Sepsis occurs when the body's response to an infection causes a cascading inflammatory reaction throughout the body. Sepsis can affect multiple organ systems and is one of the most complicated syndromes clinicians face. Despite improvements in technology and care protocols, sepsis continues to occur in American healthcare settings at epidemic levels, with more than 1.5 million U.S. residents developing the condition each year according to the CDC. About 250,000 patients with sepsis die each year, and one in three people who die in the hospital have sepsis.
This content is sponsored by Amplifire
Allied against sepsis
Amplifire Healthcare Alliance—a provider of outcome-based learning solutions for health systems—sought to improve sepsis care among its members by identifying and addressing clinical misinformation held by clinicians across healthcare. Alliance members include Aurora, Colo.-based UCHealth and Children's Hospital Colorado, Salt Lake City-based Intermountain Healthcare, Boston-based Partners HealthCare and the Duke Infection Control Outreach Network, among others.
Amplifire enlisted 23 healthcare organizations to recruit physicians and advanced practitioners from across disciplines, including hospital medicine, emergency medicine, and critical care medicine. From mid-August through early October 2017, participating clinicians took an accredited, evidencebased, 31-question sepsis training course on Amplifire's learning platform. Amplifire measures not just knowledge, but confidence in that knowledge. Among 1,245 participants, the platform identified 7,795 instances when clinicians answered questions about sepsis confidently but were wrong. By the end of the course, all of this misinformation had been eliminated.
Bob Burgin, Amplifire's CEO, spoke with Becker's about the Amplifire Healthcare Alliance and its sepsis initiative. Mr. Burgin has a long history in healthcare, including building and taking public the largest geriatric rehabilitation facility chain in the United States. Bob is one of the founding members of the Amplifire Healthcare Alliance, a collaboration of several of the largest and most advanced hospital systems in the United States. Through the use of Amplifire and other learning tools that promote effectiveness and efficiency, Bob unites members' commitment to improving patient care and reducing the incidence of avoidable harm.
Note: Responses have been lightly edited for length and clarity.
Question: This was the largest study of its kind. Why was it done?
Bob Burgin: The Amplifire Healthcare Alliance is a group of large health systems that have come together to try and solve complex clinical problems currently facing healthcare. The idea of the alliance is to use a common platform to build high-end content and conduct clinical interventions at various membership locations. Instead of standard CME compliance training, our advanced platform focuses on knowledge engineering around clinical problems that can profoundly impact financial reimbursement for health systems. Unlike most training platforms, Amplifire incorporates highly evolved data analytics to provide insight into the problem areas. We blind and benchmark the data for the benefit of all the members and share the full studies across the Alliance. A total of 23 healthcare organizations participated, and the results were sent to all members.
Q: What did you learn from this clinical intervention effort?
BB: Sepsis is a unique beast in that it's complex to diagnose and the guidelines have been rapidly evolving, so not surprisingly there's a high degree of uncertainty about the right thing to do, as well as a high degree of misinformation among physicians. If you look more closely at the clinical issues, the study identified some meaningful problem areas.
For example, one of the guidelines for treating sepsis is rapid administration of fluids. Fluid resuscitation is associated with a variety of side effects, because fluids move out of the bloodstream and into body cavities, where they can put pressure on organs. This effect is dramatically increased in septic patients, whose inflamed circulatory systems are extremely leaky. In one of our findings, roughly two-thirds of clinicians believed that a fluid bolus dissipates into the body over two hours or four hours. Actually, the process takes less than an hour and for severely septic patients it can be as short as 10 minutes. If you're monitoring organ failure that may be exacerbated by fluids, it matters how long it takes for fluids to flood the body compartments. This was one of about six clinical findings where two-thirds of clinicians had incorrect knowledge that created risk for patients.
Other findings were also notable. Patients over the age of 85 are 30 times more likely to get sepsis than patients under age 65. Clinicians strongly underestimated this difference, believing the risk was one half or one quarter what it really is. This can lead to a form of bias in which incorrect knowledge about the prevalence of a disease can make them slow to suspect sepsis. Another finding is related to the source of infection. A little over half of clinicians thought the most common infection source for sepsis was the urinary tract. Actually, the source in 50% of cases is the lungs, with the urinary tract implicated in just 10% of cases. In sepsis, finding and eliminating the source of infection quickly is critical. That includes knowing where to start looking.
Q: The findings indicated wide variance in the type and amount of misinformation among clinicians. How does this relate to clinical variation in healthcare organizations?
BB: Brent James, MD, former chief quality officer of Intermountain Healthcare and the inspirational leader in the movement that produced the landmark book To Err is Human:
Building a Safer Health System, helped us in the early days at Intermountain in developing the mission and scope of this effort. Something he said early on has proven accurate—aggregated data in complex healthcare environments has a very high propensity to hide the problem. If you use a bar graph to represent data across all 23 healthcare organizations, they look like they all have about the same level of problems, but if you drill down within a single unit on the same floor of the same hospital, there are significant disparities in knowledge between individual clinicians. Something this platform provides that others don't, is groundbreaking analytics that transform data into patterns that matter and action items that benefit individual clinicians.
One of the most common and correctable causes of patient harm and financial loss is confidently held misinformation in the minds of clinicians. The purpose of the healthcare alliance is to find and fix that problem at a granular level. To be clear, the information that we expect physicians to hold in their brains is overwhelming. They are real heroes on the front lines, and we have an obligation to provide them with highly focused and targeted knowledge engineering tools to help them succeed.
Analytics can also reveal systemic problems. Another training we did found that a single care location performed very poorly in one area. The data looked so anomalous that we sent someone to see what was going on. Sure enough, we found that they were cleaning a piece of equipment following instructions for the wrong machine. Their confidently held misinformation was actually a signal for a systemic or environmental problem that was easy to fix once identified by the system.
Q: What actions can infection control leaders take to improve sepsis outcomes?
BB: The biggest thing that is missing—the reason why this alliance exists in the first place—is the infection control leader must acknowledge they're dealing with variations of misinformation and knowledge in individual caregivers. You have to start with that assumption and deploy training and clinical knowledge engineering efforts that are adaptive to individuals who may struggle on one topic while performing beautifully on another. We have to deploy knowledge engineering initiatives that are adaptable at the individual level, which lets us aggregate data, which then leads to actionable interventions. You see systems spending a lot of money to solve sepsis, but they're treating it as if it's a single problem across the organization, and that's not how humans work. One provider needs a completely different intervention than another. What our alliance does and what we tell new members when they come in is that the first step is realizing that you're dealing with a diverse set of problems linked to knowledge gaps unique to each individual. Your clinical knowledge engineering has to start with that assumption.
Q: Are these findings transferable to other healthcare problems?
BB: At the aggregate level the findings are highly transferable, but the clinical findings in this study are unique to sepsis. We have about 60 clinical interventions comparable to the sepsis initiative. At Intermountain, we've built CAUTI and CLABSI interventions. We're finalizing a C. diff intervention at Partners HealthCare with Mass General. We've rolled out a safe injection practices course through the Duke Infection Control Outreach Network. The findings are all unique. For example, the findings for safe injection practices suggest that on-the-job training propagates misinformation about injections from one person to the next in a very scattered and unpredictable way. The consistent variable here is that this type of clinical intervention is highly applicable to virtually every clinical topic. A year ago we had two interventions; now we have 60 and all of them are built by the member health systems. The overall approach is intensely scalable.
A growing alliance
To help address the myriad clinical and fiscal challenges facing the American healthcare system in the era of value-based care, Amplifire's leadership and founding members decided to expand the reach of the organization for the purpose of advancing efforts to combat clinical misinformation.
"We've made the decision to open up the alliance nationwide, so at this point any health system in the country is welcome to participate in these clinical knowledge engineering initiatives," Mr. Burgin said. "This is the only alliance of its kind and the broader the footprint we have across the country, the better."
The sepsis intervention white paper and sample findings are available as a free download here.
More articles on infection control:
New Jersey physician's license suspended after accusations of reusing anal catheters
CDC investigates 13-state E coli outbreak: 5 things to know
3 pediatric flu deaths reported as flu season continues to intensify: 5 things to know