Human interaction between individuals and systems does not occur in a vacuum, rather it occurs in a dynamic and multidimensional setting. From a structural and procedural system performance perspective, the nursing care environment “is perfectly designed to get the results it gets” (LLoyd, Murray, & Provost, 2015). When mistakes happen in healthcare, all Joint Commission accredited healthcare organizations are obligated to analyze the care environment to assess for opportunities to improve the structural and procedural elements that lead to care failures, as in the fictitious sentinel event case of Mr. B who presented to the emergency department for a manual reduction under moderate sedation of a dislocated left hip after sustaining a ground level fall. This review focuses on root causes of the sentinel event including errors and hazards present in the care environment, provides recommendations for an improvement plan to reduce the likelihood of a recurrence of this event in the future, evaluates the likelihood of success in reducing recurrence after the recommendations are implemented, and discusses the key roles nurses at all levels play in improving and achieving sustained quality patient care and outcomes.
A. Root Cause Analysis
When a patient with a relatively benign medical history, such as Mr. B, presents to an acute care setting for treatment of a dislocated hip and subsequently dies within a week there is cause for concern and a need for a root cause analysis. A root cause analysis (RCA) is the direct application of quality improvement principles and methods focused on uncovering system and process deficiencies that lead to care failures (Werner, 2011). An RCA begins by systematically answering the following four questions (Huber & Ogrinc, 2015): What happened? Why? How will we prevent it in the future? Will our changes actually cause improvement? To begin answering those questions for the care and outcome of Mr. B, an interprofessional RCA team will be formed consisting of direct-care staff (Nurse J., Doctor T., unnamed LPN), department leadership (clinical manager or supervisor), and a risk manager or quality improvement specialist. In the event that direct care staff would not be a good fit for reasons of bias, emotional state, or otherwise, other team members would be selected who fill equivalent roles (another emergency RN, another emergency MD, etc.). It is also possible to have a family member involved if “they are able to analyze the factors contributing to an error without becoming overwhelmed by their emotions” because “their perspectives can be extremely valuable” (Huber & Ogrinc, 2015, p. 3).
After the team is formed, the group will, as a first step, develop as detailed as possible accounting of the event, normally in chronological order with details on a flow chart (who, what, when, where) documented for each occurrence in the chain of events as well as the environment of care (staffing, policies in effect, equipment available) in which the event occurred1. In summary, Mr. B presented to local, rural emergency department for reduction of a dislocated left hip under moderate sedation and his medical history included impaired glucose intolerance, prostate cancer, and hyperlipidemia. His medications included atorvastatin for cholesterol and oxycodone for chronic back pain. The hospital had appropriate staffing (quantity and skill set) consisting of an RN, LPN, unit secretary, MD, and on-call respiratory therapist; equipment including pulse oximetry, ECG with respirations, and supplemental O2; and policies, procedures, and qualifications related to conscious sedation practice in place to perform the procedure.
Beginning at 16:05 Mr. B underwent the procedure and received a total of 4 mg IVP hydromorphone, 10 mg IVP diazepam, and after the procedure was placed on continuous pulse oximetry and blood pressure monitoring every 5 minutes at 16:30 without note of ECG or respirations monitoring or supplemental O2 per policy by Nurse J. Mr. B’s vital signs are noted as stable at 16:35, an unspecified amount of time later the LPN responds to an alarm indicating Mr. B’s low SpO2 of 85% and intervenes by resetting the alarm and initiating another blood pressure reading without noting the results or addressing the SpO2 value. At 16:43, the patient’s son alerts Nurse J to another alarm and the RN notes vital signs of blood pressure 58/30, SpO2 79%, respirations undetected and no pulse palpable. A code is called, Mr. B is first noted to be in ventricular fibrillation, chest compressions are then initiated, and the subsequent ACLS interventions result in Mr. B being ventilator dependent with fixed/dilated pupils, nonreactive to noxious stimuli, and blood pressure 110/70 with sinus rhythm and regular pulse.
He is transferred to a higher level of care and passes away 7 days later. The second step on which the RCA team will embark is a detailed second flow chart of what should have happened. “By placing the two flowcharts side by side, the team can clearly see the differences between the ideal process and the actual process that led to the adverse event. This can help further pinpoint contributing factors later in the process” (Huber & Ogrinc, 2015, p. 9). In the case of Mr. B, several hazards and errors would be uncovered during this part of the RCA. As a select example, the comparison of the flow charts would reveal that the manual reduction of Mr. B’s hip using conscious sedation had known hazards, namely the use of narcotics and benzodiazepines which are associated with respiratory depression. Additionally, the care environment developed a hazard when the patient population increased both in number and acuity with the admission of the acute respiratory distress patient and increasing patient load in the lobby without note of available back up staff being called in. Examples of errors from the flow chart comparison might include failure to assess and monitor when Nurse J initiates blood pressure and SpO2 measurements, fails to initiate ECG with respiration monitoring, fails to administer supplemental O2, and leaves the room without apparently noting the baseline of the patient2.
Furthermore, there appears to be an error in the lack of communication collaboration between the RN and LPN regarding Mr. B’s post procedure status and monitoring needs, and there is a failure to rescue when the LPN notes the low SpO2 value, fails to respond, and instead re-initiates another blood pressure reading without noting the results. As Mr. B’s condition deteriorates and a code is called, an ACLS error is observed in the timeline when the patient is noted first to have absent pulse and respirations and that a monitor is next applied and the patient and displays ventricular fibrillation. Chest compressions appear to not have been the first action in this scenario, nor is end tidal CO2 monitoring noted as initiated to monitor the quality of compressions. These are examples of hazards and errors in the care of Mr. B and in an actual RCA the level of detail would likely turn up additional gaps in his treatment. The third step of an RCA is the goal of determining causes of Mr. B’s sentinel event.
Often an Ishikawa (aka: fishbone) diagram that categorizes the types of causal factors can be used, such as Charles Vincent’s categories of 1) patient characteristics, 2) task factors, 3), individual staff member, 4) team factors, 5) work environment, 6) organizational and management factors, and 7) institutional context (Huber & Ogrinc, 2015, p. 11). In the case study of this sentinel event the following can be identified as the most relevant causal factors leading to the outcome: patient age, chronic prescription narcotic use, repeat sedation/analgesia doses and drug selection, adherence to standards for procedural/post procedural assessment and monitoring, team communication, failure to rescue at the first indication of clinical decline, and quality of emergency cardio resuscitation efforts. To further clarify the causes, a series of “whys” must be asked to drill down into the true cause of the failure, keeping in mind that people are not causes and that human error generally must have a preceding cause. Without access to the detail of policies and procedures available in the setting where Mr. B’s procedure took place nor the institutional conditions, the following is a constructed example of what the RCA team might find after asking a series of “whys” to get to the root of the event.
Consider the policy on safe dosing ranges and drug selection for conscious sedation. Per orders, Nurse J administered 10 mg of diazepam IVP and 4 mg hydromorphone within a 15 minute timeframe, divided into two doses spaced roughly 10 minutes apart which is earlier that most general guidelines for repeat dosing during moderate sedation procedures (Egan, 2007). Why was that? Nurse J was aware of the policy. The MD ordered it and had used the drug selection and doses often throughout his/her career without incident even though the moderate sedation policy specifies reducing opioid dosing as the dosing of benzodiazepines increases. Why did Nurse J administer the medications with knowledge of the policy? Nurse J has experienced Dr. T’s verbal outbursts when nursing staff asks for clarification of orders or concern for patient safety.
Why did Nurse J not attempt to collaborate with Dr. T? The institutional culture permits providers to disregard policy based on personal clinical experience and has a provider-focused rather than patient-focused model of care. Why does the institution have a provider-centered model of care? This rural hospital has had a hard time retaining emergency room physicians. In this scenario one causal factor can be poor communication fostered by a provider-centered care environment in which policies have historically been superseded by physician preferences in an effort to retain providers. As the RCA team looks at other factors and drills down, they may find that the patient was not placed on an ECG and supplemental oxygen because the equipment, although available, was not easily accessible in the care area and that there was a lack of standard work for checking that all post procedure monitoring was in fact occurring and documented.
The team may also find that the failure to rescue at the first indication of clinical decline was a lack of communication between the Nurse J and the LPN regarding the status of the patient and his monitoring needs and potential a knowledge gap in the LPN skill set when he/she failed to address the alarm, gather data appropriate to the reason for the alarm, or summon assistance. Lastly, in the code situation, the team may find that lack of adherence to ACLS protocols for ventricular fibrillation may have occurred because of the infrequency of code situations in this setting and no regular mock drills to practice adherence and teamwork between recertification dates.
B. Improvement Plan
A patient who dies from acute respiratory failure after a procedure involving moderate sedation is cause enough to first investigate the reasons leading to the event, and second (perhaps most important) utilizing change theory to prevent a similar reoccurrence. Mr. B’s death was accidental, no one person or contributing event was the single direct cause of his demise, but to prevent it from happening again must involve purposeful analysis of the causes in the beginning and result in a newer, safer standard of practice in the end. In Lewin’s classic 1951 work on change, he identifies specific caveats when initiating change and describes a simply elegant three step model for how change occurs (Cherry, 2011). Change should always be implemented for good, be planned and implemented gradually, never be unexpected or abrupt, and everyone affected by the change should be involved in planning for and executing the change. Once an organization has decided that change is necessary, the organization enters an unfreezing stage when the need for change is identified and communicated, the moving stage when changes occur, and the refreezing stage when the changes implemented become the new status quo.
To reduce the likelihood of a recurrence of Mr. B’s sentinel event, the following procedural changes would be implemented using change theory to guide implementation. Standard work and documentation and communication on all immediate post-sedation patients would be developed to ensure compliance with the institutions policy on post sedation monitoring. Within 5 minutes of the conclusion of the moderate sedation procedure, the Registered Nurse will remain with the patient for a minimum of 15 additional minutes and complete the following, 1) assess vital signs, respiratory status, and document baseline findings within first five minutes, 2) ensure that patient is actively monitored by continuous ECG with respirations, continuous pulse oximetry, and automatic blood pressure reading set to a five-minute interval, 3) verify that the patient is on 2L per minute per nasal cannula supplemental O2 until able to maintain wakeful state without stimulation, and 4) verify alarm setting volume, sensitivity, and respiratory consult is initiated prior to leaving patient and informing other care team members of patient status and need for frequency of status checks until patient fully recovered per policy definitions.
C. Failure Mode and Effects Analysis
The purpose of a Failure Mode and Effects Analysis (FMEA) is to avoid what caused the need for its application in the first place. When there are undesirable events in healthcare a proactive organization seeks to reduce or eliminate those same events, and change is often needed. At times it is as direct as uncovering why staff are not adhering to policy, other times the actual patient care process can be broken which necessitates immediate change for patient safety due to current design for failure. There are multiple other variants in between. The first two pre-steps in an FMEA are selecting a process to analyze and a constructing a team to do the work. The first step involves selecting a process that does not involve excessive sub-processes (Institute for Healthcare Improvement, 2015) In the case of Mr. B, the proposed monitoring (section B, page 8) is aligned with this goal as it is largely the responsibility of a single staff member post-procedure. The obligation to patient care and safety does not transfer to another staff member, although there is a clear need for intra-team communication.
The RCA/FMEA team should be comprised of relevant staff to analyze, implement, test, and modify the proposed post-procedural patient outlined in section B above. This would likely include the direct care RN, the support LPN, the respiratory therapist, the emergency MD, the equipment supply department, the clinical supervisor, clinical nurse educators, and the department in charge of policy and procedure design/review/implementation/maintenance. Once the scope of the analysis has been determined and the team has been assembled the process continues. In step three, the team gets together and walks through the proposed new process scenario step-by-step with the goal of accurately documenting the ideal process. “The team should agree that the steps enumerated in the FMEA accurately describe the process” (Institute for Healthcare Improvement, 2015). In step four, the team should attempt to break (e.g. cause failure in) the new system and focus on potential care failures in the new process. Identification of care failures in the proposed process lays the groundwork for determining potential adverse events. When failures are detected in this step, corrections are made and retested. In this way, the new process is tested for safety and reliability before use in the field, which increases patient safety.
Severity is calculated in step five as the team assigns a numerical risk value to each possible failure in the newly created ideal process and determines the potentials for likelihood of a failure event occurring, and likelihood of a failure event being detected, and if the failure occurs, how severe is the potential harm to the patient. Numeric values are applied to rank the significance of potential care failures, and the lower the score the safer the process. In the proposed model for the care of a patient after moderate sedation, the new standard work of the RN consists of four steps. The second step for the RN is to “ensure that patient is actively monitored by continuous ECG with respirations, continuous pulse oximetry, and automatic blood pressure reading set to a five-minute interval” (section B, page 8). In step six of the FMEA the team may assign a numeric value of 2 for the likelihood a failure would occur in the second step of the new process, a numeric value of 4 that a failure would be detected, and a numeric value of 10 for the severity of a failure occurrence, yielding a total risk priority number (RPN) of 80.
The lowest RPN value is 1 and the highest value is 1,000. When all of the steps in the new procedure are ranked, the steps with the highest values “are the ones the team should consider first as improvement opportunities” (Institute for Healthcare Improvement, 2015). In the final phase as the team continues to use the FMEA to error-proof each step in the process, the team will take actions to reduce the RPN values to improve safety and reliability. Suggested actions for each of the failure modes (occurrence, detection, severity) include options to add a verification step such as a double sign off if the occurrence of a failure is likely, use of technological alerts if the occurrence of a detection failure is likely, and the use of drills and training to identify early warning signs of impending failure mode if the severity of an occurrence is likely to cause severe harm (Institute for Healthcare Improvement, 2015). Interventions identified in the FMEA would then be tested on a small scale, observed for defects, modifications made, and then gradually increased in scope before full-scale implementation.
For example, the new four step standard for the RN when caring for a post-moderate sedation patient may be tested on the next five patients to come into the emergency department to whom moderate sedation will be administered. Feedback from the staff involved in the new work flow would be gathered, and the process would be adjusted as necessary, and then retested with a larger sample size. In the case of Mr. B, appropriate monitoring and staff communication were significant root causes of the sentinel event, and the proposed standard work includes specified monitoring, documentation, and team communication with verification by a respiratory therapist, all of which will significantly reduce the likelihood of a recurrence. Furthermore, once the new workflow is determined to be successful (e.g. ideally, number of patient deaths following moderate sedation in the emergency department reaches zero!), this process could be taken to other departments where moderate sedation takes place, such as same-day-surgery centers and inpatient settings in the hospital. A similar approach for starting small, making adjustments as needed for individual departments, and repeat testing would be used.
D. Nursing’s Role in Quality Improvement
Nurses are in a prime position from the bedside to the boardroom to participate and actively influence the quality of care across many practice environments, and patient safety and positive outcomes are the core of all nursing care. “Because nurses are the key caregivers in hospitals, they can significantly influence the quality of care provided and, ultimately, treatment and patient outcomes. Consequently, hospitals’ pursuit of high-quality patient care is dependent, at least in part, on their ability to engage and use nursing resources effectively” (Draper, Felland, Liebhaber, & Melichar, 2008). In the case of Mr. B, the bedside nurse is in a unique position to see first-hand the positive and negative effects of processes, policies, and procedures on individual patients and their outcomes. Often these staff nurses have relevant recommendations to improve patient care, safety, and satisfaction, and for that reason they are a valuable asset to any quality improvement team.
These nurses are also skilled at providing input regarding the authentic work flow and documenting the current state of nursing at the point of care. In addition to the staff nurses, leaders in the healthcare delivery system such as Clinical Nurse Leaders (CNLs) have additional “expertise in quality improvement and cost-effective resource utilization” (Lenburg, 2011, p. 67). With the additional education and training, CNLs are qualified to identify areas for care improvement and lead quality improvement projects. In the case of Mr. B, a CNL may also have certifications in quality improvement frameworks such as Six-Sigma © to lead a team in documenting, revising, testing, and implementing quality improvement changes with the goal of reaching a zero-defect process.
Nurses in management roles play a valuable role in promoting care environments that support the direct-care staff and promoting and managing processes that support positive patient outcomes. A nurse manager for a department, in the case of Mr. B, brings expertise to a quality improvement team regarding broader issues that affect the care environment including staffing concerns, education concerns, and budgetary concerns. From an executive management perspective, the chief nursing officer (CNO) has a role in supporting a safe environment as well. Health organizations and their leaders focused on safety “use constant surveillance to manage the potential for humans (health professionals) to make unintentional mistakes. Organizations with a just-culture promote learning from adverse events” (Sherwood & Hicks, 2011, p. 473). Nurses at all levels have valuable insight into what works to promoted safe patient care, and what detracts from it.
Failure to monitor, failure to rescue, and failure to communicate are frequent themes in adverse patient events. In virtually any other industry, defects are considered to be part of the cost of production or part of the cost of providing a service. In healthcare, defects can and do cause serious harm, irreparable damage, and death. In the case of Mr. B, his son will have to continue with the loss of his father and the knowledge that it could have been prevented. The healthcare providers who were attending Mr. B in the emergency department that day will have to continue with the knowledge that a combination of their human error and system defects resulted in the code and eventual death of a patient. Learning from adverse events and improving the safety and quality of patient care and outcomes is at the heart of the nursing profession because human beings are the focus of nursing care. With continued quality improvement efforts, led and influenced by nurses at every level, patient outcomes will continue to be improved and adverse events reduced.
Cherry, B. (2011). Nursing leadership and management. In B. Cherry, & S. Jacob, Contemporary nursing: Issues, trends, and management (pp. 333-363). St. Louis: Mosby. Draper, D., Felland, L., Liebhaber, A., & Melichar, L. (2008, March). The role of nurses in hospital quality improvement. Retrieved from Robert Wood Johnson Foundation: http://www.rwjf.org/en/research-publications/find-rwjf-research/2008/03/the-role-of-nurses-in-hospital-quality-improvement.html Egan, T. (2007). Moderate sedation administration for nonanesthesiologists: Scientific basis of safe clinical practices. Retrieved from Medscape Nurses: http://www.medscape.org/viewarticle/567803 Huber, S., & Ogrinc, G. (2015). How a root cause analaysis works. (K. Vega, Ed.) Retrieved from Institute for Healthcare Improvement: http://www.ihi.org/education/IHIOpenSchool/courses/Pages/default.aspx Institute for Healthcare Improvement. (2015). Failure modes and effects analysis tool. Retrieved from Institute for Healthcare Improvement: http://www.ihi.org/resources/Pages/Tools/FailureModesandEffectsAnalysisTool.aspx Lenburg, C. (2011). The influence of contemporary trends and issues on nursing education. In B. Cherry, & S. Jacob, Contemporary nursing: Issues, trends and management (5th ed., pp. 41-70). St. Louis: Mosby. LLoyd, R.,