Buzzwords and Trends
Every year Healthcare Information and Management Systems Society (HIMSS) Conference and Exhibition has different buzzwords and promising trends in healthcare. Year 2015 was "Big Data analytics", while year 2016 was "population health management". This year it was the next phase of the analytics evolution: "predictive analytics, precision medicine, machine learning, anomaly detection and artificial intelligence" to augment clinical decision making. This can change the nature of healthcare from reactive to proactive and preventive.
Precision medicine, also known as personalized medicine, requires huge volumes of data and enormous computer processing power to sift through the masses of information available. Precision medicine redefines our understanding of disease onset and progression, treatment response, and health outcomes. Machine learning, anomaly detection, and artificial intelligence are the next natural evolution of predictive analytics. For example, the Real-time Outbreak and Disease Surveillance (RODS) system is a public health surveillance system for early detection of disease outbreaks using Health Level 7 (HL7) message data from clinical encounters in real time.
Big Data or Small Precise Data
Are population health management and precision medicine complements or competitors? In order to improve the health of patients across the care continuum, do we think big or think very, very small? We can think of very small data as a progression of Big Data: to get to the small precise data we first need to analyze the Big Data.
There has been a wide implementation of electronic health records (EHRs) in 95% of U.S. hospitals and use of healthcare IT to electronically store, retrieve, and analyze clinical information. There are disparate pieces of data in the EHR about each patient's condition and lab results, making it harder to use patient data for anomaly detection. The next healthcare IT transformation is the change from an encounter-based delivery of care to a "value-based care" model, which is a risk-based model with reimbursements increasingly linked to patient outcomes. Patient-generated health data (PGHD) by health apps, activity trackers, and connected health devices will be part of the "value-based care" or "fee-for-value" increasing the patient engagement. In the future, healthcare IT platforms will augment pathological and radiological diagnosis and assessment of cancer treatment.
Making Sense of the Trove of Patient Data
The next question is how do we make sense of the immense patient data that is flowing real time and simplify it so that the end users can easily understand and be able to come up with proactive prevention plans. To best achieve patient outcomes there is an urgent need for the healthcare professionals to be alerted to health issues that nobody yet knows the patient has. There is a tremendous potential in using the patient data to learn real time about chronic disease like Type 2 Diabetes (T2D), healthcare-acquired infection (HAI) or nosocomial infection like Septic shock, infectious diseases like Lyme Disease, and water poisoning like lead poisoning.
Below are a few real life examples of how healthcare will benefit if there was a healthcare IT platform that would detect and alert real time the diagnostic findings as soon as an anomaly in the lab test results is detected and also create a chronic disease risk score.
In many cases T2D is detected when it is a full-blown disease. T2D should never be a surprise. There are many biomarkers that when tracked can show the high risk of developing T2D years before T2D is diagnosed. More than 29 million Americans have T2D and more than 80 million are pre-diabetic, and it is now the third leading cause of death.
Septic shock which kills nearly 40% of the 750,000 people it affects each year, if diagnosed early can save lives. There is currently no effective therapy for septic shock and getting a diagnosis early on can make a huge difference.
Detecting infectious diseases like Lyme Disease before overt symptoms manifest. Early diagnosis and treatment improves chances of full recovery. According to the CDC, Lyme Disease is present in more than 260 counties in the northeastern United States and the true burden of Lyme disease in the U.S. is about 300,000 cases.
Detecting a water contamination crisis like Flint, Michigan by analyzing patient data to discover unsafe levels of lead in blood as soon as high lead observations are seen. In fact, the Flint lead drinking water contamination crisis was discovered by a provider using the blood lead test results from the EHR and seeing an emerging pattern of high lead blood values.
Healthcare IT Platform
Based on many of the real life cases in healthcare there is a clear need for a healthcare IT platform that can detect and flag specific biomarkers early on before any overt signs are seen and before anyone even suspects something is wrong. What if an IT platform that works with Big Data and generates precise small data, detects anomalies and generates alerts would be soon available? That would be just what healthcare IT needs.
ExtraHop will soon be releasing Addy, a cloud IT operations monitoring service. Addy applies machine learning to Big Data for anomaly detection and provides real time IT analytics. ExtraHop can track different types of operational data and uses machine learning models in the cloud to quickly detect data that seems out of step with the norm.
Modern healthcare is increasingly dependent on digital and information technology, which needs an extensive and complex network. Virtual desktop infrastructure (VDI) is essential for healthcare providers to provide patient care. If nurses and doctors are not able to quickly login to their virtual desktops, longer wait times directly affect patient care. Troubleshooting VDI performance problems in large enterprise environments is complicated for the IT Operations or the Citrix team. Many issues on the network may be responsible, like slow Active Directory server, DNS server, Database server, file share, etc. Here is where Addy can help as soon as Addy becomes available, since Addy will be collecting and analyzing many metrics that affect the network server and client latency, and applications like Citrix ICA, Database, LDAP, Kerberos, HTTP, CIFS, FTP, DNS.
ExtraHop Addy is coming soon in April 2017, this is just the beginning with endless possibilities to come in the future that will augment healthcare needs!