 2016; 101: 71-76
        • Kaye W
        • Mancini ME
        • Truitt TL
        When minutes count—the fallacy of accurate time documentation during in-hospital resuscitation.
        Resuscitation. 2005; 65: 285-290
        • Cordell WH
        • et al.
        Does anybody really know what time it is? Does anybody really care?.
        Ann Emerg Med. 1994; 23: 1032-1036