lecture C

Editor's Notes

• #2: Welcome to Quality Improvement: Learning from Mistakes: Error Reporting and Analysis and HIT. This is Lecture c. Our focus will be on several quality improvement tools that are used often in health care and may prove helpful in learning from mistakes.
• #3: By the end of Learning from Mistakes: Error Reporting and Analysis and HIT, you will be able to apply QI tools to the analysis of HIT errors.
• #4: This lecture will discuss quality improvement tools. There are two quality improvement tools that are used in all organizations accredited by the Joint Commission. They are known as mantras of modern risk management because they have the power to transform patient safety through learning. Both are systematic and team based. One is called RCA, or Root Cause Analysis, and it looks to the past to learn from mistakes. The second, Failure Modes and Effects Analysis, or FMEA, takes the opposite approach and looks to the future by anticipating errors and putting processes in place to either prevent or mitigate potential harm. However, there are additional tools that can be used to improve safety, such as flowchart analysis or cause-and-effect fishbone diagram.
• #5: Root cause analysis is a structured, problem-solving process used to identify the causal or contributing factors underlying adverse events. In a methodical way, team members consider all possible causes of events (both human and system factors) as potential contributors to the event, rather than attributing the event to the first cause they examine. The premise underlying RCA is that we can learn from our mistakes and the mistakes of others. According to the Agency for Healthcare Research and Quality, the single most important product of an RCA is the detailed account of the events that led up to the incident. The beauty of the RCA is that the tool is based on the premise that one can learn from one’s mistakes. Therefore, although the approach is predominantly retrospective (in that we are trying to find out what went wrong in something that really occurred), changes as a result of an RCA benefit from the prospective analysis of what could go wrong if we institute this fix.
• #6: Human factors to take into consideration include knowledge, skill, and abilities of those involved in the event, including adequacy of training. Procedural factors include the complexity of the procedure itself and any dependencies that exist. For example, if use of a central line safety checklist is required, but the lists are paper based and cannot be found at the time they are needed, then this could have led to an adverse event under review. Use of a central line cart that is stocked with checklists, or the availability of an electronic checklist at the point of care, would be appropriate interventions. Defects or failures in equipment or facilities, work conditions, and communication are also examined.
• #7: RCA, or Root Cause Analysis, is used whenever something really bad happens and we want to understand why so that we can prevent a similar event from occurring in the future. We know that if we dissect the error by developing a chronological description of the who, what, when, where, and how of the situation, we can find out the probable cause and put something in place to prevent the event from occurring again. RCA is based on the belief that we can learn from our mistakes. Let’s take a practical example from everyday life. A 17-year-old was taught how to do laundry, in anticipation of her living in a college dormitory. She returned home one weekend with a total body rash and oily clothes. After taking her to the dermatologist and getting her prescriptions filled, her concerned parent wanted to try to uncover what led to this situation. Obviously, something happened between the time the teenager lived at home and “practiced” doing laundry successfully and the time she returned in misery. She left for school with clear skin, a bottle of unscented Tide, and a suitcase filled with five pairs of jeans and a multitude of tops. What could be done in order to get to the root of this problem? The answer: an RCA.
• #8: Root cause analyses are very structured. In this case, the team assembled for the inquiry included the student, her concerned parent, and the teen’s dermatologist. First, they needed to describe what happened. The student had arrived home with a total-body rash and oily jeans in her suitcase. Next, they had to look at all areas that may have affected the problem. In our example, this included the college dorm’s laundry facilities, the student’s home laundry facilities, and how each of these were used by the teen.
• #9: A diagram was drawn of the steps that the daughter had been taught to use when doing laundry. Here it was important to look at the conditions in her home, with respect to the laundry room. Next, a second diagram was drawn about how she actually did laundry in her home. And finally, a diagram was drawn of the steps she took in the dorm. Again, it was important to also look at the dorm’s laundry facilities. In examining these steps, it was discovered that the dorm washing machine was much smaller than the home machine, was a front loader versus a top loader, required much less detergent and a different concentration of detergent, and had only a single rinse cycle (while the machine at home had an automatic double rinse cycle). The teenager had followed her parent’s instructions to the letter; unfortunately, the equipment was different, and she did not know to alter her procedure in response to these differences. Her clothes retained soap, rubbed against her skin, and caused the rash.
• #10: So what plans could be made? Definitely use less soap, and use the kind that is made for use in front loaders. Set the rinse cycle so that the clothes could be double rinsed. Of course, the responsibility belonged to the daughter and not the parent. The new plan had to happen as soon as she returned to school. The concerned parent would be able to determine if the plan was effective when the student reported her skin status and by examining her clothes on her next visit home.
• #11: Now that we’ve had a practical example, let’s talk about a health care example that you may encounter as an HIT professional. Mrs. A. received blood in the emergency room. Within 15 minutes, she experienced a bad reaction. Her nurse realized that she had received blood intended for another patient. Mrs. A. was transferred to the intensive care unit to be stabilized. The Emergency Department staff wanted to know how this could have happened, so they assembled a team to identify possible causes. As you can see, the ramifications of patient identification errors can be catastrophic in the health care setting. It’s essential that the health care team work together to understand the root causes of errors such as these so that similar events can be decreased in the future.
• #12: So let’s walk through the steps of an RCA for this event. As soon as clinical leaders became aware of the event, they would discuss what happened. Since this is a very serious event, they would decide to do a Root Cause Analysis. To start, they would identify the affected areas or services. In this case, clinicians and blood transfusion services, as well as quality improvement staff or the risk manager would be selected to join the RCA team. In addition, since HIT has shown promise in helping to prevent patient identification errors, as an HIT professional, you might be requested to contribute to this analysis.
• #13: Early on, the team will take a “deep dive” into the details of the event. They will develop flowcharts on how the blood product transfusion patient identification process was designed, how it is usually done in this particular Emergency Department, and how it was done on the day that this event occurred. Invariably, we would find that these three workflow processes differ in some way or another. The team might uncover flawed processes for patient identification. They might also find something wrong with the patient verification process against the blood product. They may uncover that inadequate staffing levels led to workarounds because nurses were too busy to do the usual safety checks. Or that nurses who were managing the blood transfusion were not fully trained in the facility’s blood product transfusion practices.
• #14: The team may find several root causes. Here we can learn lessons from the aviation industry. We can increase the potential for success by prioritizing each cause and selecting interventions, based on their ability to prevent future harm as well as our ability to successfully implement these interventions, given our available resources. Luckily, the team thought to add you, an HIT professional. You might point out the merits of barcode technology for blood product verification. This technology interfaces with both the electronic health record and the transfusion medicine information system. You might be assigned responsibility and a timeline to develop a project plan. As a team, you would all determine a measurement strategy that would show that you have improved the ability to accurately identify patients who receive blood transfusions. As you can see, this is a very effective strategy for learning from our mistakes. But we don’t have to wait for mistakes to happen before we take action. There are other quality improvement tools that anticipate events and try to either stop them from happening or minimize harm if they do happen.
• #15: Failure Modes and Effects Analysis looks to the future in a prospective attempt to predict where errors could possibly, potentially occur. FMEA combines the likelihood of a particular process failure with an estimate of the relative impact of that error to produce a "criticality index." By combining the probability of failure with the consequences of failure, this index allows for the prioritization of specific processes as quality improvement targets. This is one of the tools that is used in the third phase of reliability work.
• #16: During an FMEA, the team analyzes ways in which a process or design can fail, why it might fail, and how it can be made safer or better. The FMEA is based on the premise that we do not have to wait for errors to happen. We can anticipate errors and adverse events, and we can do things to reduce the risk of harm.
• #17: Let’s look at a health care example. After reading several articles about laboratory specimen errors that result in lab tests being done on the wrong patients, doctors at a community office practice decide to examine the potential for this problem to happen in their office laboratory. Let’s walk through the steps of an FMEA.
• #18: The first thing to do is to select a high-risk process. In this instance, it is patient identification but could be any process that affects large numbers of patients, carries a high risk for patients, or has known process problems identified by other organizations. Here, the provider members of this group practice are alerted to a high-risk process by reading articles written by other doctors. They wonder about the risk of this particular problem occurring in their office lab and decide to do a Failure Modes and Effects Analysis. They also wonder whether there is an IT solution that would minimize the risk of laboratory patient misidentification errors in their practice. When assembling the team, you want to include people who are closest to the issue involved, those who are critical to implementing potential changes. You want to have a respected, credible leader, someone with decision-making authority, and people with diverse knowledge bases. So our hypothetical provider group would ask you, as an HIT professional, to participate in this FMEA. They also include the lab supervisor, an office nurse, and two lab technicians.
• #19: The flow diagram starts with CPOE orders entered. It is connected to “labels print in lab area.” This one is connected to “phlebotomist asks patient to state name and birthdate.” This one is connected to “phlebotomist checks patient labels with patient's response.” This is connected to a diamond that reads “correct patient.” If the answer is “yes,” it connects to a box that reads “perform collection.” If the answer is “no,” it connects to a box that reads “follow clinic procedure.” Together, they diagram the process and look for ways in which the process could potentially fail. They identify three places that the process could fail: there could be a problem with the label printing; the phlebotomist could forget to ask the patient to state his name and birthdate, or the phlebotomist could lead the patient to affirm an incorrect identity by phrasing the question to include the incorrect name, “Mr. Jones, right?”; finally, the phlebotomist could fail to check the patient’s lab requisition label with the patient’s response.
• #20: During the hazard analysis, the team may find that most often, phlebotomists forget to request the patient’s name and birthdate. This could lead to inappropriate or missed lab draws. On your advice, they decide to investigate barcode technology as a method to increase the safety of the laboratory specimen identification process in their office lab.
• #21: Both RCA and FMEA are powerful quality improvement tools that can assist teams in understanding errors and implementing improvement actions. RCA looks to past errors for answers, while FMEA anticipates error. RCA asks the questions: how, why, and who; FMEA asks: what if? RCA is predominantly a retrospective dissecting of events that have occurred, while FMEA is a prospective, proactive anticipation of error that could occur in the future. RCA looks for causes and contributors; FMEA looks for possible failure modes. While RCA is subject to hindsight bias, this is not the case with FMEA.
• #22: Another area of interest at the intersection of HIT and patient safety is that of measurement. The National Quality Forum (NQF) led a project to develop a set of recommendations around the measurement of HIT-related safety issues. To accomplish this task, NQF: Conducted an environmental scan of relevant HIT measures and measure concepts; Convened a multi-stakeholder committee to guide and provide input on all phases of the project as well as engage NQF members and public stakeholders at key points throughout the project; Developed a conceptual framework for HIT safety as a way to categorize measure concepts and gaps in measurement; Identified challenges to the measurement of HIT-related safety events and adoption of best practices to strengthen measurement efforts; and Identified and prioritized key measurement areas and potential measure concepts related to HIT safety.
• #23: In addition, the NQF convened a national HIT Safety Committee. Expert members of the committee identified nine key measurement areas for HIT safety, each of which include several measure concepts that could potentially reflect performance in that area, possible data sources or data collection strategies for each measurement topic, as well as the entities that could potentially be held accountable for performance in each area. The final list of key measurement areas is as follows: Clinical decision support. System interoperability. Patient identification. User-centered design and use of testing, evaluation, and simulation to promote safety across the HIT lifecycle. System downtime (data availability). Feedback and information sharing. Use of HIT to facilitate timely and high-quality documentation. Patient engagement. HIT-focused risk management infrastructure. Additional details on each measurement area are available in the full report, which was finalized in February 2016 and is available on the NQF website.
• #24: This concludes Lecture c of Learning From Mistakes: Error Reporting and Analysis and HIT. In summary, we covered two quality improvement tools that can be used to detect and analyze HIT errors — the Root Cause Analysis and the Failure Modes and Effects Analysis. We discussed the Root Cause Analysis as an effective QI tool that looks to the past for the cause of adverse events and seeks to prevent these causes from resurfacing in the future. Failure Modes and Effects Analysis was presented as another equally effective QI tool that uses a different approach; it anticipates adverse events, conducts a hazard analysis, and looks to the future to prevent failure modes with the highest criticality. We did a deeper dive into some of the dimensions of these two analytic methods — discussing the concept of a hazards analysis and aspects of flowcharting. We made the point that the ability to construct and use flow diagrams is an important skill for professionals involved in these processes. With an understanding of QI tools such as RCA and FMEA, health IT professionals can be key contributors to improvement efforts.
• #25: No audio.
• #26: No audio.