What is Hemodynamic Instability

Hemodynamic instability is defined as one or more out-of-range vital sign measurements like low blood pressure or abnormal heart rate (arrhythmias). It can result in chest pain, cold extremities, loss of consciousness, shortness of breath, and even death.

Being a common COVID-19 complication in adult patients, the importance of recognizing patients at risk of hemodynamic instability has hit a more critical point than ever.

Clinicians favor using the most common vital signs of heart rate and blood pressure to diagnosis their patients.

However, they are often busy doing multiple patients at a time and can miss the disease.

Making it one of the most common causes of death in critically injured patients.

But multiple companies are working on AI solutions to prevent that, with Fifth Eye creating “Analytic for Hemodynamic Instability (AHI)” and Clew Med creating “CLEWICU”

Let’s take a dive into the two.

Analytic for Hemodynamic Instability (AHI)

Fifth Eye is a venture capital-backed startup that created the non-invasive clinical tool AHI.

The system uses a single lead of electrocardiogram (ECG) to recognize hemodynamic deterioration well before it is outwardly evident.

According to the company, it’s the first solution to establish real-time streaming data for early onset hemodynamic instability.

A research study in the International Journal of Medical Health Sciences was awarded “Best Paper” at the 2020 International Conference on Health Analytics.

The company is now awaiting approval from the U.S. Food & Drug Administration before marketing it to other health systems.

CLEWICU

In February 2021, the U.S. Food and Drug Administration (FDA) approved “CLEWICU,” an AI-based, first-of-its-kind solution for ICUs.

Developed by Tel Aviv-based CLEW Medical with claims that the system provides notification of clinical deterioration up to eight hours in advance – enabling prompt, proactive patient care.

CLEWICU constantly monitors and categorizes a patient’s risk level, providing doctors with physiological insight into a patient’s likelihood of future hemodynamic instability.

What Does This Mean for Hemodynamic Instability

These two AI-based solutions provide preemptive and potentially life-saving information that reduces alarm fatigue and improves clinical outcomes.

The machine-learning models are trained to identify the likelihood of occurrence of significant clinical events for ICU patients using patient data input from various sources and run through an AI-based algorithm.

The data is analyzed in near real-time to present calculated insights and notifications for dedicated AI models which provide a picture of a unit’s overall status.

Low-risk patients unlikely to deteriorate are also identified, enabling more optimized ICU resource management.

AI, machine learning, and deep learning tools allow physicians to manage the massive amount of patient data, then diagnose health problems like hemodynamic instability, atrial fibrillation and arrhythmia.

This is just one example of the strides being made in the healthcare industry. Check out 3 ways AI is revolutionizing intensive care of critically ill patients

Healthcare isn’t the only area where AI is taking massive strides. Check out our article on how AI can make anyone a pro dancer.

RediMinds are working on numerous exciting AI projects right now, helping businesses and healthcare provides change the face of their industries.

Stay up to date with our latest projects and exciting AI news by joining our monthly newsletter.

The Issues with MELD

Patients on the waitlist who are judged to be at the highest risk of death are identified using current MELD and MELD-Na procedures. However, there is some question about whether this process is effectively and fairly identifying these patients.

For example, there are some gender-based disparities. Gender disparities in liver transplantation have been identified but are poorly understood. Women are 30% less likely than males to undergo liver transplantation, and this discrepancy has grown with the implementation of the MELD-based allocation system. Women are more likely than males to die while waiting for a donor organ in the MELD era.

This research discusses some of the gender disparities. The differences are most likely multifactorial and go beyond height and listing MELD score. One proposed factor of this disparity is a systematic bias in MELD score, which disadvantages women due to lower muscle mass and, hence, serum creatinine. However, no study could show that sex-adjustment of MELD score would eliminate this inequity.

Other factors like the differences in liver disease etiology and progression and the prevalence of hepatocellular carcinoma (HCC) have not been considered. Patients with HCC, typically males, get additional MELD points to help them get lung transplants when their biological score alone does not indicate the urgency of the need for surgery. However, it’s unclear if women’s increased risk of dying while waiting for a liver transplant is due to lack of access or biological factors.

Using Neural Networks to Predict Mortality

RediMinds used neural networks to determine if more effective models could be used alternatively to the MELD and MELD-Na methods to predict mortality rates of waitlisted patients. An artificial neural network is an effort to replicate the network of neurons that comprise the human brain in order for the computer to learn and make choices similarly to a human.

The data was obtained from UNOS, which comprised pre-transplant and post-transplant data on patients registered with OPTN. MELD score is a 90-day severity score for liver transplant patients, so using Neural Networks and a subset of the UNOS data, we used the performance of our 90 Day model against the MELD score for comparison.

After preprocessing data and building Neural Networks, for each outcome, model performance was assessed. The assessments used different subsets of data depending on ethnicity, gender, age, diagnosis, region, and year. To examine the difference in performance, the 90-day mortality prediction model was compared to the MELD score.

The models developed in this study worked well in predicted the mortality of waitlisted liver transplant patients. In fact, the 90-day mortality model significantly outperformed the MELD score in its predictions. The team also built models to estimate the probability of waitlist death at 180 days and 365 days. These other models also performed well.

Cross-validation is essential because it ensures the model is not biased, reliant, and can be used outside of one section or subset of data. To avoid the biases of MELD scores, the finished model is cross validated to check the performance of the model with different age groups, ethnicities, and various other factors.

Implications for the Future

These models can be used to identify individuals on the waitlist who have a high risk of mortality across different time periods and can be used as an alternative to the MELD score since it predicts outcomes with more accuracy. In addition, the methodology, techniques, and tools used to create these models can also be applied to many other surgical outcomes.

To learn more, check out this case study.
 

Appendix:

American Society of Transplant Surgeons (ASTS)

Organ Procurement and Transplant Network (OPTN)

United Network for Organ Sharing (UNOS)

Model for End-Stage Liver Disease (MELD)

MELD-Na (MELD-Sodium)

Hepatocellular carcinoma (HCC)

This isn’t the only way that our research into AI in healthcare is improving patient outcomes. Check out our guide to 6 exciting applications of AI in healthcare.

The Current Problem with HL7 and Electronic Health Data Exchange

Currently, there are several types of software and systems for storing patient data in place in the healthcare industry. Even though HL7v2 is commonly used in these software and systems, many different institutions use different methods for implementing HL7v2. When two systems that support HL7v2 communicate, an analysis procedure is typically involved.

Although patient information can be electronically exchanged via different systems and software using the same HL7v2 standard, it’s often incompatible, hard to read, time-consuming, or generally just confusing. The issue is not due to the usage of various versions of HL7v2 but rather how the standard has been implemented in different systems. With the nature of patient medical charts, this information should be quick and easy to exchange and read when it’s transferred between institutions.

FHIR As a Solution

Following the issues that come with HL7v2, FHIR was developed to help create a standard format of sharing data easily without confusion. FHIR is more standardized in its implementation, meaning that it will redefine how information is stored and exchanged between medical institutions. Overall, FHIR will make the exchange of information very easy since the layout and the organization of information within the file is standardized.

Using a deidentified ICU dataset available for public use, RediMinds created models using the FHIR file format. With the FHIR file format, RediMinds was able to develop models efficiently. Because FHIR data is standardized, data processing for model creation is easier and shorter. It also helps create more sophisticated analytical models.

Creating models with other file formats such as HL7v2 is not as simple since there are different variables and information in the file stored in different ways. On the other hand, with the FHIR file format, developing models becomes backend technology-agnostic since the file is standardized, and data will always be stored in the same fashion. Therefore, no matter the technology and software used by the hospital, the data will always be labeled the same.

Other Benefits of FHIR

There are a few other benefits of storing data in FHIR format. First of all, it is based on modern web standards, making it simple to use in mobile apps, online applications, and data sharing across different platforms. Since FHIR Resources are standardized, it has the same meaning across hospitals. On the contrary, despite HL7v2 being the primary standard of many medical institutions, its implementation varies across software, which causes confusion with the data.

With standardized data, insurance firms may use clinical data to enhance claims data to evaluate risks better, reduce costs, and optimize outcomes. In addition, patients may have more control over their health by obtaining medical information through applications that operate on smart devices. Because of its standard format, it cuts down on the amount of time spent on understanding different hospitals’ data and allows more time for decision-making. Efficiency and speed are both important features due to efficiency being necessary in the case of medical emergencies.

Furthermore, with minimal adjustments, Data Engineering and other activities that are done with FHIR data may be utilized across hospitals. Lastly, unlike HL7v2 and other standards, it is easy for humans to read and interpret without the computer running an analysis on the data.

Implications Of FHIR

Despite FHIR being a better choice for storing and sharing data, adopting FHIR may present challenges for intuition. FHIR was designed to simplify implementation but implementing an in-house FHIR implementation may mean acquiring new skills. Developing an interface or app is only the first step in the process. In the future, a team will need to maintain and grow a backend FHIR service, which may require extra skill in managing additional infrastructure.

Developing a FHIR-based cloud service may aid in overcoming adoption barriers. With the appropriate cloud implementation, you can speed up provisioning, personalize the implementation by integrating cloud-based services, and achieve adequate scalability for the future. Furthermore, the standardized file format of FHIR supports the development of AI solutions or advanced analytical solutions.

To learn more, check out this case study.