Improving Handoff Communications in Critical Care* Utilizing Simulation-Based Training Toward Process Improvement in Managing Patient Risk

Transcription

1 CHEST Transparency in Health Care Improving Handoff Communications in Critical Care* Utilizing Simulation-Based Training Toward Process Improvement in Managing Patient Risk Haim Berkenstadt, MD; Yael Haviv, MD; Atalia Tuval, MSc; Yael Shemesh, RN; Alexander Megrill, MD; Amir Perry, MSc; Orit Rubin, PhD; and Amitai Ziv, MD Background: A patient admitted to the medical step-down unit experienced severe hypoglycemia due to an infusion of a higher-than-ordered insulin dose. The event could have been prevented if the insulin syringe pump was checked during the nursing shift handoff. Methods: Risk management exploration included direct observations of nursing shift handoffs, which highlighted common deficiencies in the process. This led to the development and implementation of a handoff protocol and the incorporation of handoff training into a simulation-based teamwork and communication workshop. A second round of observations took place 6 to 8 weeks following training. Results: The intervention demonstrated an increase in the incidence of nurses communicating crucial information during handoffs, including patient name, events that had occurred during the previous shift, and treatment goals for the next shift. However, there was no change in the incidence of checking the monitor alarms and the mechanical ventilator. Conclusions: Simulation-based training can be incorporated into the risk management process and can contribute to patient safety practice. (CHEST 2008; 134: ) Key words: miscellaneous; practice; training in internal medicine Abbreviation: SBME simulation-based medical education Editor s note: This article is a retrospective investigation of an apparently minor incident related to a nursing shift handoff that led to a prospective evaluation of the medical process and resulted in a proactive quality and safety intervention that included a simulationbased training component. The authors describe and discuss the possible contribution and limitations of simulation-based training as part of risk management processes. *From the Israel Center for Medical Simulation (Dr. Berkenstadt, Megrill, Rubin, and Ziv), Department of Internal Medicine A (Dr. Haviv and Ms. Shemesh), Sheba Medical Center, Tel Hashomer, Israel; Organizational Behavior Program (Ms. Tuval), the Faculty of Management, Tel Aviv University, Tel Aviv, Israel; the Industrial Engineering and Management Department (Mr. Perry), Ben Gurion University, Beer-Sheba, Israel; and the National Institution for Testing and Evaluation (Dr. Rubin), Jerusalem, Israel. The authors have reported to the ACCP that no significant conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article. Manuscript received April 4, 2008; revision accepted April 21, Reproduction of this article is prohibited without written permission from the American College of Chest Physicians ( org/misc/reprints.shtml). Correspondence to: Haim Berkenstadt, MD, The Israel Center for Medical Simulation, Sheba Medical Center, Tel Hashomer Ramat Gan 52621, Israel; berken@netvision.net.il DOI: /chest Reporting adverse events in medical systems typically requires exploring beyond the initial description of the event in an attempt to identify the wide variety of contributing factors that led to the For editorial comment see page 9 eventual incident. It is necessary to reflect on what the incident reveals regarding the safety gaps and inadequacies in the health-care system in which the incident occurred in order to learn and potentially prevent future similar events. 1 Although descriptions of system analysis have previously been reported, 2,3 the aim of this article was to describe a process by which the retrospective investigation of an apparently minor incident led to a prospective evaluation of the medical process, resulting in a proactive quality and safety intervention that included a simulation-based training component. In this event, a nurse at a medical step-down unit 158 Transparency in Health Care

2 administered insulin in continuous infusion to a patient at a dosage higher than that ordered by the physician and documented it. During the handoff between nursing shifts, the insulin syringe and the syringe pump were not checked, and severe hypoglycemia was recognized only 45 min later. Reflecting the increased recognition of the importance of the handoff process for safer medical practice, handoff was introduced recently 4 by the Joint Commission International Center for Patient Safety in the United States as a national patient safety goal. As shown in the Appendix, handoff communication must include up-to-date crucial information regarding patient status and care, treatment and service, and any recent or anticipated changes in the condition of the patient. The handoff process should be interactive in order to enable active discussion between the health professionals involved in the handoff on both sides (ie, the provider and the receiver of the information regarding every patient). 5 Major barriers to effective handoffs include the following: the physical setting (including background noise and lack of privacy); the social and professional setting, such as pressure on medical personnel at the time of handoff; organizational hierarchy and status issues; language and cultural barriers; and the limitations of communications via telephone, , paper, or computerized records. 6 All of these contributing factors make the handoff process prone to safety gaps and mistakes, but at the same time raise the opportunity for intervention aimed at improving team members handoff communication skills. Materials and Methods Risk management exploration was conducted following the reporting of the event through the institutional incidentreporting system. According to accepted policy, 7 the chronology of the event was first established and revealed the following facts: (1) the patient was admitted to the department at 3:00 pm with sepsis; (2) the physician ordered insulin administration in a continuous infusion using a sliding scale of insulin dose adjusted to blood glucose levels, which were measured every 3 h; (3) the staff nurse prepared the insulin syringe pump and, according to the physician s order, began the infusion of 6 U of insulin per hour at 4:00 pm, but the starting insulin dose was not indicated in the nursing notes section of the patient s chart; (4) according to the blood glucose levels measured at 10:00 pm, the correct insulin dose should have been 2 U per hour, but the actual dose administered was still 6 U per hour (similar to the starting dose); (5) at 10:45 pm, the handoff between nursing shifts took place at the nurses station (and not by the patient s bed, as per policy); (6) at 11:45 pm the patient was found by the staff nurse sweating and pale with blood glucose levels of 35 mg/dl, and a check of the syringe pump revealed that the patient had been given insulin in a dose higher than that indicated; and (7) the insulin infusion was stopped, and IV glucose was administered. Following the facts collection step, care-delivery problems that represented departures from safe margins and institutional protocols were identified. The problems that were identified included the following: (1) missing documentation of insulin dose in the nursing notes sections of the patient s chart; (2) although blood glucose levels were monitored every 3 h as ordered by the physician, insulin doses were not adjusted accordingly; and (3) the nursing shift handoff took place at the nursing station and not by the patient s bed, as per policy. In the last step of the risk management analysis process, factors that contributed to the occurrence of the incident and its underlying problems, as well as other deficiencies not directly related to the incident, were listed and categorized. These factors included a lack of institutional protocols for shift handoff, a lack of training on handoff skills, and interpersonal problems among members of the nursing staff in the department. Based on the above analysis and following approval of the institutional review board ethics committee, a study was designed and conducted at one of the hospital s medical step-down units, a five-bed unit that is part of a 35-bed medical department, which is equipped to treat ventilated and nonventilated patients. The study observed all 25 members of the nursing staff, who signed an informed consent approving passive observation of their practice during handoffs and active participation in the simulation-based training that followed. In the first step of the study, 224 observations on nursing shift handoffs were completed by a single observer (A.M.) using a preconfigured checklist. The checklist was developed on the basis of expert opinion and was modified according to the findings of the first 24 observations. The observer focused on the verbal communication between nurses, the actions performed by the nurses, and the use of the written patient charts during the handoff process. The observer witnessed the handoff process of one patient at a time, and the observations were distributed in a balanced manner between shifts and nurses. Based on the analysis of the data collected through the observations and by structured interviews with five senior nurses, an intervention aimed at improving the quality and safety of the handoff process was developed. The intervention included the following two layers: (1) the development and implementation of a written checklist/protocol for the handoff (Appendix); and (2) the development and incorporation of simulation-based handoff training into a full-day teamwork and communication workshop, in which the teams also had an opportunity to practice the proper use of the handoff protocol. The simulation-based training was conducted at MSR, The Israel Center for Medical Simulation. 8 Three simulators were utilized during the training, including the Human Patient Simulator (Medical Education Technologies, Inc; Sarasota, FL), the SimMan (Laerdal; Stavenger, Norway), and the AirMan (Laerdal). The simulated setting was of a fully equipped, three-bed, step-down unit. In addition, a standardized/simulated patient (called the SP actress) was trained to play the role of a patient s family member. The simulation training scenarios were videotaped and used for debriefing purposes immediately following each simulation session. Three cameras (one of which had pan-tilt-zoom capabilities) were connected to a digital recording system that fed a four-quadrant screen with software that enabled reflective debriefing. Relevant handoff tasks that were incorporated into the scenarios included the following: the handoff of the three patients between nursing shifts; recognizing a mistake in the continuous infusion of heparin administration; recognizing a mistake in the set up of mechanical ventilation; and recognizing a mistake in blood administration (same patient name, similar identification number, and different blood type). Debriefing sessions following the scenarios (facilitated by H.B., A.T., and Y.H.) focused on the introduction of the handoff protocol, communication between nurses, and the identification and prevention of mistakes during the handoff. All 25 members of the nursing staff in the step-down unit, in groups of four nurses each, participated in the simulationwww.chestjournal.org CHEST / 134 / 1/ JULY,

3 based training. The training lasted 6 h and included four simulated scenarios and four debriefing sessions. Six to eight weeks following the simulation-based training, a second round of observations took place in the same step-down unit, which included 166 observations of nursing shift handoffs. The observations were performed by the same observer (A.M.) using the same checklist, and the data were compared to the preintervention data using 2 analysis. Results A total of 390 observations were completed in this study, 224 before the intervention, and 166 following the intervention. The distributions of observations per nurse and between nursing shifts were similar before and after the intervention. Following the intervention, there was an increase, during the handoff process, in the number of nurses indicating the patients names, ages, diseases, and the reasons for step-down unit admission (Table 1). There was also an increase in the number of nurses indicating the occurrence of events during the previous shift (from 199 of 224 nurses [88%] before intervention to 166 of 166 nurses [100%] following the intervention; 2, 19.8), and in the number of nurses presenting treatment goals for the next shift (from 97 of 224 nurses [43%] before the intervention to 115 of 166 [69%] following the intervention; 2, 25.9). Following the intervention, there was an increase in the number of nurses referring during the handoff process to the patients physiologic parameters presented by the monitoring system, to patients fluid balance (Table 2), and to physicians medical orders (Table 3). However, following the intervention, the monitor alarms were checked and adapted to the patient s status in only 4 of 166 cases (2%; 0 of 224 before the intervention; 2, 5.4), the mechanical ventilator was checked in 25 of 38 cases (66%; 7 of 88 cases before the intervention; 2, 46.8), and the medications being administered in continuous infusion were checked in 20 of 31 cases (65%; 12 of 109 cases before the intervention; 2, 39.2). Discussion Incident reporting is considered to be the cornerstone of most initiatives to improve patient safety. However, reports that are not followed up by incident analysis contribute little to the understanding of causes and to the prevention of future incidents. 1 In this study, we have described a case report of a proactive risk management process coupled with simulation-based training. The reported incident described was analyzed with the traditional retrospective approach, but also motivated a prospective evaluation of safety gaps and inadequacies in the process of handoff in the hospital s step-down units and led to an interactive simulationbased training involving the nursing staff of the stepdown-unit in an attempt to improve their clinical skills in general, with an emphasis on their handoff skills. This approach was inspired by the British systems analysis methodology, assuming that the purpose of the risk management process is to reflect on what the incident reveals about the gaps and inadequacies in the health-care system in which it occurred 7,9 ;itis different from the commonly used American root cause analysis approach. 10,11 In the intervention phase of this study, simulationbased medical education (SBME) was chosen as the educational approach. SBME provides a hands-on experiential educational opportunity, enabling controlled and proactive exposure of trainees to both regular and complex uncommon clinical scenarios. SBME training is trainee oriented, as it is conducted in a safe and mistake-forgiving environment, where trainees can learn from their errors and training can be adjusted to their needs and deficiencies. 12 SBME further provides a unique opportunity for training in team and interpersonal communication skills, which constitute a well-recognized patient safety factor that is seldom addressed in traditional medical education. Training is performed without the ethically disturbing duality of patient care and medical training, which is associated with traditional bedside teaching. Another important benefit of SBME is the reproducible, standardized, objective setting that it provides for both formative assessment (debriefing) 13,14 and summative assessment (testing). 15,16 Over the past decade, advanced medical simulation has been widely used in teaching curricula, 17 in team crisis resource management training, 18,19 in the evaluation and accreditation of medical trainees and graduates, 20,21 and in detecting gaps in medical knowledge. 22,23 Although the use of SBME for these applications may improve patient Table 1 Patient Demographic Data and Medical History During Handoff: Comparison Between Observations Made Before and Following the Intervention Variables Before Intervention Following Intervention 2 Statistic p Value Indicating patient s name 108/224 (48%) 159/166 (96%) Indicating patient s age 42/224 (19%) 105/166 (63%) Indicating patient s diseases 134/224 (60%) 162/166 (98%) Indicating the reason for step-down unit admission 136/224 (61%) 166/166 (100%) Transparency in Health Care

4 Table 2 Comparison of Patient Physiologic Parameters and Fluid Balance During Handoff Between Observations Made Before and Following the Intervention Reference Before Intervention Following Intervention 2 Statistic p Value Heart rate during handoff 85/224 (38%) 115/166 (69%) BP during handoff 109/224 (49%) 110/166 (66%) Oxygen saturation during handoff 120/224 (54%) 140/166 (84%) Fluids balance 146/224 (65%) 151/166 (91%) Laboratory results 59/224 (26%) 124/166 (75%) safety, to our knowledge, this study is the first to demonstrate the value of its use as part of prospective risk management intervention. Although the prospective evaluation and intervention used in this study decreased the incidents of mistakes performed during handoff, as indicated by the observations, this methodology cannot be used by itself on a routine basis. Unlike a regular risk management process that takes hours or days, the process described in this study, including the observations and intervention, lasted a few months. Moreover, the cost of the process in terms of manpower, working hours, simulation time, and other resources cannot be neglected. This aspect is further emphasized by the fact that the costly simulation-based intervention was not compared to a simple intervention such as lectures to the nursing staff and implementation of a handoff protocol without training. Indeed, a recent study 24 demonstrated a reduction in technical errors and information omissions in the handoff of infants after complex congenital heart surgery from the operating room team to the intensive care team following the development and introduction of a protocol without simulation-based training. Moreover, in this study, improvements were demonstrated in some handoff behaviors but not in others. While communication behaviors using interpersonal verbal skills (eg, indicating patient name and events, presenting treatment goals, and referring to patient physiologic parameters) were improved, actual physical behaviors (eg, checking the monitor or ventilator) were minimally changed or were not changed at all. Although both kinds off behaviors were part of the training scheme and were acknowledged by the participants as important for patient safety, this difference might imply that improvements in training are indicated and environmental constraints during clinical work such as time pressure, work overload, or conflicting demands should be addressed in the implementation of new behaviors. Another major aspect in any training program aiming at improving work behaviors generally, and eliminating risky behaviors specifically, is the worker s motivation. While the intention to work toward a goal was indicated as a major source of work motivation, it follows that in order for a goal to induce motivating and augmenting performance efforts, the presented goal should be specific, perceived as realistic, and accepted by the individual. 25 Respectively, simulation-based training that is aimed at changing the worker s behavior should consider these aspects. The intervention in this training exercise was not proven to optimize all of the handoff parameters, and some gaps, mainly in the actual checks of the mechanical ventilators, syringe pumps, and the monitoring system alarms, require further intervention and implementation in order to achieve satisfactory improvement. However, despite the limitations indicated above, unlike most risk management processes, we were able to prospectively demonstrate the value of the intervention (ie, the positive influence of simulation-based training on most postintervention observations). Furthermore, we were able to define the objectives of the training program on the basis of documented common deficiencies, thus making the intervention more valid as well as better linked to the actual medical practice. In addition, this study is unique in its choice to apply simulationbased training to common daily routines that have strong implications for patient safety, and in its Table 3 Comparison of Patient Medical Orders During Handoff Between Observations Made Before and Following the Intervention Reference Before Intervention Following Intervention 2 Statistic p Value Mechanical ventilation orders 25/88 (28%) 34/38 (89%) Sedation orders 48/75 (64%) 30/30 (100%) Orders for medications administered in continuous infusion 71/109 (65%) 31/31 (100%) Feeding orders 120/224 (54%) 145/166 (87%) CHEST / 134 / 1/ JULY,

5 encouraging results in terms of the proven effectiveness of the intervention, as reflected by the nurses improved actual practice during handoffs. Appendix: Handoff Protocol/Checklist Administrative data Patient name/age/gender/family Date of admission to unit Problem list Medical history Reason for admission to unit Current medical problems Current status Neurology: consciousness and sedation score Cardiovascular status: BP, heart rate, and rhythm Respiratory status: respiratory rate, oxygen saturation, oxygen supplementation, and mechanical ventilation Medications Regular medications Medications in continuous infusion Feeding Lines and invasive devices IV lines and fluids input Nasogastric tube and input/output Urinary catheter and urine output Endotracheal tube and secretions Results Laboratory results Radiology results Events during the last shift Hemodynamic Respiratory Infection Other Hands-on checking Running fluids Medications in continuous infusion Mechanical ventilation Monitor alarms Dressings Tasks expected to be done Laboratory tests and imaging studies (pending results or need to order) Procedures Consultations Others References 1 Vincent CA. Analysis of clinical incidents: a window on the system not a search for root causes. Qual Saf Health Care 2004; 13: Wachter RM, Shojania KG, Saint S, et al. Learning from our mistakes: quality grand rounds, a new case-based series on medical errors and patient safety. Ann Intern Med 2002; 136: Garnerin P, Huchet-Belouard A, Diby M, et al. Using system analysis to build a safety culture: improving the reliability of epidural analgesia. Acta Anaesthesiol Scand 2006; 50: Joint Commission International Center for Patient Safety. Improving handoff communications: meeting national patient safety goal 2E. Available at: /. Accessed May 22, Committee on Patient Safety and Quality Improvement. Communication strategies for patient handoffs. Obstet Gynecol 2007; 109: Solet DJ, Norvell JM, Rutan GH, et al. Lost in translation: challenges and opportunities in physician-to-physician communication during patient handoffs. Acad Med 2005; 80: Vincent CA, Taylor-Adams S, Chapman EJ, et al. How to investigate and analyse clinical incidents: clinical risk unit and association of litigation and risk management protocol. BMJ 2000; 320: Ziv A, Erez D, Munz Y, et al. The Israel Center for Medical Simulation: a paradigm for cultural change in medical education. Acad Med 2006; 81: Clinical Safety Research Unit. Who are we and what is our mission? Available at: Accessed May 22, Vincent C. Understanding and responding to adverse events. N Engl J Med 2003; 348: Stalhandske E, Bagian JP, Gosbee J. Department of Veterans Affairs patient safety program. Am J Infect Control 2002; 30: Issenberg SB, McGaghie WC, Hart IR, et al. Simulation technology for health care professional skills training and assessment. JAMA 1999; 282: Ende J. Feedback in clinical medical education. JAMA 1983; 250: Rooks L, Watson RT, Harris JO. A primary care preceptorship for first-year medical students coordinated by an area health education center program: a six-year review. Acad Med 2001; 76: MacRae H, Regehr G, Leadbetter W, et al. A comprehensive examination for senior surgical residents. Am J Surg 2000; 179: Weller JM, Bloch M, Young S, et al. Evaluation of high fidelity patient simulator in assessment of performance of anaesthetists. Br J Anaesth 2003; 90: Seropian MA. General concepts in full scale simulation: getting started. Anesth Analg 2003; 97: Blum RH, Raemer DB, Carroll JS, et al. Crisis resource management training for an anaesthesia faculty: a new approach to continuing education. Med Educ 2004; 38: Weller J, Wilson L, Robinson B. Survey of change in practice following simulation-based training in crisis management. Anaesthesia 2003; 58: Devitt JH, Kurrek MM, Cohen MM, et al. The validity of performance assessments using simulation. Anesthesiology 2001; 95: Murray DJ, Boulet JR, Kras JF, et al. Acute care skills in anesthesia practice: a simulation based resident performance assessment. Anesthesiology 2004; 101: Morgan PJ, Cleave-Hogg D, DeSousa S, et al. Identification of gaps in the achievement of undergraduate anesthesia educational objectives using high-fidelity patient simulation. Anesth Analg 2003; 97: Morgan PJ, Cleave-Hogg D, DeSousa S, et al. High-fidelity patient simulation: validation of performance checklists. Br J Anaesth 2004; 92: Catchpole KR, De Leval MR, McEwan A, et al. Patient handover from surgery to intensive care: using formula 1 pit-stop and aviation models to improve safety and quality. Paediatr Anaesth 2007; 17: Erez M, Early PC, Hulin CL. The impact of participation on goal acceptance and performance: a two-step model. Acad Manag J 1985; 2(suppl): Transparency in Health Care