Category: Big Data

Thinking About Algorithms

Take 7 minutes of your day and read this “primer”  on algorithms. Models that underlie Wall Street investing, Amazon advertising, disease diagnosis and even park our cars. Their role in our lives is getting bigger and bigger, so it is important to spend a couple of minutes to think about their strengths and their weaknesses.

Their primary strength is efficiency. They operate quickly, with little or no human effort involved in making them sing and dance. They can incorporate huge quantities of information seamlessly and eliminate the systematic “biases” to which Behavioral Economics introduced us.

BUT. Garbage in, garbage out! Many algorithms, for example, are programmed for men, leaving women by the wayside. As the author points out, algorithms can actually teach social biases by automating  disparity.  

Bottom Line. The message of this little piece is clear. The best outcomes are produced not by unchecked human behavior or by algorithms running thoughtlessly, but rather through a thoughtful combination of Artificial Intelligence and human judgement.

Think about it!

In Primary Care, “Data” Is A Four-Letter Word

Check this out.  What you will see is a brilliant article about how data should, and shouldn’t, be used in primary care. Pretty simple actually.  The three key principles that the article elucidates are shown here.  They include:

Data should be collected to inform doctors and their patients, not to pander to some MBA bureaucrats who want to tell physicians how to practice medicine. 

Similarly, data should never be allowed to get in the way of the flow of treatment. They should enhance treatment, rather than decrease its efficiency.

Finally, data should be treated as treatment information only. Ancillary to the attainment of a treatment goal.  To the extent that hitting a metric becomes the goal, it is time to rethink the role that the data are serving.

Example of all of this? Colon cancer screening. Virtually every eligible patient claims that they will comply with a directive to get screened for colon cancer. BUT. Only 50% actually do so. If used to inform doctors and their patients, that’s useful data. If used to beat PCP’s over the head because they are not practicing “ideal” medicine, that’s data as a four-letter word.

Bottom Line. The key idea here is really a simple one. Data will always be a four-letter word to PCP’s if they are employed as a tool to appease regulators, payors, etc.  We need data to serve as a tool for doctors to use, not to be used against them.

Easy to say, tough to implement in 2019.  

Whither Goes Marketing Research???

GfK Read this Investor Relations Communication from GFK. Those of you who “knew me when” know that I used to run the GFK healthcare business in the U.S. Throughout my entire experience with the company, I was proud to be associated with a team that had such high standards, such a great work ethic and so much potential for the future. BUT. Recent days have seen the “departure” of several top executives by mutual agreement, and have seen communications to investors that can best be summarized in the company’s own words: “GfK adjusted its outlook on August 5, 2016 in view of the challenges mentioned and the figures reported for the first half of the year. Depending on the development of the order intake in the Consumer Experiences sector and the progress of the growth initiatives in Consumer Choices sector, sales growth below the market and a lower margin than last year cannot be ruled out for 2016 as a whole.” Several articles in the trade press have voiced a concern that if this is happening at GfK, that has seen a drop in the price of its stock of 25% in the last year, the financial future of the top 5 marketing research companies may be far from a rosy one. Smaller “boutique” firms, with lower overhead and more nimble technological development skills, may be the wave of the future. Bottom Line. We have seen it in other sectors. Fundamental changes can make the biggest leaders go poof. I’m thinking that the new focus on “big data,” which involves the analysis of existing data rather than marketing research “field work,” and other trends in our digital age like do-it-yourself on-line questionnaires might make the large marketing research companies become dinosaurs.  Stay tuned!

The Medium Is Not The Message!!!

Health Care Blog No, folks. That is not a doubly bastardized typo of the famous Marshal McLuhan quote about media. Rather, it is the message of a recent post in the Health Care Blog. It is sort of brilliant, actually. The point of the article is that while Evidence-Based Medicine and much medical treatment is based on measures of central tendency (mean, median, mode), it is the shape of the distribution that should be of interest to Physicians and diagnosed patients alike. Bottom Line. Read this post and think about its message  carefully. When hope, reality and statistics get thrown into the pot and stirred, what should we be paying attention to as we develop treatment policies? Or when we choose treatments for ourselves and for our loved ones?

Big Data And ER Frequent Flyers

Sharing ER data Here is an idea that just makes sense as a way to solve a problem that doesn’t seem to make much sense. HUH? Well, it seems that there is a group of people, large enough to matter, that spends its days traveling from ER to ER, hospital to hospital. In many cases, food, a warm bed or someone to talk to are the driving forces here. Problem. Historically, the various hospitals being visited had no way of knowing that they were part of a multiple stop chain. Thus, they repeated the taking of medical histories, ordered redundant laboratory tests, assigned duplicative case managers, etc. Anybody figure out what to do here? Yup, set up a “big data” system that shares patient/visit information across neighboring hospitals in real time.  And in 2016, that is actually starting to happen. Why hasn’t it happened before?  Getting all of the hospitals to agree to a common platform has been, to put it mildly, a significant challenge! Bottom Line. Ah! But if you read the article closely, you will see a countervailing force at work here. Hospitals in a geographic region often see themselves as competing for patients and for funding dollars, so they are not necessarily supportive of sharing information that would give any kind of break to the competition. Good grief!!!

Transparency, No Action

Congress Blog   Who couldn’t see this one coming? The Sunshine Act, which requires pharmaceutical companies to report the amounts of $ that they give to physicians in everything from lunches to cash “honoraria,” has in its half a decade of existence produced scads of data. Look up your doctor and you can see to the nearest penny how much she has received from each company. BUT. As this blog points out, there is little or no action that is resulting from the availability of these data. Bottom Line. Think about it. What am I going to do if I find out that my physician received a baloney sandwich from Merck? How about two baloney sandwiches? Yell at her? Switch doctors? It’s an old thought but a true one. Data are simply the beginning of a food chain that goes down to information, insights and action. As we have said before, the stories in many areas of healthcare today are the same. Lots of data, not a lot of action. Yup, you could see this one coming!

Googling UK Health Data

Google Health in the UK Check out this article. In it, you will see the announcement of an agreement that will provide Google’s artificial intelligence subsidiary with “huge” amounts of clinical data on patients passing through three of the UK’s largest hospitals. Apparently, the NHS believes that, in return, it will receive clinical insights from “big data” analytics that will help them to improve the quality of care and to reduce treatment costs. Maybe. But read between the lines and you will find that the parties involved in this deal were initially less than forthcoming in describing its breadth and depth to the public. While Google is without a doubt a leader in its ability to wring information/insights/action out of data, their aspirations in the healthcare space are also rather lofty, thus causing me some pause in terms of their motivations here. Bottom Line. I have a question. Who is going to be monitoring the legality and ethics of this arrangement, and others like it, as we go forward into the future? I really hope that we are pleased with the answer that we get on this one!

Connected Health

  Connected HealthCheck out this piece on feedback loops in healthcare. What you will see is a reminder of something that we all know. Digital feedback, in and of itself, doesn’t necessarily lead to better patient outcomes. What you need is feedback that provides guidance, motivation, connection with the rest of a patient’s treatment protocol, etc. Bottom Line. SO. While it’s fun to watch your “numbers” on your Fitbit, Apple Watch, etc., you still need somebody to help you to interpret these numbers in a motivational way if they are to make any difference in what you do about your medical condition, health and wellness, etc.  

Debunking HIPAA In The Digital Age

Debunking HIPAA Check out this post. It provides a major heads up for all of us. Ask most people what HIPAA was supposed to accomplish, and you will get an answer that is some version of “protect the confidentiality of patient data.” In this piece, we are reminded that the HIPAA’s real purpose was to make medical data flow more smoothly among the various parties who have legitimate needs to access this information. That’s different! Given the misperception of the intent of the law, it is little wonder that, despite multiple attempts at clarification and correction, many policies and practices have been put into place to reinforce the notion that “HIPAA means NO!” Bottom Line. What’s the heads up? Simple! It’s the law of paradoxes. Pass a law to accomplish one thing and have exactly the opposite effect. We have to watch out for that one in everything we do.

$4 BILLION Invested in IBM’s Watson Health

Dr. Jain I’m going to ask for a big hunk (20 minutes) of your time today to listen to this podcast from MEDCITY NEWS. In it, you will here the story of the “cognitive” use of Watson to sort through tons of medical information. You will here Dr. Jain, CMO of Watson Health, meaningfully describe AI as it pertains to Watson as “augmented intelligence” rather than “artificial intelligence.” Translated, this means that the vision of Watson Health is to assist, rather than to replace, clinician judgment. Data, in this “ecosystem,” become “raw material.” Insights are not the final product here. Rather, the output must be a meaningful change in clinical intervention and workflow. And the venture is called Watson Health, rather than Watson Healthcare, to cut a far bigger swath, including wellness. Bottom Line. With its powerful investments, acquisitions, partnerships and technological head start, Watson could well serve a central role in the future of the health and wellness spectrum. I will be fascinated to watch, and to keep you informed, as this juggernaut continues to move forward and evolve. But in the meantime, is listening to this update worth 20 minutes of your time? I’m betting it is!