Requesting crowd Peer Review: Visual Primer on Social Determinants of Health

Hello, crowd.

If you have a few moments, I am putting up a piece that will potentially be published in the Innovation Learning Network’s ( @HealthcareILN ) most awesome annual report, Insights 2012 ( you can see the 2011 version here).

I am not a fan of closed peer review, I am one of open peer review, and since ILN happily doesn’t follow the Ingelfinger Rule, I am posting a draft of my piece here. Can you help me by letting me know:

  • Do you understand it?
  • Does it help you understand social determinants of health?
  • Would it help someone who’s not close to this topic understand it better?
  • Does the visual thinking approach connect well?

Please add your ideas / improvements / corrections / random and organized thoughts in the comments. Thank you!

Changing Behavior: Thinking visually about social determinants

For ILN Insights 2012 (draft, rev 4, 8/21/12) • Ted Eytan, MD MS MPH

The Visual Thinking Journey

At the 2011 Innovation Learning Network InPerson meeting, my group was having trouble ideating until an ILNner came by and said, “try using drawings in addition to words.” Instantly our world opened up. Departing, he recommended the book “Back of the Napkin,” by Dan Roam.

I read that as well as the follow-up book by Roam, “Blah Blah Blah: What To Do When Words Don’t Work.” And a few months later,, I was inspired to use some of these visual techniques to lay out a different kind of behavior change – from thinking about the individual’s lifestyle choices to thinking about a society’s choices that make populations healthy or not.

We are relying too much on modifying a person’s individual choices to improve health. I see it when we talk about mobile apps, challenges and games that don’t address the “causes of the causes” of a person’s behavior that are the social determinants of health.

So let’s dive in!

WHO: Many citizens, few health professionals

If you look visually at a health system, like this example based on Kaiser Permanente, you can see that we are out-personned. There isn’t enough people working in health care to reach every person we serve to guide their daily choices. People and organizations outside of health care are needed to improve health.

HOW MUCH: What we measure should matter to people and society

In a social determinants approach, a measurable goal such as life expectancy and/or disability-free life expectancy (DFLE) is used. And not only is the number important, but the inequality between the most and least is important.

The above figure (licensed under the Open Government License v1.0) is from the groundbreaking Marmot report, published in 2010 (See: Now Reading: Why a focus on lifestyle behavior change may not improve health: The Marmot Review | Ted Eytan, MD). It shows that the vast majority of the English population will never make it to retirement age without some form of disability from illness or another reason.

That’s a more meaningful, measurable goal for a person and a society – the ability to lead a long, healthy life, regardless of background.


Let’s condense all those dots

We’ll show the “spread” from healthiest to unhealthiest using smooth bars. Let’s also keep track of our own staff separately. Health care staff should be as healthy or healthier than the people they are serving, right? As best we can tell today, they are as healthy as the members they are serving, no more, no less.

WHAT: The Results We Want

Future we want

Future that scares us

There are two alternate futures. The future we want is the one where everyone is healthier. There is less of a disparity between the most healthy and the least healthy. Our staff shows even greater health gains, leading our members.

The future that scares us is the one where the most advantaged are much healthier, the least advantaged are only a little healthier.


HOW: Proportionate Universalism

“To reduce the steepness of the social gradient in health, actions must be universal, but with a scale and intensity that is proportionate to the level of disadvantage. We call this proportionate universalism.” (Marmot review)

Making investments at all levels (individual, family, community, AND society)

We should invest across all of these areas to be successful. Working at the family, community, and society level is much harder, unfortunately.

Investing too much at the level of the individual: Lifestyle Drift

Mobile apps that remind us to record our calories or promote competition among our friends are designed to change our individual choices. Why do we focus our attention here? Probably because it feels less complicated.

Part of the explanation for this emphasis lies with the comparative ease of identifying action to address behavior, rather than the complexity of addressing social inequalities shaping such behaviors. This has led to the seemingly less challenging route of lifestyle interventions – this tendency has been described as ‘lifestyle drift’. (Marmot review)

Putting it all together, The tomorrow we want

Using the visual thinking approach, this is the one image that includes the WHO/WHAT and HOW of behavior change and social determinants to improve health.

If we spend the right amount of time at the individual level as well as the societal level, people at all levels of the health gradient will improve their health, and the gap between them will decrease. We will model the improvement in our own workforce.

Putting it all together – The tomorrow we don’t want

If we invest poorly, and do what seems easiest, we’ll get the outcome we don’t want. The most advantaged, who didn’t need as much help in the first place, will get healthier. The least advantaged may or may not get healthier. Our workforce will not achieve greater health gains. Our costs, and more importantly their costs (money, time, lives) will not be manageable. That will make all of us unhappy.

Let’s go from cartoon to reality, because we’re not trying to fix a cartoon – we’re trying to fix this…

Data and images from present day Washington, DC


  • We need to understand the role of health care in improving health; we cannot do it alone •
  • We need to have a measurable goal in mind – not just pounds lost or blood pressure lowered, but longer, healthier life, and less inequality between the most healthy and the least healthy
  • When we talk about innovating in health, we need to think beyond individual interventions – this is in the scope of health care and a health care Innovation Learning Network.

“(We were taught) first, that we were not in the business of giving patients health. That our responsibility was to treat the excesses of human folly or the misfortune caused by human genes or frailty. Second, that we were not in the business of preventing death. Third, we were to intervene only if we were certain that we would make a difference. Otherwise we would simply watch and hope that tincture of time would make a difference…If we violate these three rules … we end up with unrealistic expectations from health care, much futile care, high health care costs and poor health.” – from Baicker K, Chandra A. The Health Care Jobs Fallacy. New England Journal of Medicine. 2012;366(26):2433–2435

“Morning in Eckington. Washington, DC, USA

All of the images used in this piece are below, with a few bonuses.


Love it! There is a typo here as the “the” should be “they” in “they” should be as healthy….: We’ll show the “spread” from healthiest to unhealthiest using smooth bars. Let’s also keep track of our own staff separately – the should be as healthy or healthier of the people they are serving, right? As best we can tell, they are as healthy as the members they are serving, no more, no less.

On second thought it looks like whole sentence needs to be rewritten.”…the should be as healthy or healthier of the people they are serving, right?”

katellington Kate thanks for marking up the Marmot image and providing an alternate. I agree, it’s really hard to pick up, because it’s not how we think in the United States. Let me see if I can explain it a bit better. I think it’s important to show that some people based on where they live and what their circumstances are, will not (a) get to a certain age and (b) if they get to that age, are unlikely to get there without a disability, so they won’t enjoy life. That’s what the diagram shows. Does that make it easier to figure out, or do you have any suggestions? Thanks again, Ted

tedeytan what you’ve said is clear and to the point and should be a caption/annotation of the chart. I appreciate the open crowdsource approach.  Thanks for considering my comments.

Reposting helpful feedback, from Megan Ranney, MD, MPH, FACEP (@meganranney ) – I am going to work on these and will post a reply. Thank YOU!!–Ted:I have a few thoughts on your piece; feel free to take them or leave them. I readily admit that I come from a very biased perspective (MD/MPH, strong believer in/researcher of social determinants of health). Overall, I think it’s a great overview of the topic, and I love the graphics – they make the whole article much more accessible. However, I just took a look at the ILN 2011 Insights (had never heard of this group, but very cool pieces, and I’m registering for their website!): it looks like they go for short & sweet articles, yes? With that in mind, there are a few things I found obscure, puzzling, or unnecessary. For instance:- the first 2 paras, describing why you are using “visual thinking,” distract a bit from the overall topic. (I’d say that the use of visual thinking, a la Edward Tufte, is a whole different topic, no?)- images 4/5/6 were tough to grasp. In image 4, you need an x & y axis, I think? And then you need to keep the same axis in images 5 & 6? I didn’t get how your picture your words (the image did show health disparities between providers & patients, but I didn’t “get” visually how this disparity translates to cost variability, etc)…… Could you have the width of the columns = the cost for that group? And the y axis = median “healthiness” (defined however you’d like)?….. also, I’d use a different word rather than “health gradient”? maybe “health disparities”? (more commonly used)- as I mentioned on twitter, would credit the “social ecological model” for image 7 -> this is *such* a powerful model, and folks who read this deserve to be able to learn more about it- images 12 & 13 are very powerful BUT if someone doesn’t know DC they’re kindof tough to decipher. I’m an injury prevention person and love the assault/homicide reference – but…. consider including UPenn’s “Trauma Maps” (Brendan Carr MD MS = PI) to show the disparity between injury death versus access to trauma centers?,Megan Ranney*************Megan L. Ranney, MD, MPH, FACEPAssistant Professor, Injury Prevention Ctr, Dept of Emergency Medicine

Ted, had a hard time tracking with narrative. Graphics (a lot of them) seem disconnected from associated copy, and at the end, am left with dissonance cuz I really want to get your message. I think I get the conclusion, yet the simplification message from the back of the napkin approach where images tell the story does not come through for me.Thanks for sharing….Just an ‘N’ of 1 ‘reflection, as per your request and my commitment to community. 

2healthguru Thank you Greg! This is why I choose crowdsourced peer review, to find this out before publication. I’ll gather more opinions and make changes, of course,Ted

[…] Innovation Learning Network (@HealthcareILN) Insights 2013 is here! And thanks to to the graphical stylings of Tim Rawson (@noswar) and the editing ambrosia of Chris McCarthy (@McCarthyChris), I get to be a part of it. The article as published (on page 58), entitled “7 Visual Insights of Social Determinants” was actually posted on this blog in draft form to get community feedback. That feedback was super helpful, as it exposed the first iteration as a bit of a jumbled mess (see: Requesting crowd Peer Review: Visual Primer on Social Determinants of Health | Ted Eytan, MD). […]

Ted Eytan, MD