Protocol Failure Detection: The Conflation of Acute Respiratory Distress Syndrome, SARS-CoV-2 Pneumonia and Respiratory Dysfunction.

In medicine, protocols are applied to assure the provision of the treatment with the greatest probability of success. However, the development of protocols is based on the determination of the best intervention for the group. If the group is heterogeneous, there will always be a subset of patients for which the protocol will fail. Furthermore, over time, heterogeneity of the group may not be stable, so the percentage of patients for which a given protocol may fail may change depending on the dynamic patient mix in the group. This was thrown into stark focus during the severe acute respiratory syndrome-2 coronavirus (SARS-CoV-2) pandemic. When a COVID-19 patient presented meeting SIRS or the Berlin Criteria, these patients met the criteria for entry into the sepsis protocol and/or acute respiratory distress syndrome (ARDS) protocol, respectively and were treated accordingly. This was perceived to be the correct response because these patients met the criteria for the "group" definitions of sepsis and/or ARDS. However, the application of these protocols to patients with SARS-CoV-2 infection had never been studied. Initially, poor outcomes were blamed on protocol noncompliance or some unknown patient factor. This initial perception is not surprising as these protocols are standards and were perceived as comprising the best possible evidence-based care. While the academic response to the pandemic was robust, recognition that existing protocols were failing might have been detected sooner if protocol failure detection had been integrated with the protocols themselves. In this review, we propose that, while protocols are necessary to ensure that minimum standards of care are met, protocols need an additional feature, integrated protocol failure detection, which provides an output responsive to protocol failure in real time so other treatment options can be considered and research efforts rapidly focused.

Artificial intelligence systems for complex decision-making in acute care medicine: a review.

The integration of artificial intelligence (AI) into acute care brings a new source of intellectual thought to the bedside. This offers great potential for synergy between AI systems and the human intellect already delivering care. This much needed help should be embraced, if proven effective. However, there is a risk that the present role of physicians and nurses as the primary arbiters of acute care in hospitals may be overtaken by computers. While many argue that this transition is inevitable, the process of developing a formal plan to prevent the need to pass control of patient care to computers should not be further delayed. The first step in the interdiction process is to recognize; the limitations of existing hospital protocols, why we need AI in acute care, and finally how the focus of medical decision making will change with the integration of AI based analysis. The second step is to develop a strategy for changing the focus of medical education to empower physicians to maintain oversight of AI. Physicians, nurses, and experts in the field of safe hospital communication must control the transition to AI integrated care because there is significant risk during the transition period and much of this risk is subtle, unique to the hospital environment, and outside the expertise of AI designers. AI is needed in acute care because AI detects complex relational time-series patterns within datasets and this level of analysis transcends conventional threshold based analysis applied in hospital protocols in use today. For this reason medical education will have to change to provide healthcare workers with the ability to understand and over-read relational time pattern centered communications from AI. Medical education will need to place less emphasis on threshold decision making and a greater focus on detection, analysis, and the pathophysiologic basis of relational time patterns. This should be an early part of a medical student's education because this is what their hospital companion (the AI) will be doing. Effective communication between human and artificial intelligence requires a common pattern centered knowledge base. Experts in safety focused human to human communication in hospitals should lead during this transition process.

Combined endothelial and epithelial barrier disruption of the colon may be a contributing factor to the Ebola sepsis-like syndrome.

After an initial febrile viral syndrome, infection with Ebola virus often induces an explosive late "Ebola sepsis-like syndrome" which appears very similar to some phenotypes of bacterial sepsis and is commonly fatal. It is possible that direct and diffuse viral infection of both the endothelium and epithelium of the colon may cause sufficient disruption of both the endothelial and epithelial barriers to induce exposure or leakage of endotoxin and bacterial antigens to, or into, the vascular system precipitating or exacerbating the Ebola sepsis-like syndrome. If colonic barrier disruption or vascular exposure of bacterial antigens from the colon is found to comprise an important mechanism of the Ebola sepsis-like syndrome, protocolized timed decontamination of the bowel with or without timed prophylactic antibiotics might warrant investigation.

Dynamic diagnostic relationism: a new diagnostic paradigm for complex rapidly changing clinical conditions.

Decades of large, apparently well-designed clinical trials have failed to generate reproducible results in the investigation of many complex rapidly evolving and changing conditions such as sepsis. One possibility for the failure is that 20th century threshold science may be too simplistic to apply to complex rapidly changing conditions, especially those with unknown times of onset. There is an acute need to reconsider the fundamental validity of the application of simple threshold science in the study of complex rapidly evolving and changing conditions. In this letter, four potential axioms are presented which define a new science which assesses the probability of disease as a function of motion images of all the available clinical data.

The diagnosis of sepsis revisited - a challenge for young medical scientists in the 21st century.

In 1991, a well-meaning consensus group of thought leaders derived a simple definition for sepsis which required the breach of only a few static thresholds. More than 20 years later, this simple definition has calcified to become the gold standard for sepsis protocols and research. Yet sepsis clearly comprises a complex, dynamic, and relational distortion of human life. Given the profound scope of the loss of life worldwide, there is a need to disengage from the simple concepts of the past. There is an acute need to develop 21st century approaches which engage sepsis in its true form, as a complex, dynamic, and relational pattern of death.

Patterns of unexpected in-hospital deaths: a root cause analysis.

BACKGROUND: Respiratory alarm monitoring and rapid response team alerts on hospital general floors are based on detection of simple numeric threshold breaches. Although some uncontrolled observation trials in select patient populations have been encouraging, randomized controlled trials suggest that this simplistic approach may not reduce the unexpected death rate in this complex environment. The purpose of this review is to examine the history and scientific basis for threshold alarms and to compare thresholds with the actual pathophysiologic patterns of evolving death which must be timely detected. METHODS: The Pubmed database was searched for articles relating to methods for triggering rapid response teams and respiratory alarms and these were contrasted with the fundamental timed pathophysiologic patterns of death which evolve due to sepsis, congestive heart failure, pulmonary embolism, hypoventilation, narcotic overdose, and sleep apnea. RESULTS: In contrast to the simplicity of the numeric threshold breach method of generating alerts, the actual patterns of evolving death are complex and do not share common features until near death. On hospital general floors, unexpected clinical instability leading to death often progresses along three distinct patterns which can be designated as Types I, II and III. Type I is a pattern comprised of hyperventilation compensated respiratory failure typical of congestive heart failure and sepsis. Here, early hyperventilation and respiratory alkalosis can conceal the onset of instability. Type II is the pattern of classic CO2 narcosis. Type III occurs only during sleep and is a pattern of ventilation and SPO2 cycling caused by instability of ventilation and/or upper airway control followed by precipitous and fatal oxygen desaturation if arousal failure is induced by narcotics and/or sedation. CONCLUSION: The traditional threshold breach method of detecting instability on hospital wards was not scientifically derived; explaining the failure of threshold based monitoring and rapid response team activation in randomized trials. Furthermore, the thresholds themselves are arbitrary and capricious. There are three common fundamental pathophysiologic patterns of unexpected hospital death. These patterns are too complex for early detection by any unifying numeric threshold. New methods and technologies which detect and identify the actual patterns of evolving death should be investigated.