Rethinking How to Measure "Risk" in Healthcare - A more technical post for those evaluating risk scores for care management, coupled with a real-world example.

Risk Scores in Clinical Care - You're Not From Around Here Are You? - A quick history of how Risk Scores developed and why they are often out of place in care management.

The Dangers of Claims Based on Claims - A quick overview of why healthcare needs to use more than Claims data for good decisions.

What Datat Scientists Need to Learn to Work in Healthcare - a few questions and insights for aspiring healthcare data scientists.

Hey Machine Learning...if That's Even Your Real Name - How the overhyping of AI is distracting people from the real opportunities to improve healthcare with machine learning.

Dartmouth Researcher Detect Falls Without the Data Scientists - A short story of how Cyft is empowering researchers to detect falls in inpatient notes.

When Life Gives You MACRA - Is it really the specific policies that make healthcare reform so challenging or the fact that we're trying to implement them all at once? [article - The Healthcare Blog]

Why We Created Cyft - Cyft's first company post explaining why we created a for-profit predictive analytics company.

Data Thinking in Healthcare - A framework for thinking through how to put data to work improving clinical care and operations. [article - The Healthcare Blog]

Healthcare Startups - Why Now and So What? - Why the sudden dramatic increase in interest in healthcare by both startups and retail orgs and what does it mean for the field? [article - The Healthcare Blog]

Health IT: How Developing Markets Can Trump US - With the US Market Hostile Toward Health IT Innovation Today, Entrepreneurs Should Look Toward the Developing World for Markets and Opportunities. [article - InformationWeek]

Why We Need Design Thinking in Healthcare - Healthcare, to date, has been largely "engineered," not designed. Designers begin by understanding how people work in the real world, and then create the best IT system that's technically feasible. [article - InformationWeek]

Healthcare's Deadly Data Problem Until our information systems are designed to measure and improve care we will continue to have to guess the answer to many important questions, including how many people we kill each year by accident. [article - The Healthcare Blog]

Electronic medical records at a crossroads; Impetus for change or missed opportunity?  After decades of stunted growth the adoption rates for electronic medical records has finally begun to take off.  But will the US government's enormous investment pay off?  Not if certain criteria aren't demanded by customers and delivered by vendors. [manuscript - JAMA]. 

Medicine's 4th Paradigm - Healthcare is about to undergo a major paradigmatic shift that will affect all of its stakeholders.  Read more about what's happening and the implications. [blog post] 

Comparative Effectiveness Research and Medical Informatics - What is this field known as "medical informatics" and what is its role in healthcare?  A simple use case helps present a model of the areas of focus for the field.  Although geared toward its contribution to a specific type of research known as "comparative effectiveness research," this is the model I use to design informatics curriculum, select projects, staff teams, etc.  [manuscript - AJM].  

What Exactly is Personalized Medicine and Why Should I Care? - Anything ending in 'omics' seems to be quite popular these days.  What is genomics really and just how close are we to turning the hype into actual improved care? [blog post]

The $100 Genome: Preparing for the Personalized Medicine Revolution - All technologies have a potential tipping point - the point at which the product becomes ubiquitous in the market.  For molecular medicine (e.g., genomics, proteomics, etc) to reach the promise of "personalized medicine" many speculate that a price point of $1,000 for the sequencing of the entire genome will push molecular medicine over that tipping point.  How soon / realistic is this? [blog post]

Detours on the Road to Personalized Medicine - The use of an individual's biology to determine how best to keep that individual healthy is going to require some pretty drastic rethinking of how medicine is conducted.  We outline specific areas that the medical industry will be forced to reconsider in order to get a dialog started.  [manuscript - JAMA]

Why Sequencing the Genome Shouldn't Explain Disease - Some have argued that genomic sequencing has been somewhat of a disappointment because early scientific forays do not explain disease in the way that we had hoped.  Genomic sequencingshouldn't explain disease.  But that doesn't mean we're not about to undergo a medical and scientific revolution as a result of this work. [blog post]

Barrier to Personalized Medicine #1: The Free Rider Dilemma - The amount of genomic information demands that discovery of appropriate interventions based on biology is dependent on enormous numbers of subjects participating in research.  Unfortunately, while everyone wants access to the latest and greatest treatments, not everyone expects to contribute to research. [blog post] 

Personalized Medicine's Next Frontier - Next Gen Phenotyping - When the cost of a whole genome sequence hits $1,000 the barrier to personalized medicine will no longer be genotyping .  Instead success in delivering the right intervention to the right person based on biology will rely on accurately and comprehensively characterizing the patient and disease population, otherwise known as phenotyping.  Why this will be no small feat and what we can do to prepare. [blog post]

Lessons Learned Building the Infrastructure for the Million Veteran Program - For the last five years I had the great fortune of leading the information technology portion of what has been called one of the "5 Most Innovative Government Big Data Projects." Here I describe some of the lessons learned in doing so. [article - Information Week]

6 Questions to Guide Natural Language Processing Strategy - Clinical NLP is gaining the attention of vendors, CIO's, and hospitalists.  This commentary lays out important considerations any organization should think through before investing. [article - Information Week Health]

A Point-of-Care Clinical Trial - We embedded a clinical trial in the VA's electronic medical record system.  As a result, we're answering important clinical questions using the power of randomization at the cost of an observational study.  The reason for this approach and the design of the first POC clinical trial is described. [manuscript - J of Clinical Trials]. 

Implementation of the First Point-of-Care Clinical Trial - In an informatics journal I describe in more detail how exactly we incorporated a clinical trial into the VA's medical record system.  Also described are the considerations one must keep in mind in attempting to combine clinical and research systems in such ways. [manuscript - JAMIA].

Business Intelligence ≠ Healthcare Intelligence  - A response to the growing need for quality measurement and improvement is the use of business intelligence products and methods.  While learning from data is a must it's important to understand the fundamental differences between the natures of business versus healthcare intelligence. [article - Health 2.0 news]

The Surprising Story of the Original Luddite - 27 hours of travel from India gives one plenty of time to research life's great mysteries.