Physician Diversity

Identifying medicine’s gender equity gaps is just the start

Identifying gender equity gaps in medicine is an important first step toward reaching an unbiased system. But coming up with solutions—and following through on those solutions—is quite another.

AMA Moving Medicine Podcast

AMA Moving Medicine highlights innovation and the emerging issues that impact physicians and public health today.

Listen Now

In the second of two “AMA Moving Medicine” podcast episodes about gender bias in medicine, Esther Choo, MD, MPH, discusses practices and processes that reduce gender bias in hiring, promotion, leadership and inclusivity.

Below is a lightly edited full transcript of the presentation. You can tune in on Apple Podcasts, Google Play or Spotify.

Dr. Choo: I'm going to present to you some broad solutions. And again, in this session, we can't solve everything. But I hope to give you a few ideas for how we might, as a field, move forward on this issue and see if you can take those back to your workplaces and hopefully disseminate them.

I think, first of all, we can't go very far without having some understanding of why inequities happen. And I think we should acknowledge all the time that this is a complicated issue. Otherwise we would've solved it already.

Related Coverage

AMA principles will further gender equity among physicians

And we tend to really focus on implicit gender bias, and trust me, more than anybody, I believe that this is a major factor in the inequities that happen. But there are a range of things, and I think we need to consider all of them, including systems or structural elements in health care that keep us from behaving differently despite our good intentions. But I do think that, no matter what, we cannot ignore the really profound role of implicit gender bias.

I think gender bias sessions are so available that I didn't attempt to replicate them in any way here. I do challenge you to seek one out if you haven't already taken one. There are also really wonderful educational materials online.

I think for this session, suffice it to say that in social science studies, we can actually randomize people by gender. And when we do, it is evident that the same exact person is evaluated differently in terms of their commitment, competence, and worth in dollars if you just change gender.

And the same thing goes for parental status, actually. If you just change parental status, the same person is undervalued if they are female who is a mother. And if you're a man who's a father, actually, your value goes up. This is kind of well-established in the literature.

So, if you ever doubt why these things happen, these are powerful forces. ...

I think what I'd like you all to do once you leave here, if these concepts of bias are not completely familiar to you, is just be more attuned at how these biases might play out day to day in the workplace. Because once you put on a bias lens, you kind of see it everywhere. Unfortunately, the literature really backs this up. At Mayo Clinic, they did this study where they looked at recordings of internal medicine grand rounds, and they realized they captured them from the moment the person was introduced.

When they broke it down by gender, they found that men were almost always introduced by their formal title, Dr. “So and So.” Whereas women, when introduced by men, were often called just by their names without their professional titles. Less than 50% of the time were their professional titles used.

I also want to talk about inclusivity because if you just pull up any journal—I challenge you, just pick any journal in your specialty, and look at the teams of people who are writing the major studies. I look at this all the time, and it is something that is obvious from across the room that men publish with other men, and they build each other's productivity like that by always sustaining these networks that create consistent productivity.

This was actually documented in a study from Brazil where they took advantage of the fact that if you are nationally funded for research, you have to enter all your work into a database, including your co-authors and your gender, for whatever reason.

Women include women at about the proportion to which they are represented in the field. The men tend to work disproportionately with all-male teams. And this is something I mention because we can all fix this tomorrow. You may not have federally funded research, but you all have some groups, committees, task forces.

Over time, everyone's responsible for a group that gives you productivity and influence.

Whatever that group is, you can stop now and consider the diversity in that group and consider moving forward, ensuring that you have representation of voices that will contribute to your project meaningfully because they come from diverse backgrounds. Of course, we are familiar with the way that we tend to put men up on stage, even when the topic really lends itself to having at least one woman, or you have the token woman.

Of course, there's a whole website devoted to this. It's a Tumblr called, “Congrats, you have an all-male panel.” For whatever reason, David Hasselhoff is the face of this.

But it's so obvious and so simple to try to include some visible diversity. Remember, people are watching, including our younger generation who are deciding if they have a seat at the table across our specialties and our influential groups. Every group I go to, I pick on a little bit about awards. So, you are not unique. But I just pulled up on the website, on AMA's website, the most recent award winners, and I wasn't cherry picking.

These were the first ones that came up, and I would say exclusive of awards that are specifically for women, these awards are male.

The student impact awards, seven out of eight of the Nathan Davis Award. The medal of valor, the AMA's medical executive lifetime achievement award, five out of six people were male. And this is again multifactorial but something that we should consider when we think about how our awards are promoted, who gets them year after year and how they’re named so that they draw a certain gender or not.

This is a moveable area of improvement for all our societies.

I think another element to solving this problem is understanding that we all gain by fixing it. Again, this is too often framed as an issue of simple fairness or integrity, and an issue that simply affects women and not others. And I want to make sure that everyone is oriented to the fact that there is a strong business case for equity, and there's a strong clinical case for it.

The business case is kind of extrapolated from the corporate world where they have looked, this is data from McKinsey, and they've looked at how diversity, both gender diversity and racial and ethnic diversity, correlates with better financial performance. And it is pretty clear, across sectors, that if you have a more diverse workforce and executive board of your company, that it actually correlates to increased financial returns compared to the median.

And conversely, if you have little diversity in your workforce and in your executive board, it actually brings you down to the lowest tertile of financial performance.

Related Coverage

Women in medicine are trailblazers, advocates and leaders

And when McKinsey has looked across elements that relate to organizational performance and success, things like bringing on the best talent, strengthening customer relationship, increasing employee satisfaction, improving their decision making, enhancing the company's image, all these things that all companies care for in addition to financial returns, diverse groups do better across the board. And this is why, in non-health care corporate America, they're going after gender and racial and ethnic diversity and equity very aggressively in a way that I have not seen come to health care.

I also want to touch on the clinical case for equity. There have been a handful of studies, really delightful over the past few years, looking at how male physician-led teams compare to female physician-led teams. Have you guys seen this study in JAMA Internal Medicine? I’m guessing it's one of the most read studies ever and certainly one of the most controversial studies ever published in this journal, in any of the JAMA journals.

They looked at a large database of Medicare patients admitted to the hospital. Of course, if you're admitted to the hospital, you don't choose whether you're going to be admitted to a male or a female hospitalist. It's a quasi-randomized study. And they found that when they looked over this huge patient database, patients treated by female physicians had lower odds of death and readmission compared with patients cared for by male physicians.

And, you know, a lot of these “gender of your treating physician” studies for years have really been about patient satisfaction. Did you like your doctor? Did you feel understood or connected? And this study was so bold because it actually looked to see if people died, which is a much more compelling outcome. It's not a one-off. An almost an identical study was published for patients who went into surgery at Canadian hospitals and found very similar outcomes.

This one is from within emergency medicine, my specialty, so I can't help but mention it.

This is from a large Florida database that looked at patients who presented to the hospital having an MI. And they found that if a female patient with MI presented at the hospital was treated by a female physician, their likelihood of survival was much higher than if they were treated by a male physician. The female-female concordance are the lines off to the right, whereas a male to female treated patient is off to the left. And this study did something else rather than just look at those outcomes. It looked at the proportion of women in a group, in an individual emergency group, and they looked at outcomes based on that.

And what they found is that female patients experienced better outcomes in EDs that a had a higher percentage of female physicians, even if, actually particularly so, for patients treated by male physicians. In other words, if you were just in a setting where there were lots of other women, where there was gender diversity in that group, everybody performed better and it gave a huge boost to the performance of male physicians. And I think that is the bottom line.

Do I think that female physicians are better than male physicians? Of course I don't think that. There's no basis for that.

But I do think that having a female physician at the lead is a marker of a more diverse health care team. And when we look at just having more females in the overall group, it will improve collective intelligence. And this is what many, many studies of diversity have shown us.

Another solution. When we talk about sexual harassment and gender bias, I find that people get uncomfortable, and the self-efficacy and confidence of high-achieving physicians drops to the floor. I can't explain this, but I can combat it a little. I think what we as a group of pretty smart and accomplished people who have done amazing things in health care need to do is use the same mental frameworks as we do for any other problem.

One is a chronic disease framework, and this I bring up just to try to get us to aim a little bit higher when we address these issues. So, in the chronic disease model we try to detect disease early. First prevent it before it happens, and then if we're not able to prevent it, then at least detect it early so that we can influence outcomes.

I mean, I'm an emergency medicine doctor. I think many of you probably know this process a lot better than I do. I would say in the case of sexual harassment in gender and equity, we basically wait for full-blown metastatic disease before we try to detect it.

And that is really my analogy for waiting until there are formal complaints or lawsuits. By the time there's a successful lawsuit or an article in the media, that has been going on for many, many years, often decades, and affecting hundreds of people. But that's what we respond to as organizations and as a field.

I would like us to really aim for excellence here and move our detection way upstream so that we can actually head off problems before they get to that point. I think this is achievable and in the realm of what we do every single day.

The other thing I want to suggest is that we take a health care quality approach to these issues.

In the Donabedian framework, we talk about putting appropriate structures in place, having process measures, and then having many carefully measured outcomes. And so, I'm at a stroke center of excellence, I'm sure many of you are, and the amount of infrastructure and monitoring of this program is mind boggling.

We put structures in place including policies, procedures, a ton of training and specialized staff. We do process measures all the time to make sure we're implementing policies and procedures appropriately, and then we track outcomes, minute outcomes that start actually before they even get into the hospital from the prehospital care, through the ED, into the inpatient ward and out the door, to 90-day outcomes and beyond. And I will say that, if I do not make my door to CT time or my CT to tPA time, I know within days I get that feedback and I am engaged in corrective action so that doesn't happen again.

And we also hold the entire community accountable for any measure, for any step in the system where that process could have broken down so that we cannot meet our target outcomes. All these things are very, very carefully defined and we work as a team to meet the target outcomes at every step of the way for the best possible patient outcomes. What if we did this for other topics in this realm? What if we did this for sexual harassment and discrimination? What would that look like?

Related Coverage

12 women using their platforms to advocate for physicians, patients

And I think it involves just the same things. It involves putting in the appropriate policies, procedures and resources. Adequate human resources. We create safe reporting pathways. We routinize training, particularly for our health care leaders, and then we invest in organizational restructuring that will change the culture that allows this problem to thrive.

What about the process measures? We need to make sure we don't just have things in place, but we're actually implementing the policies and procedures as we are saying, that they're implemented effectively, and in a timely manner.

We need to provide better supports for the reporters of harassment and discrimination during the entire process. We need to treat these not as one-off events but actually as organizational failures that deserve root cause analysis. And we need to get iterative feedback on all our processes and be continually improving them. And then we need to clearly describe outcomes that go well beyond number of formal complaints, investigations and lawsuits. We need to measure the amount of equity we have in compensation promotion and leadership and the types of sanctions that occur against confirmed transgressors.

I think we also really need to be thoughtful about the attrition rates for women in underrepresented minorities because we know that harassment and discrimination are major drivers of attrition from organizations and from the field.

What if we applied this framework to recruitment, hiring and retention of a diverse workforce? Once again, we put the structures, the policies, the procedures in place that will promote diversity. You can imagine the types of process measures that we'll put in place to make sure that our policies and procedures are working and are guarding against bias.

And you can also imagine the types of outcome measures that we'll implement at every step of the way so we can see where our diversity efforts are failing and we can extend our examination not just to successful recruitment and hiring, but to all the things that are required to retain a diverse workforce, including appropriate compensation, promotion and leadership opportunities. Does that sound fair? Of course it's all going to be the same no matter what we're looking at, no matter what individual element we're looking at.

So, I would just say, let's all try to put good stuff in place, make sure we're doing it well, and then measure, measure, and measure our target outcomes in very specific ways, just as we are in health care problems. And, of course, this is happening all around us. Last year I was so excited to follow from a distance when AMA passed its policy regarding gender equity in medicine, and we are seeing other major houses of science and medicine adopt new policies and procedures that will advance safety and equity in our workplace.

I just want to say for following the Donabedian framework, of course, the policies and procedures fall in the first bin. Then there's all the hard work of process measures and targeting and achieving outcomes and the iterative improvement of all those things, so let's celebrate the wins that include new policies, and then let's start the really hard work of making sure that we're following through with the promise of those policies.

I would love to see medicine go from being reactive to proactive on these issues. People always walk away from these sessions and they're like, "What can I do at my own shop?" And I would say there's a lot that every single person, even the medical students in the room, can do every step of the way. As you go about your day-to-day business of being engaged in the health care workforce, you can always consider the language that you use to discuss, address and recommend other people. See how your biases creep in.

How do you assess men versus women versus others for competency and value? What are the assumptions you make about commitment for men versus women versus others? How do we confer responsibility in compensation? What is the diversity of any of the teams that you build and sustain? What is the representation of speakers whenever you're involved in that process? Who is recognized and rewarded and up on stage?

All of us over the course of our careers are involved in these processes in some ways, and you can actually leave this room and start to change the way that you do these things. I also think these topics can feel uncomfortable, and we actually need to routinize them if we are actually going to solve them.

What in health care do we solve by avoiding it and failing to look under the hood? Conversations like this one feel good and actually are part of the process of getting to where we want to be. I also want to see us do structured community-wide and systems-focused solution building and step away from trying to find individual scapegoats and taking a person-level approach to problems of harassment and inequity. This is a long conversation. There's lots of work. I hope you all follow the work of TIME’S UP Healthcare. We are a large group of women nationally working very hard on all these issues and can use your support.

You can listen to this episode on Apple Podcasts, Google Play or Spotify.