Blog Viewer

What if you could ask your favorite voice-activated device these questions:

• “How many Adverse Drug Events (ADE) were associated with warfarin vs. Direct Oral Anticoagulants (DOACs) in our organization? … How do our ADEs compare with the national average?”
• “How do the ADEs compare between patient specific insulin syringes made in pharmacy versus using multi-dose vials in nursing units?”
• “What are the most common errors in the medical/surgical units across our system?”

Imagine you are shopping online for beauty products, electronics, or anything else you normally buy online.  Using various filtering tools, virtual carts, and confirmation steps online allow you to quickly and accurately order what you’re looking for; if you selected a laptop,  accessories such as a mouse, mouse pad, headphones, and other related items are recommended based on what others have ordered in the past. Many have dabbled or even embraced using voice controlled ordering (e.g. “OK Google” or “Alexa”), without even having to log into a cell phone or computer.  In 2017 alone, the United States Postal Service shipped approximately 5.7 billion packages, with the large majority of the packages arriving to the right people and places around the right times ^1,2.

In 2016, there were approximately 4.4 billion prescriptions written ³.  Overall the healthcare system has gotten better with getting the right medication or treatment to the right patient. In the past 10 years, the increase of Electronic Medical Record (EMR) adoption (at least tripling since 2009) decreased the total number of errors due to handwriting or transcription.  Secure electronic systems for controlled substances transactions may eliminate the risk of prescription tampering. There is a lot of potential with increasing safety in healthcare by applying many of the same tools and technology as the online ordering process.

Internet of Things (IoT) usage is estimated at 37 million consumers and is rapidly growing; Boston Children’s has used IoT integration to allow providers to ask such questions as “Who is the charge nurse on 7 South?”, and also answer basic questions such as “What is a formulary?” for patients.  The use of IoT is allowing patients to know Urgent Care and ED visits wait times for nearby facilities and make their choice for care accordingly. There is tremendous potential in facilitating higher quality care both inside organizations as well as in patients’ homes. Perhaps sooner than later patients will be able to say “Order all the supplies and medications I need for my colonoscopy [and have it shipped to me]”.

Blockchain technology for healthcare organizations is analogous to companies’ ability to share de-identified customer information to other companies for targeted marketing purposes.  This allows for a secure but common denominator to compare similar patients, thus enabling better regulations and guideline development. Imagine being able to analyze and trend past medical histories for patient profiles (without patient identifiers) with a rare disease state at a national or even global level - this is the theoretical benefit of blockchain, which is part of the foundation needed by many of the other technologies mentioned in this blog.

Highly accurate prescriptive analytics integrated into daily workflows is the holy grail of big data, data mining, and deep learning.  For example, a provider that starts ordering a DOAC would be prompted with a recommendation to order warfarin instead, due to previous history of non-compliance with high cost medications and adverse effects from another DOAC prescribed at a different hospital.  

Big data ⁴

• The premise of “big data” is the availability of large amounts of clinical data that allows for analysis of associations, patterns, and trends that may not be as easily discernible with limited data.

Analytics

• Extensive use of data, statistical and quantitative analysis, explanatory and predictive models, and fact-based management to drive decision to actions.
• Types of Analytics:
• Descriptive ^5,6
• The examination of data or content, usually manually performed, to answer a question.
• Use in HealthCare:
• Describing current situation and problems.
• For example, describing a heart failure patient population.
• Predictive ^6,7
• Encompasses a variety of statistical techniques from predictive modeling, machine learning, and data mining that analyze current and historical facts to make predictions about future or otherwise unknown events.
• Use in HealthCare:
• Identifying trends and outcomes of actions taken.
• Predicting patients at risk for readmissions.
• For example, predicting outcomes in heart failure patients.

• Prescriptive ^6,8
  • Prescriptive analytics combines the application of mathematical and computational sciences with the results of descriptive and predictive analytics to suggests decision options.
  • Use in HealthCare:
    • Optimizing clinical, financial, and other outcomes.
    • For example, recommending the best care plan for heart failure patients (taking patient specific information into consideration).

Data mining ⁹

Using methods that intersect with machine learning, statistics, and database systems to discover patterns in large data sets.

Artificial Intelligence (AI)¹⁰

• Ability to simulate intelligent human behavior.
• Machine learning ¹¹
• Core AI methodology that builds systems and algorithms that learn from data.
• Machine learning relies on neural networks (a computer system modeled on the human brain).
• By utilizing multilevel probabilistic analysis, such applications allow computers to simulate and even expand on the way the human mind processes data.
• As a result, not even the programmers can be sure how their computer programs will derive solutions.
• Deep learning ¹²
• Software utilizes deep learning to recognize patterns in distinct layers.
• Use in HealthCare:     
  • Diagnosis using imaging data.
  • This mechanism is becoming increasingly useful. It allows the different layers within the system to work both independently and collectively, for example separating different aspects such as color, size, shape and then integrating the outcomes.
  • These newer visual tools hold the promise of transforming diagnostic medicine and can even search for cancer at the individual cell level.

Here is a link to a short video explaining the difference between three key concepts of artificial intelligence vs machine learning vs deep learning.

Blockchain ^13,14

Blockchain is a peer-to-peer (P2P) distributed ledger technology for a new generation of transactional applications that establishes transparency and trust. The blockchain design consists of three main gears: shared ledger, distributed network, and digital transactions.

• Use in HealthCare:
• Blockchain technology has the potential to address the interoperability challenges currently present in health IT systems and be the technical standard that enables individuals, health care providers, health care entities, and medical researchers to securely share electronic health data.

Internet of things (IoT) ¹⁵

• Internet of thing is a concept reflecting a connected set of anyone, anything, anytime, anyplace, any service, and any network.
• Use in HealthCare:
• IoT in healthcare is a heterogeneous computing, wirelessly communicating system of apps and devices that connect patients and health care providers together to diagnose, monitor, track, and store vital statistics and medical information.  
• Compliance with treatment and medication at home and by healthcare providers.
• Various medical devices, sensors, and diagnostic and imaging devices can be viewed as smart devices or objects constituting a core part of the IoT.

References:

1-        Alexa gets a bigger say in healthcare. Retrieved from: https://www.digitalcommerce360.com/2018/03/14/alexa-gets-bigger-say-healthcare/. Accessed 5/13/18

2-        Postal Facts: A Decade of Facts & Figures. Retrieved from: https://facts.usps.com/table-facts/. Accessed 5/13/18

3-        Total number of medical prescriptions dispensed in the U.S. from 2009 to 2016 (in millions). Retrieved from: https://www.statista.com/statistics/238702/us-total-medical-prescriptions-issued.  Accessed 5/13/18

4-        Davenport TH, Competing on Analytics, 2007, Harvard Business School Press

5-        Descriptive Analytics. Retrieved from: https://www.gartner.com/it-glossary/descriptive-analytics/. Accessed 5/08/18

6-        Healthcare Big Data Analytics: From Description to Prescription. Retrieved from: https://healthitanalytics.com/news/healthcare-big-data-analytics-from-description-to-prescription . Accessed 5/08/18

7-        Predictive Analytics. Retrieved from: https://www.gartner.com/it-glossary/predictive-analytics/. Accessed 5/08/18

8-        Prescriptive analytics. Retrieved from: https://en.wikipedia.org/wiki/Prescriptive_analytics .  Accessed 5/09/18

9-        Data Mining. Retrieved from: https://en.wikipedia.org/wiki/Data_mining. Accessed 5/09/18

10-     Artificial Intelligence in Healthcare: Separating Reality from Hype https://www.forbes.com/sites/robertpearl/2018/03/13/artificial-intelligence-in-healthcare/#6def35d61d75. Accessed 5/09/18

11-     Machine Learning. Retrieved from: https://en.wikipedia.org/wiki/Machine_learning . Accessed 5/10/18

12-     Deep Learning. Retrieved from:  https://en.wikipedia.org/wiki/Deep_learning . Accessed 5/10/18

13-     This Is Why Blockchains Will Transform Healthcare. Retrieved from: https://www.forbes.com/sites/bernardmarr/2017/11/29/this-is-why-blockchains-will-transform-healthcare/#5aaf52b31ebe. Accessed 5/11/18

14-     Blockchain For Health Data and Its Potential Use in Health IT and Health Care Related Research. Retrieved from:  https://www.healthit.gov/sites/default/files/11-74-ablockchainforhealthcare.pdf  Accessed 5/11/18

15-     The Internet of Things for Health Care: A Comprehensive Survey. Retrieved from: https://ieeexplore.ieee.org/document/7113786/ Accessed 5/11/18