Advertisement

A Simulation Systems Testing Program Using HFMEA Methodology Can Effectively Identify and Mitigate Latent Safety Threats for a New On-Site Helipad

Published:September 27, 2021DOI:https://doi.org/10.1016/j.jcjq.2021.09.013

      Background

      Fundamental changes in critical systems within hospitals present safety risks. Some threats can be identified prospectively, others are only uncovered when the system goes live. Simulation and Healthcare Failure Mode and Effect Analysis (HFMEA) can be used together to prospectively test a system without endangering patients. The research team combined iterative simulations and HFMEA methodologies to conduct simulation-based clinical systems testing (SbCST) to detect and mitigate latent safety threats (LSTs) prior to opening a hospital helipad.

      Methods

      This study was conducted in three phases. In Phase I, an interprofessional team created a process map and conducted a tabletop exercise, identifying LSTs that could theoretically occur during patient transfer from the new helipad. Using HFMEA methodology, steps predicted to be affected by the new helipad were probed. Identified LSTs were assigned a hazard score. Mitigation solutions were proposed. Results from Phase I were used to plan Phase II, which used low-fidelity simulation to test communication processes and travel paths. High-fidelity simulation was used in Phase III to test previously identified LSTs.

      Results

      Over three testing phases, 31 LSTs were identified: 15 in Phase I, 7 in Phase II, and 9 in Phase III. LSTs fell under the categories of care coordination, facilities, and equipment, and devices. Eighteen (58.1%) were designated “critical” (hazard score ≥ 8).

      Conclusion

      A three-phase SbCST program using HFMEA methodology was an effective tool to identify LSTs. An iterative approach, using results of each phase to inform the structure of the next, facilitated testing of proposed mitigation strategies.
      Fundamental changes in critical hospital systems, such as the creation of new spaces or a change in patient transport protocols, present safety risks. Although some of these threats can be identified prospectively, others are latent and discovered only when the system goes live. Simulation-based clinical systems testing (SbCST) provides the opportunity to test a system without endangering patients. Simulation has long been used in other industries to uncover barriers to efficiency and safety threats prior to operationalization.
      • Henriksen K
      • Moss F.
      From the runway to the airway and beyond.
      A growing body of health care literature describes the use of simulation programs to test and improve new hospitals or units within hospitals prior to opening.
      • Adler MD
      • et al.
      Use of simulation to test systems and prepare staff for a new hospital transition.
      • Kerner Jr., RL
      • et al.
      Simulation for operational readiness in a new freestanding emergency department: strategy and tactics.
      • Gardner AK
      • et al.
      In situ simulation to assess workplace attitudes and effectiveness in a new facility.
      • Ventre KM
      • et al.
      Using in situ simulation to evaluate operational readiness of a children's hospital-based obstetrics unit.
      • Bender GJ.
      In situ simulation for systems testing in newly constructed perinatal facilities.
      • Geis GL
      • et al.
      Simulation to assess the safety of new healthcare teams and new facilities.
      Although resources exist to consider safety in the planning and construction of helipads, none describe the use of simulation to test patient flow from newly constructed helipads to clinical destination units prior to use for patients.
      • Militello PR
      • Ramzy AI.
      Safety by design. Shock trauma center's helipad received special consideration during its planning and construction.
      ,

      Federal Aviation Administration. Chapter 4: Hospital Heliports. Sep 30, 2004. Accessed Oct 8, 2021. https://www.faa.gov/documentLibrary/media/advisory_circular/150-5390-2B/150_5390_2b_part3.pdf.

      Healthcare Failure Mode and Effect Analysis (HFMEA) is a framework designed to prospectively identify safety threats, termed “failure modes.”
      • DeRosier J
      • et al.
      Using health care Failure Mode and Effect Analysis: the VA National Center for Patient Safety's prospective risk analysis system.
      Failure modes are the ways that a process can fail. Effects are the consequences of each failure mode. HFMEA assigns a hazard score, the product of two indices: severity and probability (see Figure 1). HFMEA enables teams to prospectively identify, categorize, prioritize, and mitigate failure modes before harm occurs. The weakness of the HFMEA methodology is that assessment of potential risks and their underlying causes is based solely on domain experts’ memories and knowledge. Simulation can address this weakness, allowing realistic enactment and reflection by the entire patient care team, an improvement when compared to theoretical exploration of processes (for example, “brainstorming”) alone.
      • Davis S
      • et al.
      Failure modes and effects analysis based on in situ simulations: a methodology to improve understanding of risks and failures.
      ,
      • Nielsen DS
      • et al.
      Augmenting health care failure modes and effects analysis with simulation.
      Figure 1
      Figure 1Shown here is an analysis key for assigning probability, severity rating, and hazard score to each failure mode. Healthcare Failure Mode and Effect Analysis (HFMEA) assigns a hazard score to each failure mode discovered during testing. Each hazard score is the product of two indices: severity of impact to the patient or clinical team and probability of that failure mode occurring. Categorizing failure modes based on hazard scores enables a team to prioritize resources. Any failure mode with a score of eight (8) or higher required prioritization for remediation and retesting.
      Figure 1
      Figure 1Shown here is an analysis key for assigning probability, severity rating, and hazard score to each failure mode. Healthcare Failure Mode and Effect Analysis (HFMEA) assigns a hazard score to each failure mode discovered during testing. Each hazard score is the product of two indices: severity of impact to the patient or clinical team and probability of that failure mode occurring. Categorizing failure modes based on hazard scores enables a team to prioritize resources. Any failure mode with a score of eight (8) or higher required prioritization for remediation and retesting.
      In January 2020 Maine Medical Center opened a new double helipad on top of a newly constructed patient care tower. This new helipad replaced an older, single helipad positioned atop an adjacent parking garage. In the preoccupancy, post-construction phase, our simulation center was tasked with conducting SbCST to identify and mitigate latent safety threats (LSTs) that could occur during patient arrival and transport to destinations within the hospital (emergency department [ED], ICU, pediatric ICU [PICU], and cardiac catheterization laboratory). We hypothesized that by using iterative phases of simulation-supported HFMEA, we could identify and mitigate failure modes identified in each round of testing.

      Methods

      Participants and Setting

      This prospective investigation occurred over five months at the Maine Medical Center (MMC), a 637-bed academic, Level 1 trauma center located in Portland, Maine. MMC accepts approximately 500 helicopter transports per year, most often interfacility transfers from affiliated hospitals via LifeFlight of Maine (LOM), and less frequently arrivals from the US Coast Guard or other entities. From 2018 to 2019, MMC constructed a new dual helipad. The new helipad was built to accommodate simultaneous helicopter landings, a need historically noted to arise several times per year.

      Prebriefing and Debriefing

      In December 2018 the Simulation Center was tasked with SbCST of the helipad, prior to opening in January 2020. Our first task was to recruit an interprofessional team of investigators comprised of Simulation Center staff (two directors, a curriculum specialist, and two simulation specialists), an emergency medicine physician, a patient safety specialist, and a performance improvement specialist. This team reviewed existing literature and designed a plan for iterative rounds of SbCST for patient transport from our new helipad to common hospital destinations. We planned that each phase of testing would inform the time frame, content, and simulation fidelity for the next. Testing would be deemed complete when no new significant failure modes were uncovered and/or mitigation strategies could be implemented successfully without the need for additional testing. Simulation debriefings would be facilitated using HFMEA methodology by our patient safety specialist (Figure 2).
      Figure 2
      Figure 2This figure provides an overview of steps in a Healthcare Failure Mode and Effect Analysis (HFMEA).

      Healthcare Failure Mode and Effect Analysis

      Assemble the Team

      Prior to initiating testing, our study team obtained support from institutional executives and operational leaders to help ensure project success. This included direct meetings and involvement in the planning process. These leaders included representatives from LOM, hospital executives responsible for the expansion project, Interventional Cardiology, Emergency Medicine, the Simulation Center, adult and pediatric ICUs, regional communications center, construction team, and the Performance Improvement department. Leaders recruited frontline staff directly responsible for patient care and transportation: MDs, RNs, registration staff, flight nurses and pilots, communications center representatives, and security personnel. This engagement provided the opportunity to understand stakeholder concerns about the new helipad and identify the specific process to be studied.

      Graphically Describe the Process

      We created a process flow diagram to graphically describe the process of interest, beginning with the “patient identified for helicopter transfer and ending with arrival on the destination unit where the “patient receives care.” The investigator team identified steps predicted to be most affected by the new helipad due to its distinct geographic location as compared to the prior helipad (highlighted in Figure 3).
      Figure 3
      Figure 3Shown is a flow diagram of our process map, from identification of a patient in need of transfer to arrival at the destination care unit. Steps that were significantly affected by the new helipad were tested and are highlighted in yellow/starred. Steps that were not expected to change are denoted in gray. OSH, outside hospital; MMC, Maine Medical Center; ETA, estimated time of arrival.

      Conduct a Hazard Analysis

      Phase I was designed as a tabletop exercise of approximately 120 minutes duration and held in August 2019. The patient safety facilitator led the group through an in-depth analysis of each step affected by the new helipad location, eliciting input from participants on potential failure modes during each step. Failure modes were catalogued on whiteboards. Group consensus was used to assign each failure mode a category, probability, and severity scores, which were then used to calculate a hazard score for each failure mode identified (Figure 1). Any threat with a score of eight (8) or higher required prioritization for remediation and retesting. Mitigation solutions were proposed and discussed by the group. Action items were identified, and responsible parties were given one month to complete them before Phase II testing.
      Based on the findings from Phase I, we identified key areas of testing for Phase II to include communication processes to receiving units, registration processes, and optimal paths of travel to four common destinations (ED, adult special (intensive) care unit [SCU], PICU, and cardiac catheterization lab). We determined that low-fidelity simulation (an adult unpowered manikin on a stretcher with a cardiac monitor/defibrillator and intravenous [IV] pole) was sufficient to accurately mimic the physical space required by the transport unit to test mitigation strategies from failure modes uncovered in Phase I (Figure 4). Phase II was conducted over a four-hour period in September 2019. Experienced simulation instructors co-facilitated a prebrief planning session identifying communications and travel paths to the four key destinations of interest. Each of these transport paths was tested using low-fidelity simulation, including time-keeping and tracking of patients’ exposure to additional hallway traffic during transport (to document privacy lapses). Each simulated transport was then co-debriefed by a patient safety specialist and a simulation instructor using a Plus-Delta model.
      • Schertzer K
      • In Patti L.
      Situ Debriefing in Medical Simulation.
      The group identified failure modes, agreed on severity and probability ratings to calculate a hazard score, and proposed mitigation solutions (Figure 1). Because of the urgency of patients being transferred to the catheterization lab, two reasonable paths of travel (identified during Phase I) were tested for comparison. Action items were identified, and responsible parties were given one month to complete them before Phase III testing began.
      Figure 4
      Figure 4Phase II used low-fidelity simulation with an unpowered mannequin, a stretcher, a cardiac monitor/defibrillator, and an IV pole.
      After the team's review of LSTs from Phases I and II, three objectives for Phase III testing were identified: (1) Refine processes to support near-simultaneous landing of two helicopters; (2) Validate proposed solutions to address communication gaps during PICU transfer; and (3) Test the transfer process for a patient who decompensates after helicopter landing but prior to arrival at the destination care unit. To best address these objectives, we determined that high-fidelity simulation testing would optimally support the exercise. This included the use of powered adult and pediatric mannequins on stretchers with an IV pole and cardiac monitors displaying vital signs. The actual landing of two helicopters on the new helipad with full LOM flight crews was also employed to realistically simulate simultaneous landing and the team's response to a cardiac arrest during transport. On December 11, 2019, high-fidelity manikins were prepared and loaded into two helicopters at our local airport. The helicopters then flew to the hospital using standard communication practices and landed 60 seconds apart. Simulated patients were sequentially unloaded and simultaneously transported to their respective accepting units (Figure 5). For the adult simulated patient, a basic simulation scenario involving cardiac arrest requiring cardiopulmonary resuscitation (CPR) post-landing and during transport was scripted. A pediatric manikin was transported to the PICU without a change in clinical status. The dual landing simulation was debriefed as single group, with the discussion co-facilitated by a patient safety specialist and a simulation instructor using a Plus-Delta model. The group again identified failure modes detected, agreed on severity and probability ratings to calculate a hazard score, and proposed mitigation solutions.
      Figure 5
      Figure 5Maine Medical Center security and LifeFlight of Maine crew transport an adult cardiac patient during phase 3 testing.

      Results

      We conducted three phases of SbCST over a five-month period, which included 49 total participants. Stakeholders are listed for each Phase in Table 1. Over the three phases, 31 potential failure modes were identified: 15 in Phase I, 7 in Phase II, and 9 in Phase III (Table 2). Failure modes were identified in three major domains: care coordination (18), facilities (10), and equipment and devices (3). Of the failure modes identified, 32.3% were classified as both high severity and high probability. Eighteen (58.1%) of the 31 failure modes were designated “critical” (hazard score ≥ 8). Table 3 provides a summary of key failure modes discovered.
      Table 1Stakeholder Groups That Participated in Helipad Testing by Phase
      DisciplinePhase IPhase IIPhase III
      Emergency MedicineCenter for Performance and ImprovementAdult Intensive CareHannaford Simulation CenterInterventional CardiologyLifeFlight of MainePatient Access and RegistrationPediatric Intensive CarePatient Safety/Risk ManagementSecurityConstructionRegional Communications CenterXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
      Total Participants1816
      Eight individuals had also participated in Phase I.
      15
      Eight individuals had previously participated in Phase I or II.
      low asterisk Eight individuals had also participated in Phase I.
      Eight individuals had previously participated in Phase I or II.
      Table 2Number of Latent Safety Threats Identified per Phase and Stratified into Categories of Care Coordination, Equipment and Devices, and Facilities
      Care CoordinationEquipment and DevicesFacilities
      Phase I—Low Fidelity735
      Phase II—Medium Fidelity403
      Phase III—High Fidelity702
      Table 3Summary of Latent Safety Threats Identified in Phases I, II, and III
      Phase I: Tabletop Exercise; Phase II: Low-Fidelity Simulation; Phase III: High-Fidelity Simulation.
      Steps in theProcessFailureModeFailureCauseFailureEffectsSeverity(S)Probability(P)Hazard Score(S x P)ActionFollow-Up
      1. Communications Center pages the appropriate care team.Center staff are saturated with large amounts of pages and phone calls.Communications Center is understaffed.Team members are not informed of incoming patient—delayed care.428Continue to monitor frequency.No action other than to monitor.
      Safety threats posed by communication failures within the paging system, highlighted by our testing, resulted in a separate Healthcare Failure Mode and Effect Analysis (HFMEA)—led by Patient Safety, not involving simulation—to address these issues. The focus of this HFMEA was to understand the current limitations of our paging system and consider the steps needed to improve communication software. Resulting solutions included expansion of the Wi-Fi in older sections of the hospital, planning for upgraded paging software, and assignment of responsibility and accountability for updating emergency paging lists. PICU, pediatric ICU; LOM, LifeFlight of Maine; ED, emergency department.
      Short or no pre-notification of incoming patientShort travel timeTeam members are not mobilized in time, causing delayed care.428Continue to monitor frequency.No action other than to monitor.
      Safety threats posed by communication failures within the paging system, highlighted by our testing, resulted in a separate Healthcare Failure Mode and Effect Analysis (HFMEA)—led by Patient Safety, not involving simulation—to address these issues. The focus of this HFMEA was to understand the current limitations of our paging system and consider the steps needed to improve communication software. Resulting solutions included expansion of the Wi-Fi in older sections of the hospital, planning for upgraded paging software, and assignment of responsibility and accountability for updating emergency paging lists. PICU, pediatric ICU; LOM, LifeFlight of Maine; ED, emergency department.
      PICU team not routinely notified of patient arrival.The communication is not always about pediatric patients so PICU team was left off.PICU team not prepared for patient, causing delayed care.4416Add PICU charge RN to notification list.Determined in Phase II that adding PICU to notification list would cause false alerts for adult patients. Solution is to have Patient Registration call the PICU team when the patient is a confirmed pediatric patient.
      Safety threats posed by communication failures within the paging system, highlighted by our testing, resulted in a separate Healthcare Failure Mode and Effect Analysis (HFMEA)—led by Patient Safety, not involving simulation—to address these issues. The focus of this HFMEA was to understand the current limitations of our paging system and consider the steps needed to improve communication software. Resulting solutions included expansion of the Wi-Fi in older sections of the hospital, planning for upgraded paging software, and assignment of responsibility and accountability for updating emergency paging lists. PICU, pediatric ICU; LOM, LifeFlight of Maine; ED, emergency department.
      Pagers are caught in queue.Older technology is not supported.Delay in team notification4416Review backup phone tree/secondary call plans for effectiveness.Phone tree is correct and maintained by the service.
      Safety threats posed by communication failures within the paging system, highlighted by our testing, resulted in a separate Healthcare Failure Mode and Effect Analysis (HFMEA)—led by Patient Safety, not involving simulation—to address these issues. The focus of this HFMEA was to understand the current limitations of our paging system and consider the steps needed to improve communication software. Resulting solutions included expansion of the Wi-Fi in older sections of the hospital, planning for upgraded paging software, and assignment of responsibility and accountability for updating emergency paging lists. PICU, pediatric ICU; LOM, LifeFlight of Maine; ED, emergency department.
      Wi-Fi isn't available.Older sections of the hospital have no Wi-Fi signal.Pages do not send in some areas—delayed care.428Review backup phone tree/secondary call plans for effectiveness.Phone tree is correct and maintained by the service.
      Safety threats posed by communication failures within the paging system, highlighted by our testing, resulted in a separate Healthcare Failure Mode and Effect Analysis (HFMEA)—led by Patient Safety, not involving simulation—to address these issues. The focus of this HFMEA was to understand the current limitations of our paging system and consider the steps needed to improve communication software. Resulting solutions included expansion of the Wi-Fi in older sections of the hospital, planning for upgraded paging software, and assignment of responsibility and accountability for updating emergency paging lists. PICU, pediatric ICU; LOM, LifeFlight of Maine; ED, emergency department.
      Vendor takes the system down without notice.Systemwide paging outageTeam members are not paged—delayed care.4312Review backup phone tree/secondary call plans for effectiveness.Phone tree is correct and maintained by the service.
      Safety threats posed by communication failures within the paging system, highlighted by our testing, resulted in a separate Healthcare Failure Mode and Effect Analysis (HFMEA)—led by Patient Safety, not involving simulation—to address these issues. The focus of this HFMEA was to understand the current limitations of our paging system and consider the steps needed to improve communication software. Resulting solutions included expansion of the Wi-Fi in older sections of the hospital, planning for upgraded paging software, and assignment of responsibility and accountability for updating emergency paging lists. PICU, pediatric ICU; LOM, LifeFlight of Maine; ED, emergency department.
      Pager batteries die.Not routinely checked.Some team members don't receive the page.428Assign team member to check batteries and replace when 50% full.Completed.
      2. LOM crew meets Security in helipad vestibule.Helicopters land simultaneously requiring more Security assistance.Increased capacity with two helipadsCrew and patient may have to wait in helipad vestibule for additional assistance to transport.428Confirm and test Transport Services as backup for Security.Phase III—confirmed and tested.
      Coast Guard doesn't accompany the patient along with Security.Coast Guard protocolRequires an ED provider to be present at the helipad.414Coast Guard is required to wait until an ED physician is present.No action other than to monitor.
      Difficulty maneuvering transport stretcher in helipad vestibuleTight corner in vestibuleDamage to facility144Install protective barrier on wall.Completed.
      3. Patient transported to appropriate area.Helipad elevator failureMechanical failureDelay in travel time414Review protocol for backup landing zone and patient transport.Completed.
      Hospital elevator(s) failureMechanical failureDelay in travel time414Plan backup path of travel via alternative patient towers and lateral paths of travel.Completed.
      Patient does not get registered.New path does not transport patient by Registration desk.No procedures/treatment rendered if patient not registered.4312Engage with Registration staff; change process for registration during transport.Phase II trial and revised in Phase III for ideal meeting location.
      Both Registration page numbers (triage and direct admit) included on pages for patients going directly to unit.Correct Registration page number not known to Communications Team.Wrong team member activated; confusion in process takes unneeded team member out of work unnecessarily.144Identify ideal Registration team member and only page them.Identified in Phase II; tested and confirmed in Phase III.
      Balloon Pump doesn't fit into elevator.Elevators in some patient towers are smaller than others.Delay in travel time326Plan route using adequate sized elevator.Phase III test confirmed the use of the larger elevator.
      Elevators overridden.Elevator inappropriately overridden for nonurgent patient or staff transport.Delay in travel time326Security to modify badge reader.No action other than to monitor.
      Path obstructed by supplies, equipment, hallway patients in ED hallway.ED overcrowdedDelay in travel time414Review backup paths of travel when needed.No action other than to monitor.
      Public hallways decrease patient privacy.Patient transported through public hallways.Patient loses privacy.248Optimized paths of travel for both speed and privacy.Tested in Phases II and III for optimal paths.
      Travel path to Cardiac Cath lab is longer from new helipad.The new helipad is at the other end of the hospital.Delay in patient transport/definitive care3412Test and time multiple paths of travel for shortest time.Tested in Phase II and Phase III—confirmed shortest and safest route.
      Patient decompensates pre-arrival.Patient condition changes—not always within our controlDelay in definitive care414Protocol developed and tested. Reroute to ED where trained staff and equipment are located.Tested in Phase III—appropriate plan.
      Patient decompensates during hospital transport.Patient condition changes- not always within our control.Delay in patient transport/definitive care414Protocol developed and tested. Reroute to ED where trained staff and equipment are located.Tested in Phase III—appropriate plan.
      Helipad can become slippery when wet/icy.Surface is nontextured.Crew members injure themselves or patient during transport.4312Repaint lines with textured paint, diligent salt application in winter months.Completed.
      Rear exit from aircraft difficult on helipad close to vestibule.Smaller second helipad and wind pattern affect rear exit.Delay in patient transport/definitive care326LOM crew will exit the patient on the other side of the aircraft or hover and wait until the other helipad is available.Discussed in Phase III—Med Flight crew will share findings.
      low asterisk Phase I: Tabletop Exercise; Phase II: Low-Fidelity Simulation; Phase III: High-Fidelity Simulation.
      Safety threats posed by communication failures within the paging system, highlighted by our testing, resulted in a separate Healthcare Failure Mode and Effect Analysis (HFMEA)—led by Patient Safety, not involving simulation—to address these issues. The focus of this HFMEA was to understand the current limitations of our paging system and consider the steps needed to improve communication software. Resulting solutions included expansion of the Wi-Fi in older sections of the hospital, planning for upgraded paging software, and assignment of responsibility and accountability for updating emergency paging lists. PICU, pediatric ICU; LOM, LifeFlight of Maine; ED, emergency department.
      In Phase I, the three highest hazard scores involved communication and path of travel. Failure modes with hazard scores of 16, the highest score possible, were “Pagers are caught in queue” and “PICU team not routinely notified of patient arrival.” Possible effects of pager malfunction included failure of team member notification and a delay in care coordination. Possible effects associated with the PICU not being notified of patient arrival included team members not prepared for patient arrival and delay of care.
      The second-highest-ranking failure modes—with a hazard score of 12—included “Patient does not get registered.” A possible effect of this failure mode was a delay in treatment. Phase I testing also identified that representatives from our Patient Access Team, who register patients on arrival, were not included in the stakeholder group, an error that was remedied in Phase II. A second failure mode, “Vendor take the paging system down without notice,” also had a hazard score of 12. The effect of this failure mode was that team members would not be paged and there would be a delay in treatment.
      Communication and path of travel testing were identified as key objectives for Phase II testing, including determination of the most expeditious path of travel, with consideration of patient privacy, to each respective destination unit.
      In Phase II, the highest-ranking failure mode was again “PICU team not routinely notified of patient arrival,” as the mitigation strategy identified after Phase I testing had not been operationalized. The second highest hazard score was “Travel path to Cardiac Cath lab is longer from new helipad.” The potential effect of this failure mode was a possible delay of time-sensitive interventions for patients with urgent cardiac conditions. By testing and comparing two routes of travel in Phase II, we were able to determine an optimal path. With representatives from Patient Access now involved, we again noted that the failure mode “Patient does not get registered,” could result in a critical delay in potentially lifesaving procedures, as the patient would not be located in the electronic health record. During this phase, we trialed a new process for registration and determined the mitigation strategy was effective for some, but not all destinations. An additional mitigation strategy was then proposed and tested in Phase III.
      Phase III was designed to include actual helicopter landings and test a failure mode identified in Phase I: insufficient security personnel to meet two helicopters landing simultaneously. For this purpose, we opted to use high-fidelity simulation in an effort to stress communication processes identified in Phases I and II. During this simulation, the patient condition worsened during transport to the original destination. This allowed us to evaluate how teams would communicate this critical information and redirect the patient if necessary. Phase III also included a pediatric transport to the PICU to again test PICU communication that had proven problematic in Phases I and II.
      Highest hazard scores in Phase III included a lack of page to PICU teams when the helicopter had landed, lack of efficient patient registration upon arrival, and discovery that the helipad can become dangerously slippery when wet or icy. The potential effects of a slippery helipad during inclement weather include potential injury to crew members or the patient during transport. Solutions for PICU notification and adjustment of the location of patient registration personnel were again modified. The proposed mitigation strategies for the dangerous helipad conditions were to repaint slippery lines with sand-textured paint and to ensure timely and liberal salt application during winter storms.

      Discussion

      We successfully used iterative, three-phase, SbCST in conjunction with HFMEA methodology to identify and mitigate failure modes in the post-construction, preoccupancy phase of a new helipad. Review of failure modes, including hazard scores and degree of process change necessary to implement proposed solutions, was used to identify the objectives and structure for each subsequent phase. Each iterative phase was also used to test proposed solutions to address failure modes identified in earlier phases. Several challenging issues took more than one phase to successfully mitigate. Focus on identifying the minimum and necessary level of fidelity needed to achieve the objectives of each phase allowed us to use the fewest resources possible to achieve our goals. With that in mind, the level of simulation fidelity escalated, beginning with a tabletop exercise and culminating with a high-fidelity event. Increasing the level of fidelity allowed us to more realistically test mitigation solutions for failure modes discovered during earlier phases. Using an HFMEA framework in conjunction with simulation leveraged two effective tools, each designed to prospectively detect and address failure modes prior to the new helipad activation.
      Simulation can replace or amplify real-world experience for the purpose of facilitating reflective learning and has been used as an effective tool for complex system improvement without risking harm to patients. Previous studies have successfully demonstrated that simulation can be used as a vehicle for evaluating new health care facilities for LSTs.
      • Adler MD
      • et al.
      Use of simulation to test systems and prepare staff for a new hospital transition.
      ,
      • Ventre KM
      • et al.
      Using in situ simulation to evaluate operational readiness of a children's hospital-based obstetrics unit.
      ,
      • Geis GL
      • et al.
      Simulation to assess the safety of new healthcare teams and new facilities.
      This body of work recognizes that LSTs are inherent in new design and that simulation can be used to help mitigate risk before patient use. Indeed, the process often identifies LSTs with implications well beyond the process being studied. In our case, one such example was that elevators descending from the helipad could be overridden by nonemergent patient transport procedures, a threat previously undetected and applicable to any emergent transport (see Table 3). Detection of this LST could not have been detected via brainstorming and resulted in security personnel changing the elevator badge reader to prevent this from occurring again in the future.
      One of the less tangible, but essential, benefits of SbCST is that it requires the assembly and engagement of multidisciplinary stakeholders to examine a complex problem or system. This enables discovery of failure modes, which may be very specific and known only to certain perspectives in the process. One example from our project was that only helicopter pilots could appreciate that wind from a certain direction would require landing on the smaller helipad in a position that would render it impossible to safely unload the patient. This type of collaborative problem solving is inherent to simulation and is one reason it is so beneficial to patient safety systems.
      • Phrampus PE.
      Simulation and integration into patient safety systems.
      ,
      • Woods DM
      • et al.
      Improving clinical communication and patient safety: clinician-recommended solutions.
      Use of the HFMEA rubric is a well-recognized tool to test systems and improve patient safety.
      • Cronrath P
      • et al.
      PCA oversedation: application of Healthcare Failure Mode Effect Analysis (HFMEA™).
      ,
      • Ouellette-Piazzo K
      • Asfaw B
      • Cowen J.
      CT Healthcare Failure Mode Effect Analysis (HFMEA®): the misadministration of IV contrast in outpatients.
      The initial steps of an HFMEA process (assembling the team, graphically depicting a process, and breaking it down into subprocesses for a hazard analysis) are essentially equivalent to a tabletop exercise and served as our Phase I. HFMEA adds to traditional simulation a structure that allows for clear identification, categorization, and quantification of risk for safety threats. Tabletop exercises are inherently limited, however. They rely on the memories of the participants to identify failure modes and don't take into account all the complex interactions of the care team during the process and are not accurate reflections of the all the conditions that affect patient care. In situ simulation amplifies the scope of hazard analysis by including action-based realism. Nielsen and colleagues compared HFMEA alone with HFMEA plus in situ simulation to test a new process for response to an obstructed breech delivery. They found HFMEA plus simulation detected 33% more LSTs and more hazardous failure modes when compared to HFMEA alone.
      • Nielsen DS
      • et al.
      Augmenting health care failure modes and effects analysis with simulation.
      These two safety tools synergize well, and their successful pairing has been described by several groups to test clinical systems or spaces. In 2019 Colman et al. leveraged these two safety tools, using failure mode and effects analysis with SbCST to test a newly constructed pediatric subspecialty outpatient center before opening, identifying more than 300 LSTs.
      • Colman N
      • et al.
      Prevent safety threats in new construction through integration of simulation and fmea.
      Simulation also accentuates the power of HFMEA by enabling testing of mitigation solutions. Our iterative approach also expanded on the synergistic strengths of these two systems testing methods, allowing for testing and refinement of mitigation solutions for LSTs. In our study, LSTs from Phase I helped shape the focus and fidelity needed in Phases II and III. Our Phase I tabletop exercise was well suited for initial identification of problem areas. Themes emerged, including communication, paths of travel, and patient registration processes. One example was a failure to include Patient Access and Registration personnel in our Phase I assembly of stakeholders. In the Phase I exercise it became obvious that geographic considerations affected our existing registration processes (patients were physically transported by the patient access desk from the base of the elevator below the tower housing the original helipad; however, the new tower elevator did not) and that without a means to register a patient, it was nearly impossible to deliver safe and timely care. Geography was also relevant to our communication processes to destination units. The PICU and cardiac catheterization lab looked directly out on the existing helipad, and therefore used direct observation of a helicopter landing to prepare for patient arrival. Due to these units facing away from the new helipad, a new process for notification of imminent patient arrival to the unit was required. We therefore designed Phase II to test paths of travel, including patient registration and arrival notification protocols, objectives that could be achieved via low-fidelity simulation. Lee et al. described a similar use of iterative simulation (tabletop followed by higher fidelity) to adapt processes to care for COVID-19 patients.
      • Lee M
      • et al.
      Talk-through walk-through”—a simulation approach adapted during preparation for COVID-19.
      Their group used just two phases of testing, but we found that certain problem areas, such as registration processes and communication to ICUs, required more than one additional phase of testing. Simulating the mitigation solution from the HFMEA process after Phases I and II illustrated that the modification had not entirely solved each problem and additional modifications were needed. By the time of the Phase III high-fidelity event, we had made two rounds of improvements to communication processes to destination units based on identified failure modes. For our final phase, we designed two “stress” tests of those systems, including near simultaneous landing of two helicopters, and a deterioration in patient condition (cardiac arrest) mid-transport to the destination unit to deliberately probe protocols. For this objective, high-fidelity simulation best allowed us to prospectively re-create those high-stakes “stresses.”
      LSTs related to communication proved challenging to mitigate. Issues of communication failure related to technology (Wi-Fi coverage, pager system software) prompted a separate HFMEA exercise lead by Patient Safety. Communication to the PICU were particularly difficult to mitigate. In Phase II we found that the mitigation solution proposed after Phase I had not been operationalized. In Phase III we implemented the mitigation solution—for the Regional Communications Center to page the PICU directly upon helicopter landing—but found that the pager identified was often left at the nursing station and thus did not alert the team. After Phase III the process was changed to directly phone the PICU. This example supports the value of iterative rounds of testing. In fact, an additional round would have been useful to confirm resolution of the issue. Instead we had to confirm mitigation after the first pediatric patient transport.
      We cannot overstate the importance of engaging and obtaining support from institutional executives, operational leaders, and care team stakeholders early in the planning stages. Doing so enabled us to recruit and access the time of numerous care team stakeholders. By involving a thorough breadth of stakeholders, we gained a more complete assessment of the process, executed a resource-intensive Phase III with dual helicopter landings, and better ensured accountability for implementing mitigation solutions. After testing, we demonstrated value added to the system by cataloging and quantifying LSTs, including mitigation strategies, and reporting results to the hospital's Quality and Safety Committee. In addition, we partnered with the hospital Communication Department to arrange for press coverage, resulting in a favorable story in the city newspaper. Demonstration of return on investment (patient safety and favorable publicity) for this SbCST example has led to further requests for SbCST from hospital leadership.

      Limitations

      There are several limitations to our study and its generalizability to other settings. First, this endeavor required a large amount of personnel and equipment resources. The project would not have been possible without adequate stakeholder endorsement and participation, and a simulation center well-resourced with expertise and equipment. In particular, the simultaneous landing of two helicopters, enabling high-fidelity testing, was possible because of strong advocacy from LOM personnel who believed that this testing in a controlled setting (fair weather, planned, with no patient aboard) was critical to ensure patient and flight crew safety. This attests to the value of engaging all stakeholders. In addition, our simulation team became involved with the project after physical plant design and construction. This limited our ability to modify or alter major structural elements of the helicopter landing system. For example, post-construction it was discovered that our second landing pad is not large enough to land and safely unload a patient if the wind was coming from a particular direction. This may have biased our search for LSTs, as we tended to focus more on care coordination and equipment than facilities. A growing body of literature suggests that involving SbCST from early planning phases can optimize safety.
      • Colman N
      • et al.
      Simulation-based clinical systems testing for healthcare spaces: from intake through implementation.
      Our study used a multiphase approach to discover LSTs with a new helipad and set of transport protocols, but we performed each phase only once. It is likely that repeating each phase with different members of each stakeholder group could have revealed more LSTs and better ensured secure solutions to problem areas.
      Our testing was deemed complete when no new significant LSTs were uncovered and/or mitigation strategies for any remaining high-priority LSTs could be implemented successfully without the need for additional testing. Upon reflection, this goal may not have been very feasible, given the downstream effects of mitigation measures employed after Phase III. As we found from earlier phases, some LSTs required multiple phases of testing to optimize the mitigation strategy (specifically, the registration process). It is possible that LSTs discovered during Phase III could require additional testing and revision of initial mitigation strategies.
      Finally, although we observed improvements in care coordination and workflows over the study phases, actual translation of these improvements into meaningful patient metrics was not measured. Certain systematic LSTs, such as geographic holes in pager coverage, are larger issues that require additional personnel and procedures to solve.

      Conclusion

      We found an iterative, three-phase SbCST program using HFMEA methodology to be an effective tool to identify and mitigate latent safety threats prior to opening a new on-site helicopter landing pad. We also found that the two methods of safety testing we employed (simulation and HFMEA) complemented each other well. Simulation enhanced our HFMEA, identifying additional failure modes through realistic enactment compared to theoretical exploration of processes (“brainstorming”) alone. HFMEA engaged the spectrum of stakeholders from frontline workers to operational and clinical leaders, and provided a structure for LSTs to be categorized, prioritized, and mitigated. Iterative, multiphased simulation enabled testing of proposed safety solutions.

      Acknowledgments

      The study team would like to gratefully acknowledge the support of the following people, without whom this important work could not have been completed: Michael Baumann, MD; Jamie Grant; Keith Friedrich; Chris Pare; Thomas Judge and LifeFlight of Maine; and the team at the Hannaford Center for Safety, Innovation and Simulation, including Susie Lane, Todd Dadaleares, Tyler Johnson, Christine Mallar, Bethany Rocheleau, Erin Siebers, Christyna McCormack, Mike Shepherd, Stephanie Salamone, Mariah Wheeler, and Tania Strout, RN.

      Conflicts of Interest

      All authors report no conflicts of interest.

      References

        • Henriksen K
        • Moss F.
        From the runway to the airway and beyond.
        Qual Saf Health Care. 2004; 13 Suppl 1: i1
        • Adler MD
        • et al.
        Use of simulation to test systems and prepare staff for a new hospital transition.
        J Patient Saf. 2018; 14: 143-147
        • Kerner Jr., RL
        • et al.
        Simulation for operational readiness in a new freestanding emergency department: strategy and tactics.
        Simul Healthc. 2016; 11: 345-356
        • Gardner AK
        • et al.
        In situ simulation to assess workplace attitudes and effectiveness in a new facility.
        Simul Healthc. 2013; 8: 351-358
        • Ventre KM
        • et al.
        Using in situ simulation to evaluate operational readiness of a children's hospital-based obstetrics unit.
        Simul Healthc. 2014; 9: 102-111
        • Bender GJ.
        In situ simulation for systems testing in newly constructed perinatal facilities.
        Semin Perinatol. 2011; 35: 80-83
        • Geis GL
        • et al.
        Simulation to assess the safety of new healthcare teams and new facilities.
        Simul Healthc. 2011; 6: 125-133
        • Militello PR
        • Ramzy AI.
        Safety by design. Shock trauma center's helipad received special consideration during its planning and construction.
        J Air Med Transp. 1990; 9: 15-17
      1. Federal Aviation Administration. Chapter 4: Hospital Heliports. Sep 30, 2004. Accessed Oct 8, 2021. https://www.faa.gov/documentLibrary/media/advisory_circular/150-5390-2B/150_5390_2b_part3.pdf.

        • DeRosier J
        • et al.
        Using health care Failure Mode and Effect Analysis: the VA National Center for Patient Safety's prospective risk analysis system.
        Jt Comm J Qual Improv. 2002; 28 (209): 248-267
        • Davis S
        • et al.
        Failure modes and effects analysis based on in situ simulations: a methodology to improve understanding of risks and failures.
        in: Henriksen K Advances in Patient Safety: New Directions and Alternative Approaches, vol. 3: Performance and Tools. Agency for Healthcare Research and Quality, Rockville, MD2008 (Accessed Oct 8, 2021)
        • Nielsen DS
        • et al.
        Augmenting health care failure modes and effects analysis with simulation.
        Simul Healthc. 2014; 9: 48-55
        • Schertzer K
        • In Patti L.
        Situ Debriefing in Medical Simulation.
        StatPearls Publishing, Treasure Island, FL2020 (Updated: Sep 23, 2020.) Accessed Oct 8, 2021
        • Phrampus PE.
        Simulation and integration into patient safety systems.
        Simul Healthc. 2018; 13: 225-226
        • Woods DM
        • et al.
        Improving clinical communication and patient safety: clinician-recommended solutions.
        in: Henriksen K Advances in Patient Safety: New Directions and Alternative Approaches, vol. 3: Performance and Tools. Agency for Healthcare Research and Quality, Rockville, MD2008 (Accessed Oct 8, 2021)
        • Cronrath P
        • et al.
        PCA oversedation: application of Healthcare Failure Mode Effect Analysis (HFMEA™).
        Nurs Econ. 2011; 29: 79-87
        • Ouellette-Piazzo K
        • Asfaw B
        • Cowen J.
        CT Healthcare Failure Mode Effect Analysis (HFMEA®): the misadministration of IV contrast in outpatients.
        Radiol Manage. 2007; 29: 36-44
        • Colman N
        • et al.
        Prevent safety threats in new construction through integration of simulation and fmea.
        Pediatr Qual Saf. 2019 Jun 24; 4: e189
        • Lee M
        • et al.
        Talk-through walk-through”—a simulation approach adapted during preparation for COVID-19.
        Simul Healthc. Epub. 2020 Oct 27;
        • Colman N
        • et al.
        Simulation-based clinical systems testing for healthcare spaces: from intake through implementation.
        Adv Simul (Lond). 2019 Aug 2; 4: 19